computational intelligence ‐ information processing ... · fpga for a parallel image processing...
TRANSCRIPT
1 | P a g e
CIIPS is a multidisciplinary research group which brings together researchers from engineering, science, mathematics, physics and computer science. The group combines an active undergraduate and postgraduate teaching programme with pure and applied research to provide an environment in which innovative theoretical developments can be rapidly turned into technologies that provide solutions to a range of real‐world problems. The focus of the research is the development of intelligent information processing systems and their applications. The group is active in the areas of artificial neural networks, embedded systems, digital signal processing, pattern recognition, image processing, parallel and reconfigurable computing, mobile robots, software engineering, electromobility and automotive systems. The staff members of the group include: Ms Linda Barbour (Administrative Assistant), Room 4.14, Email: [email protected], Telephone: 6488 3897 Dr Adrian Boeing (Adjunct Research Fellow), Email: [email protected] Professor Thomas Bräunl (Head), Room 4.15, Lab 3.13, Email: [email protected], Telephone: 6488 1763 Professor Gary Bundell (Adjunct Professor), Room 3.11, Email: [email protected], Telephone: 6488 3897 Mr Chris Croft (Adjunct Research Fellow), Room 3.11, Email: [email protected], Telephone: 6488 3897 Professor David Harries (Adjunct Professor), Room 3.11, [email protected], Telephone: 6488 3897 Professor Terry Woodings (Adjunct Associate Professor ‐ also Computer Science and Software Engineering), Room 4.18, Lab 3.02, Email: [email protected], Telephone: 6488 2618. Associate Professor Kevin Vinsen (Research Associate Professor ICRAR), Email: [email protected]. Professor Anthony Zaknich (Adjunct Professor), Room 4.04, Email: [email protected], Telephone: 6488 1764.
AUTOMOTIVE LAB (Prof. Thomas Bräunl)
The Automotive Lab was established in 2008 and is dedicated to research in alternative drive systems, such as plug‐in electric vehicles, as well as active driving safety, such as driver‐assistance systems. The Automotive Lab currently houses five vehicles, a BMW X5, a Hyundai Getz, a Lotus Elise S2 and two Formula SAE race cars. A list of all projects is available at: http://robotics.ee.uwa.edu.au/students/projects.html
ComputationalIntelligence‐InformationProcessingSystems(CIIPS)
2 | P a g e
REV (Renewable Energy Vehicle) This is a Faculty‐wide project and looks at finding alternatives to petrol‐based cars. These projects are suitable for students in Mechanical, Mechatronics, Electrical, Computer Engineering and Computer Science. REV has converted/built the following cars to electric drive:
REV Eco (2008): Conversion of a 2008 Hyundai Getz to electric drive using DC technology
REV Racer (2010) : Conversion of a 2002 Lotus Elise S2 to electric drive using AC technology
REV SAE‐2010: Conversion of an the 2001 UWA Motorsport car to electric drive (dual motor)
REV SAE‐2012: New EV design from scratch for Formula SAE Electric (quad wheel‐hub motors)
REVski 2013: Conversion of a 2008 Sea‐Doo Jet Ski to Electric Drive Web: theREVproject.com Projects available: 1. Group Project Electric Jet Ski
a. Mechanical Students: Motor and battery mounts, force coupling, cooling system,
evaluation
b. Electrical Students: Power and instrumentation circuitry, safety systems
c. Computer Students: Motor control, driver information system
2. Group Project Formula SAE Electric 2014
Preparing the FSAE car for competitions in Europe and/or Australia and implementing intelligent
sensor‐based motor control for the 4‐wheel hub motors. Performance measurements and
improvements
3. Intelligent EV Charging
IT project requiring C/C++ skills to implement a simulated HEMS (Home Energy Management
System), which allows EV drivers to select their charging preferences according to an assumed
dynamic hourly energy tariff (minimising costs) or for assumed available solar/wind green power
(maximising environmental benefits).
4. Driver Information systems for road‐licenced REV Vehicles
a. REV Eco: Programming an EyeBot M6 (Linux) in C/C++ to display vehicle/battery status,
drive statistics and GPS navigation.
b. REV Racer: Programming a Windows PC with graphical user interface to display
individual battery status, drive statistics, GPS navigation, generate artificial engine
sound
3 | P a g e
5. Drive‐by‐Wire: SAE Electric
Change brake and steering on the Formula SAE car to optional drive‐by wire by adding a motor to
the steering column and a servo on the brake.
6. Autonomously Driving SAE Electric
Follow a pre‐set race course by using GPS, radar and imaging sensors for a fully autonomous drive.
7. Real‐Driving Simulator
Link the SAE car's controls (steering, brake, accelerator) to the inputs of a commercial driving
simulator system, so the simulation can be driven from the real SAE car. Further build a safe roll
stand, so the car can spin its wheels at a safe speed while linked to the simulator.
ROBOTICS AND AUTOMATION LAB (Professor Thomas Bräunl)
The Robotics and Automation lab has been active since 1998 undertaking research on all types of autonomous mobile robots, including intelligent driving and walking robots, autonomous underwater vehicles, and unmanned aerial vehicles. We also work on the design of embedded controllers and embedded operating systems, as well as on simulation systems. A list of all projects is available at: http://robotics.ee.uwa.edu.au/students/projects.html Projects available: 1. Autonomous Groups of Robots
Our 5 Pioneer AT outdoor robots are equipped with advanced range and positioning sensors. They can work as a team through communicating with each other or with one of several base stations.
Group Project available: In this project, a group of students will implement a navigation/exploration task for a group of autonomous mobile robots, based on the tasks of the MAGIC2010 robotics competition, based on the ROS robot operating system. These are IT projects and require good C/C++ programming skills.
Individual tasks are—
a. Robot navigation
b. Map generation, loop closing (individual robot) and map fusion (between robots)
c. Object detection using laser scanner and camera
d. Human‐robot interface through console at ground station, displaying all robot information
4 | P a g e
2. Advanced Embedded Systems
Projects available:
a. Supercomputer on a Chip: Design a SIMD parallel computer on an FPGA for a parallel image processing application. Define individual parallel processing elements and test them with Retro before implementation.
b. Development of image pre‐processing routines in VHDL for FPGA
c. Development of RoBIOS actuator and sensor routines (motor and sensor drivers) for embedded controller
3. EyeSim Simulation Systems
For the new high‐performance embedded controller, we need a new version of the EyeSim simulator, which can also emulate the new controller features such as:
widescreen, color LCD touch‐screen dual (stereo) cameras updated RoBIOS functions
Project available:
Implement new EyeSim simulator with advanced features
SYSTEMS ENGINEERING ANALYSIS MANAGEMENT (Mr Chris Croft and Professor Thomas Bräunl)
Projects Available: 1. Real time analysis of music to determine the chord progression: The aim of the project is to sample music and determining the chord progression in real time. The project will be in two stages, first stage to break down the musical sound into the harmonics and determine the root frequency as well as the harmonics. The key will be known for this process. The second stage will be to sample the whole song and determine the key, then using this information in real time to determine the chord progression. 2. Renewable energy powered unmanned aerial vehicle (UAV)
Feasibility of a long flight at high altitude UAV using renewable energy sources to provide a platform
at an altitude above commercial aviation. This project will involve investigation of sources of power
over long periods of time (in the scale of quarters if not years).
3. Tethered flight of a UAV
5 | P a g e
Tethered flight of a UAV where a UAV flies in a predetermined path as tethered from a ground
station. The AR Drone will be used for this project and will require both hardware and software.
Two projects may be offered in this topic.
6 | P a g e
Signals and Systems Engineering
The Signals and Systems Engineering (SSE) has several research laboratories outlined below: The Signals and Information Processing Laboratory undertakes key research in the areas of
speech processing and speech and speaker recognition leading to developments in voice‐activated technologies, robust speech recognition in real environments and biometric security by speaker verification.
The Control Systems Research Laboratory undertakes theoretical and applied research in the areas of mathematical modeling, state estimation, robust control and sliding mode control.
The Biomedical Engineering Laboratory undertakes research into applying theoretical control techniques to biomedical systems such as blood glucose control in diabetics, closed‐loop control of mechanical ventilation in critically ill and patient controlled analgesia.
The Signal Processing for Wireless Communications Laboratory undertakes fundamental and applied research into broadband radio communications and underwater acoustic communications leading to applications in areas such as broadband wireless to the bush, oceanographic data collection, and offshore pipeline monitoring.
The Renewable Energy Laboratory undertakes fundamental and applied research into renewable energy technologies.
Key areas of expertise
Biomedical Systems
Control Systems
Power Electronics Applications
Renewable Energy
Signal Processing for Wireless Communication
Speech and Image Processing and Recognition
Underwater Acoustic Communications
Facilities Software: Matlab, PSIM, Pspice, Nuance SDK, NS/2, Mathematica, and LabView Hardware: Specialized hardware for control, communication, power electronics, renewable energy and signal processing
7 | P a g e
Year 2014 Final‐Year Projects in SIGNALS AND INFORMATION PROCESSING SYSTEMS
Prof Roberto Togneri We are currently in the first century of the information age and the new era of information systems engineering. The Signals and Information Processing (SIP) Lab is offering exciting and challenging final‐year projects to students who can demonstrate the required motivation and passion. If you are interested in any of the projects on offer please email <[email protected]> or drop by Room 4.10 to discuss further, your thoughts and concerns, and to help you make the right choice. For reading material and resources please also have a look at the online version http://www.ee.uwa.edu.au/~roberto/research/projects2014.html. 1. Speech Enhancement and Intelligibility Speech enhancement encompasses a range of approaches which attempt to take a speech signal which has been degraded by additive noise and reverberations and by use of clever spectral and temporal signal processing is able to make the speech more intelligible for human listeners. In this project you will investigate one or more spectral and temporal approaches that are designed to deal with either or both additive or reverberant speech with particular emphasis on speech intelligibility and how it compares to the usual measures of speech quality. For example, classical speech enhancement (spectral subtraction, Wiener filtering, etc.) for additive noise and/or more novel solutions (spectral subtraction of late reverberations, inverse filtering, etc.) for reverberant speech. This can either be an experimental or more theoretical project involving analysis, implementation and evaluation of relevant signal processing theory and algorithms. Check it out: Speech Enhancement Tutorial , Review of Speech Enhancement Paper , Speech Enhancement: Theory and Practice Book , Speech Intelligibility Measures Paper 2. Microphone Arrays for Speaker Localisation and Separation Microphone arrays consist of multiple microphones geometrically arranged so as to capture the directional information of speech and interfering sources. In this way it is possible to separate different speakers and speakers from interfering noise based on their spatial location. In this project you will investigate the application of microphone array technology in the SIP Lab to the task of speech source separation. More than one project is possible covering several interdependent investigations: setup and configuration of the microphone array hardware, software and processing for microphone arrays, simple practical beamforming (BF) to separate two (or more!) actual speakers, sophisticated signal processing algorithms for the estimation of Direction of Arrival (DOA) of the desired speech source, or Blind Source Separation (BSS) of individual spatially diverse sources. Projects can be hands‐on, experimental or highly mathematically theoretical. Team projects also quite possible. Check it out: Iain McCowan's Home Page , Beamforming Tutorial , BSS Software/Data , Conv BSS , Roomsim. 3. The Cone of Silence: Active Noise Cancellation With advances in fast digital signal processing it has become possible to do quite sophisticated signal filtering, especially with audio signals and acoustics. By clever feedback signal processing and adaptive filtering one can create any virtual sound environment over a so‐called sweet spot. In this project you will examine the application of the LMS adaptive filter and the Filtered‐X variant proposed over a decade ago and build a real working active noise cancellation (ANC) system and perhaps one and for all create that cone of silence (one of failed props used in the comedy series Get Smart, that was the 60s, the problem was solved in the 90s). You can other do this at home using only two freestanding microphones and stereo loudspeaker systems or the microphone array technology in the SIP Lab. This is high‐level systems and experimental engineering project where you
8 | P a g e
will build and test how effective ANC is and the varied limitations to make it work properly. Check it out: Filtered‐X LMS Algorithm Paper , Active Noise Control in MATLAB 4. Deep Belief Networks for Object Recognition Back in the 90s came the artificial neural network (ANN) and now we get the deep learning network (DBN). Neural networks are information processing systems which attempt to learn mappings/transfer functions/processes in the same way the human neural system works: by complex non‐linear mappings and templates/patterns. The DBN first came to light in 2009 and has now spread like wild‐fire throughout the computing community, with companies like Microsoft and Google driving the technological impacts in human computer interaction and data processing. In this project for the highly motivated student looking for cutting edge research in information and signal processing you will investigate DBNs and implement a prototype working system for image/object recognition. And if you think you have what it takes then consider speech recognition (the jury is out whether DBNs will replace HMMs, ask Microsoft and Google what they think). Check it out: 3D Object Recognition with DBN Paper , Learning Deep Architectures Chapter , Deep Learning Resources Page 5. Audio‐Visual Speech and Speaker Recognition In current speech recognition, only the audio information is used, and yet it is well known that visual lip reading also works for speech recognition, especially in noisy conditions maybe the only means to understand and for hearing impaired the only way to communicate. For biometric identification, speaker recognition usually implies audio information only, and yet face recognition is just as effective, so why not combine the two together? In this project you will investigate the fusion of audio‐visual information for either speech or speaker recognition. For speaker recognition you can implement a basic audio‐visual speaker recognition prototype using standard tools for face recognition and speaker recognition and investigate different fusion strategies. You can do this by direct capture of audio‐visual features of friends and family, recordings of pertinent TV broadcasts (e.g. newsreader broadcasts) or make use of available AV corpora. Check it out: Audio‐Visual Speech Recognition Overview Paper , Audio‐Visual Recognition Overview Paper , Audio‐Visual Recognition Application Paper ,VidTIMIT Corpus , AVOZES Corpus 6. Compressive Sensing (CS) for Audio Enhancement Compressed sensing (CS) is a novel technique used to reconstruct a signal from few training examples, possibly below the Nyquist sampling rate. This has profound implications for source compression and signal acquisition. However the application of CS to audio signals has been somewhat limited. One potentially exciting approach is to use sparsity to enhance blind source separation (BSS). In this project for the highly motivated student with a solid mathematics background and keen interest in signal processing you will investigate how CS can be applied to audio separation and enhancement for both speech and music signals. Or if you have any other ideas for a project on compressed sensing / sparse representations we can consider. Check it out: CS: The Big Picture , Compressive Sensing Resources , Compressed Sensing for Audio Paper ,Using CS for BSS Paper 7. Empirical Mode Decomposition (EMD) and the Wavelet Transform For the student who wants another project which is mathematically challenging and explores novel signal processing paradigms look no further! Empirical Model Decomposition (EMD) decomposes a time‐series signal into its constituent intrinsic mode functions (IMFs) directly in the time‐domain and is especially useful for complex biomedical, financial and meteorological signals. A very recent work (still under review) has put forward an alternative intrinsic wavelet function (IWF) based on recent work on matched wavelet functions. In this project for the highly motivated student with a strong mathematics background and penchant for all things signal processing looking to be challenged by
9 | P a g e
the latest research in the area, you will be tasked to investigate the IWF and how it compares to the IMF. Check it out: EMD Demo , EMD Presentation , EMD Sound Paper , Matched Wavelet Function Paper 8. Non‐Extensive Statistics for Feature Normalisation One of the frustrations of speech recognition is the sensitivity of systems to the acquisition environment. This arises from mismatches between where the voice recognition was trained and where it is actually used. The quickest solution is adaptation and retraining to new environments. Another solution is to use speech features which are more robust to the environment. Many such robust features have been proposed. In this project you will investigate a recently proposed q‐log spectral normalisation technique based on the concept of non‐extensive statistics. For the highly motivated student with a keen interest in mathematics and algorithms the concept of non‐extensive statistics can be explored, or for the student interested in latest techniques the q‐LSMN algorithm can be implemented and evaluated for speech recognition in the presence of both additive noise and reverberant speech. Check it out: HMM Toolkit (HTK) Book , q‐LSMN Paper 9. Build Your own Speech Recognition System This is a systems engineering project where you will build and investigate the technology that underpins speech recognition system. Possible systems you may like to build include: limited vocabulary (e.g. financial transactions, control commands, etc.), English alphabet recognition (for dictation and spelling), recognition of complete phrases rather than just words, recognition in another language, recognition of connected speech (speaking a limited set of words with deliberate pauses), and real‐time voice‐activated applications (e.g. design of a reliable voice‐activated TV remote). For a more challenging project you can investigate advanced issues like: keyword spotting, task independent phone models, continuous speech recognition, tone and syllable recognition (e.g. spoken Mandarin), recognition of the confusable /e/ set of alphabets: "b", "d", "e", "g", "p", "t", and real‐time implementation with minimum memory and computational requirements, etc. Team projects also quite possible. Check it out: Speech Recognition Resources , HMM ToolKit (HTK) , CMU Sphinx ToolKit , FBDTW , WebASR
10 | P a g e
Year 2014 Final‐Year Projects in SIGNAL PROCESSING FOR WIRELESS COMMUNICATIONS LABORATORY
Prof. David Huang
Final‐year projects offered in SPWCL are focused on the implementation, simulation and analysis of various future generation communication systems.
1. Underwater Acoustic Communications
Acoustics is the primary means to achieve wireless communications in the oceans. Underwater acoustic communications could play a key role in many subsea applications such as oceanographic data collection, environmental monitoring, and offshore hydrocarbon exploration and production. In this project, you will investigate various signal processing techniques for underwater acoustic communications. Potentially, you can also build an acoustic modem (the key component in underwater acoustic communications) using a Digital Signal Processing (DSP) platform.
2. Broadband Wireless to the Bush
Wireless communications, due to its potentially low initial deployment cost, high scalability and flexibility, will play a key role in providing broadband communications to sparsely populated areas of Australia. This project focuses on promising technologies for future broadband wireless communications especially to rural areas: Multiple‐Input and Multiple‐Output (MIMO) Systems Orthogonal Frequency Division Multiplexing (OFDM) Systems Fountain Codes Multiple Access Techniques Low Density Parity Check Codes
3. Build Your Own Radio Station Using USRP Can you imagine using a single gadget to produce and receive all the possible radio signals (e.g., digital radio, TV signal, CDMA signal, GSM signal, GPS signal, to mention a few)? The USRP (Universal Software Radio Peripheral), aided with a general purpose computer and the “software‐defined radio” technology, can achieve this task. So using the USRP, you can easily build your own radio station. Potentially, you can also use the box to eavesdrop your mates’ mobile phone conversation, though we do not recommend you to do so.
11 | P a g e
Power system emulation hardware platform with interactive student interface
A/Prof. Lawrence Borle, Prof. Victor Sreeram, and A/Prof. Farid Boussaid
This project aims at developing an innovative state‐of‐the‐art power system emulation hardware platform that integrates real‐time touchscreen control/monitoring of hardware in the emulation loop. This exciting platform with interactive student interface will improve the learning experience in power systems for students at UWA and Australian universities, thereby stimulating renewed interest in this area. The proposed platform, intended for use in undergraduate and postgraduate labs, will provide students with invaluable hands on experience on the operation and real‐time response of real‐world power systems. While other educational lab equipment exist (e.g. Lab‐Volt), they are very limited in scope (power transmission/series compensation only). The scope of the proposed platform is much broader as it constitutes a contribution towards making power engineering education more attractive, modern, and effective in preparing students for power engineering careers. The proposed demonstration power system emulation platform (Figure 1) will consist of a single phase 240 volt power system complete with:
Two independent generation sources, with one variable in voltage magnitude and phase
Representative transmission line impedances on each source
Variable linear loads at the distribution end
Variable non‐linear, harmonic rich (diode rectifier) loads at the distribution end
Solar photovoltaic energy source and inverter, emulating the effect of rooftop PV in residential areas
A unified power flow controller (UPFC) to demonstrate shunt and series injection of voltage, current or impedance (to emulate various FACTS devices). The UPFC will be used to vary power flow between branches, regulate voltages and provide active filtering of harmonics.
Power system protection systems can be added in the future.
Considering UPFC mode of operation, they include operating as a series voltage source, current source or impedance (inductive or capacitive). Possible shunt modes include operating as a shunt voltage regulator either through reactive current control or through emulating a shunt inductance or shunt capacitance. In addition, in series current source mode, the UPFC will act as an active harmonic filter, demonstrating the effects of removing the current harmonics. Through the graphical user interface, students will be able to set various operating parameters, including which of the sources, loads, rooftop PV, UPFC are turned on/off. Examples of other parameters students can control include:
the active and reactive components of the linear load
the power of the non‐linear load
the mode of operation of the UPFC
the magnitude and phase of the variable source, which determines the sharing of active and reactive power between the transmission lines
12 | P a g e
Bi-D
irect
ion
al
Inve
rter
Examples of practical experimentation that could be undertaken include:
Vary the magnitude and phase of the second generator relative to the first, and observe the changes in active and reactive power flow in the two branches.
Use the UPFC as a shunt impedance, and observe the corresponding effect on voltage regulation, power flow, stability, etc.
Use the UPFC as a series voltage injection or series impedance, and observe the corresponding effect on voltage regulation, power flow, stability, etc.
Vary the magnitude of the non‐linear load, and observe the harmonic penetration into the branch circuits. Then turn on the active filtering in the UPFC and observe the change in harmonics in each branch.
Vary the rooftop PV power and observe the effects on the power system, including the case where the PV output exceeds the load.
The student interface (Figure 2) will consist of data acquisition and control hardware, two embedded computer/touchscreen displays with interactive monitoring and control software. The data acquisition will collect and display data (e.g. voltages, currents, active and reactive power flow, harmonics, etc…), in real‐time, in a user friendly graphical display. The computer‐displays will be permanently and securely mounted on the outside of the emulated power system enclosure and will be dedicated as human interface devices only. Students will interact with the system through the touchscreen, controlling power system parameters and observing values and waveforms on the graphical displays. Two 4 channel embedded oscilloscopes will be connected to the first embedded computer and will show the voltage and current waveforms in real‐time at four nodes in the power system (generator1, generator2, UPFC input, and load input). Several project students will be given the task of determining how to generate and display instantaneous active and reactive power.
13 | P a g e
Power system emulation hardware platform with interactive student interface This project aims at developing an innovative state‐of‐the‐art power system emulation hardware platform that integrates real‐time touchscreen control/monitoring of hardware in the emulation loop. This exciting platform with interactive student interface will improve the learning experience in power systems for students at UWA and Australian universities, thereby stimulating renewed interest in this area. The proposed platform, intended for use in undergraduate and postgraduate labs, will provide students with invaluable hands on experience on the operation and real‐time response of real‐world power systems. While other educational lab equipment exist (e.g. Lab‐Volt), they are very limited in scope (power transmission/series compensation only). The scope of the proposed platform is much broader as it constitutes a contribution towards making power engineering education more attractive, modern, and effective in preparing students for power engineering careers. The development of the proposed power system emulation hardware platform will involve around fifteen undergraduate design/research projects and 3 PhD students over the coming year. The proposed demonstration power system emulation platform (Figure 1) will consist of a single phase 240 volt power system complete with:
14 | P a g e
Two independent generation sources, with one variable in voltage magnitude and phase
Representative transmission line impedances on each source
Variable linear loads at the distribution end
Variable non‐linear, harmonic rich (diode rectifier) loads at the distribution end
Solar photovoltaic energy source and inverter, emulating the effect of rooftop PV in residential areas
A unified power flow controller (UPFC) to demonstrate shunt and series injection of voltage, current or impedance (to emulate various FACTS devices). The UPFC will be used to vary power flow between branches, regulate voltages and provide active filtering of harmonics.
Power system protection systems can be added in the future.
Considering UPFC mode of operation, they include operating as a series voltage source, current source or impedance (inductive or capacitive). Possible shunt modes include operating as a shunt voltage regulator either through reactive current control or through emulating a shunt inductance or shunt capacitance. In addition, in series current source mode, the UPFC will act as an active harmonic filter, demonstrating the effects of removing the current harmonics.
Figure 1: Emulated Power System Single Line Diagram
Through the graphical user interface, students will be able to set various operating parameters, including which of the sources, loads, rooftop PV, UPFC are turned on/off. Examples of other parameters students can control include:
the active and reactive components of the linear load
the power of the non‐linear load
the mode of operation of the UPFC
the magnitude and phase of the variable source, which determines the sharing of active and reactive power between the transmission lines
Examples of practical experimentation that could be undertaken include:
Vary the magnitude and phase of the second generator relative to the first, and observe the changes in active and reactive power flow in the two branches.
Use the UPFC as a shunt impedance, and observe the corresponding effect on voltage regulation, power flow, stability, etc.
15 | P a g e
Use the UPFC as a series voltage injection or series impedance, and observe the corresponding effect on voltage regulation, power flow, stability, etc.
Vary the magnitude of the non‐linear load, and observe the harmonic penetration into the branch circuits. Then turn on the active filtering in the UPFC and observe the change in harmonics in each branch.
Vary the rooftop PV power and observe the effects on the power system, including the case where the PV output exceeds the load.
The student interface (Figure 2) will consist of data acquisition and control hardware, two embedded computer/touchscreen displays with interactive monitoring and control software. The data acquisition will collect and display data (e.g. voltages, currents, active and reactive power flow, harmonics, etc…), in real‐time, in a user friendly graphical display. The computer‐displays will be permanently and securely mounted on the outside of the emulated power system enclosure and will be dedicated as human interface devices only. Students will interact with the system through the touchscreen, controlling power system parameters and observing values and waveforms on the graphical displays. Two 4 channel embedded oscilloscopes will be connected to the first embedded computer and will show the voltage and current waveforms in real‐time at four nodes in the power system (generator1, generator2, UPFC input, and load input). Several project students will be given the task of determining how to generate and display instantaneous active and reactive power.
Figure 2: Emulated Power System Interface Diagram
16 | P a g e
The second embedded computer (Figure 2) will be used to allow the students to configure the power system and components as they see fit, and view what is happening in real‐time. The SCADA system will receive system parameter values (voltages, currents) through the readily available CompactRIO hardware. The SCADA system will communicate with the power converters (UPFC, VSI and configurable loads) through a serial modbus protocol to send commands and receive converter operating parameters. System voltage and current (obtained from the CompactRIO hardware and power converters) magnitudes and phases will be displayed for several nodes in the power system circuit. The SCADA software, Vision Module from Ignition, will provide powerful HMI/SCADA clients that launch anywhere on the network. Specific Final Year Project Descriptions 1. SCADA system hardware: ‐ choosing the appropriate equipment; touchscreen display, RTU, measurement, communications? etc. ‐ designing the wiring, and getting it done. ‐ putting the system together with the rest of the project. ‐ preparing user and maintenance documentation. 2. SCADA system software: ‐ choosing the appropriate software etc. ‐ designing the user interface, and getting it done. ‐ setting the software up so that different lab exercises can be implemented. ‐ putting the system together with the rest of the project. ‐ preparing user and maintenance documentation. 3. Voltage and Current waveform display: ‐ choosing the appropriate equipment; display, oscilloscope, measurement, etc. ‐ designing the wiring, and getting it done. ‐ putting the system together with the rest of the project. ‐ preparing user and maintenance documentation. 4. Power system emulation: ‐ determining appropriate power system parameters (inductors, resistors, capacitors, transformers, switchgear) ‐ choosing the appropriate passive components, including any variable or stepped component values, etc. Coordinate with others and power converter supplier. ‐ coordinate power ratings of all components in the whole system. ‐ designing the wiring, and getting it done. ‐ putting the system together with the rest of the project. ‐ preparing user and maintenance documentation. 5. PV system integration: ‐ determine how the PV system can be implemented into the power system emulation ‐ choosing any required equipment (transformer? control interface, communications?), etc. Coordinate with others and power converter supplier. ‐ designing any wiring, and getting it done. ‐ putting the system together with the rest of the project. ‐ preparing user and maintenance documentation.
17 | P a g e
6. Lab exercise preparation: ‐ determining what lab exercises should be undertaken by students when the system is operational, considering power flow control, active and reactive power, harmonics, stability, etc. ‐ ensure the other student designs will be suitable for the lab exercises. ‐ SCADA system programming to implement the lab exercises ‐ putting the system together with the rest of the project. ‐ preparing user and maintenance documentation. 7. Project Coordinator ‐ One student could do their project on project management, and make sure everything gets done in an organized coordinated way. (The REV project has this kind of project)
18 | P a g e
Investigation into the use of low voltage inverters to provide a UPFC function on SWER lines
A/Prof. Lawrence Borle, Prof. Victor Sreeram, A/Prof. Farid Boussaid, Le Truc (PhD student)
Single Wire Earth Return (SWER) systems (see figure in appendix) are used as an economical power transmission in rural areas of the world where loads are sparse. Invented by Lloyd Mandeno in New Zealand in 1925 to be used for electrifying New Zealand’s rural areas, today, we have over 200,000 km of SWER systems installed around Australia and New Zealand. These lines are subject to large voltage variances due to the relative length and high impedance of the line, resulting in high voltages under light loading and low voltages under heavy loading. As loads continue to grow in rural distribution networks reaching its capacity some form of upgrade is necessary to provide reliability and power quality expected in the 21st century. Due to low load densities and long distances involved, the conventional upgrades of SWER such as conversion to three‐phase power may be expensive and difficult to justify economically. The project funded by ASTP investigates the use of Unified Power Flow Converter (UPFC) to provide cost effective alternatives to the conventional SWER upgrades. The specific final year projects on this topic include:
1. Complete a study into UPFC power flow at 50 Hz and at harmonic frequencies. Simulate the operation of the UPFC on SWER lines. Demonstrate the effectiveness of the UPFC conceptually to produce voltage regulation and act as an active filter.
2. Design and assemble the SWER emulation hardware (Variac and configurable Pi network components) ensuring electrical safety. Assemble configurable linear and nonlinear loads. Connect the UPFC to the SWER emulation and loads, ensuring electrical safety.
3. Implement control algorithms to operate back‐to‐back inverters as a UPFC with voltage
19 | P a g e
regulation with the SWER emulation and loads. Procure custom‐made fully programmable back‐to‐back inverters which are suitable for this purpose. Implement a user interface which will allow monitoring and control of the UPFC from a computer. Demonstrate the effectiveness and analyse the operation in terms of energy efficiency and SWER line capacity improvements.
4. Simulate the effect of adding energy storage to the UPFC. Demonstrate whether battery charging and management can be achieved without any adverse effect on the voltage regulation and active filtering.
5. Implement control algorithms to operate back‐to‐back inverters as a revised UPFC which includes additional requirements for transformer‐less operation with access to the dc‐link. Procure custom‐made fully programmable back‐to‐back inverters which are suitable for this purpose. Demonstrate the effectiveness and analyse the operation in terms of energy efficiency and SWER line capacity improvements.
6. Add the batteries to the UPFC. Demonstrate the effectiveness of the developed voltage regulation, active filtering and battery management. Add monitoring and control of the battery management to the user interface. Analyse the operation in terms of energy efficiency and SWER line capacity improvements.
Appendix: How does SWER work?
A diagram of a SWER system is shown in the figure. The system uses a distribution isolation transformer feeding a single wire or a cable with earth used as a return path for the system. The wire has multiple termination points along its length and can extend for hundreds of kilometres. At each termination point a step down transformer is used to step the voltage down to two 240 V supplies via a centre tap or a single 480 V supply.
20 | P a g e
Year 2014 Final‐Year Projects in RENEWABLE ENERGY LABORATORY
Prof. Kit Po Wong, Prof. Herbert Iu and Prof. Tyrone Fernando
1. Stability Analysis of a DFIG Wind Turbine System (2 Projects) Presently there is a global concern about the economic downturn and a green earth which in turn is related to a better and efficient method to generate and transmit electric power. Wind energy systems are becoming popular. Doubly fed induction generator (DFIG) is a popular is a popular wind turbine system due to its high energy efficiency, reduced mechanical stress on the wind turbine, and relatively low power rating of the connected power electronics converter. The DFIG is also complex involving aerodynamical, electrical, and mechanical systems. With increasing penetration level of DFIG‐type wind turbines into the grid, the stability issue of DFIG is of great importance to be properly investigated. The aim of this project is to study the small signal stability of the DFIG wind turbine system.
2. Control Strategy of DFIG Wind Turbines for Power System Fault Ride Through (2 Projects) Doubly fed induction generator (DFIG) is a popular wind turbine (WT) system due to its high energy efficiency, reduced mechanical stress on the WT, and relatively low power rating of the connected power electronics converter of low costs. With increasing penetration level of WTs into the grid, the wind power grid connection codes in most countries require that WTs should remain connected to the grid to maintain the reliability during and after a short‐term fault. The ability of WT to stay connected to the grid during voltage dips is termed as the low‐voltage ride‐through (LVRT) capability. The aim of this project is to develop a control strategy for both the rotor and grid side converters to enhance the LVRT capacity of the DFIG WT.
3. Functional Observers (2 Projects) Functional observers can estimate a linear function of the state vector without having to estimate all the individual states. Direct estimation of any required linear functions of the state vector results in reduced order observers. The aim of this project is apply the theory of functional observers to the problem of Load Frequency Control in Power systems and also use the theory in processing measurements from Phasor Measurement Units in Wide Area Monitoring of Power systems.
4. Three‐port dc/dc converter design (2 Projects)
Due to the growth of usage of renewable energy, multi‐input single‐output converters become
popular. The aim of this project is to design 3‐port dc/dc converters using the fundamental power
flow and circuit theory. Good mathematical skills and simulation skills are essential in this project.
21 | P a g e
MICROELECTRONICS RESEARCH GROUP
The Microelectronics Research Group (MRG) is headed by Professor Laurie Faraone, and consists of 6 academic staff, 12 research staff, and 21 postgraduates. The MRG has built up to be one of Australia’s largest and most respected microelectronics research groups. This has led to collaboration with a wide range of international and Australian industry and research organisations1. In 2008, the Microelectronics Research Group was awarded a prestigious Eureka Prize for Science in Support of Defence and Security for its world leading work.
The projects undertaken by the MRG cover:
Microelectronics,
Optoelectronics,
Micro‐electromechanical systems (MEMS),
Nanotechnology, and
VLSI
with applications in:
Agriculture,
Defence,
Manufacturing,
Medicine,
Remote sensing and environmental monitoring,
Spectroscopy,
Surveillance, and much more
The work covers areas from semiconductor device modelling, fabrication, and fundamental material characterisation, through instrumentation and control electronics, to systems integration and analysis. The support for the research and engineering projects undertaken by the MRG comes from a mixture of Government and industrial funding.
1 Some of the organisations with which the MRG has ongoing links include: Defence Science and Technology Organisation (DSTO, Australia), Tenix Defence Systems (Australia), CSIRO Division of Telecommunications and Industrial Physics (Australia), Australian Institute of Nuclear Science and Engineering (AINSE, Australia), Integrated Spectronics (Australia), Grain Research and Development Corporation (GRDC Australia), IMEC (Belgium), CEA-LETI (France), CNRS (France), University of Tabriz (Iran), Technion (Israel), Korean Advanced Institute of Science and Technology (KAIST, Korea), Vigo Systems (Poland), US Navy Space and Naval Warfare Center (SPAWAR, USA), Lakeshore Cryotronics (USA), University of California at Santa Barbara (UCSB, USA), University of Illinois at Chicago (UIC, USA), Naval Research Laboratories (NRL, USA), DRS Infrared Technologies (Dallas, USA), Defence Advanced Research Project Agency (DARPA, USA), Army Research Labs (ARL, USA), Night Vision Electronic Sensor Directorate (NVESD, Washington, USA), Raytheon Vision Systems (Santa Barbara, USA), University of New Mexico (Albuquerque), Charles University, Prague (Czech Republic), Parma University (Italy), Agilent Technologies (Australia and USA), Goodrich ISR (USA), Panorama Synergy (Australia)
22 | P a g e
The facilities available are among the best in Australia in the area of semiconductor material and device fabrication, characterisation and modelling. The group operates a nanofabrication facility for fabrication of semiconductor devices, and has a multitude of test and characterisation equipment, as well as commercial packages used for device and circuit simulations, modelling and layout. For many projects in this area students will use state‐of‐the‐art equipment and techniques to measure a number of important semiconductor or device parameters. In some projects students will also undertake data analysis and help develop explanations of the observed semiconductor material/device behaviour, much of which will be new and never before reported. There is also scope to develop automated instrumentation control software in some areas.
All the final year projects offered by the MRG are self‐contained but are related to, or support, larger ongoing projects (for more information about these visit the MRG web page at http://mrg.ee.uwa.edu.au/ ). The type of work involved varies from fundamental and theoretical research through to applied instrumentation control and development. Most projects include some mixture of experimental and theoretical work. Final year students will work alongside postgraduate students and research staff with all projects conducted in a collegiate atmosphere of collaboration between students and staff.
The following sections describe the general areas of research undertaken by the MRG. There are a number of projects available in each area. Most projects tend to be somewhat open‐ended as would be expected with a research program. The exact limits and expected outcomes of your project depend to some extent on your interests, and will be defined in consultation with you. You will be part of a team of staff and postgraduate students, each working on different aspects of a larger project and who are interested and keen to see your project succeed. In return for this support, you will be expected to regularly report your results to the group and take a wider interest in the overall project. For more information about projects, please contact any of the members of the MRG. Particular experience and/or skills are advantageous to achieve the requirements of some of the projects described. However, the scope of projects is sufficient to allow the student to obtain these skills during the course of project.
Academic Staff
Prof. Lorenzo Faraone (Head, MRG) [email protected] Rm: 1.78 Associate Prof. Farid Boussaid (VLSI) [email protected] Rm: 4.20 Prof. John Dell [email protected] Rm: 1.68 Prof. Brett Nener [email protected] Rm: 1.72 Prof. Adrian Keating (MechEng) [email protected] Rm: 1.02B (ENCM) Prof. Gia Parish [email protected] Rm: 1.76 Research Staff
Prof. Jarek Antoszewski [email protected] Rm: 1.71 Dr Robert Basedow [email protected] Rm: 2.02 Dr John Bumgarner [email protected] Rm: 1.66 Dr Nima Dehdashtiakhavan [email protected] Rm: 1.21B Dr Venkatesh Chenniappan [email protected] Rm: 1.21B Dr Renjie Gu [email protected] Rm: 1.21B Dr Fei Jiang [email protected] Rm: 1.02 Dr Gregory Jolley [email protected] Rm: 1.21B Associate Prof. Wen Lei [email protected] Rm: 1.58 Prof. Mariusz Martyniuk [email protected] Rm: 4.17 Associate Prof. Thuyen Nguyen [email protected] Rm: 2.02 Associate Prof. Ramin Rafiei [email protected] Rm: 2.02 Associate Prof. Yongling Ren [email protected] Rm: 2.02
23 | P a g e
Prof. Dilusha Silva [email protected] Rm: 1.67 Prof. Gilberto Umana‐Membreno [email protected] Rm: 1.69 Mr Nir Zvison (Lab Technician) [email protected] Rm: 2.02 The following Postgraduate Students are doing their research with the MRG:
Roozbeh Anvari Wonjae Lee Anna Podolska Yan Wang Kirsten Brookshire Shoufeng Lin Gino Putrino Danny Wee Ben Cheah Hao Liu Balaji Sankarshanan Jing Zhang Amit Choudhary Shaohua Lu Rohit Sharda Hemendra Kala Imtiaz Madni James Sharp Farah L. Muhammad Khir Haifeng Mao Mohamad Susli Ashok Kumar Kurapati Radha Krishnan Nachimuthu Dhirendra Kumar Tripathi
International Visitors to the MRG (2013/2014):
Prof. Charles R. Becker RIBER SA, University of Illinois at Chicago, USA
Prof. Antoni Rogalski Military University of Technology, Poland
Important note: Many of the project descriptions below are deliberately broad. In most cases projects can be tailored to student preferences such as for
Theory
Modelling (in‐house or using commercial packages such as Sentaurus or ANSYS)
Measurements using high technology instrumentation, metrology tools and/or optics.
Hands‐on (“build”)‐type projects, including mechanical or electrical hardware, optics, software and/or interfacing
If in doubt, just ask the relevant supervisor(s).
A. INFRARED SENSORS
The ability of IR detectors to directly sense the thermal output of an object has found wide application in thermal imaging for medical diagnostics, bushfire detection, satellite remote sensing, search and rescue, thermal loss budget estimation, as well as the more traditional defence and aerospace applications. In addition, emerging applications of IR detectors are found in spectroscopic systems for mineral exploration, pipeline monitoring, pollution detection and identification, and gas monitoring systems. Specific examples include; detection of tumours and tissue damage, detecting illegal waste disposal by ships in harbours, preventative maintenance in electrical switchgear such as high voltage transformers.
A1. HgCdTe‐based detectors (L Faraone, J Antoszewski)
Characterisation of leading edge photovoltaic devices and/or arrays. For the fabrication of sensitive IR detectors, the highest performance is achieved in devices using the semiconductor material mercury cadmium telluride (HgCdTe or MCT). Modelling and characterisation of molecular beam epitaxially grown HgCdTe materials and heterostructures for high‐performance IR detectors.
A2. Fabry‐Perot filters for micro‐spectrometers (L Faraone, J Antoszewski, D Silva)
The Microelectronics Research Group is working on optical MEMS technology to make optical detectors which are only sensitive over a small, electrically tuneable wavelength range. In
24 | P a g e
essence, a spectrometer on a chip that has widespread applications in food security, precision agriculture, defence & security, and biomedical diagnostics.
A3. Sensor‐on‐board system for field deployable micro‐spectrometers (R Rafiei, D Silva, J Bumgarner)
The Microelectronics Research Group microspectrometer project aims to design a field deployable infrared MEMS spectrometer which will be utilised for the assessment of soil and grain parameters. The goal of this project is to design and construct a compact electronic system (i.e. sensor‐on‐board) for the MEMS spectrometer. The electronic system will include the hardware and software needed for I/O interface with the MEMS spectrometer, signal conditioning, CPU communication, and power supplies.
A4. Long‐wave infrared filter and mirror characterization (R Rafiei, D Silva, H Mao)
The Microelectronics Research Group (MRG) has recently commenced research into novel materials and fabrication technologies for long wave infrared (LWIR) Bragg mirrors with the ultimate aim of developing a compact, robust, and relatively low cost LWIR Fabry‐Perot microspectrometer. LWIR spectrometer technology has broad ranging applications in environmental monitoring and agriculture while also offering important advantages for military applications, such as night vision, and target identification in clutter. The final year project will involve radiometric modelling of the LWIR measurement system, improving the LWIR measurement system capabilities, and LWIR filter and mirror characterization studies.
A5. Silicon carbide for MEMS microspectrometer applications (J Bumgarner, M Martyniuk, D Silva)
As an expansion of the Microelectronics Research Group microspectrometer for field deployable infrared MEMS spectrometer, a project to look into using silicon carbide for the membrane material of MEMS microspectrometer would be investigated. This project would include optical modelling of the performance of the device with this material and basic optical and mechanical characterization of SiC films acquired from another university up to the fabrication of fixed Fabry‐Perot microdevices.
A6. IR Microspectrometer for viniculture (J Bumgarner, D Silva, R Rafiei)
The Microelectronics Research Group microspectrometer project aims to design a field deployable infrared MEMS spectrometer. An extension of the applications from grain into viniculture is an obvious future direction. This portion of the project would be to examine the optical spectra of juices and wines and their variations with additives to determine the measurement capabilities of the microspectrometer in this application. Potentially building an unattended measurement and reporting system would be the final deliverable of such a project building on the results from A3 above.
B. CHEMICAL SENSORS
Small, inexpensive and robust chemical sensing is now a major business, with applications from security (detection of explosives at an airport for example), to global warming (determination of total carbon in soil is important for carbon sequestration and development of carbon‐credit schemes), water recycling (monitoring of contaminants), and so on. Within MRG chemical sensors are being investigated in a variety of materials systems and device architectures. These projects are quite open and could be anything from a broad scoping study to investigating a particular type of device in detail.
25 | P a g e
B1. AlGaN/GaN based ion sensors (G Parish, BD Nener)
Characterisation and optimisation of ion sensors for environmental monitoring applications. AlGaN/GaN‐based transistor structures offer high sensitivity, robust sensors.
B2. MEMS based chemical/biological sensors (J Dell. L Faraone, A Keating, M Martyniuk, D Silva)
Design and testing of cantilever‐based sensors. and novel optical read‐out technologies based on integrated waveguides, gratings and MEMS cantilevers
B3. Lab‐on‐a‐chip (A Keating)
Lab‐on‐chip research within MRG focuses on techniques which can allow rapid analysis of ultra‐small volumes of fluid. One approach being considered is the use of acoustophoresis on‐chip (ultrasonics) to setup a standing wave within the microchannel.
C. BIOMEDICAL SENSORS
The ability to monitor biological and chemical signals with an electronic device is a tremendously innovative approach for cell research and process control in pharmaceutical and microbiological production, and chemical sensing applications.
C1. MEMS‐based biosensors (J Dell, L Faraone, A Keating, M Martyniuk, D. Silva)
Microcantilever based biosensors are a novel next generation approach to building high sensitivity sensor arrays. The aim of this project is to create a computer controlled system which focuses a pulsed high power laser onto an absorbing thin film.
C2. AlGaN/GaN based biosensors (G Parish, B Nener)
A bio‐friendly, chemically inert and stable III‐Nitride‐transistor‐based bio/chem‐sensor will be developed to detect responses to various specific compounds/chemicals, particularly through cell receptors.
26 | P a g e
D. FUNDAMENTAL MATERIAL STUDIES
The group studies a variety of materials including:
HgCdTe: The MRG has an established Molecular Beam Epitaxy (MBE) growth facility for mercury cadmium telluride (HgCdTe) semiconductor structures for high‐performance infrared detectors. MBE, a state of the art technology for semiconductor crystal growth, allows growth of layers of different semiconductors, from as thin as a single atomic layer, to layers tens of microns thick. MBE technology is very important for fabrication and design of ultra‐high performance electronics and optoelectronics devices using bandgap engineering.
III‐nitrides: The group III‐nitride semiconductor materials (GaN, AlN, InN, and their alloys), a relatively new material system, are commercially successful in short wavelength optoelectronics covering wavelengths ranging from green to ultraviolet. Research is now focussed on microelectronics. III‐nitride‐based FET technology offers significantly improved performance for applications in RF/microwave power amplifiers, high speed switching for power electronics, and operation in harsh electrically noise environments, such as the automotive industry, space applications, and switch mode power supplies.
Porous silicon: Porous silicon is a novel nano‐material with the capability to perform as a mirror, waveguide, light emitting diode, photodetector, and sensor. Aside from optoelectronics, other applications include: photonic bandgap structures in micro‐optics, solar cells for energy conversion, gas sensing for environmental monitoring, high etch selectivity for wafer technology, highly controllable etching parameters for micromachining, biosensors, and enzyme immobilization in biotechnology.
Silicon‐on‐insulator and advanced group III‐V nanostructured devices : In the present bulk‐Si nanoelectronics technology, where individual transistors are already approaching the size comparable with silicon layer thickness (tens of nanometres), the size/thickness related issues lead to fundamental problems in the process of scaling devices. It is generally believed that further miniaturisation will be achieved through Silicon On Insulator (SOI) and advanced group III‐V semiconductors, which are nanostrucure‐based technologies with material thickness approaching less than ten nanometers, thus allowing further scaling without compromising the size/thickness ratio.
Other materials systems are also investigated in collaboration with leading international researchers and industry partners.
27 | P a g e
D1. Hall measurements for transport characterization (L Faraone, G Umana‐Membreno, J Antoszewski)
This project involves characterisation of electrical transport properties of state of the art materials in a variety of technologies such as HgCdTe for infrared sensors, wide‐bandgap semiconductors transistor structures for high power electronics (SiC and GaN based), and next generation group III‐V semiconductor structures for ultrascaled electronic devices, using Hall Effect and Magnetoresistance techniques.
D2. Material quality characterization (J Antoszewski, L Faraone, G Parish, B Nener, G Umana‐Membreno, M Martyniuk, A Keating, W
Lei)
Across the various materials systems used for devices in the MRG (particularly HgCdTe, (Al,Ga,In)N and porous silicon), fundamental material studies into electronic and optical material properties are required for continued development of state of the art device technology. The MRG has some of Australia’s best facilities for undertaking materials characterisation (such as minority carrier lifetime, photoresponse, Laser Beam Induced Current, x‐ray diffraction, infrared spectroscopy) for a wide range of semiconductors. A number of projects are available in this area, working on the development of new techniques, as well as using techniques we have already established to measure the properties of semiconductor layers.
E. ELECTRONIC/OPTIC SYSTEMS
The projects in this area involve the design and production of electronic and optics systems that assist in characterising MRG devices or demonstrating the capabilities of MRG work.
E1. Electronic/Optic Systems (A Keating)
Examples could include instrumentation for sensor operation, or position control systems for material/device fabrication.
E2. A versatile optical system for MEMS device characterization (R Rafiei, D Silva, L Faraone)
MEMS optical devices are well suited for use with pixelated arrays. As such, their wider application will benefit from accurate characterization of their optical properties at high spatial resolution and over a wide range of operating conditions. At The University of Western Australia we are in the process of building an experimental system for accomplishing this. This system will enable measurements to be carried out across a wide range of temperatures and with a spectral sensitivity ranging from visible to long‐wave infrared. The final year project will involve system radiometric and ray‐trace modelling, hands‐on testing and performance characterization, followed by device measurements.
F. ATMOSPHERIC PROPAGATION (N Fowkes, B Nener)
The ultimate performance of an EO System is determined by the atmosphere. The atmosphere can degrade the signal through scattering and absorption by aerosols, background radiation, scintillation and refraction. Projects in this area are both theoretical and/or involve experimental work. The experimental work is at sites like Rottnest measuring atmospheric parameters important to EO propagation and involves the design and installation of the instruments, and the analysis and modelling of atmospheric data and effects. The theoretical work involves mathematical and numerical modelling of atmospheric effects relevant to EO systems, particularly refraction and scintillation.
28 | P a g e
F1. Refractive Index Change in Atmosphere
Modelling of the effects on light and microwave propagation of refractive index changes due to temperature gradients in the atmosphere over the ocean; modelling of mirages and other image distortions of objects seen at large distances; scintillation.
G. INTEGRATED CIRCUIT DESIGN (F Boussaid)
G1. Camera‐on‐chip
The current trend in Digital Imaging Technology is towards building camera‐on‐a‐chip imaging systems, i.e., CMOS imagers. The fully integrated product results in significant manufacturing cost savings, reduced system size, but also in lower power consumption. The unique concept of CMOS imagers offers the opportunity to integrate photo‐sensing array and signal processing circuitry on a single silicon chip, enabling the development of a new generation of smart mobile imaging systems. Half the size of a small postage stamp, a CMOS imager chip can even be swallowed (pill‐camera) to transmit images from inside the body. Besides biomedical, CMOS imagers have numerous commercial applications in cell phones, PC notebooks or any application for which a “micro‐camera” can be requested.
Microphotograph of a fabricated CMOS imager (3.5×3.5mm2)
Proposed final year projects will involve building such a camera, and optimize its performance in terms of dynamic range, resolution and/or power consumption. A research project is also available on 3D image sensors. During these projects, you will further develop your analog/digital electronic design skills.
G2. Electronic nose
Sniffing‐dogs are able to detect thousands of chemicals with high sensitivity and selectivity using only biological components. These nasal powerhouses have been successfully used to search for pipeline leaks, drugs, or explosives. You will develop a biologically inspired Electronic Nose (or ENose for short), that mimics the organization and neural processing of the olfactory bulb. The Enose will comprise a chemical sensor array and a gas recognition engine, integrated on a single chip. Projects offer an opportunity to discover and apply neuroscience principles into made‐made engineering systems. Projects will be tailored around your interests, whether neuroscience and/or integrated circuit design.
29 | P a g e
Microphotograph of a fabricated electronic nose
G3. UWA Unmanned Aerial Vehicle (UAV
We seek to develop, design and manufacture a UWA‐made UAV (Unmanned Aerial Vehicle). UAVs are man‐made flying vehicles capable of operating without a person on board. UAVs come in a large variety of sizes and shapes and applications. This project will focus on the ‘miniature’ UAV category, which is defined as having a maximum take‐off weight of 30Kg, maximum flight time of 2 hours and a maximum altitude of 300 meters. This category represents a good trade‐off between size, complexity and cost when compared to larger UAVs. Mini‐UAVs are large enough to carry useful payloads such as digital cameras, sensors and other equipment. They are also small enough that they can be built relatively cheaply and do not come with the strict regulation requirements for operating and testing larger aircrafts.
The core components of a UAV are: the airframe, the propulsion system, the avionics, the radio data link, the base‐station and the payload. Given that the UAV project has just started, help is sought at all levels and disciplines.
G4. Efficient energy harvesting interface circuits
The ever increasing demand for portable and miniature yet computationally powerful electronic devices has put stringent size and weight requirements on the power source or battery, whose capacity is in turn being increasingly limited. The proposed projects will tackle the issue and explore ways to design efficient interface circuits to extract (harvest) the maximum power from available ambient energy sources (e.g., solar power, thermal energy, or kinetic energy). During the project, you will further develop your analog circuit design skills in the area of energy harvesting and power management.
G5. Wearable Wireless Sensors
Wearable wireless sensors enable the real‐time capture of biophysical and kinematic data for a wide range of application areas from remote monitoring of epilepsy, stroke and other chronic disease patients to motion‐analysis for sports or gait analysis for the early detection of Parkinson. You will acquire and process data using wearable wireless sensors for one of the aforementioned applications. The project will involve hardware interfacing and the development of signal processing algorithms.
30 | P a g e
G6. Propose your own project
If you have a project idea for your final year project, please write a small paragraph identifying the problem/challenge you would like to address. Your proposal will then be evaluated to assess that it meets the degree requirements. The scope of the project may be reviewed accordingly. Please feel free to book an appointment with me to discuss any relevant project ideas you may have.
31 | P a g e
http://obel.ee.uwa.edu.au
Outcome focused, interdisciplinary research using light to perform imaging, sensing and
diagnosis in biomedical applications.
Our projects are designed to contribute to our research which has an international profile. As a
result they can be challenging and extremely rewarding. We offer projects in one or more of the
following subjects:
Optical engineering – design and realization of optical systems.
Theoretical optics and electromagnetic theory – development of theory underlying the imaging techniques we employ.
Instrumentation, electronics and system integration.
Image processing – improving the quality of images and extracting new types of information.
Numerical modelling – modelling of image formation.
Biology and medicine – interpretation of images and application of techniques.
Software engineering – development of robust software systems to drive our imaging systems.
It is not a prerequisite that you be experienced in these fields to undertake a project. You would be
supervised by one of OBEL’s research staff and given the opportunity to learn the valuable
transferable skills required to complete your project. OBEL has a team ethos and you would be
actively encouraged to work as part of a team.
Head of group: Winthrop Professor David Sampson
32 | P a g e
OBEL has a tradition of undertaking quality research in optical and biomedical engineering. We
attract high calibre honours students who go on to be employed by leading companies. For example,
recent graduates of OBEL have gone on to work in companies such as Google, Delloite,
Schlumberger and Finisar and universities such as MIT, University of Illinois and ANU. Recent
honours students have also published original work in academic journals such as:
B. Lau et. al, “Imaging true 3D endoscopic anatomy by incorporating magnetic tracking with optical
coherence tomography: proof‐of‐principle for airways,” Optics Express, 18(26), 7173‐27180, 2010
B.F. Kennedy et. al, “Strain estimation in phase‐sensitive optical coherence elastography,”
Biomedical Optics Express, 3(8), 1865‐1879, 2012.
OBEL’s microscope in a needle was a finalist at the 2012 Australian Museum Eureka Prizes for
Innovative Use of Technology. In actively seeking to commercialise out work, we regularly patent
novel technology developed within the group.
The OBEL team is composed of a mixture of research staff and PhD students:
Research staff: W/Prof. David Sampson, Dr Robert McLaughlin, Dr Brendan Kennedy, Dr Dirk
Lorenser, Dr Peter Munro, Mr Bryden Quirk, Mr Rodney Kirk.
PhD Students: Loretta Scolaro, Blake Klyen, Boon Yew (Teddy) Yeo, Xiaojie Yang, Peijun Gong, Kelsey
Kennedy, Lixin Chin, Andrea Curatolo, Shaghayegh Eshaghian.
In addition to our staff and students, we collaborate with international researchers based in, for
example, the USA, Italy, Poland and the UK.
All project areas can accommodate two or more students. It may not be completely clear to you
what you would actually do from the project description alone. Please take the opportunity to come
and talk to us so that we can define a project around your skills. Contact details can be found on our
webpage.
1. Medical imaging with optical coherence tomography
Optical coherence tomography (OCT) is an ultra‐high resolution medical imaging modality. Conceptually, it is similar to ultrasound imaging, except that reflections of light are detected rather than sound. This enables a much finer scale of image than is possible with ultrasound. OCT is providing images of unprecedented clarity of living biological entities and is providing new information on a variety of diseases and conditions, including cancer and muscular dystrophy.
OCT research at OBEL aims at understanding and improving the technique and in designing and building instruments for various applications, including breast cancer (with surgeons at Sir Charles Gairdner Hospital and Royal Perth Hospital), skin (scar assessment with Royal Perth Hospital), and
33 | P a g e
animal muscle tissue (for muscular dystrophy research with Miranda Grounds at UWA Anatomy & Human Biology).
Examples of possible projects:
i) Instrumentation Design: Design and construct a compact OCT system for clinical use. The focus is on miniaturization of the current optical fibre‐based system and involves the design and construction of portable/compact electronic and optical modules. The portability and reliability of the system will be tested under clinical conditions;
ii) Image Processing of medical images: OBEL is currently exploring the use of several image processing techniques, including segmentation and registration, to extract new types of information from high resolution medical images, Projects include developing new algorithms to analyse images of breast cancer, lung disease and burn scars. This is a software‐based project and will require knowledge of either Matlab or C++.
2. OCT needle probes
We have developed a number of prototype needle probes, where the miniaturised optics of the OCT system are encased in a medical hypodermic needle. These probes will enable surgeons to more accurately detect cancer during surgery, and provide new ways to assess lung diseases.
We are actively exploring new probe designs, and possible projects will focus on the optics and mechanical design of the probe itself. These are hardware‐based projects that involve researching a new probe design, fabrication of the optics and assessment of the probe.
3. High resolution elastography
Elastography is a new imaging technique which creates an image of what tissue ‘feels’ like. It can be used to differentiate between healthy and diseased tissue by measuring the elastic properties of the tissue. We have combined this technique with optical coherence tomography to achieve high resolution elastography. Examples of possible projects are:
(left): Histology of human breast cancer tissue. (right) Optical coherence tomography scan the same tissue, acquired
with an OCT needle probe.
34 | P a g e
Signal processing to estimate strain: We implement elastography by placing the tissue under mechanical load and measuring the resulting displacement using optical coherence tomography. The strain introduced to the tissue is then determined from the spatial derivative of displacement. We are exploring new algorithms to improve the strain estimation and to remove artefacts from our images. Possible projects in this area include the development of phase unwrapping algorithms and the acquisition and analysis of the Doppler spectrum.
Handheld probe design and implementation: We are developing handheld probes to allow clinicians to perform elastography measurements on patients. This involves design of an optical fibre probe and loading mechanism. The loading mechanism must be synchronised with the image acquisition. Also, it is important to measure the force exerted by the probe on the tissue. Your project could involve working in part of the team aiming to develop the first handheld optical elastography probe.
4. Optical theory and modelling
We have developed the most realistic models of image formation in optical coherence tomography
currently available. We are very keen to put the models to work to make new discoveries about
imaging in biological tissue. We offer the following projects in this area:
i) Rigorous modelling of light tissue interaction for large volumes of tissue. We have expertise in using rigorous techniques such as the finite difference time domain method to model light scattering however we would like to extend this technique to be able to model larger volumes of tissue. We intend to make use of a technique which makes use of Fourier transforms to calculate spatial and temporal derivatives which allows a substantial reduction in the field sampling density. This project would require a student with, or a desire to develop, skills in electromagnetic theory and C/C++ programming.
(a) Optical image; and (b) Strain image of a tumor-simulating material.
A focused beam propagating in air (top) and a focused beam travelling through tissue (lower)
35 | P a g e
ii) Another way of extending the volumes of tissue which we can model is to parallelise our current finite difference time domain code. This code could run on iVEC’s cluster computer. Again, this project would require a student with, or a desire to develop, skills in electromagnetic theory and C/C++ programming.
iii) The main factor limiting optical coherence tomography’s imaging depth is scattering by biological tissue Two images on this page demonstrate how scattering effects the beam used for imaging (see above) and also how it results in image distortion (see right). We have a project to try out a number of different types of beams theoretically and experimentally. This is an exciting do some cutting edge research.
iv) We have a number of other modelling problems to solve, please come and discuss with us if you would like to do a modelling project but don’t particularly like those mentioned above.
5. Investigation of optical properties of biological samples
One goal in OCT is to measure the optical properties of different tissue samples. This includes measuring how rapidly light attenuates and the directionality of optical scattering. It is important to benchmark optical property measurements in OCT using independent techniques. A possible project would involve building optical setups using components such as lasers, optical fibres, goniometers and integrating spheres to allow this benchmarking to be performed. Another possible project would involve characterisation of nanocrystals for use in optical spectroscopy.
6. Visualisation and rendering
A large component of our research involves
efficient handling of large 3D datasets.
Possible project in this area are related to
visualisation, feature extraction or GPGPU
(General Purpose Graphics Processing Unit)
development. This could involve real‐time 3D
rendering for our optical imaging systems.
3D rendering of an OCT scan of lung tissue, showing alveoli
and small airways (bronchioles).
Experimental OCT image (a) and simulated images (b) and (c) with increasing amounts of
36 | P a g e
7. Improving OCT image quality
OCT images are subject to a granular or mottled appearance, similar to ultrasound, due to speckle.
We are interested in developing techniques to reduce speckle and improve image quality, whilst
maintaining high resolution. A possible project would involve developing new techniques and
comparing existing techniques to achieve this.
One technique that we are keen to investigate uses sparse image representations. This technique
has already been demonstrated in photography and has great potential for denoising and feature
extraction in optical imaging.
8. Fluorescence imaging and tomography for small animal imaging
The use of fluorescence tomography and planar imaging is increasing rapidly in medical research. It
allows for molecular imaging, albeit at low spatial resolution, in live small animals in such a way that
the animal is not harmed. We are at the very beginning of building a fluorescent tomography system
and are offering projects in the design and construction of the system. Broadly speaking, the system
will be composed of a broadband light source with an array of spectral filters, a photographic
camera with a low noise CCD and a tuneable spectral filter to filter light which enters the lens of the
camera. The camera and tuneable filter will all need to be computer controlled. We will also need to
develop a means of determining the three‐dimensional surface of the imaged object, which in the
first instance will be a phantom, not an animal. This project will also have significant signal
processing and information extraction aspects.
This will be an excellent project for student wishing to work with various types of hardware which
need to combine into a single system. It is also an opportunity to join a project at the ground level.
37 | P a g e
CEED (Co‐operative Education for Enterprise Development) is a formal program designed to link the abilities and training of undergraduate and postgraduate students with research and development needs of progressive organisations within the wider community. The CEED program at UWA gives access to any discipline within the University CEED joint research projects include a period of full time work on the project at the client's site at some stage during the project, as well as the normal time invested on campus. CEED clients determine the topics, which are set up as extended undergraduate or postgraduate projects. Though seldom required, project content can be adapted to increase in academic demands to ensure high grade students can demonstrate their academic ability. Clients provide resources particular to their project, and contribute a fixed sum as their share of other costs. Two DTSO atmospheric propagation CEED projects are offered and are listed below. Students will need
to apply for these projects through the CEED program. Full details of the projects and the application
procedures are available at www.ceed.uwa.edu.au. Applications close on Friday October 4th.
Refractive Effects in the Maritime Boundary Layer (Supervisor – Brett Nener)
Modelling and simulation improvements are sought in the following areas:
• Effects of sea‐spray on periscope windows
• Effects of camera movement (orientation and position) on simulated video (motion blur)
• Refractive effects (e.g. mirage) in the maritime boundary layer
Sea spray impacting on the periscope and the ensuing drain‐down over the periscope window can
impede periscope operations. Not only in the visible band but also in the thermal wave bands where
transmission and emissivity effects may impact on performance and small droplets on the window
may be falsely perceived as detections.
Further research is also sort in the area of refractive effects in the maritime boundary layer. Such
effects can cause false horizons, obscure surface objects, and render recognition tasks more difficult.
Previous work on analytical forward and inverse solutions based on linear, quadratic, or exponential
refractive index height profiles provides a solid foundation for continued study.
38 | P a g e
The Inverse Solution of Refractive Effects in the Maritime Boundary Layer Near the Sea Surface
(Supervisor – Brett Nener)
Modelling and simulation improvements are sought in the following areas:
• Effects of sea‐spray on periscope windows
• Effects of camera movement (orientation and position) on simulated video (motion blur)
• Refractive effects (e.g. mirage) in the maritime boundary layer
Sea spray impacting on the periscope and the ensuing drain‐down over the periscope window can
impede periscope operations. Not only in the visible band but also in the thermal wave bands where
transmission and emissivity effects may impact on performance and small droplets on the window
may be falsely perceived as detections.
A refraction model is needed which accepts meteorological input data and estimates the refractive
index profile over a given wavelength band (including visible and thermal bands). The model would
then use analytical solutions to predict the associated ray paths for a given line of sight and field of
view. This research will also involve development of the exponential analytical solution. Secondly,
further study is sort to develop the inverse solution using real imagery of known maritime surface
objects in known meterological conditions, and estimate the effective refractive index profile over a
range of weather conditions. DSTO can assist with provision of reference data sets if required.