a development of hybrid genetic algorithm 11 & type-2...
TRANSCRIPT
Phayung Meesad
The objectives of this research are to
1. Propose a prototype system combining genetic algorithm with type 2 fuzzy logic system.
2. Apply type 2 fuzzy optimized by genetic algorithm to work with the FPGA for applications in control systems.
3. Evaluate the performance of the proposed system.
King Mongkut’s University of Technology North Bangkok
OBJECTIVE
Kanchana Viriyapant
METHADOLOGY RESULTS & DISCUSSION
The proposed interval type 2 fuzzy logic system, which can
well handle uncertain data, will be embedded in hardware FPGA
that can be apply to control applications that need fast operations.
FPGAs allow for the implementation of an ideal mix of peripherals
and system infrastructure. Genetic algorithm is used to fine tune the parameters of interval type 2 fuzzy system.
RESULTS & DISCUSSION
Figure 2: Fuzzy Logic Controller for DC Motor Simulation
Control Rule of Fuzzy Logic Controller for DC Motor
GUIDELINES FOR
THE INNOVATION
Acknowledgement
This research work is financially supported by Office of the Higher Education
Commission, and King Mongkut's University of Technology North Bangkok (contract no. 2554A11962031)
Contact Phayung Meesad, tel: 0898918466, email: [email protected]
A Development of Hybrid Genetic Algorithm
& Type-2 Fuzzy Logic System on FPGA and
Applications for Control Systems
Suwannee Thubjeen
0
1
0
1
0
1
1x
2x
px
y0
1
1 1
1 1 1
1
2 2
2 1 1
2
1 1
: ,
: ,
: ,
p p
p p
l l
l p p
Rule If x is F and and x is F
Then y is G
Rule If x is F and and x is F
Then y is G
Rule If x is F and and x is F
T l
hen y is G
Interval Type 2 Fuzzy Logic System
Genetic Algorithm
Chromosome Encoding
t
Targets
Inputs
Outputs
Chromosome
Fitness Evaluation
Mate Select
CrossOver
Mutation
Chromosome Decoding
Chromosome
Criteria Met
Yes
No
Figure 3: Step Response of DC Motor Speed Control
1. If (error is NB) and (derror is NB) then (du is PB) (1)
9. If (error is NM) and (derror is NM) then (du is PM) (1)
15. If (error is NS) and (derror is NB) then (du is PB) (1)
22. If (error is Z) and (derror is NB) then (du is PM) (1)
29. If (error is PS) and (derror is NB) then (du is Z) (1)
35. If (error is PS) and (derror is PB) then (du is NB) (1)
36. If (error is PM) and (derror is NB) then (du is Z) (1)
42. If (error is PM) and (derror is PB) then (du is NB) (1)
43. If (error is PB) and (derror is NB) then (du is Z) (1)
49. If (error is PB) and (derror is PB) then (du is NB) (1)
A DC motor Simulink Model with load is used in the
experiment for Fuzzy Controller. The step response simulation results is
shown in Figure 3. The data will be collected from simulation and
used for Interval Type 2 Fuzzy Logic Controller, which is automaticaly
generated and optimized by the genetic algorithm. Fuzzy logic
controller is implemented suing Simulink of Matlab and will be
transformed to VHDL and downloaded to Sparttan 6 Xilink GPGA for real control situations.
Figure 1: Hybrid Interval Type 2 Fuzzy and Genetic Algorithm