research article measurement of international roughness ...with -axis accelerometers and a gps...

Research ArticleMeasurement of International Roughness Index byUsing 119885-Axis Accelerometers and GPS

Yuchuan Du Chenglong Liu Difei Wu and Shengchuan Jiang

Key Laboratory of Road and Traffic Engineering of the Ministry of Education Tongji University Shanghai 201804 China

Correspondence should be addressed to Shengchuan Jiang jiangsc87gmailcom

Received 3 March 2014 Revised 21 May 2014 Accepted 29 May 2014 Published 30 June 2014

Academic Editor Andy H F Chow

Copyright copy 2014 Yuchuan Du et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

The International Roughness Index (IRI) is a well-recognized standard in the field of pavement management Many different typesof devices can be used to measure the IRI but these devices are mainly mounted on a full-size automobile and are complicated tooperate In addition these devices are expensive The development of methods for IRI measurement is a prerequisite for pavementmanagement systems and other parts of the road management industry Based on the quarter-car model and the vehicle vibrationcaused by road roughness there is a strong correlation between the in-car 119885-axis acceleration and the IRI The variation of speedof the car during the measurement process has a large influence on IRI estimation A measurement system equipped with 119885-axisaccelerometers and aGPS devicewas developedUsing the self-designingmeasurement systembased on themethodology proposedin this study we performed a small-scale field test We used a one-wheel linear model and two-wheel model to fit the variation ofthe119885-axis accelerationThe test results demonstrated that the low-cost measurement system has good accuracy and could enhancethe efficiency of IRI measurement

1 Introduction

The IRI was developed in 1986 using the results of theInternational Road Roughness Experiment performed inBrazil in 1982 [1] Since then the IRI has become a well-recognized standard for the measurement of road roughnessThe main advantages of the IRI are that it is stable over timeand transferable throughout the world

The IRI is an index defined by applying the algorithmproposed by Sayers [2] to a measured realization of thelongitudinal profile The measurement of roughness is quitedifficult and complex because it depends on the vehicularcharacteristics in addition to the actual pavement situation[3] Moreover the road roughness levels are readily affectedby vehicle structures and driving speed During the course ofhalf a century of development engineers and scientists haveinvented several techniques andmethods for measuring roadroughnessThemeasurement devices can be divided into fourgeneral types [4] response-type road roughness measuringsystems (RTRRMS) direct profile measurements indirectprofile measurements and subjective rating panels TypeI devices measure the pavement roughness by correlating

the RTRRMS measurements with the IRI calculated from aprofile for example using a bump integrator or NAASRAroughness meter Type II devices measure the road profiledirectly which involves measuring each wheel trackseparately for example using a 3-meter long beam or laserroad surface tester (LRST) Type III devices measure thelongitudinal profile over the wavelength range of interestfor example using a General Motors Research (GMR)profilometer Type IV devices evaluate the pavement qualitybased on assessment guidance and personal experience Thecommon measurement methods are compared in Table 1

Most highway agencies collect IRI data using a laserroad surface tester or GMR profilometer This equipmentmeasures surface profiles at traffic speeds and providesexcellent results for use in network analysis for pavementmanagement systems However because these devices aremounted on a full-size van automobile or trailer it is difficultto use them on the roadway for short periods of time Inaddition these devices are rather expensive and delicate Forthese reasons they are not effective for providing feedbackto contractorsrsquo crews Hajek et al [5] analyzed the influenceof several different factors on the IRI data that was collected

Hindawi Publishing CorporationMathematical Problems in EngineeringVolume 2014 Article ID 928980 10 pageshttpdxdoiorg1011552014928980

2 Mathematical Problems in Engineering

Table 1 Characteristics of common road roughness measurement methods

Method of measurement Principal of operation Measurementdevice Advantage Disadvantage

3-Meter Long Beam Direct profilemeasurement Aluminum beam Simple reliable data

collectionInefficientlower mechanization

Level Direct profilemeasurement Level and rod Time-stable

straightforward Time-consuming inconvenient

Bump integrator RTRRMS LVDT displayer and acartrailer Simple fast

Time-instable affected by vehiclevibration performs at the samespeed

Laser road surface tester Direct profilemeasurement

Car with laser deviceand calculators Straightforward efficient High cost of operation and

maintenance

GMR profilometer Indirect profilemeasurement

LVDT accelerometerpotentiometer and a van Convenient efficient

Precise instrument requiredhigh cost of operation andmaintenance

and made several recommendations for IRI measurementIRI measurement at the network level has become a routinepractice for many road agencies in recent years On the otherhand IRI measurement at the project level is also requiredprimarily for accepting or price-adjusting paving contractorsrsquoproducts The development of methods for IRI measurementis a prerequisite for a pavement management system (PMS)and other parts of the road management industry

An accelerometer is a device that measures the acceler-ation in one two or three orthogonal axes and they areused widely in the fields of civil engineering biology andindustry Accelerometers can be used to measure the vehicledriving status where they facilitate evaluations of the overallvehicle performance and responseThis information can thenbe used tomake adjustments to various vehicle subsystems asnecessary [6ndash8] Accelerometers can also be used to measureseismic activity inclination machine vibration dynamicdistance and speed with or without the influence of gravity[9 10]

This study focused on building amodel for estimating theIRI as well as developing an effective and low-cost system formeasuring the IRIThe steps of this study which are reflectedby this organization of this paper are as follows

(i) Introduction to the principle of using 119885-axisaccelerometers to measure IRI

(ii) Modeling the relationship between the variation inthe in-car 119885-axis acceleration and the IRI

(iii) Development of a self-designed measurement systemwith 119885-axis accelerometers and a GPS device

(iv) Field testing results

2 The Principle of Using 119885-AxisAccelerometers to Measure IRI

21 International Roughness Index To study the effects ofthe road pavement characteristics on the ride quality weneed a valid measurement of the pavement roughness and acomprehensive index to evaluate both road roughness and theride quality as perceived by road users Tomeet this objective

a fundamental index (IRI) was established by theWorld Bankin 1986

The most often employed and most useful model of avehicle suspension system for developing a low-level con-troller for a vehicle suspension is the quarter-car modelin which only one quarter of the vehicle is taken intoconsideration The model is two-dimensional because onlymovement in the 119885 direction is taken into consideration Itconsists basically of a single wheel which is represented in theform of a spring A general representation of a two-degree-of-freedom quarter-car model is shown in Figure 1

In this model the sprung and unsprung masses thatcorrespond to one corner of the vehicle are denoted by 119898

119904

and 119898119906 respectively The suspension system is represented

by a linear spring of stiffness 119870119904and a linear damper with a

damping rate 119862119904 while the tire is modeled by a linear spring

of stiffness119870119905 119884 is the input By drawing free body diagrams

and applying Newtonrsquos Second law we obtain the followingdifferential equations [11]

119898119904

119885119904+ 119862119904( 119885119904minus 119885119906) + 119870119904(119885119904minus 119885119906) = 0

119898119906

119885119906+ 119862119904( 119885119906minus 119885119904) + 119870119904(119885119906minus 119885119904) + 119870119905(119885119906minus 119884) = 0

(1)

We can eliminate the masses from the equations leavingthe equations in this form

119885119904+ 119862119904( 119885119904minus 119885119906) + 1198701(119885119904minus 119885119906) = 0

119906 119885119906+ 119862 ( 119885

119906minus 119885119904) + 1198702(119885119906minus 119885119904) + 1198701119885119906= 1198701119884

(2)

Using the response of the quarter-car model at a travelspeed of 80 kmh calculated for each point along the distanceof travel the IRI can be defined as follows [1]

IRI = 1119871int

119871

0

1003816100381610038161003816119885119904 minus 1198851199061003816100381610038161003816 119889119909

(3)

where 119871 is the distance along the road onwhichmeasurementis performed

Mathematical Problems in Engineering 3

ms

mumu

Ks

Kt

(a)

Zs

Zu

Y

CsKs

mu

ms

Kt

(b)

Figure 1 Quarter-car vehicle model (a) Vehicle representation (b) simplified representation

22 Power Spectral Density Power spectral density is aprobabilisticmethod which is ameasure of themean squaredvalue of a random variable In general thismethod is used forrandom vibration analysis which describes how the power ofa signal or time series is distributed over different frequencies

If we regard the pavement as a continuous surface the IRIsequence of road profiles is a randomphenomenon that obeysa zero-mean Gaussian distribution This can be regarded asa stationary stochastic process Therefore it is appropriateto describe the pavement characteristics using the powerspectral density

Define 119883(119905) as a stationary stochastic process and 119877119909(119905)

as its autocorrelation function and if the Fourier transformof 119877119909(119905) exists then

119878119909(120596) =

1

2120587int

+infin

minusinfin

119877119909(120591) 119890minus119894120596120591

119889120591 (4)

119878119909(120596) is the PSD of 119883(119905) 120596 is angular frequency and

119877119909(120591)

can be described by the inverse Fourier transform of thepower spectral density

119877119909(120591) = int

+infin

minusinfin

119878119909(120596) 119890119894120596120591

119889120596 (5)

These two equations form a Fourier pair called theWiener-Khintchine formula [12 13]

23 Correlation between IRI and PSD To evaluate thedynamic actions transmitted from a vehiclersquos movement onthe road surface we need to develop equations that expressthe physical state of the systems According to our definitiona quarter-car model meets the condition of a LTI (lineartime invariant) system Regarding 119884(119905) as systematic excita-tion 119885

119904(119905) and 119885

119906(119905) as systematic response based on the

transmission property of LTI system the frequency responsefunction can be solved by means of Laplace transform [13]

119867119885119878119884

(120596) =119896119905(119895119862119904120596 + 119896119904)

Δ (120596)

119867119885119906119884

(120596) =

119896119905(minus1198981199041205962

119895119862119904120596 + 119896119904)

Δ (120596)

Δ (120596) = (minus1198981199041205962

119895119862119904120596 + 119896119904) (minus119898

1199061205962

119895119862119904120596 + 119896119904+ 119896119906)

minus (119895119862119904120596 + 119896119904)2

(6)

The pavement roughness is assumed to be a randomstationary variable and using the exciting force of thequarter-car model the systematic response can be expressedas follows

119885119904(119905) = 119867

119885119878119884(120596) 119884 (119905)

119885119906(119905) = 119867

119885119906119884(120596) 119884 (119905)

(7)

Define 119885(119905) = 119885119904(119905) minus 119885

119906(119905) and regard 119885(119905) as the

response of the system One has

119885 (119905) = 119867119885119878119884

(120596) 119884 (119905) minus 119867119885119906119884

(120596) 119884 (119905)

=1198961199051198981199041205962

Δ (120596)119884 (119905) = 119867

119885119884(120596) 119884 (119905)

(8)

As mentioned previously large volumes of measurementdata show that the pavement roughness conforms to thefundamental hypothesis of a vibration source random fieldwhich is a zero-mean local ergodic Gaussian random fieldAccording to the principles of a LTI system the systematicresponse is also a random stationary variable so we may alsoset 119878119884(120596) as the PSD of the pavement roughness and thus we

can determine the PSD and the mean squared value of 119885(119905)Consider

119878119911(120596) =

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596)

1205932

119911= 119877119911(0) = int

infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596) 119889119908

(9)

The mean squared value represents the equivalent ampli-tude of signals and thus it can represent the size of the signalamplitude because they are approximately equal and theyconform to a linear correlation Thus we may derive theexpression for the IRI as follows [11]

IRI = 119860 times 120593119911= 119860 times radicint

infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596) 119889119908 (10)

According to the random process theory 120596 = 2120587119891 119899 =119891V and 119909 = V119905 and thus a different expression for the PSDcan be obtained

119878119909(119899) = 2120587V119878

119909(120596) (11)


Acceleration is the second derivative of vertical displace-ment Consider

119878119886(120596) = (

120596

V)

4

119878119909(120596)


infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

(120596

V)

4

119878119886(120596) 119889119908

(12)

Therefore it can be seen that there is a linear relationshipbetween the IRI and the square root of pavement power spec-tral density function Matlab was used to calculate the mea-sured acceleration values Then the measured accelerationvalues were fitted to the known actual road surface roughnessvalues Finally we obtained the conversion formula for themeasured acceleration values and the IRI

3 Correction Model Based on Hybrid Sensors

31 Measurement of IRI by Multiple 119885-Axis AccelerometersBecause the work described in this paper was aimed atestablishing a connection between the 119885-axis acceleration(in the direction of gravity) and the IRI we selected fortesting and modeling a section of the road network ofShanghai for which the IRI had already been measured bythe highway administration in 2012 A strategy of averagingmultiplemeasurements was applied to reduce the influence ofmeasurement errors In the tests the 119885-axis accelerometerswere placed flat in the car to eliminate the effect of gravitycaused by the weight of the components at the same timethe accelerometers were completely fixed inside the car sothat they would fully reflect the sprung vibration of the car Inaccordance with the quarter-car model accelerometers werefixed separately at the center of the vehicle and over the fourwheels Each link in the road network was tested four times

311 Single-WheelModel Thesquare root of each value of thepower spectral density of the road-induced acceleration wascalculated using Matlab and then values that were obviouslytoo small or large were eliminatedThe averages of the squareroot were calculated for the left and right wheels The resultsare shown in Table 2

The square root values for the left and right wheels werecalculated separately and compared with the standard IRIvalues It was found that for both the left and the right wheelsthere was a linear relationship between the square root valueand the standard IRIThemodel obtained for the right wheelswas

IRI = 21204119883119903minus 28401 (13)

where 119883119903is the square root of the power spectral density of

the acceleration for the right wheels on the link 1198772 is 09173the fit of the model is very good Similarly the model for theleft wheels was

IRI = 17887119883119897minus 33525 (14)


the acceleration for the left wheels on the link 1198772 is 08707the fit of the model is again very good

Table 2 Results for the left and right wheels

Number oflinks selected

Left-wheel squareroot value

Right-wheel squareroot value

StandardIRI

1 51634 42223 57042 63474 52826 90623 40195 32460 34514 39012 32666 38745 43624 34117 39596 35366 22714 22817 52493 41268 59158 43359 35055 48389 42007 30775 434010 50837 37241 540311 44185 31880 399112 44069 35530 429013 50762 41863 561414 49560 42084 570115 37991 32827 362916 40984 32920 396317 39900 33378 486118 30861 25950 251719 35595 33126 353720 32307 25859 2697

a998400y1a998400y1 ay1

ayr

Figure 2 Mutual effects of left and right wheel acceleration

312 Two-Wheel Model During the observation process wefound that there was a relationship between the accelerationsof the left and right wheels The car body is a rigid structureand thus when driving over the continuous surface of a pave-ment if a change is produced on one side a correspondingchange will inevitably appear on the other side as shownin Figure 2 Thus although the tires and the suspensionsystem can reduce the effects of the two wheels on eachother the quarter-car model used in the model describedabove does not provide an accurate simulationTherefore wecan improve the accuracy of the model by considering theinfluences of the wheel accelerations on each other

First we consider the relationship between the acceler-ations of the left and right wheels We used SPSS to analyzethe correlations between the accelerations of the left and rightwheels and the results are shown in Table 3

Although there is a nice relationship between the accel-erations of the left and right wheels the relationship isrelatively complex so we cannot directly establish a fittingmodel However because of the linear relationship between


Table 3 Correlations

Right wheel acceleration Left wheel acceleration

Right wheel accelerationCorrelation coefficient 1000 851lowastlowast

Significance (2-tailed) sdot 000119873 31 31

Left wheel accelerationCorrelation coefficient 851lowastlowast 1000Significance (2-tailed) 000 sdot

119873 31 31lowastlowastCorrelation is significant at the 001 level (2-tailed)

Table 4 Results of multivariate linear fitting using SPSS

(a) Model Summary

Model 119877 119877 Square Adjusted119877 Square

Std error of theestimate

Change statistics119877 square change 119865 change df1 df2 Sig 119865 change

1 971a 942 936 3869706 942 139067 2 17 000aPredictors (constant) right left

(b) ANOVAb

Sum of squares df Mean square 119865 Sig1

Regression 41650 2 20825 139067 000aResidual 2546 17 150Total 44195 19

aPredictors (constant) right leftbDependent variable iri

(c) Coefficientsa

ModelUnstandardizedCoefficients

Standardizedcoefficients 119905 Sig

119861 Std error Beta1

(Constant) minus3442 495 minus6948 000left 782 318 405 2456 025right 1300 369 581 3524 003

aDependent variable iri

the square root of the power spectral density of the single-wheel accelerations and the IRI one might guess that there isalso a linear correlation between the power spectral densityof the square root of the two-wheel accelerations and thestandard IRI This means that

IRI = 1198861119883119897+ 1198862119883119903+ 1198863+ Δ (15)

Therefore we used multiple linear fitting of the left- andright-wheel square root values with the standard IRI

We used the package spss170 to perform multiple linearregression The regression results were as follows

IRI = 0782119883119897+ 1300119883

119903minus 3442 (16)

1198772 is 0942 which satisfies the requirement for precision

Therefore the goodness of fit for the two-wheel linear model

is better than that for the one-wheel linear model so the two-wheel model can improve the fit (Table 4)

32 Velocity Correction In the actual measurements becauseof the limitations imposed by the conditions the speed of thevehicle was not equal to the specified speed of 80 kmh At thesame time the relationship between the IRI and the speed isnot simply a monotonic increase or decrease but it is verycomplex and it depends on features of the pavement surfacesuch as the waveform Perera et al [14] and Xiao-qing andLi-jun [11] measured the IRI values of pavements over a longperiod at different speeds and compared the measurementswith the IRI value at the specified speed of 80 kmh Theresults showed that the relationship between the speed andthe IRI value is complex and it does not follow any definiteproportionality relationshipThe experiment showed that the


12

10

08

06

04

02

00 20018016014012010080604020

l = 3m

l = 30ml = 10m

(kmmiddothminus1)

IRI (

mmiddotm

minus1)

Figure 3 IRI values for different speeds and different waveforms (119897is the road wavelength)

IRI variedwith the change in velocity (from 0ndash200 kmh) andwavelength (divided into long wave 119897 = 30m medium wave119897 = 10m and short wave 119897 = 3m) as derived in Figure 3

Because of the nonlinear relationship between IRI valueand speed the variation of speed during the measurementprocess has a large influence on the IRI measurementTherefore when 119885-axis accelerometers are used to measurethe IRI amodification for speed is required At the same timeit is also important in an IRI measurement to consider thelength of road over whichmeasurement is performed and thestart and end points Consequently we chose a GPS system toallow us to perform modification for speed and so on

To analyze the influence of speed on the results of themodel we chose an experimental roadwith light trafficwherethe traffic speed could reach 80 kmh It has been foundfrom field tests that when the vehicle speed reaches 60 kmhcertain areas of the vehicle tire are not in contact with theroad pavement This violates our earlier assumption that thedistance between the tire and pavement is zero Thereforein our experiment we chose speeds of 20 30 40 50 and60 kmh on the test road measured the 119885-axis accelerationof the car wheels and calculated the IRI for the link foreach speed separately using the model If the results wereconsistent this would indicate that the model did not dependon themeasurement speed If not we couldmodify themodelby fitting the calculated results to the vehicle speed Becausethe same model was used to calculate the IRI value in eachcase the different IRI values for different speeds were relatedonly to the power spectral density values of the accelerationTherefore we fitted the power spectral density and speeddirectly The results are shown in Table 5

When the power spectral density value and the speed Vare fitted the model is

PSD = 00263V2 + 06027V (17)

1198772 is 09991 this model has a very good fit

Using this model we can consider the reason whydifferent speedswill lead to different results of the calculation

Table 5 Values of power spectral density for different speeds

Speed (kmh) 20 30 40 50 60Power spectraldensity 209189 417322 68044 1156324 1304095

The spatial frequency (in units of mminus1) that characterizes thecollection frequency of the numerical data is the number ofsamples permeter Because the frequency at which the sensorreads data is constant the spatial frequency of the sequenceof acceleration data is related only to the speed At the sametime because of the small variation of the measuring speeda quadratic fitting curve is better than a quartic curve As aresult aftermodification for the effect of speed the IRImodelis

IRI = 0782120572119883119897+ 1300120572119883

119903minus 3442 (18)

where 120572 is the correction coefficient for the effect of speedgiven by

120572 =1

radic00003865V2 + 00009125V (19)

4 System Design

According to the fitting model we constructed in Section 3we developed a coupled system for measuring the IRI Thesystem comprised two sections a hardware platform and asoftware system

41 Description of theHardware Platform Thehardware plat-form was used to collect 119885-acceleration data for pavementsas well as GPS data which was transmitted to the softwaresystem The hardware platform comprised the following(Figure 4)

(i) 119885-axis accelerometer sensors type MMA8451Q(ii) GPS module type Ublox NEO-6M(iii) ZigBee modules(iv) Microcontroller units (MCU) type TC12C5608AD

The 119885-axis accelerometer is a triaxial intelligent low-power mechanical acceleration sensor This type of sensorcan access data from both a low-pass filter and a high-passfilter and therefore greatly reduces the peak data require-ments for data analysis and achieves faster data transfer Thesample frequency of the accelerometers was set to 01 secondwhich made the pavement appear more like a continuoussurface and this ensured the data processing efficiency TheGPS module contained an integrated audio frequency chipa baseband chip and a core CPU The core controller couldconnect multiple 119885-axis gravity accelerometers and GPSdevices The positional accuracy of GPS is 1 meter which isadequate for field experiments and road tests

42 Description of the Software Platform Thesoftware systemwas used to obtain real-time data from hardware devices andit comprised two main components (Figure 5) as follows


P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0

P10ADC0P11ADC1P12ADC2P13ADC3P14ADC4P15ADC5P16ADC6P17ADC7

P20PWM2

P21 P21P22P22P23P23RESETRSTP30RXDRXDP31TXDTXDXTAL2XTAL2XTAL1XTAL1

P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +

Figure 4 Schematic diagram of the system circuit

(i) Data acquisition software based on C for obtainingthe acceleration rates from various devices and stor-ing these data

(ii) AMatlab GUI data processor (based onMatlab GUI)for calculating the IRI (or RQI) using the modelconstructed above

43 Description of the System Design The system couldrecord real-time 119885-axis accelerations in different pavementconditions at various frequencies as well as using differentparameters such as those used to describe the drivingconditions (Figure 6)

The system design had a two-tier construction Theaccelerometers obtained the real-time acceleration and trans-mitted them to the data acquisition software via a Zigbeemodule GPS was also received as geographic informationfrom satellites and transmitted to the software via BluetoothWhen the data flowed into software tier the acquisitionsoftware could match the GPS information and accelerationwith the time data as well as reading the real-time changesbased on the wave patterns and locating the position wemeasured on the e-map The data obtained from the dataacquisition software were used by the Matlab GUI dataprocessor The processor removed any abnormal data basedon residual analysis and we then used the model described

above to calculate the PSD of pavements and to evaluate theIRI value

5 Field Test

To verify the accuracy of the model we used the abovemeasurement system to perform a field test The accelerom-eters were placed flat and fixed completely inside the vehicleso they fully reflected the sprung vibration of the vehicleAs described in Section 3 the accelerometers were fixedseparately over the wheels Given that the practical tracks onwhich both the front and back wheels travel were basicallythe same it was not necessary to collect acceleration datarepeatedly By contrast the tracks of the left and right wheelswere totally different so their mutual effects could not beneglected At the same time the back wheels were affectedless by the engine so the accelerators were set immediatelyabove the right back wheel and left back wheel in the test car

Eight typical roads in Shanghai for which the IRIs wereprovided by the Shanghai Highway Administration Bureauwere selected for the field testThe specific method employedwas as follows

(i) We confirm the stake mark of the testing origin anddestination and recorded the positional data


(a) (b)

Figure 5 (a) Data acquisition software (b) Matlab GUI data processor

AccelerometerSatellites

Data processing Multiple

linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer

Bluetooth transmitter

GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data

acquisition software

Data

Matlab GUI data processor

Hardware platform

Match by time data

Figure 6 Schematic diagram of the measurement system

(ii) The experimental car with the measurement systemwas used to acquire the vertical acceleration (accu-racy 001mg frequency 10Hz) and GPS (accuracy1m frequency 1 Hz) at about 60 kmh where datawere recorded for both the left and right wheels

(iii) The acceleration datawere imported intoMatlabGUIbefore calculating the approximate IRIs of these roadsusing the linear fitting model

We selected themeasurement data for links 21 to 28 as theinput of the model to calculate the IRI We then calculatedthe relative error between the approximate IRI obtained

from the model and the actual IRI (detected by the vehiclebearing road laser profilometer Highway Administration ofShanghai 2012) to test the suitability of themodelThe resultsare shown in Table 6

When we used the model presented in Section 3 to fit the119885-axis acceleration to the IRI the results showed that therelative error of the approximate IRI value was lower than15 and the standardized residual was between minus2 and +2Therefore this model can meet the needs of the majority ofpavement measurements Thus this method based on 119885-axisaccelerometers and GPS devices is feasible for measuring theIRI


Table 6 Approximate values of the IRIs obtained using the model

Link number Right-wheel squareroot value

Left-wheel squareroot value Actual IRI Approximate IRI Relative error

()Standardized

residual21 30182 43766 39041 4109 498 07877322 37904 43842 49140 4566 761 minus07644323 34982 40119 42430 4292 114 035007124 31138 39016 36570 4086 1049 141684625 44060 54115 65001 6137 592 minus08068226 50087 55317 64629 6020 736 minus10308427 57322 42199 39496 4315 846 123830228 49631 43613 44039 3904 1282 minus119086

6 Conclusions

To address the problems of pavement roughness measure-ment we established an IRI estimation model based onregression analysis Based on themultiple linear fittingmodeland velocity correction model we developed a coupledsystem that can record the real-time 119885-axis acceleration indifferent pavement conditions at different times and withdifferent values for various other parameters

The variation in the in-car 119885-axis acceleration causedby road roughness can be regarded as a combination of thevibration produced by different mechanical componentsand thus the vertical acceleration is strongly correlated withthe IRI The quarter-car model was a LTI system and themean squared value of the power spectral density couldrepresent the equivalent amplitude of signals which canrepresent the size of the signal amplitude and thus we useda regression method to model the variation in the 119885-axisacceleration and the IRI We used the power spectral densitysequence of the 119885-axis acceleration to model the IRI Aninnovative feature of the measurement process was thatmultiple local accelerations were considered in order toimprove the goodness of fit of the model

Because the relationship between the IRI value andthe speed is nonlinear variation of the speed during themeasurement process has a large influence on the measuredIRI value The length of road along which the measurementis performed and the start and end points are also importantin this measurement so we used a GPS device to allow us totake account of speedThe influence of speed on the results ofthemodel was analyzed andwe then put forward the conceptof a speed correction coefficient to improve the reliability ofthe model

We used the IRI evaluationmodel and system to measurethe IRI of some typical roads in Shanghai When our modelwas used to fit the 119885-axis acceleration to the IRI the resultsshowed that the relative error of the estimated IRI was lessthan 15

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

This work was based on the results of a research projectwhich was supported by a research Grant no 2012AA112402from the Ministry of Science and Technology of the PeoplersquosRepublic of China and a research Grant no 11511501100 fromShanghai Science and Technology Committee The work ofthe last author was supported by Program for ChangjiangScholars and Innovative Research Team in University andShanghai Pujiang Program (11PJD022) The authors take soleresponsibility for all views and opinions expressed in thispaper

References

[1] MW Sayers TDGillespie andWD Paterson ldquoGuidelines forthe conduct and calibration of road roughness measurementsrdquoWorld Bank Technical Paper 46TheWorld Bank WashingtonDC USA 1986

[2] MW Sayers ldquoOn the calculation of IRI from longitudinal roadprofile TRB Paper No 95 0842rdquo in Proceedings of the 74thAnnual Meeting on Transportation Research Board (TRB 95)Washington DC USA 1995

[3] J R Prasad S Kanuganti P N Bhanegaonkar A K Sarkar andS Arkatkar ldquoDevelopment of relationship between roughness(IRI) and visible surface distresses a study on PMGSY roadsrdquoProcediamdashSocial and Behavioral Sciences vol 104 pp 322ndash3312013

[4] M W Sayers T D Gillespie and A V Queiroz ldquoThe interna-tional road roughness experiment Establishing correlation anda calibration standard for measurementsrdquo Tech Rep WTP451986

[5] J J Hajek T J Kazmierowski and G Musgrove ldquoInternationalroughness index as a measure of customer satisfaction rdquoin Proceedings of the Annual Meeting of the TransportationAssociation of Canada Victoria Canada 1995

[6] T J Kwon M Gerla V K Varma M Barton and T R HsingldquoEfficient flooding with passive clusteringmdashan overhead-freeselective forward mechanism for ad hocsensor networksrdquoProceedings of the IEEE vol 91 no 8 pp 1210ndash1220 2003

[7] H Sabbineni and K Chakrabarty ldquoLocation-aided flooding anenergy-efficient data dissemination protocol for wireless sensornetworksrdquo IEEE Transactions on Computers vol 54 no 1 pp36ndash46 2005


[8] J E Jefferies R W DeMay and G L Lachinyan ldquoRentalcar-share vehicle access andmanagement system andmethodrdquo USPatent Application 13830754 2013

[9] T Dishongh F Guilak andMMorris ldquoApparatus for monitor-ing physiological activity and environmental datardquo US PatentApplication 11641973[P] 2006

[10] A Prakash B N Sharma and T J Kazmierowski ldquoInves-tigation into observational variations in pavement conditionsurveyrdquo in Proceedings of the 3rd International Conference onManaging Pavements vol 2 pp 290ndash301 Washington DCUSA 1994

[11] Z Xiao-qing and S Li-jun ldquoRelationship between internationalroughness index and velocity of quarter carrdquo Journal of TongjiUniversity vol 33 no 10 pp 1323ndash1327 2005

[12] P Stoica and R L Moses Introduction to Spectral AnalysisPrentice-Hall Upper Saddle River NJ USA 1997

[13] X Zhou L Yan and L Sun ldquoStudy and validation of therelationship between international roughness index and powerspectral densityrdquo China Civil Engineering Journal vol 40 no 1pp 99ndash104 2007

[14] R W Perera C Byrum and S D Kohn Investigation ofDevelopment of Pavement Roughness 1998

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MathematicsJournal of


Mathematical Problems in Engineering

Hindawi Publishing Corporationhttpwwwhindawicom

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of


Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of


Mathematical PhysicsAdvances in

Complex AnalysisJournal of


OptimizationJournal of


CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of


Operations ResearchAdvances in

Journal of


Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of Mathematics and Mathematical Sciences


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Algebra

Discrete Dynamics in Nature and Society



Decision SciencesAdvances in

Discrete MathematicsJournal of


Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Stochastic AnalysisInternational Journal of


Table 1 Characteristics of common road roughness measurement methods

Method of measurement Principal of operation Measurementdevice Advantage Disadvantage

3-Meter Long Beam Direct profilemeasurement Aluminum beam Simple reliable data

collectionInefficientlower mechanization

Level Direct profilemeasurement Level and rod Time-stable

straightforward Time-consuming inconvenient

Bump integrator RTRRMS LVDT displayer and acartrailer Simple fast

Time-instable affected by vehiclevibration performs at the samespeed

Laser road surface tester Direct profilemeasurement

Car with laser deviceand calculators Straightforward efficient High cost of operation and

maintenance

GMR profilometer Indirect profilemeasurement

LVDT accelerometerpotentiometer and a van Convenient efficient

Precise instrument requiredhigh cost of operation andmaintenance

and made several recommendations for IRI measurementIRI measurement at the network level has become a routinepractice for many road agencies in recent years On the otherhand IRI measurement at the project level is also requiredprimarily for accepting or price-adjusting paving contractorsrsquoproducts The development of methods for IRI measurementis a prerequisite for a pavement management system (PMS)and other parts of the road management industry

An accelerometer is a device that measures the acceler-ation in one two or three orthogonal axes and they areused widely in the fields of civil engineering biology andindustry Accelerometers can be used to measure the vehicledriving status where they facilitate evaluations of the overallvehicle performance and responseThis information can thenbe used tomake adjustments to various vehicle subsystems asnecessary [6ndash8] Accelerometers can also be used to measureseismic activity inclination machine vibration dynamicdistance and speed with or without the influence of gravity[9 10]

This study focused on building amodel for estimating theIRI as well as developing an effective and low-cost system formeasuring the IRIThe steps of this study which are reflectedby this organization of this paper are as follows

(i) Introduction to the principle of using 119885-axisaccelerometers to measure IRI

(ii) Modeling the relationship between the variation inthe in-car 119885-axis acceleration and the IRI

(iii) Development of a self-designed measurement systemwith 119885-axis accelerometers and a GPS device

(iv) Field testing results

2 The Principle of Using 119885-AxisAccelerometers to Measure IRI

21 International Roughness Index To study the effects ofthe road pavement characteristics on the ride quality weneed a valid measurement of the pavement roughness and acomprehensive index to evaluate both road roughness and theride quality as perceived by road users Tomeet this objective

a fundamental index (IRI) was established by theWorld Bankin 1986

The most often employed and most useful model of avehicle suspension system for developing a low-level con-troller for a vehicle suspension is the quarter-car modelin which only one quarter of the vehicle is taken intoconsideration The model is two-dimensional because onlymovement in the 119885 direction is taken into consideration Itconsists basically of a single wheel which is represented in theform of a spring A general representation of a two-degree-of-freedom quarter-car model is shown in Figure 1

In this model the sprung and unsprung masses thatcorrespond to one corner of the vehicle are denoted by 119898

119904

and 119898119906 respectively The suspension system is represented

by a linear spring of stiffness 119870119904and a linear damper with a

damping rate 119862119904 while the tire is modeled by a linear spring

of stiffness119870119905 119884 is the input By drawing free body diagrams

and applying Newtonrsquos Second law we obtain the followingdifferential equations [11]

119898119904

119885119904+ 119862119904( 119885119904minus 119885119906) + 119870119904(119885119904minus 119885119906) = 0

119898119906

119885119906+ 119862119904( 119885119906minus 119885119904) + 119870119904(119885119906minus 119885119904) + 119870119905(119885119906minus 119884) = 0

(1)

We can eliminate the masses from the equations leavingthe equations in this form

119885119904+ 119862119904( 119885119904minus 119885119906) + 1198701(119885119904minus 119885119906) = 0

119906 119885119906+ 119862 ( 119885

119906minus 119885119904) + 1198702(119885119906minus 119885119904) + 1198701119885119906= 1198701119884

(2)

Using the response of the quarter-car model at a travelspeed of 80 kmh calculated for each point along the distanceof travel the IRI can be defined as follows [1]

IRI = 1119871int

119871

0

1003816100381610038161003816119885119904 minus 1198851199061003816100381610038161003816 119889119909

(3)

where 119871 is the distance along the road onwhichmeasurementis performed


ms

mumu

Ks

Kt

(a)

Zs

Zu

Y

CsKs

mu

ms

Kt

(b)






119878119909(120596) =

1

2120587int

+infin

minusinfin

119877119909(120591) 119890minus119894120596120591

119889120591 (4)


119877119909(120591)


119877119909(120591) = int

+infin

minusinfin

119878119909(120596) 119890119894120596120591

119889120596 (5)



119904(119905) and 119885



119867119885119878119884

(120596) =119896119905(119895119862119904120596 + 119896119904)

Δ (120596)

119867119885119906119884

(120596) =

119896119905(minus1198981199041205962

119895119862119904120596 + 119896119904)

Δ (120596)

Δ (120596) = (minus1198981199041205962

119895119862119904120596 + 119896119904) (minus119898

1199061205962

119895119862119904120596 + 119896119904+ 119896119906)

minus (119895119862119904120596 + 119896119904)2

(6)


119885119904(119905) = 119867

119885119878119884(120596) 119884 (119905)

119885119906(119905) = 119867

119885119906119884(120596) 119884 (119905)

(7)

Define 119885(119905) = 119885119904(119905) minus 119885

119906(119905) and regard 119885(119905) as the


119885 (119905) = 119867119885119878119884

(120596) 119884 (119905) minus 119867119885119906119884

(120596) 119884 (119905)

=1198961199051198981199041205962

Δ (120596)119884 (119905) = 119867

119885119884(120596) 119884 (119905)

(8)



119878119911(120596) =

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596)

1205932

119911= 119877119911(0) = int

infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596) 119889119908

(9)



infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596) 119889119908 (10)


119878119909(119899) = 2120587V119878

119909(120596) (11)



119878119886(120596) = (

120596

V)

4

119878119909(120596)


infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

(120596

V)

4

119878119886(120596) 119889119908

(12)






IRI = 21204119883119903minus 28401 (13)



IRI = 17887119883119897minus 33525 (14)







StandardIRI

1 51634 42223 57042 63474 52826 90623 40195 32460 34514 39012 32666 38745 43624 34117 39596 35366 22714 22817 52493 41268 59158 43359 35055 48389 42007 30775 434010 50837 37241 540311 44185 31880 399112 44069 35530 429013 50762 41863 561414 49560 42084 570115 37991 32827 362916 40984 32920 396317 39900 33378 486118 30861 25950 251719 35595 33126 353720 32307 25859 2697

a998400y1a998400y1 ay1

ayr













(a) Model Summary





(b) ANOVAb




(c) Coefficientsa







IRI = 1198861119883119897+ 1198862119883119903+ 1198863+ Δ (15)



IRI = 0782119883119897+ 1300119883

119903minus 3442 (16)






12

10

08

06

04

02

00 20018016014012010080604020

l = 3m

l = 30ml = 10m

(kmmiddothminus1)

IRI (

mmiddotm

minus1)






PSD = 00263V2 + 06027V (17)






IRI = 0782120572119883119897+ 1300120572119883

119903minus 3442 (18)


120572 =1

radic00003865V2 + 00009125V (19)

4 System Design







P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0


P20PWM2


P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +







5 Field Test





(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










ms

mumu

Ks

Kt

(a)

Zs

Zu

Y

CsKs

mu

ms

Kt

(b)






119878119909(120596) =

1

2120587int

+infin

minusinfin

119877119909(120591) 119890minus119894120596120591

119889120591 (4)


119877119909(120591)


119877119909(120591) = int

+infin

minusinfin

119878119909(120596) 119890119894120596120591

119889120596 (5)



119904(119905) and 119885



119867119885119878119884

(120596) =119896119905(119895119862119904120596 + 119896119904)

Δ (120596)

119867119885119906119884

(120596) =

119896119905(minus1198981199041205962

119895119862119904120596 + 119896119904)

Δ (120596)

Δ (120596) = (minus1198981199041205962

119895119862119904120596 + 119896119904) (minus119898

1199061205962

119895119862119904120596 + 119896119904+ 119896119906)

minus (119895119862119904120596 + 119896119904)2

(6)


119885119904(119905) = 119867

119885119878119884(120596) 119884 (119905)

119885119906(119905) = 119867

119885119906119884(120596) 119884 (119905)

(7)

Define 119885(119905) = 119885119904(119905) minus 119885

119906(119905) and regard 119885(119905) as the


119885 (119905) = 119867119885119878119884

(120596) 119884 (119905) minus 119867119885119906119884

(120596) 119884 (119905)

=1198961199051198981199041205962

Δ (120596)119884 (119905) = 119867

119885119884(120596) 119884 (119905)

(8)



119878119911(120596) =

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596)

1205932

119911= 119877119911(0) = int

infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596) 119889119908

(9)



infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

119878119910(120596) 119889119908 (10)


119878119909(119899) = 2120587V119878

119909(120596) (11)



119878119886(120596) = (

120596

V)

4

119878119909(120596)


infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

(120596

V)

4

119878119886(120596) 119889119908

(12)






IRI = 21204119883119903minus 28401 (13)



IRI = 17887119883119897minus 33525 (14)







StandardIRI

1 51634 42223 57042 63474 52826 90623 40195 32460 34514 39012 32666 38745 43624 34117 39596 35366 22714 22817 52493 41268 59158 43359 35055 48389 42007 30775 434010 50837 37241 540311 44185 31880 399112 44069 35530 429013 50762 41863 561414 49560 42084 570115 37991 32827 362916 40984 32920 396317 39900 33378 486118 30861 25950 251719 35595 33126 353720 32307 25859 2697

a998400y1a998400y1 ay1

ayr













(a) Model Summary





(b) ANOVAb




(c) Coefficientsa







IRI = 1198861119883119897+ 1198862119883119903+ 1198863+ Δ (15)



IRI = 0782119883119897+ 1300119883

119903minus 3442 (16)






12

10

08

06

04

02

00 20018016014012010080604020

l = 3m

l = 30ml = 10m

(kmmiddothminus1)

IRI (

mmiddotm

minus1)






PSD = 00263V2 + 06027V (17)






IRI = 0782120572119883119897+ 1300120572119883

119903minus 3442 (18)


120572 =1

radic00003865V2 + 00009125V (19)

4 System Design







P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0


P20PWM2


P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +







5 Field Test





(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra











119878119886(120596) = (

120596

V)

4

119878119909(120596)


infin

minusinfin

1003816100381610038161003816119867119885119884 (120596)1003816100381610038161003816

2

(120596

V)

4

119878119886(120596) 119889119908

(12)






IRI = 21204119883119903minus 28401 (13)



IRI = 17887119883119897minus 33525 (14)







StandardIRI

1 51634 42223 57042 63474 52826 90623 40195 32460 34514 39012 32666 38745 43624 34117 39596 35366 22714 22817 52493 41268 59158 43359 35055 48389 42007 30775 434010 50837 37241 540311 44185 31880 399112 44069 35530 429013 50762 41863 561414 49560 42084 570115 37991 32827 362916 40984 32920 396317 39900 33378 486118 30861 25950 251719 35595 33126 353720 32307 25859 2697

a998400y1a998400y1 ay1

ayr













(a) Model Summary





(b) ANOVAb




(c) Coefficientsa







IRI = 1198861119883119897+ 1198862119883119903+ 1198863+ Δ (15)



IRI = 0782119883119897+ 1300119883

119903minus 3442 (16)






12

10

08

06

04

02

00 20018016014012010080604020

l = 3m

l = 30ml = 10m

(kmmiddothminus1)

IRI (

mmiddotm

minus1)






PSD = 00263V2 + 06027V (17)






IRI = 0782120572119883119897+ 1300120572119883

119903minus 3442 (18)


120572 =1

radic00003865V2 + 00009125V (19)

4 System Design







P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0


P20PWM2


P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +







5 Field Test





(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra

















(a) Model Summary





(b) ANOVAb




(c) Coefficientsa







IRI = 1198861119883119897+ 1198862119883119903+ 1198863+ Δ (15)



IRI = 0782119883119897+ 1300119883

119903minus 3442 (16)






12

10

08

06

04

02

00 20018016014012010080604020

l = 3m

l = 30ml = 10m

(kmmiddothminus1)

IRI (

mmiddotm

minus1)






PSD = 00263V2 + 06027V (17)






IRI = 0782120572119883119897+ 1300120572119883

119903minus 3442 (18)


120572 =1

radic00003865V2 + 00009125V (19)

4 System Design







P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0


P20PWM2


P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +







5 Field Test





(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










12

10

08

06

04

02

00 20018016014012010080604020

l = 3m

l = 30ml = 10m

(kmmiddothminus1)

IRI (

mmiddotm

minus1)






PSD = 00263V2 + 06027V (17)






IRI = 0782120572119883119897+ 1300120572119883

119903minus 3442 (18)


120572 =1

radic00003865V2 + 00009125V (19)

4 System Design







P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0


P20PWM2


P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +







5 Field Test





(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










P32INT0P33INT1

GND

M4

M3

M2

M1

GND

STC12C5608AD

GND

GNDCON2

110592M

22pF

STC12C5608AD

22pF

GND

12

SDASCL

GND

GND

GND

GND

GND

GND

GND

+5

+5

+5

+5

SDASCL

SDASCL

SDASCL

P26P27PWM0

P10P11P12P13P14P15P16P17PWM2

10120583F

10120583F

VDD

VDD

VDD

VDD

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

MMA8451

MMA8451

SA0

SA0

SCL GNDVDDSDA

VCC

VCC

VCCVCCVCC

P26P27

P37PWM0


P20PWM2


P34T0ECIP35T1PWM1

P24PWM3

P25GND

P32P33P34

PWM1

PWM3

P25

GND

Y1

J1

C1

C2

C3U1

01UF

R110K

+

C5C4 +







5 Field Test





(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra










(a) (b)




linear fitting

PSD calculation

Velocity correction

Calculate IRI

Accelerometer

Accelerometer

Accelerometer


GPS device

Zigbee trasmitter

Bluetooth receiver

Z-acceleration

GPS data

Zigbee receiver

Initial Data


Data


Hardware platform

Match by time data











()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra













()Standardized


6 Conclusions







Acknowledgments


References























Volume 2014




Journal of











Journal of


Function Spaces






Algebra
























Volume 2014




Journal of











Journal of


Function Spaces






Algebra









research article measurement of international roughness ...with -axis accelerometers and a gps...

Documents