Traffic simulation models calibration using speed-density relationship: An automated procedure based on genetic algorithm

Chiappone, Sandro; Giuffrè, Orazio; Granà, Anna; Mauro, Raffaele; Sferlazza, Antonino

doi:10.1016/j.eswa.2015.09.024

Thispaperpresentsthefirstresultsofaresearchwhichappliedageneticalgorithmtocalibrateamicroscopictrafficsimulationmodelbasedonspeed–densityrelationships.AlargesetoftrafficdatacollectedfromtheA22Freeway,Italy,wasusedandacomparisonwasperformedbetweenthefieldmeasurementsandthesimulationoutputsobtainedforatestfreewaysegmentbyusingtheAimsunmicroscopicsimulator.Thecalibrationwasformulatedasanoptimizationproblemtobesolvedbasedonageneticalgorithm;theobjectivefunctionwasdefinedinordertominimizethedifferencesbetweenthesimulatedandrealdatasetsinthespeed–densitygraphs.Forthispurpose,thegeneticalgorithmtoolinMATLAB®wasapplied.Keepinginmindtheobjectivetoautomatizethisprocess,theoptimizationtechniquewasattachedtoAimsunviaasubroutine,sothatthedatatransferbetweenthetwoprogramscouldautomaticallyhappen.AnexternalscriptwritteninPythonallowedtheMATLAB®softwaretointeractwithAimsunsoftware.Abettermatchtothefielddatawasreachedwiththeoptimizationparameterssetwiththegenetical-algorithm.Inordertochecktowhatextentthemodelreplicatedreality,modelvalidationwasalsoaddressed.Resultsshowedthatageneticalgorithmisusefullyapplicableinthecalibrationprocessofthemicroscopictrafficsimulationmodel.Beneficialeffectsareexpectedbyapplyingthesuggestedoptimizationtechniquesinceitsearchesforanoptimumsetofparametersthroughanefficientsearchmethod.

Traffic simulation models calibration using speed-density relationship: An automated procedure based on genetic algorithm / Chiappone, Sandro; Giuffrè, Orazio; Granà, Anna; Mauro, Raffaele; Sferlazza, Antonino. - In: EXPERT SYSTEMS WITH APPLICATIONS. - ISSN 0957-4174. - 44:(2016), pp. 147-155. [10.1016/j.eswa.2015.09.024]