Research On Simulation Of Pedestrian Swarm At Signalized Intersection

Pedestrian traffic at signal intersections often gathered into a swarm due to timeconstraints and the impact of the prevailing environment.Based on analysis of thecharacteristics of pedestrian swarm at signalized intersections，the general rules ofpedestrian crossing the signalized intersections can be concluded.And then,asimulation software based on the rules which can simulate the movement of thepedestrian swarms will provides a new approach to traffic management department.The Beijing West Da Wang Road,Wu ShengRoad and other sections of thetypical signalized intersection are taking as the research object. A large number ofcoordinates, velocity and acceleration dataof different pedestrian swarms are collectedby video analysis software. And based on the measured data，the characteristicanalysisof the pedestrian individuals, between individuals and pedestrian swarms is done.Based on the analysis and summary of the domestic and foreign pedestrian trafficmodels,Social Force Model is selected as the basis model for the individual pedestrianmovement behavior model. Then the traditional Social Force Model is optimized inorder to make it applicable to the signalized intersection. Nonlinear regressionanalysis is used to calibrate the optimized model parameters, and the accuracyanalysis of the model is done based on the experimental data.First step of establish swarming prediction model is the correlation analysis ofswarming center coordinates between adjacenttimes,swarming area parametersbetween adjacenttimes,and between swarming center speed and swarmingdistance.Then, the corresponding parameter prediction model is established by usingregression equation which based on the results of correlation analysis, and theaccuracy analysis of the model is done by using the error test between theexperimental data and predicted data.The swarming center coordinates control model, swarming area control modeland the swarming center speed control model is established by using mechanicsanalysis, which is in order to reduce the gap between the predicted value and thecorresponding parameters of the actual value of the parameter. The final parametercontrol modelwill regulate the pedestrian movement coordinately and cohesively.The signalized intersection pedestrian traffic simulation software is developed bylanguage of C#, which is based on theindividual pedestrian model and the swarming model.The software can simulatethat the individuals will form a cohesive swarm atthe green time, and output the data of pedestrain speed and coordinate.Accuracyanalysis of the software is done by using the error test between the experimental dataand output data. The software can evaluate the effect of pedestrian traffic organizationprogram and the pedestrian signal timing plan, thiswill be a new means fororganizaing pedestrian trafficeffectivelyand enhancing the safety of pedestrianscrossing the street.