Platoon Dispersion Prediction Based on Transformation Kernel Estimate

Vehicle arrival rate is the key parameter for signal planning at intersections, which are determined by upstream intersection signal planning and platoon dispersion characteristic. With consideration of randomness and uncertainty of traffic flow arriving at intersections and the requiring for coordinated signal control between adjacent intersections, this paper presents a new method for predicting vehicle arrival rate through estimating platoon dispersion characteristic based on transform kernel estimate under unknown sample distribution function. Based on the simulation platform of VISSIM 4.20 combined with VC ++ and MATLAB mixed programming, the predicting accuracy of the method is proved. Furthermore, the comparison application in signal control system with Webster platoon dispersion calculating formula and the existing signal planning plan encourages the validity and potential of the method.