Study On Key Technology Of Intelligent Transportation System

Transportation problem has become an important factor which restricts urbaneconomic development, how to design a smart, reliable and feasible traffic controlalgorithm has become the focus of research in transportation field.Traffic flow forecasting is the precondition and basis for intelligent traffic control,it is difficult to establish precise mathematical model for traffic flow because of itsnon-linear characteristic. To make full use of the non-linear, high accuracycharacteristic of wavelet function and the adaptive learning ability of neural network,this thesis uses improved wavelet neural network to predict the traffic flow, improvedaccuracy and precision of traffic flow forecasting, it provides an effective way fortraffic flow forecasting.In this paper a fuzzy controller is designed to adjust the amount of green time ofthe intersection, this fuzzy controller make the predict traffic flow of major and minorroads as its input, the adjustment amount of green time as its output. In addition, inorder to coordinative control of roads, another fuzzy controller also designed to adjustthe amount of phase offset, this fuzzy controller make the predict traffic flow of majorroads as its input, the phase offset adjustment as its output. Fuzzy control algorithmcan take full advantage of fuzzy controller and reduces the parameters of controlsystem, making system easy to implement and control.Traffic control algorithm must be validated before put into operation, at present,more feasible and economical way is traffic simulation. After studying the TSIS RTEinterface, the paper gives the RTE configuration and development process, andimplemented RTE based on fuzzy control. By simulation, the algorithm based on fuzzycontrol can effectively reduce network traffic delay, traffic congestion situation hasmarkedly improved, and traffic conditions have been improved and enhanced.