An Offset Simulation Optimization Of Arterial Street Based On Neural Network And Genetic Algorithm

The problem of urban traffic congestion has been receiving much attention.The coordinated control of arterial street signals can effectively improve the efficiency of arterial street traffic and alleviate urban traffic congestion.The optimization of offset parameter has always been the emphasis and difficulty of the research on the coordination control of arterial street signals.The traditional offset optimization method is based on adjacent intersections while it ignores the internal correlation of offset among multiple continuous intersections,and for the timing optimization model of the "scheme generation",the validity of the model cannot be quickly verified.In order to solve this problem,an offset simulation optimization model based on neural network and genetic algorithm is constructed.Mainly divided into the following three aspects:(1)Based on the intersection correlation research,the mechanism of interaction of multiple intersections is analyzed.Then,the offset of the continuous intersection is analyzed,and the relationship between the delay of arterial street and the offset of the intersection is analyzed.Considering that the arterial street delay function interacting with multiple intersections has complex nonlinear characteristics,it is difficult to express with a precise mathematical formula,and a neural network is introduced to characterize the relationship between the arterial street delay and the multi-intersection offset.(2)The arterial street delay function based on neural network is designed.The simulation part is realized by the neural network,that is,input the offset scheme among intersections and output the arterial street delay value;the genetic algorithm is designed to realize the optimization function,that is,to find the offset scheme corresponding to the minimum delay function of the arterial street in the neural network.They iterate with each other to solve the optimal offset of the arterial street.(3)Taking Shenghe Road in Wuhou District of Chengdu as an example,build the Vissim simulation platform based on the survey data.At first,the secondary development of Vissim was carried out by Matlab,and different offset schemes were randomly simulated to obtain the corresponding arterial street delay sample data.Then,based on the sample data,the neural network was used to fit the relationship between the vehicle delay and offset of the arterial street,and the offset of each intersection corresponding to the optimal arterial street delay in the neural network was found by genetic algorithm.At last,the optimization results are compared with combination method,mathematics analysis method and Synchro,and the sensitivity analysis of related factors is carried out to prove the effectiveness of the simulation optimization model.This paper presents an optimization model of arterial street offset from the perspective of simulation optimization,which can accurately describe the relationship between arterial street delay and intersection offset.The results can be used to optimize the offset of arterial street and further improve the urban traffic efficiency.