Study On Traffic Flow Prediction And Route Guidance Algorithm In Traffic Guidance System

In recent years,with the development of social economy,traffic congestion,traffic accidents and other traffic problems have become increasingly prominent.In order to solve these problems,traffic guidance system has been introduced into the urban traffic management and has been developed rapidly.The short-term traffic flow prediction and route guidance algorithms are the key technologies of the traffic guidance system.To predict the future traffic flow reasonably,and give a reasonable path,not only can provide the decision-making basis for the traffic management department,but also can easily guide travelers travel,avoid getting into traffic jam and save travel time.It is very difficult to describe the change rule of urban road network because of its time variability and complexity,so it is very important to study the real-time and accurate traffic flow forecasting and route guidance algorithms.Based on the analysis of urban road traffic data,this paper takes the urban road network and intersections as the research object,and the traffic prediction of the non-detector intersections,network traffic prediction and route guidance algorithms are studied.The main research work of the paper includes the following several aspects:1.This paper introduces the methods of traffic data collection and the characteristics of the data,analyzes the feasibility of the traffic flow prediction,describes the method of identifying and repairing the abnormal data.The time series analysis method and Lyapunov exponent analysis are used to determine the predictability of traffic flow,and use the historical trend data and the weighted of measured data to repair abnormal data.2.According to some city road intersections without detector or fault detector,based on the analysis and study of several commonly used non-detector intersection traffic flow prediction methods.This paper presents a traffic flow prediction method of non-detector intersections based on fuzzy C-means clustering.In this method,the intersections are clustered into the same cluster by fuzzy clustering,and then use multiple linear regression to prediction the traffic flow.The experimental results verify the effectiveness of the algorithm.3.Based on the research of traffic flow prediction model,and gives out the traffic flow prediction model based on support vector regression method.Due to the learning speed of theparameters of SVR is slow,studied the global search characteristic of genetic algorithm which can optimize the parameters of SVR.Finally through the experiment to verify the rationality of GA-SVR model.4.Based on the analysis of the advantages and disadvantages of several traditional algorithms for solving the optimal path algorithm,and introduce an ant colony algorithm based on simulated evolution to select the optimal path of traffic.The main principle of the algorithm is that the ant colony depends on the pheromone associated with the path length to find the optimal path.At the same time,improve the shortcomings of the ant colony algorithm and compare the improved ant colony algorithm with genetic algorithm and through the experiment verify the validity of the algorithm.5.Using GA-SVR prediction model and shortest route guidance algorithm based on ant colony algorithm,design and realize the traffic guidance system based on J2 EE framework.