Dynamic Path Planning Research Based On Real Time Traffic

The timely and accurate traffic flow forecast is an important prerequisite for intelligent traffic management and it is a very important issue of traffic data mining research.It also has a very important significance for traffic management,traffic flow guidance,dynamic traffic assignment,etc.Furthermore,it plays a vital role in the design and implementation of intelligent transportation systems.Traffic speed prediction as an important direction of traffic flow forecasting,can directory reflect road traffic conditions.Traffic speed prediction is dynamic access to a certain number of traffic flow data to predict future traffic flow.Then infer the possible path of the future traffic state according to the prediction of the traffic speed information to in order to planning the travel route rationally.This kind of predictive planning can reduce the waiting time of people,ease the traffic pressure,and also helpful for the traffic management and control.In this paper,we focus on short-term traffic speed prediction and path planning in the Intelligent Transportation Systems.Short-term traffic speed prediction is an important basis for judging traffic conditions in Smart City and the basic premise of path planning,it needs to analyze historical traffic data mining.Due to the special nature of traffic data,the existing data mining algorithms cannot be directly applied to knowledge of mining transport sector,nor efficient in large-scale direct traffic speed data.So this paper proposes genetic algorithm and wavelet neural network combined model(GA-WNN)to achieve short-term traffic speed prediction.GA-WNN mainly for wavelet neural network easy to fall into local minimum problem,proposed using the global search ability of genetic algorithm to optimize the initial parameters of Wavelet Neural Networks,then using neural network with optimized parameters to predict traffic speed.Experiments show that it has high prediction accuracy base on combined model,the predicted speed and the actual speed prediction has a high degree fit,to provide data and theoretical basis for short-time traffic conditions judgment.Then base on the traffic speed prediction,according to the needs of our travelers for the travel time,distance and other factors,we use improved ant colony algorithm for users planning the best real-time dynamic travel path.And simulation experiments show that real-time dynamic path planning more in line with the needs of travelers in travel distance,travel time,travel costs.The real-time dynamic travel path planning base on short-term traffic speed predicted make a contribution in ease traffic pressure,reduce accidents and reduce emissions contributed.