Short-term Mixed Traffic Flow Forecast And Application Based On Wavelet Analysis

In the intelligent transportation system,real-time and accurate short-term traffic flow prediction has always been the focus of scholars in many countries.Real-time and accurate short-term traffic flow forecasting is the basis of intelligent traffic control and management.The problem of high complexity,randomness,and uncertainty of short-term traffic flow makes the prediction unsatisfactory.In order to improve the prediction accuracy,more and more prediction models are applied to this field.Therefore,it is of great significance to study the short-term traffic flow prediction and application for urban traffic management.In this dissertation,the key issues were discussed such as short-termtraffic data collection device,data prediction and traffic flow fusion applications.The main contents are as follows:Firstly,based on the understanding and analysis of the theoretical knowledge of traffic flow,the real-time data of traffic flow is found to be very important for forecasting and analyzing traffic flow conditions.Design cross-camera device location based on data acquisition methods and acquisition process issues.Secondly,aiming at the complexity,randomness and uncertainty of short-term traffic flow forecasting,single model prediction cannot achieve the desired accuracy.Wavelet neural network(WNN)and ant colony-WNN prediction models are established to predict traffic in the next period.Wavelet analysis is used to denoise and reconstruct the data,and the WNN model is used to predict short-term traffic flow.In order to avoid falling into a local optimum when training the WNN model,the ant colony algorithm is used to optimize the weights.Simulation results of two models through MATLAB show that,based on the error index formula calculation,the ant colony-WNN combination model can more accurately predict short-term flow data.Finally,aiming at short-term traffic flow prediction on mixed roads of motor vehicles and non-motor vehicles,the traffic flow fusion model is designed,and the conversion relationship between non-motor vehicles and motor vehicles is analyzed and calculated,and fusion parameters are obtained.By comparing the evaluation indicators,the fused forecastdata is closer to the real data.In short-term traffic flow forecasting applications,traffic capacity is an important indicator for intelligent traffic decision-making.Different models will affect the size of road capacity.This paper analyzes the influence of large,medium and small cars on the capacity of the vehicle based on the time and distance models.In the GUI environment of MATLAB,a mixed traffic flow prediction and analysis system is designed.Three aspects of integrated infrastructure,data processing and platform reflect the application of traffic flow prediction in intelligent transportation.