Short-Term Traffic Flow Prediction Of Wavelet Neural Network Based On Improved Wolf Pack Algorithm

In the intelligent transportation system,real-time and accurate short-term traffic flow prediction has always been the focus of scholars in various countries.In order to improve the accuracy of prediction,more and more combined forecasting models are applied to this field.The wavelet neural network(WNN)model combining the advantages of wavelet analysis and neural network,has good effect on the prediction of short-term traffic flow.Wolf pack algorithm(WPA)is a recently proposed optimization algorithm,with good global convergence.Therefore,this paper combines the wolf pack algorithm with the improved gradient descent algorithm for short-term traffic flow prediction based on the wavelet neural network.Firstly,five traffic flow data sets were obtained from the Transportation Data Research Laboratory at the University of Minnesota and the Transport Performance Measurement System in California.Then the traffic flow data of each data set were pretreated,including repairing anomaly data,noise reduction,phase space reconstruction.Finally,morlet wavelet neural network is used to forecast the short-term traffic flow.The simulation results show that the WNN model can predict the overall trend of short-term traffic flow,but the stability and prediction accuracy are still to be improved.A WNN model based on WPA(WPA-WNN)is proposed to overcome the shortcomings that the weight matrix and the wavelet factor of wavelet neural network are sensitive to the initial value and easy to fall into the local minimum.By using the global optimization ability of WPA,a set of optimal weights and wavelet factors are found for the WNN,and the weights and wavelet factors are optimized by the gradient descent algorithm.The simulation results show that the combination of WPA and gradient descent algorithm is effective,and the IWPA-WNN model can improve the stability and accuracy of short-term traffic flow forecasting.Then,an improved WPA(IWPA)is proposed.The short-term traffic flow prediction simulation experiment based on the IWPA-WNN shows that the stability and accuracy of short-term traffic flow prediction are improved,and the running time is shortened.Secondly,in order to further improve the accuracy of forecasting,this paper applies error compensation(EC)to WNN short-term traffic flow forecasting by Using WNN model to extract the error data of traffic flow prediction.The simulation results show that the WNN with EC can effectively improve the accuracy of short-term traffic flow prediction.Finally,the error compensation method is combined with the IWPA-WNN for short-term traffic flow prediction.The simulation results show that the short-term traffic flow prediction based on the EC-IWPA-WNN model has good performance both in stability and accuracy.