A short-term traffic flow prediction model based on EMD and GPSO-SVM

According to the time and space, randomness and volatility of traffic flow, a short-term traffic flow forecasting model based on empirical mode decomposition (EMD), genetic particle swarm optimization(GPSO) and support vector machine (SVM) is proposed. Firstly, the traffic flow sequence is decomposed into different frequency components by EMD. Then the crossover and mutation factors of the genetic algorithm are introduced into PSO to optimize the parameters of SVM, and the optimal SVM is obtained. Finally, the traffic flow components of different frequencies are input to the optimized SVM to realize the short time traffic flow prediction. A practical example is given based on the measured data of the road network in Changchun City. The results show that, the best effect and the highest prediction accuracy were obtained by the proposed model compared with PSO-SVM and GPSO-SVM.