基于VMD-HSSA-BP神经网络模型的短时交通流预测

Short⁃term traffic flow prediction based on VMD⁃HSSA⁃BP neural network model

针对交通流数据的非平稳性和短时交通流预测高精度要求的问题,提出了一种由变分模态分解(VMD)、改进麻雀搜索算法(HSSA)和BP神经网络的组合预测模型。模型利用变分模态分解降低历史交通流数据的非平稳性,使用Hammersley和自适应控制因子分别改进麻雀搜索算法种群初始化和发现者位置更新公式,提高麻雀搜索算法的收敛速度和寻优能力,使用改进后的麻雀搜索算法寻找BP神经网络的最优权值和阈值,提升BP神经网络预测的精准度。通过仿真,将模型与现有模型进行对比,模型预测结果更好,验证了模型能克服交通流数据非平稳性,并具有较好的预测精度。

Aiming at the non⁃stationarity of traffic flow data and the high accuracy requirement of short⁃term traffic flow prediction,a combined prediction model was proposed,which was composed of Variational Mode Decomposition(VMD),improved Sparrow Search Algorithm(HSSA)and BP neural network.The model uses variational mode decomposition to reduce the non⁃stationarity of historical traffic flow data.Hammersley and adaptive control factors are used to improve the population initialization and finder’s position update formulas of sparrow search algorithm respectively,so as to improve the convergence speed and optimization ability of sparrow search algorithm.The improved sparrow search algorithm is used to find the optimal weight and threshold of BP neural network.Improve the accuracy of BP neural network prediction.Through simulation,the model is compared with the existing model,and the prediction result of the model is better,which verifies that the model can overcome the non⁃stationarity of traffic flow data and has a better prediction accuracy.