基于PSO-BP的远距离物流车辆等候时间判断研究

Study on Waiting Time Judgment of Long-distance Logistics Vehicles Based on PSO-BP

综合考虑了道路的复杂性与意外事故的概率,对物流车辆安装GPS与信息处理模块,形成浮动数据库,结合历史数据与实时的路况形成预测数据集,数据集加载到神经网络中进行时间预测,预测过程中使用粒子群优化对权重系数进行优化,保证在特殊情况与意外事故中模型不会陷入局部最优。实际的测试证明,该方法对物流车辆等候时间的预测准确率高达93%,具有很强的实际意义。

In this paper, we considered comprehensively the complexity of the road conditions and the probability of accidents, then built a floating database from the data of the onboard GPS and information processing modules of the logistics enterprises and combined the historical data with the real-time road condition data to form the forecasting data sets. At the end, through an empirical testing, we demonstrated the accuracy and practical significance of the method.