物流需求预测算法的仿真研究

Simulation Research on Forecast Algorithm of Logistic Demand

研究物流需求预测问题,影响物流需求因素过多且复杂,与经济消费和价格变化相关,是一种高度非线性关系,传统预测方法采用简单的数学模型进行预测,预测精度比较低,物流需求预测复杂的非线性问题已经成了物流界研究的重点。为了提高物流需求的预测精度,提出一种支持向量机的物流需求预测方法。通过采用支持向量机的非线性能力对历史物流需求量进行学习,通过粒子群算法获得模型最优参数,对将来物流需求进行预测。采用农产品物流需求数据对模型性能进行测试,测试结果表明,支持向量机提高了物流需求预测精度,对物流管理着着重要的现实意义,为预测提供了有效的方法。

Study logistics demand forecasting problems. Logistics demand is related to many factors and has com- plicated nonlinear relation to the factors, therefore, it is difficult to use simple mathematical model to describe the tra- ditional forecasting model, and the prediction precision is low. In order to improve the prediction accuracy of logistics demand, this paper proposes a logistics demand forecasting model based on support vector machine. Firstly, the his- torical data of logistics demand is pretreated, then support vector machine is train by logistics demand data, and the parameters of SVM is optimized by particle swarm algorithm to obtain the optimal logistics demand forecasting model. Then the optimal forecasting model is used to predict logistics requirements test set, and logistics demand forecasting results are obtained. The model performance is tested with agricultural logistics demand data, which shows that the support vector machine can improve the logistics demand forecasting accuracy and has important theoretical and realis- tic significance for logistics management.