物流短期集中需求预测方法研究与仿真

Study and Simulation of Forecasting of Short-term Intensive Logistics Demand

针对短时间内区域物流配送需求集中增加的问题,提出建立BP人工神经网络预测模型对短期配送需求进行预测。通过分析影响物流需求变化的数据指标与区域实际现状,建立数学模型,并对该区域在特殊时间内的物流配送收发量、库存量以及周转总量进行分析研究,利用三层结构的BP人工神经网络对不同的预测结果进行误差比较,从而得出准确的短期物流需求量预测结果。采用MATLAB神经网络工具箱中的函数编程进行仿真实验,实验结果表明,短期物流集中需求实际量与仿真结果存在极小的误差,因此,基于BP人工神经网络的短期集中物流需求预测方法可以为物流企业提供决策依据。

In this paper, in view of the intensive increase in regional logistics and distribution demand within short periods of time, we proposed to establish a forecasting model based on BP artificial neural network. By analyzing the data index influencing the changes in logistics demand and the actual situation of a region, we established the corresponding mathematic model, analyzed the logistics distribution volume, inventory volume and total turnover of the region within a particular time, and compared the error of different forecasting results using the three-layered BP artificial neural network to find the most accurate one. At the end, we used an experimental simulation to prove the validity of the method.