基于PSO-BP的贵州省物流需求预测研究

Research on Logistics Demand Forecasting in Guizhou Province Based on PSO-BP

摘要: 合理预测物流需求对物流业高质量发展具有重要意义。为提高预测结果的准确性，以贵州省为例，构建PSO-BP模型对未来三年的物流需求进行预测。首先选取12个指标建立指标体系，并进行灰色关联度验证。然后运用粒子群算法(PSO)优化反向传播网络(BP)，实证结果显示，PSO-BP的预测效果和拟合能力均优于单一的BP模型。最后使用GM(1,1)获得12个指标未来三年的预测值，将其代入PSO-BP模型得到贵州省未来三年的物流需求量。

Abstract: Reasonable prediction of logistics demand is of great significance for the high-quality development of the logistics industry. To improve the accuracy of the prediction results, taking Guizhou Province as an example, a PSO-BP model is constructed to predict the logistics demand for the next three years. Firstly, select 12 indicators to establish an indicator system and verify it through grey correlation degree. Then, particle swarm optimization (PSO) was used to optimize the back propagation network (BP), and the empirical results showed that the prediction performance and fitting ability of PSO-BP were superior to a single BP model. Finally, use GM (1,1) to obtain the predicted values of 12 indicators for the next three years, and substitute them into the PSO-BP model to obtain the logistics demand of Guizhou Province for the next three years.