改进粒子群优化BP神经网络的旅游客流量预测  

Prediction for tourist traffic based on improved particle swarm optimization BP neural network

作　　者：于明涛[1] 叶晓彤[2] 

机构地区：[1]四川理工学院自动化与电子信息学院,四川自贡643000 [2]四川理工学院网络信息管理中心,四川自贡643000

出　　处：《微型机与应用》2015年第21期51-54,共4页Microcomputer & Its Applications

基　　金：四川省智慧旅游研究基地重点项目(ZHZ14-04)

摘　　要：旅游客流量受多种因素影响,传统的时间序列预测模型无法描述预测对象的规律,人工智能方法如BP神经网络,其结构的选择过多依赖经验,基于此提出了利用改进的粒子群算法优化BP神经网络,通过惯性因子的非线性递减来改善粒子群的寻优性能。将该预测模型应用于自贡灯会的客流量进行实际预测分析,通过对150组训练样本和50组测试样本的实验仿真,可知改进后的方法提高了预测结果的准确度,并且涉及参数少、简单有效。Tourist flow is influenced by many factors. The traditional time series prediction model cannot describe the laws of the forecasted object. Artificial intelligence methods such as BP neural network, the choice of its structure relies too much on experience. Based on these above, the improved particle swarm optimization （PSO） algorithm was used to optimize the BP neural network. It uses nonlinear decreasing inertia factor to improve the performance of particle swarm optimization. The prediction model is applied to the flow of Zigong Lantern Festival forecast analysis. Through simulation of 150 sets of training samples and 50 groups of test samples, the result shows that the improved method improves the accuracy of the prediction, and involves less parameters, simple and effective.

关 键 词：旅游客流量预测 BP神经网络 粒子群算法 非线性递减 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]