文献类型：期刊文献

中文题名：改进粒子群优化BP神经网络的洪水智能预测模型研究

英文题名：On Application of Improved PSO-BP Neural Network in Intelligent Flood Forecasting Model

作者：何勇[1];李妍琰[2]

第一作者：何勇

机构：[1]信阳农林学院计算机科学系;[2]河南财经政法大学计算机与信息工程学院

第一机构：信阳农林学院计算机科学系,河南信阳464000

年份：2014

卷号：39

期号：5

起止页码：75-80

中文期刊名：西南师范大学学报：自然科学版

收录：CSTPCD;;北大核心:【北大核心2011】；CSCD:【CSCD_E2013_2014】；

语种：中文

中文关键词：粒子群;BP神经网络;洪水预测;径流量

外文关键词：particle swarm optimization;the BP neural network;flood prediction;runoff

摘要：该文提出改进的PSO-BP算法在洪水预测应用中建立预测模型.以BP神经网络为基础,提取观测站往年平均径流量作为洪水属性.采用改进的PSO-BP算法对神经网络的各个参数进行优化,最后建立模型应用于流域观测站的洪水预报模型,叙述了PSO粒子群算法和BP神经网络算法,详细阐述粒子群算法优化BP神经网络的权值和阈值,得出最优的BP神经网络预测适应度值.通过实验仿真对比,结果表明此方法预测结果比BP神经网络算法和混沌径向基神经网络模型算法精度更高,提高了预测的效率.
The flood prediction model base on PSO‐BP algorithm has been proposed in this paper .Extrac‐tion of observation station in average runoff as flood has been conducted on the bases of BP neural net‐work .Using the improved PSO‐BP algorithm parameters of the neural network has been optimized ,the flood forecasting model for watershed observing station with the model .This paper introduces the particle swarm optimization algorithm and BP neural network algorithm ,a detailed explanation of PSO algorithm to optimize BP neural network weights and threshold .Through the simulation results ,this method fore‐casting result is higher than the BP neural network algorithm and chaos RBF neural network model accura‐cy ,is an effective method of prediction and reliable flood .