基于粒子群算法优化BP神经网络的高低温试验箱温度预测
Temperature Prediction of High and Low Temperature Test Chamber Based on BP Neural Network Optimized By Particle Swarm Optimization Algorithm
彭白雪 1陈清华 2王建刚 2王皖楠3
作者信息
- 1. 安徽理工大学 机电工程学院,淮南 23200
- 2. 安徽理工大学 机电工程学院,淮南 23200;广东立佳实业有限公司,东莞 5230000
- 3. 广东立佳实业有限公司,东莞 5230000
- 折叠
摘要
为提高高低温试验箱内部温度预测精度,通过建立粒子群算法优化后的BP神经网络(PSO-BP)模型对高低温试验箱内工作区温度变化情况进行预测,并利用试验采集的有限点温度数据进行对比分析,为高低温试验箱内温度特性的分析计算提供理论和数据支持.结果表明PSO-BP网络取得最小训练误差为 9.35×10-5,与BP神经网络相比,优化后的PSO-BP神经网络训练集和测试集拟合精度分别提高了 1.09%和 2.43%.BP网络和PSO-BP网络平均绝对误差(MAE)分别为 1.480 和 0.753,均方根误差(RMSE)分别为 1.979 和 1.842,综合表明PSO-BP神经网络预测精准度更高,可有效获得高低温试验箱内连续完整的温度情况,提高了试验箱研发工作效率.
Abstract
In order to improve the prediction accuracy of the internal temperature of the high and low temperature test chamber,the BP neural network(PSO-BP)model optimized by the particle swarm optimization algorithm is established to predict the temperature change in the working area of the high and low temperature test chamber,and the finite point temperature data collected by the test is used for comparative analysis,which provides theoretical and data support for the analysis and calculation of the temperature characteristics in the high and low temperature test chamber.The results show that the minimum training error of PSO-BP network is 9.35×10-5.Compared with BP neural network,the fitting accuracy of training set and test set of optimized PSO-BP neural network is improved by 1.09% and 2.43% respectively.The mean absolute error(MAE)of BP network and PSO-BP network were 1.480 and 0.753,respectively,and the root mean square error(RMSE)were 1.979 and 1.842,respectively.The results show that the PSO-BP neural network has higher prediction accuracy,which can effectively obtain the continuous and complete temperature situation in the high and low temperature test chamber,and improve the research and development efficiency of the test chamber.
关键词
高低温试验箱/粒子群算法/BP神经网络/温度预测Key words
high and low temperature test chamber/particle swarm optimization algorithm/BP neural network/temperature prediction引用本文复制引用
基金项目
安徽省重点研发计划(2022a05020030)
安徽理工大学环境友好材料与职业健康研究院研发专项(ALW2021YF12)
出版年
2024
NETL
NSTL
万方数据