基于GA-BP神经网络的负氧离子浓度反演模型研究  

Study on inverse model of negative oxygen ion concentration based on GA-BP neural network

作　　者：杨佳男 马飞鸿 胡斌 曾松伟[2] YANG Jianan;MA Feihong;HU Bin;ZENG Songwei(College of Mathematics and Computer Science,Zhejiang A&F University,Hangzhou 311300,China;College of Optical Mechanical and Electrical Engineering,Zhejiang A&F University,Hangzhou 311300,China)

机构地区：[1]浙江农林大学数学与计算机科学学院,浙江杭州311300 [2]浙江农林大学光机电工程学院,浙江杭州311300

出　　处：《传感器与微系统》2023年第8期62-64,77,共4页Transducer and Microsystem Technologies

基　　金：浙江省自然科学基金公益资助项目(LGN18C200017)。

摘　　要：针对负氧离子浓度监测过程中存在的手段单一,无法满足日常监测需求的问题,分析负氧离子浓度与环境参数之间的关系,以温度、湿度以及PM2.5浓度作为输入变量,通过建立遗传算法(GA)优化的反向传播(BP)神经网络模型(GA-BP),对负氧离子浓度进行反演分析。实验结果表明:基于BP神经网络的负氧离子浓度反演结果平均相对误差为11.12%。使用GA优化后的BP神经网络对负氧离子浓度的反演效果更好,平均相对误差(MRE)仅为6.51%。基于GA-BP神经网络的负氧离子浓度反演模型,可为负氧离子的深入研究提供可靠的理论依据,同时,该研究模型的应用将大幅降低负氧离子浓度的监测成本,推动负氧离子监测技术的进步。Aiming at the problem that the method of negative oxygen ion concentration monitoring is lack,which cannot meet the daily monitoring needs,the relationship between the concentration of negative oxygen ions and other enviromental parameters is analyzed.Using temperature,humidity and PM2.5 concentration as input variables,negative oxygen ion concentration inversion is analyzed,by establishing back propagation(BP)neural network model optimized by genetic algorithm(GA)(GA-BP).Experimental results show that the average relative error of the negative oxygen ion concentration inversion results based on BP neural network is 11.12%.The BP neural network optimized by GA has a better inversion effect on the concentration of negative oxygen ions,and the mean relative error(MRE)is 6.51%.Therefore,inverse model of negative oxygen ion concentration based on GA-BP neural network can provide a reliable theoretical basis for the in-depth study of negative oxygen ion.At the same time,the application of this model will greatly reduce the monitoring cost of negative oxygen ion concentration and promote the progress of negative oxygen ion monitoring technology.

关 键 词：负氧离子浓度 PM2.5浓度 反向传播神经网络 遗传算法 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP391[自动化与计算机技术—控制科学与工程]