基于改进粒子群优化BP神经网络的火灾预警方法  

Fire warning method based on improved particle swarm optimization BP neural network

作　　者：尹春杰[1,2] 赵钦 王光旭 宋其征 王强 YIN Chunjie;ZHAO Qin;WANG Guangxu;SONG Qizheng;WANG Qiang(School of Information and Electrical Engineering,Shandong Jianzhu University,Jinan 250101,China;School of Mechanical and Electrical Engineering,Weifang Vocational College,Weifang 261041,China;Shandong Zhongke Advanced Technology Co.,Ltd.,Jinan 250101,China)

机构地区：[1]山东建筑大学信息与电气工程学院,山东济南250101 [2]潍坊职业学院机电工程学院,山东潍坊261041 [3]山东中科先进技术研究院有限公司,山东济南250101

出　　处：《电子设计工程》2023年第22期6-10,共5页Electronic Design Engineering

基　　金：山东省重点研发计划资助(2021CXGC011304)。

摘　　要：针对传统仓库火灾预警系统易受外界环境因素干扰,误报漏报率高的问题,采用一种基于改进粒子群优化BP神经网络的火灾预警方法,增强仓库消防系统的安全性和智能性,为传统检测系统增加一层保险。在标准粒子群优化BP神经网络的基础上,引入Tent映射,选取新的权重更新函数,改进粒子群随机搜索能力,加速神经网络收敛,输出期望结果。由Matlab软件仿真实验结果分析可知,改进后的方法相较于标准粒子群优化方法,收敛速度和准确率均显著提升,证明了该方法的优越性。Aiming at the problem that traditional warehouse fire warning system is easy to be disturbed by external environmental factors and the false alarm rate is high,a fire warning method based on improved particle swarm optimization BP neural network is adopted to enhance the safety and intelligence of warehouse fire warning system and add a layer of insurance for the traditional detection system.Based on the standard BP neural network optimized by particle swarm optimization,Tent mapping and new weight updating function are introduced to improve the random searching ability of particle swarm optimization,accelerate the convergence of neural network,and output the desired results.According to the analysis of simulation experiment results with Matlab,compared with the standard particle swarm optimization method,the convergence speed and accuracy of the improved method are significantly improved,which proves the superiority of the method.

关 键 词：火灾预警 数据融合 改进粒子群优化算法 神经网络 TENT映射 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]