摘要 针对火灾发生的不确定性及破坏力强的特点,同时存在火灾误报率和漏报率高的问题,必须采用智能检测算法才能达到最佳效果。由于一些算法存在求解速度慢和参数稳定性不足等问题,该文提出了基于随机权重策略的改进粒子群优化极限学习机(IPSO-ELM)的火灾检测方法。通过Matlab设计的IPSO-ELM网络,对火灾数据进行训练,与粒子群算法优化极限学习机(PSO-ELM)和遗传算法优化极限学习机(GA-ELM)的火灾检测结果进行比较,发现IPSO-ELM的预测准确率最高,精度比PSO-ELM、GA-ELM分别高出3.3%和5%。 According to the characteristics of uncertainty and destructive power of fire,there are problems of high false alarm rate and missing alarm rate of fire,intelligent detection algorithm must be used to achieve the best effect.Due to the shortcomings of some algorithms,such as slow solving speed and insufficient parameter stability,a fire detection method based on improved particle swarm optimization extreme learning machine(IPSO-ELM)based on random weight strategy is proposed.The IPSO-ELM network designed by Matlab is used to train the fire data.Compared with the fire detection results of PSO-ELM and GA-ELM,it is found that IPSO-ELM has the highest prediction accuracy,which is 3.3%and 5%higher than PSO-ELM and GA-ELM.
机构地区 常州大学机械工程学院
出处 《自动化与仪表》 2021年第4期63-66,78,共5页 Automation & Instrumentation
关键词 粒子群算法 极限学习机 火灾检测 遗传算法 particle swarm optimization(PSO)algorithm extreme learning machine fire detection genetic algorithm
