基于模糊神经网络的化工生产车间火灾报警方法

Fire Alarm Method of Chemical Production Workshop Based on Fuzzy Neural Network

针对化工生产车间中无火、明火和阴燃等类型的火灾,提出了一种基于模糊神经网络的火灾报警方法。设计了化工生产车间火灾报警整体的报警流程,为提升抗干扰能力,优化数据采集电路,并对数据传输的ZigBee协议栈结构进行设计,对接收到的数据进行归一化处理,提供稳定数据源,确定模糊神经网络火灾判定的输入向量,分析内部函数和权值,构建三角型隶属函数进行计算,实现火灾的分类报警。为验证设计报警方法的有效性,设计了仿真实验,并与传统报警方法对比。仿真结果表明:与传统的火灾报警方法相比,该方法针对多种火灾类型的报警应用中训练收敛速度更快,报警精度更高。

Aiming at the fire of no fire,open fire and smoldering in the chemical production workshop,a fire alarm method based on fuzzy neural network is proposed.The overall alarm process is designed for the fire alarm in the chemical production workshop.To improve the anti-interference ability,the data acquisition circuit is optimized,the ZigBee protocol stack structure for data transmission is designed,the received data is normalized,stable data source is provided,the input vector of the fuzzy neural network fire judgment is determined,the internal function and weight value are analyzed,and a triangular membership function for calculation is constructed.The classified fire alarm is realized.To verify the effectiveness of the design alarm method,a simulation experiment is designed and compared with the traditional alarm method.The simulation results show that compared with the traditional fire alarm method,the proposed method has faster training convergence speed and higher alarm accuracy for various fire alarm applications.