基于模糊神经网络的中庭火灾探测系统

Atrium Fire Detecting System Based on Fuzzy-neural Network

将探测器的输出信号模糊化后送入神经网络进行处理,采用3层前馈BP神经网络结构,用BP算法进行网络参数的训练,然后由模糊逻辑判决火灾和非火灾。用Matlab语言进行仿真,结果表明基于模糊神经网络的中庭火灾探测系统能够准确探测各种标准试验火。

Put the fuzzy output into the fuzzy-neural network to be managed after the fuzzification fo the signals which come from detector, use three layers feedforward BP neural network, and train the network＇s parameters with BP learning algorithm, then adopt the fuzzy logic to tell fire from wrong fire. Simulate this process with Matlab,it＇s turn out that atrium fire detecting system based on fuzzy-neural network can detect kinds of standard development fires exactly, and thit system has great abilities to resest disturbs.