人工神经网络在舰船火灾探测中的应用

Application of Artificial Neural Network in the Fire Detection of Naval Vessel

舰船火灾探测系统的目的是能及时、准确地探测到舰船上发生的火灾。提出一种利用人工神经网络的舰船火灾探测方法,通过设计三层反向传播(BP)人工神经网络模型,使用改进的BP算法——LM算法,对多传感器(温度传感器、烟雾传感器和CO传感器)同时探测到的数据进行智能化处理。仿真结果表明:基于人工神经网络的舰船火灾探测系统能及时、准确地识别各种火灾信号。

The purpose of naval vessel fire detection system is to detect fire at an earlier stage on board the naval vessel as to provide an estimation as accurate as possible. In this paper, a new method for naval vessel fire detection based on artificial neural network is presented. By designing a model of three layers back propagation artificial neural network, the improved BP arithmetic-LM arithmetic, is applied to the intelligent process of data synchronously detected by the multi-sensors （e. g. temperature sensor, smog sensor and CO sensor）. The result of simulation shows that different kinds of signals related to the fire disasters can be fast and accurately recognized by the fire detection system based on artificial neural network.