Application of a fuzzy neural network in fire-detection of belt conveyer in mines

The temperature,its variation rate,concentration of carbon monoxide,and its variation rate were considered as four parameters of fire detection system of belt conveyor.A kind of fuzzy inference system based on adaptive neural network was employed in the investigation of this system.The appropriate membership functions were obtained after training up 34 groups of data samples.The simulation result showed that,when 0.5 was regarded as the alarm threshold and certain delay was employed,the system would alarm no later than about 200 s as fire took place.So,the fire detecting system of belt conveyor could effectively give out early fire-alarm.And at the same time,the alarm system possessed enhanced anti-interference ability and adoptability to the environment.