Multi-Sensor Fusion Approach for Fire Alarm using BP Neural Network

Multi-sensor information fusion algorithm based on BP neural network is applied in the system of fire alarm to realize early detecting and alarming. The fire detection method based on neural network was developed using detection information for temperature, smoke density, and CO concentration to determine the probability of three representative fire conditions. The method overcomes the drawbacks of the fire alarm system using single sensor information. Results show that the proposed method can provide fire warning more accurate and timely and can effectively reduce leak-check rates and false alarms, reaching the desired purpose.