Design of an Intelligent Alarm System Based on Multi-sensor Data Fusion

Abstract The fire alarm system plays a very important role in life, but the system has problems such as false alarms and false alarms. Therefore, this paper proposes the application of fire detection based on GA-BP neural network. Firstly, the algorithm takes temperature, smoke concentration and CO concentration as the input of BP neural network, and the output is whether there is fire or not. Secondly, it combines the characteristics of genetic algorithm with strong global search ability and strong robustness. The algorithm has achieved 100% correct classification on the test set through simulation experiments. At the same time, the absolute error of the sample prediction is only 0.006, which proves that it has strong robustness, reliability and generalization ability. Finally, the model was transplanted to STM32 to prove its feasibility. This method provides a new method for intelligent identification of fire signals for early warning of fires and accurate identification of non-fire signals.