To design a new system to improve the accuracy and efficiency of electrical fire warning,in response to the diverse types of electrical fires,varying sizes of fires,and large differences in warning thresholds.Using a distributed control system(DCS)as the hardware foundation,deploying MQ-135 smoke sensors,IR Flame Sensors,and LM35 temperature sensors,and collecting data through the DCS system.Using a BP neural network improved by temperature time series model,an electrical fire threshold model is constructed through parameter normalization and temperature time series model to achieve accurate judgment and early warning of electrical fires.The experimental results show that the system can accurately detect temperature and smoke for three types of fires:socket fires,line fires,and electrical equipment fires,and is more rapid in detection time.The designed system effectively improves the accuracy and response speed of electrical fire warning by integrating advanced sensors and optimizing data processing algorithms.
Improving BP neural networkElectrical firesEarly warning systemDCS system
维普
万方数据
