A Multi-sensor Characteristic Parameter Fusion Analysis Based Electrical Fire Detection Model

Electrical fires are mostly caused by the release of thermal energy from electrical equipment. It is more difficult to detect the cause of fire than normal fires. A fire detection model that combines characteristics of many types of electrical fire sensors is proposed. These fire detection systems are difficult to accurately monitor the cause of electrical fires. The proposed model uses smoke, CO concentration, temperature, and electrical line residual current as characteristic parameters of electrical fire. It analyzes a three-tier structure including information layer, feature layer, and decision layer. Fire-risk-factor and warning duration are defined as decision factor. When the model is working, it firstly collects the residual current signal that characterizes the fault of the electrical equipment through multiple types of sensors. It conducts multi-parameter real-time monitoring of the main characteristic signals of the early stage of the electrical fire. Then it completes the fusion of the detected characteristics of electrical fire to achieve accurate identification of electrical fires. The proposed model is simulated according to national standard fire test dataset. The simulation result shows that it can quickly and accurately forecast the electrical fire and effectively reduce false alarm rate in of electrical fire detection process.