Indoor Fire Detection System Based on Multi-sensor Information Fusion Algorithm

In order to detect fires in the first place and use modern science and technology to effectively reduce and prevent disasters, four environmental parameters are collected and the fire status is determined by information fusion algorithms in this paper. In the information fusion framework of the system, the Naive Bayesian algorithm and the D-S evidence theory algorithm are used as the first level decision layer. The fuzzy comprehensive evaluation method is used as the second level decision layer of the framework, and a percentage rating value is used to make a determination of the current environmental state. The hardware part of the system uses ESP8266 as the microcontroller for processing the collected sensing signals. The Raspberry Pi is used as the master controller to handle the input signals and the effective linkage of the information fusion framework. After testing, this system can realize the determination of fire status in real environment and can make more effective, scientific, sensitive and low-cost detection results for fire detection.