Early fire prevention in the plant

In this survey, the architecture of a wireless multisensory early fire warning system at the plant was proposed. The wireless sensor network system allows one to rapidly implement the system deployment and to easily reconfigure its topology in the enterprise premises. The difference between the proposed and the existing fire systems is the applying of sensors instead of detectors. The proposed system employs temperature sensors, gas concentration, and visibility sensors. The adaptive method based on a multilayer perceptron for the processing of measurement results in a multi-sensor system was proposed. This method allows one to adapt the system in compliance with the topology, and to detect not only a fact but also the position of the fire source. Fire detection is based on the prediction model of sensor signals. The developed methods and software have been verified on the fire model basis. Fire simulation in the enterprise premises was carried out inside the “Polytechnic” Supercomputer Center using the Fire Dynamics Simulator.