State Monitoring Method of Fire Water Network Based on Multi-Dimensional Comprehensive Feature Extraction

In order to capture the sudden water pressure step signal in the fire water network, realize the judgment of the internal state of the pipeline network, and complete the real-time dynamic information monitoring target of the fire water system, an intelligent algorithm of fire water network states monitoring based on multi-dimensional comprehensive feature extraction is proposed and applied to the intelligent water pressure sensor. Firstly, the principle and realization method of sliding filter used in water pressure sensor are theoretically analyzed. Based on sliding average filter, the filtering parameters of intelligent water pressure sensor are designed. Secondly, the time domain signal of water pressure is mainly studied. On the basis of one-dimensional mean, the intelligent algorithm of sensor is realized by multidimensional expansion technology. Finally, the algorithm simulation and the performance test of intelligent water pressure sensor with intelligent algorithm in outdoor application scenarios are introduced. The experimental results show that the event judgment accuracy of the intelligent water pressure sensor is not less than 80% when the sampling frequency is 1 Hz. It can acquire the state transition information of "normal" and "leakage" in the fire water supply system in time and accurately, meet the requirements of abnormal state judgment of the real-time monitoring system, and provide auxiliary decision-making information guarantee for fire fighting and rescue operation command.