Abstract: In order to monitor non-safety status of outdoor fire hydrant, such as water discharge, low pressure, overpressure, etc., which affect normal use of rescue water intake function, data filtering processing technology was adopted,based on one-dimensional pressure data, high-order multidimensional expansion of pressure mean value, pressure difference, pressure variance and pressure variance growth rate were carried out, and a comprehensive feature extraction model of outdoor fire hydrant status was proposed. Through combination optimization of sampling frequency and sliding filter window width, pressure data under different conditions were collected, and fitting model parameters were trained. Model was applied to integrated information acquisition intelligent terminal device and installed in outdoor fire hydrant dynamic information monitoring system. The results show that model can accurately judge status of water discharge, low pressure and overpressure, and improve dynamic information safety management level of outdoor fire hydrant.

Key words: outdoor fire hydrant, pressure data, feature extraction, multidimensional extension, dynamic information monitoring