The Analysis of Sewers Inflammable Gas Based on PSO-SVR

Abstract

Due to the non-liner, poor selectivity and cross-sensitivity of the combustible gas in the sewer, an analysis prediction model of the combustible gas in the sewer has been established based on the PSO-SVR machine, the model has introduced a new particle swarm algorithm to support the vector regression machine so that it can optimize the important parameters, realizing the automatic determination of parameters of the SVR machine, and be used for quantitative analysis of combustible gas in the sewer. The simulation results show that the model of the combustible gas in the sewer based on PSO-SVR machine is superior to the compared SVR model and it has better generalization performance and higher prediction accuracy.