Design And Implementation Of VOCs Gas Detection System In Industrial Field Based On Neural Network Method

With the rapid development of industry and the increase of pollutant emissions year by year,the online air quality monitor has gained great application in the industrial field due to its many advantages such as portability and real-time nature.At present,the traditional on-line industrial field air quality monitors generally have the problem of ignoring the cross-sensitivity of the gas sensor,and it is difficult to achieve high monitoring accuracy.Aiming at the above problems,this subject designed and developed a new type of high-precision and low-cost industrial site VOCs gas online monitoring equipment based on sensor array.This article chooses to use three common VOCs gas sensors in benzene,xylene and formaldehyde to construct a gas sensor array,and designs a hardware platform for the detection system based on embedded technology.By interconnecting with the gas sensor array module,the industrial site VOCs gas is realized Online monitoring of concentration.In this paper,the basic principles of thermodynamics are used to calculate the static gas distribution,and the static gas distribution method is used to test and establish the response curve model of the sensor array to stupid,toluene,and xylene.The response data of the sensor array to the mixed gas of three VOCs was obtained through experiments.In this paper,the raw data measured by the experiment is cleaned by the method of gross error elimination.Based on this data,the data set required to establish the nonlinear mapping relationship between the low-cost sensor array and the high-cost PID sensor is established.Based on the data set,through the BP neural network in machine learning,extreme learning machine(ELM),wavelet neural network model(WNN),establish a nonlinear regression model from low-cost sensor array to high-cost PID sensor to realize the sensor Reduced installation and maintenance costs.The verification of the model through sample data shows that the error of the nonlinear regression model is low,and the detection error is controlled within 10%,which can meet the industrial needs.Finally,by designing a visualization program based on the Flask framework,and using Alibaba Cloud’s cloud server architecture to realize remote transmission of monitoring data and visualization functions,a set of low-cost installation and maintenance of industrial site VOCs gas online monitoring system was formed.