Research On Measurement Method Of Multi Component Gas Mixture Concentration And Development Of Detector

With the development of science and technology,the role of gas concentration detection in chemical production,environmental monitoring and other fields can not be ignored.In practical application,there are many kinds of gases in the environment,and the detection of single gas can not meet the needs of complex detection environment.However,the existing gas detectors are often accompanied by cross response in the multi-component mixed gas,and most of them can only detect a single gas effectively.Therefore,the research on the detection method of multi-component gas mixture and the development of its detector are of great significance for gas detection in various complex environments.In this paper,aiming at the problem of cross response when a single sensor is used to detect gas,a multi-component gas concentration detection method based on information fusion technology is studied,which can accurately detect the concentration of any mixed gas.The detection method of multi-component gas concentration is to identify and detect multi-component gas by support vector machine and BP neural network,optimize the parameters of BP neural network by the steepest descent method,and finally get the best fitting curve,and the output result is the gas concentration of each component.After giving the detection method of multi-component mixed gas concentration,methane and hydrogen were used as test gases to develop a methane and hydrogen binary mixed gas concentration detector.The hardware design part of the detector collects the information of methane and hydrogen gas mixture by catalytic combustion and thermal conductivity sensors.In order to improve the accuracy of concentration detection,the constant temperature detection circuit is designed for the two sensors after studying the detection principle of catalytic combustion sensor and thermal conductivity sensor;The design of gas concentration detection model is programmed in Matlab environment.The other modules are programmed with C language through MPLAB,and the two parts are cross compiled.In this paper,the design of methane and hydrogen binary mixture gas concentration detector prototype is tested by using the concentration gas of laboratory proportion,the test results meet the design requirements,verify the feasibility of the multi mixed gas concentration detection method based on information fusion,and provide an important theoretical basis and implementation method for multi-component mixed gas concentration detection It is of great significance to detect the concentration of compound gas.