Sensor Array System For Sensing Respiratory Acetone And Study On Data Fusion Method

With the increasing economic level in our country,the quality of life of people has been largely improved,but the problem of diabetes is also becoming more prominent,which greatly endangers people’s health security and efficient development of society and economy.Achieving an early diagnosis of diabetes is of important sense for improving the effect of cure of the condition and ameliorating the level of life of patients.At present,the association of exhaled breath markers with specific diseases has been recognized by a wide range of research scholars,achieving the measurement of the diabetic gas marker acetone has wide application prospects and economic value.In this paper,gas sensor array and data fusion methods are the main research content.A olfactory detection system based on multi-sensor array is designed,its data fusion model is built,the response signals of sensor array are analyzed qualitatively and quantitatively for mixed gases.Simulated gases and human exhaled gases are used to test,and the interference effects of practical measurement are compared and analyzed.The hardware of the system was designed based on the system.CH32F103C8T6 was used to control the operation of the system,and n RF52832 was used to integrate data and transmit data without any radio.In addition,with the help of the external circuit such as the multiple sensor collection circuit and the SD card circuit,the system was able to collect,deal with and store multiple groups of gas.The software of the system was designed to collect,store,filter and integrate the data.The data fusion algorithm for gas analysis uses BP network,and SSA sparrow search algorithm is used to optimize initialization parameters of the BPNN,to avoid the network getting trapped in local optimization during training,to improve the performance of network model for qualitative and quantitative analysis of acetone gas,and to effectively measure trace acetone in multicomponent gas.The results show that the SSA sparrow search algorithm can effectively improve the classification and prediction ability of the model.The accuracy of qualitative identification of mixed gases reaches 96.67%,the maximum relative error of quantitative prediction of acetone does not exceed 5.61%,and the prediction effect is improved significantly.It meets the requirements of qualitative and quantitative analysis of mixed gases and has a broad application prospect in the field of respiratory analysis.Simulated gas and exhaled gas sample tests of the system show that the olfactory detection system can effectively measure the exhaled acetone gas of the human body,and verify that there is a certain relationship between the blood glucose concentration and the exhaled acetone gas concentration of the human body.