Influencing Factors And Adaptive Adjustment Of Humidity Surface Of Human Seat Contact Surface

With the development of society,the public's requirements for quality of life are also increasing.The public's health awareness has been greatly improved compared with the past.research and products for special needs groups are increasingly rich.In the study of the factors affecting the thermal comfort of the seat,the research methods and research conclusions related to the relative humidity of the seat contact surface are poor.In this paper,the factors affecting the environmental humidity of the human-seat contact surface and the influence of the human body seating motion on the data acquisition device are explored.According to the previously investigative literature,we will understand the research methods and research contents in this field at home and abroad.This paper builds a human body-seat contact surface temperature rise simulation device and a seat data acquisition device based on ATMEL Microcontroller unit and HTU21 D sensor.The experiment includes human body-seat contact surface temperature fluctuation experiment and the impact of human body seating action on human body-seat contact surface data acquisition device.The experimental data eliminates the hopping data and the abnormal data by the CEEMD algorithm to achieve the smoothing effect.According to the application environment of the human body-seat data acquisition device,the data after the human body is seated is valid data,and the human body seating motion is used as the starting point for the effective data of the human body-seat contact surface data acquisition device.The seating action includes three types: fast,medium speed and slow speed.The experimental data is segmented by the mixed overlay window method,and the data segment length is 50.The 3-dimensional features of the data segment include the time-domain features composed of mean-variance-extreme difference,the frequency domain features of the FFT low-order coefficients,and the characteristics of the AR model coefficients.Thesupport vector machine classifier is trained by using three kinds of characteristics of three seating speed experimental data,and the grid search optimization algorithm is used to optimize the classifier parameters.The highest recognition rate of the seating action in the three types of feature classifiers is the time domain feature classification.The highest accuracy rate is 97.7%,The lowest recognition rate model is the AR model coefficient feature classifier,and the recognition accuracy rate is 91.3%.Establish a short-term prediction ARMA model of experimental data.Through experimental data analysis,the prediction error of the prediction model is 4%~10%.Through experimental data comparison,the seat contact surface environment control device reduces the relative humidity of the foam material cushion surface by up to 20% by increasing the air convection according to the predicted data of the predicted model.