Evaluation Method Of Office Chair Comfort Based On BP Neural Network

Office chair comfort evaluation is a multi-dimensional, multi-factor problem by office chairs characteristics, physiological and psychological factors as well as the user's environment factors. In this paper, based on office chairs comfort theory, the use of body pressure distribution measurement office chairs and sitting comfort subjective evaluation method of combining to give objective data body pressure distribution and office chairs comfort evaluation of subjective data. Subjective and objective data and the main factors were screened using correlation analysis. Application of BP neural network evaluation model for office chair sitting comfort under different conditions evaluated.The BP neural network model for office chair comfort evaluation on the basis of a comprehensive analysis of office chair sitting, the data structure of BP neural network model to build on the quality and accuracy of prediction. The evaluation model for the office chair comfort evaluation prediction method can provide a comfortable predictability based on the corresponding input parameters of the office chair.The main results of this paper are as follows.(1) On the basis of the relevant office chair comfort theory, experimental selection of office chairs, by body pressure distribution experiments, the final objective and subjective test data obtained office chair comfort. And based on objective data and subjective evaluation, analyze the difference between different seat body pressure distribution index data, the results showed that the distribution of body pressure indicators and comfort of office chairs have a certain relationship.(2) Application Correlation analysis of the experiment measured datas was analyzed screening. At forward and upright postures, the correlation analysis of subjective and objective evaluation of the data, and ultimately selected subjects height, the seat surface maximum pressure, total pressure, mean pressure, contact area and office chair types and so on totally about 6 index factors; In back posture, by subjective and objective correlation analysis, subjects selected height, the maximum pressure, total pressure, average pressure, contact area, the total pressure of the backrest, the backrest maximum pressure, maximum pressure gradient backrest, chair type 9 factors indicators.(3) BP neural network forecasting feature to create office chairs comfort evaluation model prediction. Establish office chair comfort evaluation of forecasting model, neural network by MATLAB R2015 b version software filter for the input data correlation analysis out of the index data, the output data for the office chair comfort evaluation score. According to this paper two kinds of sitting upright and forward data, master data and three back sitting posture total data, the establishment of three kinds of BP neural network, Comparative Analysis of Three BP neural network prediction, the final results by subjects height, the seat surface maximum pressure, total pressure, mean pressure, contact area and office chair type 6 indicator data that can predict the comfortable office chair.(4) Application of BP Neural Network. Increase the sample size and the number of subjects of office chairs, pressure distribution obtained experimentally by objective data and subjective comfort evaluation office chairs scores. Success of the network, most of the relative error of the predicted value and comfort assessment value of less than 10%, the network model can predict the comfortable office chair, providing an effective way for office chair comfort evaluation prediction.