Design And Implementation Of The Acquisition And Processing System Of Wheelchair Activity Data Based On Android

The rapidly ageing population and a variety of natural and man-made epidemic have caused the continuous increase in the number of the elderly and disabled,most of which have lost capacity to act with different degrees,therefore,wheelchairs become a necessary tool in their daily life.Nowadays,traditional wheelchairs can not meet the diversified needs of the elderly and the disabled,thus the research for intelligent wheelchairs has been undertaken in many countries,China’s research in this field is at the starting stage,so there are still some gaps in the technology level compared with foreign countries.In the future,intelligent planning routines and driving assistance will become an inevitable development trend of the intelligent wheelchairs,and the premise and key to realize such functions require data related to wheelchair activity as support.Under this background,this subject studies how to accurately acquire and process the data generated by wheelchair activity so as to output the target data needed by the intelligent wheelchair R&D and medical service institutions.In this paper,the wheelchair activity data denoising method is studied and optimized firstly based on wavelet analysis.A self-adaption and determination method for the optimal number of wavelet decomposition layers is proposed and an improved threshold manipulation function is designed on the basis of the experimental rule.At the same time,the experiment is conducted with the wavelet threshold manipulation function as the dimension,and the effect of the improved denoising method is evaluated from the subjectivity and objectivity.Next,this paper studies and optimizes the wheelchair action recognition method based on K-means.A relational zoom factor improving the definition ways of distance is proposed in view of clustering process,aimed at the recognition process,a presort method of wheelchair action is proposed to improve the accuracy of action recognition and experimental verify the improved algorithm.And then the acquisition and processing system of wheelchair activity data based on Android has been designed and implemented in this paper.Rely on the electric wheelchairs,this system is made up of data acquisition client and server,which collects and transmits data of accelerated velocity,angular velocity and GPS generated by wheelchair activity,denoises and recognizes the action of wheelchair activity data in the server through controlling relevant sensor integrated in Android smart mobile devices by independently developed APP.The functional testing has been conducted finally.The testing results show that the the acquisition and processing system of wheelchair activity data based on Android can meet the design goal.In addition,the experimental research on how work pattern of APP sampling rate in data acquisition client influences power consumption of smart mobile devices and accuracy of data acquisition has been carried out,and relevant conclusion has been provided.