Research On The Method Of Hand Movement Recognition And Prosthetic Finger With An Internal Type Of Drive

Half-palm incompleteness is a special case of hand disability,and half-palm prosthetic hands are also born to solve this problem.The remaining palm poses a challenge for the design and control of the prosthetic hand,but at the same time the remaining part of the palm also provides new ideas for information collection and other tasks.The research of this paper takes this as the starting point to study the effective information collection method and recognition algorithm on the palm,and form a complete artificial hand system for experiment on this basis.This article first conducts research in the field of prosthetic hands,especially half-palm prosthetic hands,and evaluates and summarizes the common information extraction methods and recognition algorithms as well as the structure of the prosthetic fingers.Regarding the content and methods of information extraction,this paper compares the methods of key analysis and experimentation,and proposes information extraction and recognition methods based on multimodal information through the summary of the experimental results.After the physical information of the palm is extracted,this article focuses on the recognition algorithms based on this information.These algorithms include pattern recognition algorithms and continuous recognition algorithms.In the study of pattern recognition algorithms,several classifiers such as support vector machines are studied and evaluated.The evaluation criteria include indicators such as classification speed and classification accuracy.On the basis of algorithm research,the recognition effect of factors such as multiple wear-offs is determined.Further exploration of the impact of Parallel to the study of pattern recognition in this article is the study of continuous recognition.The study uses linear regression to predict the position of the finger continuously and evaluates it using indicators such as MAE.In order to form a complete prosthetic hand platform and facilitate the practical application experiment of recognition algorithms,this paper conducts research in the field of structural design of prosthetic hand fingers.Due to the particularity of half palm amputation,the prosthetic finger drive in this application scenario needs to be built into the finger.Based on the design of Prosthetic finger with an internal type of drive,an attempt was made to incorporate the characteristics of coupling and self-adaptation into the design of the prosthetic hand.Finally,this paper also combines the research of information extraction and recognition with the fingers of the prosthetic hand,and conducts experiments on the whole system based on a dedicated experimental platform.The experiments include separate experiments on the performance of the prosthetic fingers and system experiments combined with the recognition algorithm.Based on these experiments,the recognition method and the performance of the prosthetic hand fingers are evaluated.