Data Collection and Analysis of Track and Field Athletes' Behavior Based on Edge Computing and Reinforcement Learning

Abstract

With the development of multimedia technology, the computer auxiliary system has become an effective means of daily training in track and field. This paper designs a data acquisition and analysis system for track and field athletes. The system uses sensor modules attached to the athlete's body to collect movement data for analysis. The whole system is implemented by edge computing architecture. In order to reduce average response time, the DDPG algorithm is used to optimize the resource allocation of the edge layer. Experimental results show that the response time of the proposed algorithm can be controlled within 1 s. Meanwhile, the SVM algorithm on the edge server is arranged to classify the data, and the overall recognition accuracy is over 90%.