ANALYSIS OF HUMAN BIOMECHANICAL CHARACTERISTICS IN SPRINT SPORTS

The completion time of sprinting is short, and it requires high explosive power from the human body. Satisfactory results can only be achieved through scientific training. From the perspective of analyzing human biomechanical characteristics, an analysis method based on an Echo State Network (ESN) was constructed. This method utilizes the strong learning ability and good approximation ability of ESN to achieve stability under limited data training. The angles of the six joints in the lower limbs are used as inputs to the ESN, and athlete sprint acceleration data is used as outputs to the ESN. The experimental results show that the constructed ESN can effectively analyze the changes in joint angles during short-distance running. Compared with statistical analysis methods, it improves the training accuracy of each joint by about 10%, which is of great significance for assisting the scientific training of short-distance running.