Research on the Optimal Motion Mode for Aquanaut with Neural Network Method

Aquanaut motion is a collective underwater movement that combines swimming and four limbs motoricity in water. It is necessary to develop algorithm to theoretically analyze the force situation and basic laws of aquanaut athletes to maintain balanced movement in water when conducting research on the movement of four limbs in water, in order to obtain the optimized energy saving pedaling frequency at which a normal sized (fat or thin) adult needs to alternate between pedaling (treading water) with legs under the lowest body energy consumption conditions. Based on the general rules of aquanaut players treading water, this article establishes a differential equation model of the human body through dynamic and kinematic analysis. For this model, this article refers to relevant data from HANAVAN et al.'s research of information database on human dynamics, solves the unknown parameters in the model, and with implementation of the neural network algorithm, obtains a definite model. In order to obtain the lowest pedaling frequency and maintain the lowest physical energy consumption for aquanaut players, this paper also uses genetic algorithm and solves it by taking pedaling frequency as one the individual's fitness parameters. In order to verify the stability and reliability of the model, this paper conducted sensitivity analysis on the results obtained by the adopted algorithms, and found that the optimized treading frequency required to maintain human stability is 2.3 Hz.