Optimizing simulation of arm stroke in freestyle for swimmers with unilateral forearm amputation

(optimization of affected arm motion)

Abstract

For competitive swimmers with a physical disability, the ideal stroke for crawl swimming has not been clarified yet. It may be different from that of able-bodied swimmers because of differences in body structure. The objective of this study was to solve computationally the fastest arm stroke of crawl swimming for swimmers with unilateral forearm amputation, and to investigate the features of the optimal stroke. The swimming motion of a swimmer with unilateral forearm amputation was reproduced on the swimming human simulation model SWUM, using the videos obtained by a swimming experiment of the swimmer. The optimal stroke was obtained by the optimization method with Particle Swarm Optimization (PSO) and SWUM. In this method, the design variables were the joint angles of the affected limbs for three degrees-of-freedom at the shoulder joint. The objective function was to maximize swimming speed. The swimming speed of the optimal stroke was 7% faster than that of the actual swimmer's stroke. The optimal stroke was the stroke in which the affected limb moved quickly from the entry position to the finish position, and stopped at the finish position to wait the unaffected limb. In addition, a parameter study was conducted by changing the timing to pull the affected limb. It was found that the timing of the optimal stroke was the best to swim fast.