Abstract
In this paper, stroke of the front crawl swimming was optimized for swimming speed in two constraint conditions in order to clarify their effect on the results. This optimization used Genetic Algorithm (GA) which was integrated into our swimming human simulation model SWUM. As the variations of the constraint condition, the upper limits for elbow joint torque were reduced to 70% and 30% from the value of previous study. The optimized stroke whose upper limit for elbow joint torque was reduced to 30% has the feature that the elbow does not bend much. From the analysis, it was found that the elbow angle of the optimized stroke depends on the upper limits for elbow joint torque.
