Optimality Studies of Human Sprinting Motions with and Without Running-Specific Prostheses

Abstract

Due to the remarkable performances of some amputee athletes, the power of their running-specific prostheses came to the fore of the discussions. The aim of our study was to compare non-amputee and amputee sprinting motions resulting from optimization using combinations of eight optimality criteria with either fixed or free average velocity. For the description of the amputee and the non-amputee athlete, we created rigid multi-body system models with 16 degrees of freedom in the sagittal plane. Each sprinting motion is the solution of a specific optimal control problem with periodicity and dynamic constraints. We found realistic human-like sprinting motions for both the non-amputee and the amputee athlete. We compared the optimized solutions to dynamics-reconstructed solutions from motion capture data and determined similarity measures for each of them. The investigation of the amputee athlete’s joint torques and ground reaction forces revealed that the real amputee athlete does not exploit the functionality of his running-specific prosthesis as much as the model. The optimal control problems with free average velocity generated human-like sprinting motions as well. However, for specific objective functions the velocities exceed the fastest measured velocities in human sprinting.