Semi-optimization of mechanism and control based on sensitivity

Abstract

Tools will assist people and bring out their abilities. However, some tools require training, and they cannot fully perform when the human input isn’t appropriate. In this case, if the performance of a tool is severely degraded by input error, it loses its main purpose as a tool. In this research, focusing on a hopping machine, we develop semi-optimization of tool and control. The design variables are the spring constant of the hopping machine and the time to jump. In the optimization, the input that minimizes the human motion, velocity, acceleration, and jerk is obtained. In the semi-optimization, the sensitivity of the jumping height with respect to input deviations is added. This reduces the performance loss in the presence of input errors, although the performance is slightly lower than that of the optimal solution. Experiments using the proposed method will be conducted to verify the effectiveness of the sensitivity.