Abstract:
The dynamic optimization of a box delivery motion is a complex task. The key component is to achieve an optimized motion associated with the box weight, delivering speed, and location. This thesis addresses one solution for determining the optimal delivery of a box. The delivering task is divided into five subtasks: lifting, transition step, carrying, transition step, and unloading. Each task is simulated independently with appropriate boundary conditions so that they can be stitched together to render a complete delivering task. Each task is formulated as an optimization problem. The design variables are joint angle profiles. For lifting and carrying task, the objective function is the dynamic effort. The unloading task is a byproduct of the lifting task, but done in reverse, starting with holding the box and ending with it at its final position. In contrast, for transition task, the objective function is the combination of dynamic effort and joint discomfort. The various joint parameters are analyzed consisting of joint torque, joint angles, and ground reactive forces. A viable optimization motion is generated from the simulation results. It is also empirically validated. This research holds significance for professions containing heavy box lifting and delivering tasks and would like to reduce the chance of injury.
