Minimizing Energy Consumption of Elastic Robots in Repetitive Tasks
研究如何通过优化弹性驱动参数和连杆轨迹，最小化执行周期性任务的弹性机器人的能耗，实验显示相比刚性驱动可节能高达62%。
Energy consumption is an important issue in robotics. This article deals with the problem of reducing the energy consumption of compliant electro-mechanical systems while performing periodic tasks. After deriving performance indices to quantify the energy consumption of a mechanical system, we propose a method to determine both the optimal compliant actuation parameters, and link trajectories to minimize energy consumption. We show how this problem can be cast in a simpler one where the optimization regards only parameters that define the shape of periodic trajectories to be subsequently determined by using numerical optimization tools. Indeed, in our framework, the optimal stiffness and spring preload can be analytically obtained as a function of the desired link trajectories. We then provide simulations and experimental validations of the obtained results on a two-link compliant manipulator platform which performs a repetitive pick-and-place task. Our experiments show that the use of compliant actuators instead of rigid ones and the optimization of their compliant parameters give rise to an energy saving up to 62% with respect to rigid actuation. Moreover, the simultaneous optimization of the compliant parameters and link trajectories provide an additional energy saving up to 20%.