Classical and modern methods for time-constrained energy optimal motion – Application to a badminton robot

• Two energy-optimal control approaches: PEOS and EOMPC are presented.
• They are applied to a badminton robot which performs point-to-point motions.
• EOMPC outperforms PEOS – both in energy and in precision.
• Also EOMPC and PEOS are compared to time-optimal control approaches: PTOS and TOMPC.
• A strong reduction of energy is obtained.

Two control approaches are presented to improve the energy efficiency of a robot which has to perform point-to-point motions during a fixed time interval. One is called Proximate Energy Optimal Servo (PEOS) and the other is called Energy-Optimal Model Predictive Control (EOMPC). The PEOS approach is based on the Proximate Time Optimal Servo (PTOS) control algorithm whose parameters are optimized in order to achieve energy efficient behavior. The EOMPC approach developed based on the Time-Optimal Model Predictive Control (TOMPC) approach achieves energy optimal performance by minimizing the energy losses in the object function. The two developed approaches are applied to a robot playing badminton and compared to two existing control approaches: the Proximate Time Optimal Servo (PTOS) and the Time-Optimal Model Predictive Control (TOMPC). The robot is still able to intercept most of the opponent shuttles on time, while a significant reduction of the energy consumption is demonstrated in both cases.