Stochastic energy saving strategies using machine learning for badminton robots

Abstract In Sports, specifically, badminton is a dynamic and skill motions for participants. Previous robots had difficulties behaving as human beings because of their high limitations of low operational speed, heavy bodies, and basic mechanisms. This paper introduces two control approaches to improve a robot's energy effectiveness which needs to make point-to-point motions over a set period. The first method is based on an Adaptive Proximate Energy-Optimized Servo Algorithm (APEOSA) which has optimized parameters for energy efficiency. The second strategy is a Model Predictive Control Strategy (MPCS) to energy management. The technique has been created for a Badminton robot. The robot may still intercept several competitor transportation units on time and in both cases, there is a significant decrease in energy consumption which has been minimized during experimental analysis. An enormous energy-saving, about 40%, is accomplished using EOMPCS and APEOSA compared to APTOS, with the same positioning error as faced by EOMPCS and APTOS.