Adaptive optimal trajectory tracking control of delta robots with online parameter estimation

摘要

As typical parallel robots, Delta robots are widely utilized in various industrial sectors, such as medicine, electronics, automobiles, etc. In industrial production, both control performance and energy consumption are crucial. From this perspective, this article proposes an adaptive optimal trajectory tracking controller for Delta robots, which ensures a trade-off between tracking error and energy consumption. The proposed controller is composed of two parts: a dynamic behavior related term, which retains the steady-state tracking response, and an optimal term, which optimizes the control performance. More specifically, neural networks (NNs) are leveraged to approximate the solution of performance index function and the optimal input based on the action-critic framework. Besides, since the Delta robot is mainly used to implement pick-and-place (PaP) tasks, which indicates that the mass of loads is changeable frequently, a parameter estimator is designed, which can online estimate the unknown load mass effectively. Through rigorous theoretical analysis, the closed-loop system is proven to be uniformly ultimately boundedness (UUB). Finally, simulations are conducted to validate the effectiveness and performance of the proposed method.