Extended Kalman Filter-Based State Estimation and Adaptive Control of Cable-Driven Parallel Robots

Abstract

Cable-Driven Parallel Robots (CDPRs) are used in ever-changing, unstructured, and long-term autonomous operations; however, they require precise component assembly to achieve high positioning accuracy. This article presents an adaptive control framework for CDPRs that addresses actuator position uncertainty in all types of CDPRs without relying on vision-based sensing. The core concept of the developed adaptive control scheme involves employing an Extended Kalman Filter (EKF) to estimate system states, including the uncertain actuator positions and the end-effector pose, and replacing the uncertain parameters in the feedback controller with their estimates. Monte Carlo Simulations (MCSs) are also conducted to evaluate the robustness and stability of the proposed estimation method under the anchor point uncertainties. Moreover, the proposed controller incorporates a robust term to compensate for the unmodeled dynamics and payload changes. The results demonstrate that the adaptive control design effectively reduces the actuator position uncertainty, enhances the end-effector positioning accuracy, and successfully compensates for the payload changes. The performance comparisons of the proposed adaptive controller over its non-adaptive counterpart and PID controller highlight its superior performance in managing the anchor point uncertainties and adapting to the payload changes.