Robust Optimal Control of Cable-Driven Parallel Robots with Moving Average Fading Memory Kalman Filter Observer

Abstract

This paper examines the filtering and control of Cable-Driven Parallel Robots (CDPRs). To develop the idea, at first, the nonlinear dynamics of the CDPR is converted into a parameterized State-Dependent Coefficient (SDC) structure such that linear filters can be implemented on the robot model. Using the filters, the optimal control of the State-Dependent Riccati Equation (SDRE) in the presence of disturbance and uncertainty is constrained using the concept of internal force. Since the position vector is not directly available to the designer, Kalman Filter (KF), Fading Memory Kalman Filter (FMKF) with the optimal value of the fading factor, and Moving Average Fading Memory Kalman Filter (MAFMKF) are used to estimate the state variable vector of CDPR in the SDC structure and control robot. The performance and efficiency of mentioned filters are compared. Finally, the effectiveness of the proposed controller and filters is checked through several simulations on a planar CDPR.