Detection and Estimation of Fault on Planar Cable-Driven Parallel Robot: SDRE Approach

Abstract

Detecting actuator malfunctions is critical for the Cable-Driven Parallel Robots (CDPRs), as these faults can lead to total system failure. This paper presents a model-based fault detection method for planar CDPRs. To develop the idea at first, the dynamics of general CDPRs is obtained with the inclusion of additive faults and disturbances. Concerning the planar CDPRs, the dynamics is transformed into a State-Dependent Coefficient (SDC) parametrization. Next, a residual vector is generated to detect faults using a semi-Luenberger observer, which obtains adaptation laws and proves convergence of state estimation via the Lyapunov method. Moreover, disturbances and faults are estimated by considering the upper bounds of their values. Since cables in CDPRs can only apply tensile forces, a State-Dependent Riccati Equation (SDRE) approach is presented to control planar CDPR. Simulation results demonstrate the effectiveness of the proposed method in detecting single and multiple cable faults while the robot tracks the constant and time-varying trajectory references.