Abstract :
This paper addresses the problem of adaptive compensation of torque disturbances in the tracking/positioning control of mechanical systems with the aim of (1) reducing the tracking error for tracking and the position error for positioning tasks, (2) increasing the robustness for uncertain parameters, and (3) using the estimated parameters for failure detection/determination and wear indication. The compensation is incorporated in a standard adaptive nonlinear controller. It merges structural torque disturbance compensation that is important for high velocities and low velocity friction compensation. The stability of the closed loop control system is assessed. The control scheme proposed is applied and tried out on a simulation model and an experimental system. The results indicate that the compensation method proposed is viable, and the degree of verisimilitude of the compensation model is high enough to enable us to use the values of the adapted model parameters for failure detection/determination and wear indication. The last point, however, requires still some experimental evidence. It is expected that the method proposed is suitable for implementation on industrial robots, both for accurate tracking/positioning tasks alone as possibly combined with failure detection/determination. The method is not expected to perform well in the presence of large unmodeled dynamics, i.e., for mechanical systems with flexible joints or with flexible links, e.g., for space structures, without further modifications.
Keywords :
adaptive control; closed loop systems; compensation; nonlinear control systems; position control; robots; stability; torque; tracking; adaptive compensation; adaptive nonlinear controller; closed loop control system; failure detection/determination; industrial robots; low velocity friction compensation; mechanical systems; position error; robustness; stability; structural torque disturbance compensation; torque disturbances; tracking error; tracking/positioning control; uncertain parameters; wear indication;