A Simple Robust Control Scheme for Robot Manipulators With Only Joint Position Measurements

Abstract

In this article, it is shown that the point-to-point motion control of rigid robot manipulators can be achieved by a simple fixed control law that requires only the measurements of the joint positions, unlike most robot motion control, in which joint velocities are invariably needed. Apart from some commonly used assumptions about the manipulator dynamic equation, the control scheme can dispense with any modeling information and yet achieve global stability. Moreover (if the joint configuration is free from singularity), the control strategy can also be used in task space control without any extra technical difficulty, except for the on-line calculation of the Jacobian matrix. The robustness property of the control law against parasitic dynamics from the driving system is discussed. Notation

Pitchfork-Type Bifurcations in a Parametrically Excited, Pd-Controlled Pendulum or Manipulator

“…As a result, there will be an o!set error in the steady state position, proportional to the estimation error, such that B is no longer the equilibrium point. It has been shown, however, that the resulting o!set equilibrium remains globally asymptotically stable [13]. Under-actuation poses an additional set of di$culties, as discussed by Kelly [11].…”

RBF Neural Network Adaptive Control for Space Robots without Speed Feedback Signal

“…These reasons further reduce the quality of the speed signal. 16) Moreover, tachometers increase the weight of moving parts and reduce the efficiency of space robotic manipulators. Therefore, it is difficult to achieve the high control precision in practical engineering.…”

Adaptive Regulation of Manipulators Using Only Position Measurements

“…Implementing these adaptive controllers still requires knowledge of gravity-model structure and joint-velocity measurements, however. Techniques for eliminating the need for velocity measurements in manipulator controllers have been proposed in e.g., Nicosia and Tomei (1990); Yuan and Stepanenko (1991); Canudas de Wit, Fixot, and Astrom (1992); Berghuis and Nijmeijer (1993);Feng and Postlethwaite (1993). One approach to this problem is to design an observer for estimating the velocity utilizing the available (high-quality) position information, and then to use this estimated velocity in place of the actual velocity in the state-feedback component of the control law.…”

“…Alternatively, the simple structure of the PD-plus-gravity solution to the manipulator-regulation problem suggests a correspondingly simple solution to the regulation problem when only position measurements are available. Thus, elegant control laws composed of a gravity-compensation term and a linear dynamic (firstorder) compensator have been proposed recently in Berghuis and Nijmeijer (1993);Feng and Postlethwaite (1993). Note that while these schemes eliminate the need for velocity measurements, they still require knowledge of the gravity model for the manipulator and payload to provide accurate position control.…”