COMPARATIVE ANALYSIS OF INTELLIGENT MOTION CONTROL ALGORITHMS FOR AN INDUSTRIAL ROBOT
Abstract
The paper compares several approaches to trajectory tracking control for an industrial robot manipulator. The robot is described by a rigid-link dynamic model that includes inertia, Coriolis and centrifugal terms, gravity, friction and external disturbances. Within one tracking-error formulation we examine classical PID control, inverse-dynamics compensation, adaptive control and adaptive control with a radial basis function (RBF) neural network. A filtered error is used to build the control law, and a Lyapunov function is used to discuss closed-loop stability. We also separate the fast local servo loop from IoT-based monitoring and edge computing, which have different timing requirements. Since no new laboratory experiment was carried out, the methods are compared qualitatively by model requirements, computational load, ability to adapt and difficulty of proving stability. Metrics for future simulation and experimental work are proposed.
