Neural net robot controller with guaranteed stability

Abstract

A multilayer neural net (NN) controller for a general serial-link robot arm is developed. The structure of the NN controller is derived using a filtered error approach. No learning phase is needed. It is argued that standard backpropagation tuning, when used for real-time closed-loop control, can yield unbounded NN weights if: (1) the net cannot exactly reconstruct a certain required nonlinear control function; (2) there are bounded unknown disturbances in the robot dynamics; or (3) the robot arm has more than one link (i.e. nonlinear case). Novel online weight tuning algorithms given include correction terms to backpropagation, plus an added robustifying signal, and guarantee tracking as well as bounded weights.<>