Position Control of Robots by Nonlinearity Estimation and Compensation: Theory and Experiments

To improve the performance of position controllers of robots, a designapproach using the method of nonlinearity estimation and compensation ispresented. The controller designed has a similar structure to that of thelinear independent joint control. Nonlinearities in the robot dynamics areestimated by a state observer based on approximate linear models of thesystem, and then compensated by an appropriate feedback. The paper presentsthe theory how to design such a controller as well as the experimentalresults verifying its performance.