Experimental robot and payload identification with application to dynamic trajectory compensation

Industrial robot manipulators have become an indispensable means of automation to increase flexibility and productivity. The ever increasing quality standards and new applications impose higher requirements on accuracy, reliability and performance. Due to the complex nonlinear robot dynamics, the design of robot controllers should include accurate dynamic robot models. Robot identification is an experimental technique to estimate realistic and accurate dynamic robot models from motion data and actuator torques measured during ‘well-designed’ experiments. This research improves and extends the existing experimental identification methods and applies the obtained dynamic models to improving path tracking control. The influence of a kinematic calibration and more appropriate model descriptions for some additional effects on the obtained model accuracy has been investigated. Further, this work presents the experimental validation of an identification approach using both ‘internal’ and ‘external’ (i.e. a force/torque platform) sensors. The use of different types of sensors into one combined identification problem provides more information on the robot dynamics, yielding improved parameter estimation accuracy and actuator torque prediction. Due to the growing importance of the robot payload, the robot identification method is extended to the estimation of the inertial parameters of the robot payload. In this application not only the actuator torque prediction accuracy is important, but special attention is paid to the accuracy of the individual parameter estimates. This work presents a payload identification approach which does not require a full identification of the manipulator, but compensates for all known robot dynamics based on available a priori information. A sensitivity analysis