Strategies And Experimental Study Of Trajectory Tracking Control On Uncertain Robot

The control problems of robotic manipulators have received greatattention in theoretical and engineering for many years. When the robotmodel is exactly known, the technique of feedback linearization innonlinear systems can solve the problems very well. However, becauseof some uncertain factors, control system quality will be deteriorated,even the control system may be instable. The intelligent controlschemes have been introduced in the uncertain robotic trajectorytracking control system to eliminate the influence of the uncertainties.First, the fuzzy self-tuning PD trajectory tracking control scheme hasbeen proposed, which regulates PD parameters automatically to meetoptimal tracking requirements. Second, using good robustness ofsliding variable structure and learning ability of neural network, acontrol scheme based on variable structure and neural network has beenpresented. But it has a bad real-time nature, and instead of this, anothercontrol scheme based on dynamic fuzzy neural network has beenpresented to make tracking control faster, adaptive and robust. Finally,by means of fuzzy self-tuning PD control strategy, trajectory trackingexperimental study has been implemented on the SRV02(2-DOF) Robot,which demonstrates the effectiveness of the proposed approach.Several novel design methods about uncertain robot trajectorytracking controller have been mentioned, and computer simulations andpractical experimental verifications have been implement. Thecontribution of this dissertation lies in accelerating the application ofintelligent control algorithm in the industrial robot trajectory trackingcontrol field.