DYNAMIC TRAJECTORY PLANNING AND SELF-TUNING ADAPTIVE CONTROL DESIGN FOR INDUSTRIAL ROBOTS (MICROCOMPUTER CONTROL, COMPUTER AIDED DESIGN, SYSTEM IDENTIFICATION)

For industrial robots, higher precision and speed are required in order to broaden their application areas. The objective of this thesis is to improve the current technology of manipulator control.;An adequate forgetting factor in the recursive parameter estimation permits rapid adaptation when substantial plant changes occur. However, the constant forgetting factor algorithm does not work well in the application of an industrial robot. A new variable, forgetting factor algorithm has been proposed in this research to ensure parameters convergence and prevent 'blowing-up' of the covariance matrix. The effectiveness of this algorithm has been demonstrated through several simulations and the improved performance demonstrated on the T3-726 industrial robot using the developed prototype controller.;Trajectory generation is usually performed off-line based on geometric and kinematical considerations only. In this research, a new adaptive approach to generate the nominal trajectories in the joint-coordinate space is developed to obtain a control with near minimum response time and tracking error.;Adaptive control systems show inherent nonlinear behavior, therefore a CAD software package was developed to perform control system design including signal analysis, off-line and on-line system identification, controller design, simulation and direct digital control operation. The realization of this program on a hierarchical computer system for industrial robot control design has also been outlined.;A solution to realize precise and high-speed motion control of a multiaxis industrial robot is proposed by using an on-line system parameter estimation and a pole-placement self-tuning control strategy. The proposed microcomputer-based adaptive controller has the capability of increasing the speed of a robot arm and improving the stability and accuracy of the system over a wide range of motions. An experimental set-up composed of a multi-microprocessor system and a Cincinnati Milacron T3-726 industrial robot was used to implement the developed real-time control strategy.