Adaptive Switched Control For A Class Of Transport Robot Manipulators

Transport robots are a kind of industrial robots that transport goods from one place to another. With the rapid development of industry, the control of robot is demanded to be more and more accurate. The abrupt change of load and the existence of friction often make the control performance deteriorate during the transport process. So, it is always a hot topic in the field of robot control to overcome the affection of changing loads and friction.An adaptive switched controller and a robust adaptive switched controller are designed to deal with the influences of load jump and nonlinear friction. The main works are as follows:Firstly, the parameter jump caused by the load changing is regarded as a switching between different subsystems. Thus, a switched model is established for the robot system whose switching signal is depend on the load. Sub-controllers are designed for subsystems to cope with different loads. Considering the uncertainties of modeling, we use adaptive mechanism in the controller design. In order to avoid the coupling of sub-adaptive laws that caused by sharing a common state, we make the adaptive law relative to the switching signal, i.e., the estimation parameters only update their values when their corresponding subsystems are active and hold their value otherwise.Secondly, taking the friction into account, a robust adaptive switched controller is designed. Asymptotical tracking is retained in the present of nonlinear friction. The advantage of this method is that friction model and accurate friction parameters are not required except for the bounds of the friction.Finally, a kind of adaptive switched controller which does not need real-time inverse for inertia matrix is designed. Constant matrixes replace the estimated inertia matrixes in the adaptive laws, which avoid the inversion of the time varying matrixes.The effectiveness is verified by an example of two DOF robot arm using MATLAB. Results show that the position and velocity of the manipulator asymptotically track the input signal. Summarize is given at last of this paper together with a perspective of future study.