| With the progress of society and the enlargement of science and technology,the modern industrial control system is stepping forward steadily on the road of complexity and large scale.As a complex MIMO nonlinear system,robot arm integrates dynamics,the development and application of computer software,control theory and other research.It has a broad application prospect,a broad application space for aerospace,precision machinery manufacturing and control and other fields.However,no matter how idealized the system model is established,it seems to be affected by uncertainty during operation,resulting in the insecure and unstable operation of the system locally or as a whole.This requires control design engineers to take system constraints into account when designing control systems to ensure that the system as a whole can still track the ideal trajectory in a relatively stable state under limited conditions.Therefore,it is necessary to consider constrained systems during designing control systems.With the rapid development in the field of robotic arm control,relevant research achievements have been updated,but there are still many problems need to be further explored.The main contributions of this thesis are enumerated as follows:1.Aiming at a class of manipulator systems with output constraints and unmodeled dynamics,an adaptive neural tracking control method based on error compensation mechanism is proposed.The unmodeled dynamic properties and the construction of first-order dynamic signals are used to deal with dynamic uncertainties.BLF is used to handle output constraints.The unknown nonlinear smooth function generated during virtual control design is estimated by using block structure radial basis function neural network.Finally,it is converned the system is SGUUB.A 2-link flexible joint robot system is performed to evaluate the utilization of the concocted control strategy.2.This paper adopts an adaptive neural tracking control method put forward with the aid of the error compensation mechanism for robotic systems with output restrictions and input saturation,which driven by Permanent Magnet Synchronous Machine(PMSM).Effective signals made available by the first-order linear system erase the uncertainties deriving from the unmodeled dynamics.The input saturation of system is solved by using tan-function with mathematical transformation,and utilizing a BLF to address output constraints.During the virtual control design phase,RBFNNs may reliably assess the unknown nonlinear function.All the signals in the controlled system are demonstrated to be SGUUB by integrating all compensation signals into the overall Lyapunov function and using the compact set in conventional dynamic surface control(DSC)stability analysis.A 2-link flexible joint robot system is performed to evaluate the utilization of the concocted control strategy.3.Aiming at the constrained manipulator system model,a suitable nonlinear mapping function is constructed to cope with the time-varying output constraints and state constraints of system by using the property of hyperbolic tangent function.In the design process,error compensation mechanisms are introduced in each step to design the controller.In order to unify the fixed threshold and relative threshold,the event trigger condition is separated from the actual controller,which makes the structure of the controller simpler.A 2-link flexible joint robot system is performed to evaluate the utilization of the concocted control strategy. |
