Coordinating manipulator-based adaptive neural network synchronous robust controller design method

Abstract

The invention relates to a coordinating manipulator-based adaptive neural network synchronous robust controller design method, mainly to enable a clamped workpiece to accurately track a desired trajectory and an internal force in a working condition with uncertain base coordinates. The method uses adaptive sliding mode control and an RBF neural network as the basis of the control method, through combination of the two and with setting of a corresponding approximation robust control item for a manipulator coordination clamping system, design of the control method is realized. The control methodcan adaptively compensate the trajectory errors caused by uncertain base coordinate translation errors and rotation errors, uncertain manipulator system dynamic parameters and uncertain base coordinate parameters are approximated through the neural network, the neural network has a function of continuously updating a weighting factor with an input field, errors and time, the uncertain base coordinate parameters can be compensated in a short time, the trajectory and internal force tracking errors of the manipulator clamping workpiece are converged to be near the desired value, and the controlprecision is improved.

Description

technical field [0001] The invention relates to a method for coordinating clamping and clamping of manipulators. More precisely, an adaptive neural network synchronous robust controller is designed for the situation that the base translation and angle parameters of the manipulator clamping system are uncertain. Ensure the accuracy of system trajectory and internal force tracking. technical background [0002] In the current processing and manufacturing industry, such as assembly, handling, spraying and welding, more and more types of work need to use multiple robotic arms to perform complex tasks that require interaction and collaboration. Although the traditional single robotic arm can meet basic tasks Requirements, but work efficiency and finishing quality need to be improved, therefore, multi-arm coordinated processing is particularly important. However, the multi-manipulator system requires the mutual cooperation between the manipulators, so the relative position betwee...

AI Technical Summary

Problems solved by technology

This patent uses the relationship between neural network and sliding mode variable structure control to design an adaptive neural network synchronous robust controller to solve the situation of uncertain parameters of the base of the manipulator.

Images