A Method for Repeated Motion Planning of Redundant Robots Using Fast Double Power Final State Neural Networks

Abstract

A Redundant Robot Repeated Motion Planning Method Using Fast Double-Power Final-State Neural Networks, Given the Desired Trajectory r of the Robot End-Effector in Cartesian Space d (t), and give the expected closing angle θ of each joint d (0); for the repetitive movement of the robot, the asymptotic convergence performance index is adopted, and the quadratic optimization problem of redundant robot trajectory planning is transformed into a time-varying matrix equation solving problem, and the fast double-power final state neural network is used as the solver . Achieving fast finite-time convergence for repetitive motion planning tasks for redundant robots in the presence of offset initial positions. The invention provides a redundant robot motion planning method using a fast double-power final state neural network, which has fast convergence in limited time, high calculation accuracy and is easy to implement.

Description

Technical field[0001]The invention relates to a repetitive motion planning technology for an industrial robot. Specifically, it proposes a redundant robot repetitive motion planning method with fast finite time convergence and an initial position deviating from a desired trajectory, using a fast dual-power final state neural network .Background technique[0002]Redundant robots refer to a type of robots that have more active joints than the minimum number of degrees of freedom required to perform the target task. Redundant robots can avoid the defects of non-redundant robots such as low flexibility, singularity, and inability to avoid obstacles, and can complete variable tasks in a complex work environment.[0003]The inverse kinematics solution of redundant robots is the basis of redundant robot motion planning and trajectory control. The extra degrees of freedom make there are infinitely many inverse kinematics solutions, so performing a predetermined task requires an infinite number ...

AI Technical Summary

Problems solved by technology

[0025] In order to overcome the shortcomings of existing redundant robot repetitive motion planning methods that cannot converge in a limited time and have low calculation accuracy, the present invention provides a fast double-power finalization method that has fast convergence in a limited time, high calculation accuracy, and is easy to implement. Redundant robot motion planning method based on dynamic neural network

Images