Design method for Variable parameter neural solver for motion planning of redundancy mechanical arm

Abstract

The invention discloses a design method for a variable parameter neural solver for motion planning of a redundancy mechanical arm. The method comprises the following steps that the solved task is converted into a performance index and a constraint condition of the redundancy mechanical arm; the performance index and the constraint condition is converted into a time-varying quadratic programming standard form of a to-be-solved system; according to a lagrangian multiplier method, the optimal value in the time-varying quadratic programming standard form is optimized; optimized information is converted into a standard time-varying matrix equation form; a deviation function is designed according to the standard time-varying matrix equation form; according to the deviation function and a power type variable parameter recursion neural dynamics method, the variable parameter neural solver used for the motion planning of the redundancy mechanical arm on a real number field is designed; and a network state solution obtained through the variable parameter neural solver is the optimal solution for the motion planning of the redundancy mechanical arm system. The method has the advantages of being high in calculation speed, high in precision, fast in convergence, high in instantaneity, and good in robustness.

Description

technical field [0001] The patent of the invention belongs to the robot motion planning method, in particular to a design method of a variable parameter neural solver for redundant mechanical arm motion planning. Background technique [0002] Redundant manipulator refers to the degree of freedom (Degrees of Freedom, or DOF) of the manipulator is greater than the degree of freedom necessary to complete the task. With more degrees of freedom, the redundant manipulator can also complete additional tasks such as obstacle avoidance, joint angle limit constraints, and the singular state of the manipulator while completing various tasks of the end effector. The traditional method for solving the inverse kinematics problem of redundant manipulators is based on the pseudo-inverse method. This method has a large amount of calculation and cannot solve the inequality problem, which is greatly restricted in the actual application and operation of the manipulator. In recent years, a sol...

AI Technical Summary

Problems solved by technology

When the neural network is applied in the actual system, such a requirement is often difficult to meet

In addition, in the actual system, the reciprocal of the inductance parameter value and the capacitance parameter value is usually time-varying, especially in large-scale power electronic systems, AC motor control systems, power network systems, etc., the system parameters are set to fixed values is unreasonable

Images