Inverse kinematic solving method based on neural network

Abstract

The invention discloses an inverse kinematic solving method based on a neural network. The inverse kinematic solving method based on the neural network mainly solves the problems that in the prior art, in the process of inverse kinematic solving, value calculation time is long, convergence is difficult to guarantee, and a computation result is single. The inverse kinematic solving method based on the neural network comprises the following steps that the equivalent optimization problem expression of an inverse kinematic problem is worked out; a recurrent neural network solving model for an optimization problem is established, and a kinematic solving variance of the optimization problem is set; model parameters for guaranteeing the convergence of the solving model are set; the initial value disturbance of a system state is set, and the variety of a movement trace line is solved. By the adoption of the scheme, the inverse kinematic solving method based on the neural network achieves the purposes that the technical speed of the model is effectively controlled, the demand for three-dimensional rendering application is met, and the calculation stability is good, and has high practical value and popularization value.

Description

technical field [0001] The invention relates to a method for solving inverse motion based on a neural network. Background technique [0002] Inverse kinematics is a concept derived from machine kinematics. It refers to the use of kinematic equations to solve the trajectory of the robot's joint motion, and finally makes the robot's end effector reach the specified position. The solution of inverse motion problems is widely used in various applications, such as interactive computer graphics and the current emerging computer 3D game design, virtual reality, etc. In order to make the motion of animation objects more realistic and accurately simulate the response and behavior of real creatures in the game, the motion control of joints can be realized by solving the motion trajectory equation. In 3D rendering, it involves the rendering speed requirements of images, so for The solution of the equations of motion has higher computational performance requirements. [0003] In the f...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a neural network-based inverse kinematics solution method, which mainly solves the problems in the prior art that the numerical calculation time of inverse kinematics solution is long, the convergence is difficult to guarantee, and the calculation result is single

Images