Modeling method and system for robot inverse dynamic model

Abstract

The invention discloses a modeling method and system for a robot inverse dynamic model, and the method comprises the steps: building a recurrent neural network which comprises an input layer, a hiddenlayer and an output layer, wherein the hidden layer comprises an outer layer and an inner layer, the outer layer comprises an outer-layer forgetting door and an outer-layer resetting door, and the inner layer comprises an inner-layer forgetting door and an inner-layer resetting door; training an inverse dynamic model of the mechanical arm by using a recurrent neural network,wherein the input dataof the recurrent neural network is the motion state of the mechanical arm, the output data of the recurrent neural network is the torque for controlling the mechanical arm to generate the motion state, and the motion state comprises the motion position, the motion speed and the motion acceleration; and enabling the inner-layer memory unit data output by the inner layer in the hidden layer of therecurrent neural network to update the outer-layer memory unit data of the outer layer, so that the memory information of the time sequence is reserved for a longer time, and the precision of outputting the predicted torque by the inverse dynamic model is improved.

Description

technical field [0001] The invention relates to the technical field of intelligent robots, in particular to a modeling method and system for an inverse dynamics model of a robot. Background technique [0002] In today's society, robots play a very important role in our lives. They can replace people to complete some simple mechanical and even high-altitude dangerous work, and an important part of this is the control of robots. The inverse dynamics model of the robot is based on the motion state (movement position, motion speed, motion acceleration) of the robotic arm of the robot to calculate the corresponding torque to control the motion of the robotic arm to achieve ideal control. However, the current physical modeling of the inverse dynamics model of the robot has errors caused by friction, centripetal force, and Coriolis force. As a result, the established inverse dynamics model cannot adapt to the current state of motion of the robotic arm and cannot meet the precise ac...

AI Technical Summary

Problems solved by technology

But neither of them can retain memory information for a long time, so they cannot provide a more accurate time series forecasting model

Images