A method for unstructured ground motion control of multi-legged robots based on depth deterministic policy gradient

Abstract

The invention relates to the technical field of multi-legged robot motion control, in particular to a multi-legged robot motion control method based on depth deterministic strategy gradient. By constructing a deep deterministic policy gradient agent structure, and extracting the action network in the final trained agent network structure as the control strategy of the real robot, it is used for the construction of the real robot, which solves the problem that the robot lacks in the unstructured terrain. Technical problems of poor generalization ability and environmental adaptability. Furthermore, in the method of the present invention, the mathematical model of the joint power supply is introduced to distribute the power supply, and the distribution plan is defined in the motion rules of the simulated robot model, and the output torque of the motor is adjusted through the distribution of power supply power to realize the intelligent network The setting of the structure output joint torque reduces the difference between the simulation environment and the real robot, so that the robot motion control strategy in the simulation environment can be better applied to the real robot.

Description

technical field [0001] The invention relates to the technical field of motion control of legged robots, in particular to a motion control method of multi-legged robots based on depth deterministic strategy gradients. Background technique [0002] With the continuous advancement of science and technology, the development of robot intelligence industry has made many breakthroughs. In many unknown and high-risk fields that humans are involved in, robots can replace people to complete many high-risk jobs, greatly reducing the difficulty and risk factor of people's work, so research on robots is beneficial to people's life and work. Great significance. [0003] Because the contact between the footed robot and the ground is point contact, it has a high degree of freedom joint design and a multi-support point foot structure design, which can realize real-time adjustment of the height of its center of gravity. It has better environmental adaptability and stability. At the same ti...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a multi-legged robot motion control method based on depth deterministic strategy gradient to solve the technical problems of traditional multi-legged robots lacking generalization ability and poor environmental adaptability on unstructured terrain

Images