Robot skill learning method and device, electronic equipment and storage medium

Abstract

The invention discloses a robot skill learning method and device, electronic equipment and a storage medium. The method comprises the following steps that: environment states of a plurality of continuous equal-interval moments are acquired, wherein the environment states comprise robot states and task stage identifiers; the environment states of the plurality of continuous equal-interval moments are input into a trained robot skill learning model to obtain an action description information sequence of a robot learning skill; and an action sequence executed by a robot is determined according to the action description information sequence. According to the method, the environment states of the multiple continuous equal-interval moments are input into the robot skill learning model to obtain the action description information sequence of the robot learning skill, so that robot skill learning is realized, the problems of difficulty in convergence and low success rate which are easily caused in the face of multi-stage complex tasks are solved, the robustness is improved, the robot skill learning efficiency is improved, and efficient and accurate learning of complex skills of the robot is realized.

Description

Technical field [0001] The present invention relates to computer technologies, and particularly relates to a method of robot learning skills, equipment, electronic equipment and a storage medium. Background technique [0002] At present, all kinds of intelligent autonomous robots have been widely used in manufacturing, marine and defense fields. With the development of robotics and artificial intelligence, robot autonomous capabilities continue to improve, we can replace the human complex tasks in more areas. [0003] As a broader application of the skills of a robot learning algorithm, the robot reinforcement learning and interactive learning environment maps from the state into action, and under the guidance of reward function, the best strategy to optimize the network, autonomous robot guidance to complete the assigned task. The traditional control method, the current robot learning skills still exist in actual use compared to many problems and challenges, especially in the fa...

AI Technical Summary

Problems solved by technology

Compared with traditional control methods, the current robot skill learning methods still have many problems and challenges in actual use, especially when faced with multi-stage complex tasks, it is prone to long learning time, difficult convergence, and low success rate. And other issues

Images