Robot intelligent grabbing method based on virtual training

Abstract

The invention discloses a robot intelligent grabbing method based on virtual training. The method comprises: building a robot virtual operation platform, randomizing the robot virtual operation platform through a domain randomization algorithm, obtaining the state of a robot and information of the environment where the robot is located so as to be preprocessed to serve as input of robot operation;constructing a depth feature extraction network; constructing a deep reinforcement learning model based on the constructed deep feature extraction network; training the constructed deep reinforcementlearning model to obtain a trained deep reinforcement learning model; and inputting the robot joint angle state and the environment information in the real world into the trained deep reinforcement learning model, generating output of grabbing control, and counting and returning the ratio of the successful grabbing times to the total times as a result index. According to the method, the heavy problems of manual manufacturing and data cleaning of a traditional method are solved, manual processing of feature extraction and trajectory planning is avoided, and the scalability and generalization performance of robot grabbing are improved.

Description

technical field [0001] The invention belongs to the technical field of robot intelligent grasping and artificial intelligence, and in particular relates to a robot intelligent grasping method based on virtual training. Background technique [0002] In recent years, information technology, intelligent hardware and automation systems have developed rapidly. After decades of research at home and abroad, a series of major advances have been made in robot-related technologies. Robots are gradually being widely used in industry, military and life. in each scenario of the service. Especially with the development of artificial intelligence based on deep learning, new demands have been put forward for the digital and intelligent development of the robot industry. Among them, the grasping task is one of the most routine and important subsets of robot operation skills. In order to make the robot obtain more general functions, grasping is a skill that must be mastered. On the basis of...

AI Technical Summary

Problems solved by technology

This technique works by training a policy for a given number of steps and a fixed number of domains, and tends to perform poorly in the face of new environments

Images