Loading...

Optimal Process Planning for Automated Robotic Assembly of Mechanical Assembles based on Reinforcement Learning Method

Raisi, Mehran | 2021

535 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54094 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Khodaygan, Saeed
  7. Abstract:
  8. Nowadays, the assembly process is planned by an expert and requires knowledge and it is time-consuming. The flexibility and optimality of the assembly plan depend on the knowledge and creativity of the expert, and therefore expertise is an important parameter in developing the assembly plan. Therefore, the use of intelligent methods to plan the assembly process has been considered by many researchers. . The reinforcing learning approach has the potential to solve complex problems due to the use of experience gained from interacting with the environment and Has been successfully implemented in controlling many robotic tasks. However, due to the inherent complexity of the assembly, as well as the hierarchical structure of this problem, a comprehensive solution method has not been introduced so far. In this research, a new approach inspired by hierarchical reinforcement learning method to solve the assembly problem is introduced. The approach of this research to solve the assembly problem is: breaking the general problem into some independent sub-problems, teaching each sub-problem with the appropriate algorithm, and integrate them in the last step. This method is promising for solving problems that have not been implemented in the context of reinforcement learning before.To measure the performance of the proposed method, a case study was performed on the assembly of parts of a gearbox by an industrial arm in the Webots simulation environment. In this case, the parts are randomly placed on the table and the robot arm must perform the optimal assembly process by identifying the position of the parts. Three sub-problems were considered:controlling the robot tool, identifying the position of the parts with machine vision, and optimal tool path planning. In the last step, by aggregating the sub-problems, the performance of the proposed approach was evaluated in 20 tests and the success of the algorithm was confirmed
  9. Keywords:
  10. Intelligent Algorithms ; Hierarchical Reinforcement Learning ; Deep Reinforcement Learning ; Machine Vision ; Intelligent Assembly ; Soft Actor Critic (SAC)Algorithm

 Digital Object List

 Bookmark

No TOC