Loading...

Learning-based Control System Design for the Bipedal Running Robot and Development of a Two-layer Framework for Generating the Optimal Paths in Various Movement Maneuvers

Amiri, Aref | 2022

320 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54802 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Salarieh, Hassan
  7. Abstract:
  8. Foot movement is one of the most powerful and adaptable methods of movement in nature. Inspired by humans, the most intelligent creatures on earth, bipedal robots have many uses. In this research, a control method for running a bipedal robot has been designed. In the simulation part of the five-link model, the robot's motion equations for running and walking at different levels are extracted by the Lagrange method. In path generation, using the two-layer optimization method and holonomic and dynamic constraints, optimal paths are produced which are kinematically and dynamically possible (feasible). Additionally, path generation is facilitated by an invariant impact constraint to ensure the periodicity of the movement path. The motion generation algorithm generates optimal time-based trajectories for walking, climbing stairs, and running. The control system consists of a controller based on deep reinforcement learning and a balance recovery system. The controller, as an intelligent and advanced agent, has the ability to correct online in different motion conditions and is also resistant to disturbances and noise to an appropriate degree. In addition, the proposed learning-based system is able to help the robot to regain balance. The control system presented in this study offers a new perspective on the control of robotic systems of bipedal robots
  9. Keywords:
  10. Bipedal Robot ; Running ; Deep Reinforcement Learning ; Optimization ; Hybrid Dynamical System ; Underactuated Biped Robot ; Trajectory Generation

 Digital Object List

 Bookmark

No TOC