1 / 32

Autonomous Robotics Lab: Cooperative Control of a Three-Robot Formation

Autonomous Robotics Lab: Cooperative Control of a Three-Robot Formation. Autonomous Robotics Team. Texas A&M University, College Station, TX Fall Presentations 21 November 2008. Outline. Motivation Autonomous Robotics Lab Project Objectives Cooperative Control Laws

mya
Download Presentation

Autonomous Robotics Lab: Cooperative Control of a Three-Robot Formation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Autonomous Robotics Lab:Cooperative Control of a Three-Robot Formation Autonomous Robotics Team Texas A&M University, College Station, TX Fall Presentations 21 November 2008

  2. Outline • Motivation • Autonomous Robotics Lab • Project Objectives • Cooperative Control Laws • Implementation Challenges • Project Results • Conclusions

  3. Motivation • NASA’sVision for Space Exploration includes returning manned missions to the moon by 2020. • Robots are expected to be an integral part of lunar and Martian exploration. • The robots can have varying levels of autonomy: • Teleoperation from Earth (Mars Rovers) • Teleoperation from the lunar surface (Chariot) • Fully autonomous

  4. Motivation • Possible autonomous tasks include: • Transporting materials from point A to point B • Moving materials from landing sites to building sites • Cooperative manipulation of large objects by n robots • Terrain mapping • Search and rescue • The SEI Autonomous Robotics Team’s Mission is to address and understand some of the challenges encountered in the development of autonomous robotics.

  5. Outline • Motivation • Autonomous Robotics Lab • Project Objectives • Cooperative Control Laws • Implementation Challenges • Project Results • Conclusions

  6. Autonomous Robotics Lab • The Autonomous Robotics Lab has been developed to enable hardware testing of autonomous robotics theory and concepts. • The lab includes: • Three iRobot Create platforms • A global-positioning system to measure robot states. • A wireless communications network. • A central PC that manages functions including: • Sequences of autonomous tasks • Trajectory planning • Trajectory-tracking control laws

  7. Autonomous Robotics Lab • The global-positioning system is an overhead camera integrated with image processing software to measure robot states (inertial position and orientation). • On the moon or Mars, satellites or star-tracking systems may provide global positioning information.

  8. Project Objectives • The semester goals are: • A hardware demonstration of cooperative control laws for a three-robot formation. • Investigation of time-delay effects on formation stability. • Primary Tasks: • Development and testing of formation control laws in a MATLAB environment. • Integrating control-law code with the central-PC software for a three-robot formation. • Hardware demonstrations of the formation control laws.

  9. Outline • Motivation • Autonomous Robotics Lab • Project Objectives • Cooperative Control Laws • Implementation Challenges • Project Results • Conclusions

  10. Cooperative Control Laws • Cooperative control involves the control of a group of entities that are working collectively to meet a common objective. • Decentralized cooperative controllers use state information from other vehicles in order to determine control inputs. • Decentralized systems are more efficient for large numbers of vehicles. • Formation control laws used here were developed by Weitz, Hurtado, and Sinclair.

  11. Cooperative Control Laws • The robot equations of motion: • The kinematic vehicle model can be written as: Exact linear representation becomes design space Transformation from design space to robot controls

  12. Cooperative Control Laws • Cooperative control laws were designed to drive three robots to a desired formation (drive errors between vehicles to zero). • Robot 1 tracks a Reference Trajectory. • Robot 2 follows Robot 1 (and reference trajectory). • Robot 3 follows Robot 2 (and reference trajectory). • General control form: Position error wrt lead vehicle Velocity error wrt lead vehicle Velocity error wrt reference trajectory Position error wrt reference trajectory

  13. Cooperative Control Laws • If a leader-tracking scheme is preferred set cp, cv = 0. • If a reference-trajectory-tracking scheme is preferred set kp, kv = 0. Position error wrt lead vehicle Velocity error wrt lead vehicle Velocity error wrt reference trajectory Position error wrt reference trajectory

  14. Cooperative Control Laws • Three control schemes were investigated: • Full-State Measurement Control Law • Rate-Estimate Control Law: rates are estimated using an additional state, f. • Rate-Free Control Law: a different control law is developed that only requires position information relative to the reference trajectory. Commanded Velocity vs. Actual Velocity

  15. Cooperative Control Laws • MATLAB Simulations of control laws were used to: • Select control gains that meet robot performance criteria (acceleration and angular turn rate). • Design reference trajectories that fit within the lab space. Full-State Control Law Rate-Estimate Control Law Rate-Free Control Law

  16. Cooperative Control Laws Full-State Control Law Rate-Estimate Control Law Rate-Free Control Law

  17. Outline • Motivation • Autonomous Robotics Lab • Project Objectives • Cooperative Control Laws • Implementation Challenges • Project Results • Conclusions

  18. Implementation Challenges • Due to lack of computational power onboard the robots, the central PC computes control inputs based upon state information received from the camera (centralized vs. decentralized). Delays are introduced in the process flow due to both computational time and planned delays when sending velocity commands to the robot. Largest delays occur when sending velocity commands to each robot. Delays must be introduced to allow the robots’ onboard microcontrollers to parse data packets.

  19. Implementation Challenges • The control laws command and , but the robot inputs are and . • There are two approaches to implementing the control laws: • Send and commands to the robot, and the onboard microcontroller finds the velocity using a first-order approximation. • Send and commands to the robot, which are held constant until the next update from the camera.

  20. Implementation Challenges

  21. Outline • Motivation • Autonomous Robotics Lab • Project Objectives • Cooperative Control Laws • Implementation Challenges • Project Results • Conclusions

  22. Project Results • Tests run for two trajectories: • Piece-wise trajectory • Tracking both lead vehicle and reference trajectory • Reference trajectory tracking only • Lead vehicle tracking only • Circular trajectory • Tracking both lead vehicle and reference trajectory • Reference trajectory tracking only • Lead vehicle tracking only

  23. Project Results Piece-wise trajectory • 4 constant-velocity trajectories • Some aggressive velocity changes

  24. Project Results • Lead-vehicle and reference-trajectory tracking (kp=kv=cp=cv = 0.5).

  25. Project Results • Piece-wise Trajectory Test Results • Lead-vehicle tracking only was unstable for the following cases: • Gains = 0.5 • Gains = 0.5 with velocity commands sent in the order: Robot 3 → Robot 2 → Robot 1 • Gains = 0.25 with reversed order

  26. Project Results • Video demonstration: piece-wise trajectory

  27. Project Results • Circular trajectory: Lead-vehicle and reference-trajectory tracking (kp=kv=cp=cv = 0.5).

  28. Project Results • Test results

  29. Project Results Lead-Vehicle Tracking (gains = 0.5). Velocity commands sent in order Robot 3 → Robot 2 → Robot 1. Lead-Vehicle and Reference-Trajectory Tracking (gains = 0.5).

  30. Project Results • Video Demonstration: circular trajectory with poor initial conditions.

  31. Conclusions • The full-state measurement cooperative formation control laws were demonstrated in hardware. • The control-law implementation on the central PC yields good results despite the computational and communication delays and the discrete implementation. • The aggressive velocity changes in the piece-wise trajectory caused some instabilities for the lead-vehicle-only tracking schemes. • Delay effects can be seen in the results of the third vehicle. • All vehicles using reference-trajectory information greatly improves convergence and mitigates delay effects.

  32. Acknowledgments • Thank you to NASA-JSC for sponsoring the project. • Thanks to Dr. Hurtado (Faculty Advisor) and Ms. Lagoudas (SEI Director).

More Related