Time-Varying Constrained Optimization and Robot Optimal Control
Description
Focusing on error-dynamics-based neurodynamic networks (EDNNs) for optimal control of real-world robots, this book interrogates the application of neurodynamic methods to time-varying constrained optimization (TVCO) problems.
It presents a thorough examination of EDNNs and their applications in TVCO and optimal control of robots. The authors systematically introduce the theoretical foundations, design methodologies, and robotic applications of EDNNs, emphasizing their superiority to traditional optimization solvers. In doing so, this book aims to fill gaps in the application of EDNNs to constrained optimization tasks with a focus on both serial robots (e.g., the Franka Emika Panda robot) and parallel robots (e.g., the Gough-Stewart platform). Key industrial challenges, including obstacle avoidance, joint-limit avoidance, pose control, and high-precision path tracking, are addressed with groundbreaking systematic integration of EDNNs and robot control, as validated by robot simulations and physical experiments.
The book offers practical guidance for researchers and engineers while providing accessible insights for non-specialists, such as librarians and booksellers, on the value of EDNNs in advancing robotic control practices.
More details
Other editions
Additional editions
Persons
Weibing Li is currently an Associate Professor with the School of Computer Science and Engineering, Sun Yat-sen University. His research interests include robotics and neural networks. He has published more than 130 papers, including IEEE TNNLS, IEEE TMECH, IEEE TSMC, IEEE TCYB, et al.
Yehui Li is currently a Postdoctoral Fellow with the School of Computer Science and Engineering, Sun Yat-sen University. His research interests include robotics and neural networks. He has published more than 30 papers, including IEEE TMECH, IEEE TSMC, IEEE TBME, IEEE TIM, et al.
Kai Huang is currently a Full Professor with the School of Computer Science and Engineering, Sun Yat-sen University. His research interests include robotics and real-time systems. He has published more than 100 papers, including Science Robotics, Nature Machine Intelligence, International Journal of Robotics Research, et al.
Yongping Pan is currently a Full Professor with the School of Automation, Southeast University. His research interests include automatic control and machine learning for robotics. He has authored or co-authored over 200 peer-reviewed academic papers, including IEEE TRO, IEEE TAC, IEEE TNNLS, et al.
Content
Section I Time-Varying Constrained Optimization 1. Preliminary Theory 2. TVCO Solvers Section II Optimal Control of Serial Robots 3. Predefined-Time Convergent Solution to Repetitive Motion Planning 4. Strictly Predefined-Time Convergent Solution to Repetitive Motion Planning 5. Acceleration-Level Solution to Obstacle Avoidance 6. Position-Level Solution to Obstacle Avoidance 7. maxQ-Function-Based Solution to Pose Control 8. Lower-Dimension NCP-EDNN for Pose Control 9. Infinity-Norm Velocity Minimization 10. Vision-Based Control with Safety Constraints Section III Optimal Control of Parallel Robots 11. Evtushenko-Purtov Function Based Solution 12. Unification and Comparison of NCP functions Based Solutions