Trajectory Planning for Mechanical Systems Based on Time-Reversal Symmetry
Stepan Ozana,
Tomas Docekal,
Aleksandra Kawala-Sterniuk,
Jakub Mozaryn,
Milos Schlegel,
Akshaya Raj
Affiliations
Stepan Ozana
Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer Science, VSB-Technical University of Ostrava, 17. Listopadu 2172/15, 708 00 Ostrava-Poruba, Czech Republic
Tomas Docekal
Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer Science, VSB-Technical University of Ostrava, 17. Listopadu 2172/15, 708 00 Ostrava-Poruba, Czech Republic
Aleksandra Kawala-Sterniuk
Faculty of Electrical Engineering, Automatic Control and Informatics, Opole University of Technology, Prószkowska Street 76, 45-758 Opole, Poland
Jakub Mozaryn
Institute of Automatic Control and Robotics, Faculty of Mechatronic, Warsaw University of Technology, ul. Św A. Boboli 8, 02-525 Warsaw, Poland
Milos Schlegel
Department of Cybernetics, Faculty of Applied Sciences, University of West Bohemia, Technická 2967/14, 306 14 Pilsen, Czech Republic
Akshaya Raj
Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer Science, VSB-Technical University of Ostrava, 17. Listopadu 2172/15, 708 00 Ostrava-Poruba, Czech Republic
The generation of feasible trajectories poses an eminent task in the field of control design in mechanical systems. The paper demonstrates innovative approach in trajectory planning for mechanical systems via time-reversal symmetry. It also presents two case studies: mass-spring-damper and inverted pendulum on the cart. As real systems break the time-reversal symmetry, the authors of this work propose a unique method in order to overcome this drawback. It computes a feed-forward reference control signal and state trajectories. The proposed solution enables compensation for the effects of couplings, which break the time-symmetry by a special proposed measure. The method suppresses the overall open-loop accumulated error and produces high-quality favorable control and state trajectories. Furthermore, the existence of the designed control signal and state trajectories is guaranteed if the equations of the motion have a solution in the direct flow of time.