Simulation of control motion of systems in the method of the generalized dynamic balancing problem

The paper considers the problem of numerical modeling of the equations of controlled motion of multidimensional nonlinear systems, i.e., the direct problem. This problem is currently one of the key stages in the synthesis of control for multidimensional robotic systems, such as zoomorphic and anthropomorphic robots, manipulators, and others. In this work, it is assumed that the control is obtained using the method of the generalized dynamic balancing problem. The resulting control is divided into two parts. The first part is the control obtained using the method of the generalized dynamic balancing problem. The second part is the control obtained by the inverse problem method. It is assumed that each part of the controls corresponds to its own group of actuators. The controls are represented as dependencies of the effects (moments, forces in mechanical systems) in the actuators. A method is proposed for stabilizing the controlled motion by introducing feedback into the model of the program motion. The kinematic parameters of the motion are also determined. The method is applicable to systems with redundant links and actuators. A number of numerical examples of the synthesis of controlled motion of high-dimensional systems are considered. The method has been experimentally tested on prototypes of an android robot and a dog robot. On the existing computational element base, the method is a real-time method for systems with a state space dimension of 102–103.
EDN: JNQGLP, https://elibrary/jnqglp

References

[1] Kwakernaak H., Sivan R. Linear optimal control systems. Wiley, 1972. 575 p. (Russ. ed.: Lineynye optimalnye sistemy upravleniya. Moskow, Nauka Publ., 1977. 650 p.).

[2] Kolesnikov A.A. Osnovy teorii sinergeticheskogo upravleniya [Fundamentals of synergetic control theory]. Moscow, Ispo-Servis Publ., 2000. 264 p. (In Russ.).

[3] Aleksandrov A.G. Sintez regulyatorov mnogomernykh system [Synthesis of controllers for multidimensional systems]. Moscow, Mashinostroenie Publ., 1986. 272 p. (In Russ.).

[4] Gorobtsov A.S., Ryzhov E.N. Zadachi nelineynoy stabilizatsii i analiticheskiy sintez rezhimov dvizheniya mnogomernykh dinamicheskikh system [Problems of nonlinear stabilization and analytical synthesis of motion regimes for multidimensional dynamic systems]. Volgograd, VolgGTU Publ., 2008. 176 p. (In Russ.).

[5] Kim J.Y., Kim J.H. Error analysis and effective adjustment of the walking-ready posture for a biped humanoid robot. Adv. Robot., 2010, vol. 24, no. 15, pp. 2137–2169, doi: https://doi.org/10.1163/016918610X534295

[6] Pinto C.M.A., Machado J.A.T. Fractional central pattern generators for bipedal locomotion. Nonlinear Dyn., 2010, vol. 62, no. 1–2, pp. 27–37, doi: https://doi.org/10.1007/s11071-010-9696-4

[7] De Santis A., Di Gironimo G., Pelliccia L. et al. Biomechanically-based motion control for a digital human. Int. J. Interact. Des. Manuf., 2012, vol. 6, no. 1, p. 15, doi: https://doi.org/10.1007/s12008-012-0139-y

[8] Englsberger J., Werner A., Ott C. et al. Overview of the torque-controlled humanoid robot TORO. Proc. IEEE-RAS Int. Conf. on Humanoid Robots, 2014, pp. 916–923, doi: https://doi.org/10.1109/HUMANOIDS.2014.7041473

[9] Haarnoja T., Ha S., Zhou A. et al. Learning to walk via deep reinforcement learning. arXiv:1812.11103, doi: https://doi.org/10.48550/arXiv.1812.11103

[10] Siekmann J., Godse Y., Fern A. et al. Sim-to-real learning of all common bipedal gaits via periodic reward composition. IEEE ICRA, 2021, pp. 7309–7315, doi: https://doi.org/10.1109/ICRA48506.2021.9561814

[11] Surana S., Lim B., Cully A. Efficient learning of locomotion skills through the discovery of diverse environmental trajectory generator priors. arXiv:2210.04819, doi: https://doi.org/10.48550/arXiv.2210.04819

[12] Bledt M.J., Powell B., Katz J. et al. Mit cheetah 3: design and control of a robust, dynamic quadruped robot. IEEE/RSJ IROS, 2018, pp. 2245–2252, doi: https://doi.org/10.1109/IROS.2018.8593885

[13] Katz B., Di Carlo J., Kim S. Mini cheetah: a platform for pushing the limits of dynamic quadruped control. IEEE ICRA, 2019, pp. 6295–6301, doi: https://doi.org/10.1109/ICRA.2019.8793865

[14] Gehring C., Fankhauser P., Isler L. et al. ANYmal in the field: solving industrial inspection of an offshore HVDC platform with a quadrupedal robot. In: Field and service robotics. Springer, 2021, pp. 247–260, doi: https://doi.org/10.1007/978-981-15-9460-1_18

[15] Hwangbo J., Lee J., Dosovitskiy A. et al. Learning agile and dynamic motor skills for legged robots. Sci. Robot., 2019, vol. 4, no. 26, art. eaau5872, doi: https://doi.org/10.1126/scirobotics.aau5872

[16] Lee J., Hwangbo J., Wellhausen L. et al. Learning quadrupedal locomotion over challenging terrain. Sci. Robot., 2020, vol. 5, no. 47, art. eabc5986, doi: https://doi.org/10.1126/scirobotics.abc5986

[17] Gorobtsov A.S. Synthesis of parameters of controlled motion of multi-link mechanical systems of arbitrary structure by the inverse problem method. Mekhatronika, avtomatizatsiya, upravlenie [Мechatronics, Automation, Control], 2004, no. 6, pp. 43–50.

[18] Gorobtsov A., Sychev O., Orlova Yu. et al. Optimal greedy control in reinforcement learning. Sensors, 2022, vol. 22, no. 22, art. 8920, doi: https://doi.org/10.3390/s22228920 (in Russ.).

[19] Gorobtsov A.S. Dynamic balance generalized problem and the promising areas of its application. Izvestiya vysshikh uchebnykh zavedeniy. Mashinostroenie [BMSTU Journal of Mechanical Engineering], 2023, no. 3, pp. 14–24, doi: http://dx.doi.org/10.18698/0536-1044-2023-3-14-24 (in Russ.).

[20] Gorobtsov A.S., Kartsov S.K., Pletnev A.E. et al. Kompyuternye metody postroeniya i issledovaniya matematicheskikh modeley dinamiki konstruktsiy avtomobiley [Computer methods for constructing and studying mathematical models of vehicle design dynamics]. Moscow, Mashinostroenie Publ., 2011. 462 p. (In Russ.).

[21] Gorobtsov A.S. The methods of the transform of multibody systems with redundant constraints. EUROMECH Colloquium 495. Bryansk, BGTU Publ., 2008, pp. 36. (In Russ.).

[22] Budanov V.M., Danilov V.A., Kapytov D.V. et al. MORS: BLDC based small sized quadruped robot. Izvestiya RAN. Teoriya i sistemy upravleniya [Journal of Computer and System Sciences International], 2025, no. 3, pp. 153–177. (In Russ.).

[23] Gorobtsov A.S. Features of solving the inverse dynamic method equations for the synthesis of stable walking robots controlled motion. Trudy SPIIRAN [SPIIRAS Proceedings], 2019, vol. 18, no. 1, pp. 85–122, doi: https://doi.org/10.15622/sp.18.1.85-122 (in Russ.).