Synergetic synchronized oscillation by distributed neural integrators to induce dynamic equilibrium in energy...

Area of Science:

• Robotics and Biomechanics
• Control Systems Engineering
• Computational Neuroscience

Background:

• Synchronization is prevalent in natural mechanical systems, with joint friction and damping causing energy dissipation.
• Coupled oscillator models are standard for managing multi-joint torque and limit cycle generation, but designing coupling terms and phase/frequency settings is challenging for unknown dynamics.
• Predefining relative coupling relationships in oscillators is difficult for systems with unknown dynamics.

Purpose of the Study:

• To present a novel method for inducing limit cycles in unknown energy dissipation systems.
• To overcome the limitations of coupled oscillator models in managing multi-joint coordination.
• To achieve synergetic synchronized oscillation without prior knowledge of system dynamics.

Main Methods:

• A simple distributed neural integrators method was developed.
• This method induces limit cycles without relying on coupled oscillators.
• It achieves self-organized pattern generation for dynamic equilibrium.

Main Results:

• Synergetic synchronized oscillation was successfully produced, adapting to various physical environments.
• The method achieved balanced energy injection via neural inputs, forming dynamic equilibrium without prior dynamics information.
• Oscillation management occurred without explicit phase or frequency knowledge, yet phase, frequency, and amplitude modulation emerged.

Conclusions:

• The proposed distributed neural integrator method effectively induces synchronized limit cycles in unknown energy dissipation systems.
• This approach enables self-organized pattern generation and dynamic equilibrium, adapting to different mechanical systems and environments.
• The method offers a new paradigm for regulating multi-joint coordination and synergetic oscillations in natural mechanical systems.