Rhythmic body movements during animal locomotion are controlled by certain neuronal circuits in the central nervous system, which are called Central Pattern Generators (CPGs). The objective of this research is to uncover the control architecture underlying adaptive and robust locomotion by CPGs through dynamic interactions with the physical environment. Scientific challenges posed by animal locomotion in general will be addressed through focused studies of leech swimming. The leech CPG adapts waveform of body undulation when the environment changes from water to air, and robustly maintains swim capability when the nerve cord is severed or the whole body is cut in half. We pose a hypothesis on the distributed control architecture that achieves such behaviors and provide supporting evidence through model-based analyses. The initial working hypothesis is that the CPG embeds an internal model of the body-fluid mechanics and can be decomposed into a natural gait oscillator and a damping compensator. The outcome of the research will advance understanding of biological information processing and distributed feedback mechanisms for robust and adaptive behaviors. The new knowledge will lay a foundation for development of feedback control theories to achieve oscillatory movements with prescribed patterns through distributed sensing and actuation of complex dynamical systems under varying environment.
A Linear Perspective on Nonlinear Oscillations in Biological Control System for Locomotion
Y. Liu and T. Iwasaki, American Control Conference, pp.1336-1341, 2019.