Tuning of neural oscillators for the design of rhythmic motions

Neural oscillators are an elegant solution that exploits the linking between biomechanics and neuroscience. Having a highly nonlinear dynamics, their parameters are difficult to tune, and current methods are based on simulation programs. This paper proposes an innovative analytical analysis of neural oscillators, both in isolated and coupled situations, by using multiple input describing functions that allow the designer to select the parameters using algebraic equations, therefore simplifying significantly the analysis of the system motion. Furthermore, nonlinear systems are easily handled using this methodology, as well as the analysis of the oscillator internal dynamics