Modelling and simulations of a central pattern generator controlled, antagonistically actuated limb joint

Wearable assistive devices (WAD) assist rehabilitation and recuperation. However, WADs' development is impeded by traditional actuators and current control paradigms' lack of compliance and adaptability, in contrast to the mammalian central nervous system. A limb's joint's compliance is a function of muscles' level of activation and passive viscoelastic properties. Increased equal and unequal co-activation of a limb joint's muscles leads to an increasingly stable joint. Unequal limb joint muscle activation leads to an angular displacement. Muscles' activation is controlled by alpha motor neurons, which are innervated by the spinal-cord's autonomous neural networks, i.e. central pattern generators. Given a continuous excitatory supraspinal input, a two-level, mutually inhibiting, CPG network can control an antagonistically actuated limb joint's muscles' level of activation, thus the joint's compliance, and joint oscillation frequency.