Zero Dynamic

Zero dynamics research focuses on controlling underactuated systems, particularly bipedal robots, by managing their inherently unstable dynamics. Current approaches leverage Hybrid Zero Dynamics (HZD) and related methods, often incorporating control barrier functions, reinforcement learning, or Bayesian optimization to generate and stabilize walking gaits, sometimes coupled with model-based control techniques like linear inverted pendulum models. This work aims to create robust and agile locomotion, improving gait stability, energy efficiency, and adaptability to uneven terrain, with implications for both robotics and the broader understanding of dynamic control systems.