Control Methods

There are many methods to control the double pendulum/acrobot/pendubot system. Here, we categorised them in

• Trajectory Optimization

• Trajectory Stabilization

• Policy-based Control

• Model Predictive Control

even though the methods can build upon each other. E.g. a trajectory can be computed with a trajectory optimization algorithm and then be stabilized with a trajectory stabilization algorithm.

• Trajectory Optimization
  • Iterative Linear Quadratic Regulator (iLQR)
  • Direct Collocation
• Trajectory Stabilization
  • Time Varying Linear Quadrativ Regulator (TVLQR)
  • Riccati Gains Stabilization
• Policy-based Control
  • Linear Quadratic Regulator (LQR)
  • Partial Feedback Linearization (PFL)
  • Soft Actor Critic (SAC)
  • Deep-Q Learning (DQN)
• Model Predictive Control (MPC)
  • Iterative Linear Quadratic Regulator (iLQR) MPC