finite-horizon-lqr

To compute optimal control sequences for a linear system using finite-horizon LQR, this skill applies backward Riccati recursion to derive optimal gain matrices and then projects states forward through simulation over your specified control horizon.

Backward Riccati recursion is a dynamic programming technique that iterates backward through the control horizon to calculate optimal gain matrices, efficiently automating the complexity of manual control law optimization for MPC frameworks.

Yes, you can use this LQR solver for robotics applications like autonomous vehicles or robotic arms, as it minimizes the quadratic cost over a specific time horizon for linear dynamical systems to calculate optimal control trajectories.

Yes, you need NumPy to run finite-horizon LQR calculations, as the solver requires NumPy specifically for handling the underlying matrix operations and linear algebra computations involved in the Riccati recursion.

The best way to calculate an optimal control trajectory for a robotic arm in Python is to define your system matrices A and B, set a control horizon N, and use this skill to solve the linear quadratic regulator problem.