Ilqr Ddp

Use when you must run iterative LQR / differential dynamic programming (ilqr-ddp) on a nonlinear discrete-time system: roll the nominal control forward through the two-state soft-landing dynamics under gravity and quadratic drag, run the backward Riccati pass that forms the value-function quadratics Qx, Qu, Qxx, Qux, Quu and the local affine control law u = ubar + K(x - xbar) + k at each step, and execute the forward pass with backtracking line search and regularization until the total cost converges. Produces the converged control sequence, the touchdown state of the constant-descent-rate landing, the feedforward and feedback gain sequences, the per-iteration cost, alpha and mu traces, and the model-versus-actual improvement comparison gating the optimal soft-landing trajectory. Trigger: ilqr-ddp, differential dynamic programming, iterative LQR, backward Riccati pass, local affine control law, forward rollout, nonlinear soft-landing flare.

ashfordeOU Updated 5 repo stars
Aero Agent Skills — aerospace engineering · agentskills.io format · Apache-2.0

File contents

ashfordeOU/aero-agent-skills commit dbd00b593e

Frequently asked questions

npx get-skill add ashfordeou/ilqr-ddp