FastDOC uses Gauss-Newton Hessian approximation to create block-sparse positive semidefinite matrices in the differential KKT system, enabling a factor-of-two reduction in factorization complexity and up to 180% empirical speedup over prior auxiliary-system methods for differentiable NMPC.
Differentiable nonlinear model predictive control.arXiv preprint arXiv:2505.01353
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
representative citing papers
Action aliasing from safety projections harms policy-gradient estimates more severely when the projection is inside the policy than when it is outside, but a penalty term restores competitiveness.
Combining gradient-based policy optimization with recursive system identification in differentiable MPC ensures convergence to an optimal controller under model uncertainty.
citing papers explorer
-
A Gauss-Newton-Induced Structure-Exploiting Algorithm for Differentiable Optimal Control
FastDOC uses Gauss-Newton Hessian approximation to create block-sparse positive semidefinite matrices in the differential KKT system, enabling a factor-of-two reduction in factorization complexity and up to 180% empirical speedup over prior auxiliary-system methods for differentiable NMPC.
-
Safe Reinforcement Learning using Action Projection: Safeguard the Policy or the Environment?
Action aliasing from safety projections harms policy-gradient estimates more severely when the projection is inside the policy than when it is outside, but a penalty term restores competitiveness.
-
Policy Optimization with Differentiable MPC: Convergence Analysis under Uncertainty
Combining gradient-based policy optimization with recursive system identification in differentiable MPC ensures convergence to an optimal controller under model uncertainty.