paper-with-me

홈 › Papers

Revisiting Implicit Differentiation for Learning Problems in Optimal Control

2023-09-21 · NeurIPS 2023 11

This paper proposes a new method for differentiating through optimal trajectories arising from non-convex, constrained discrete-time optimal control (COC) problems using the implicit function theorem (IFT). Previous works solve a differential Karush-Kuhn-Tucker (KKT) system for the trajectory derivative, and achieve this efficiently by solving an auxiliary Linear Quadratic Regulator (LQR) problem. In contrast, we directly evaluate the matrix equations which arise from applying variable elimination on the Lagrange multiplier terms in the (differential) KKT system. By appropriately accounting for the structure of the terms within the resulting equations, we show that the trajectory derivatives scale linearly with the number of timesteps. Furthermore, our approach allows for easy parallelization, significantly improved scalability with model size, direct computation of vector-Jacobian products and improved numerical stability compared to prior works. As an additional contribution, we unify prior works, addressing claims that computing trajectory derivatives using IFT scales quadratically with the number of timesteps. We evaluate our method on a both synthetic benchmark and four challenging, learning from demonstration benchmarks including a 6-DoF maneuvering quadrotor and 6-DoF rocket powered landing.

📄 PDF Abstract BibTeX

Code (1)

mingu6/implicit-diff-optimal-control 공식 구현

Methods 이 논문이 사용한 방법론

ROCKET Linear classifier using random convolutional kernels applied to time series.

Similar Papers 제목 키워드 기반

Efficient and Modular Implicit Differentiation

2021-05-31 · NeurIPS 2021 12 · Mathieu Blondel, Quentin Berthet, Marco Cuturi, Roy Frostig 외

Automatic differentiation (autodiff) has revolutionized machine learning. It allows to express complex computations by composing elementary ones in creative ways and removes the burden of computing their derivatives by h…

Meta-Learning

Improved Marginal Unbiased Score Expansion (MUSE) via Implicit Differentiation

2022-09-21 · Marius Millea

We apply the technique of implicit differentiation to boost performance, reduce numerical error, and remove required user-tuning in the Marginal Unbiased Score Expansion (MUSE) algorithm for hierarchical Bayesian inferen…

Bayesian InferenceProbabilistic Programming

Vectorized Sparse Second-Order Forward Automatic Differentiation for Optimal Control Direct Methods

2025-06-13 · Yilin Zou, Fanghua Jiang

Direct collocation methods are widely used numerical techniques for solving optimal control problems. The discretization of continuous-time optimal control problems transforms them into large-scale nonlinear programming …

Computational Efficiency

Efficient differentiable quadratic programming layers: an ADMM approach

2021-12-14 · Andrew Butler, Roy Kwon

Recent advances in neural-network architecture allow for seamless integration of convex optimization problems as differentiable layers in an end-to-end trainable neural network. Integrating medium and large scale quadrat…

Portfolio Optimization

Control-Oriented Model-Based Reinforcement Learning with Implicit Differentiation

2021-06-06 · ICML Workshop URL 2021 7 · Evgenii Nikishin, Romina Abachi, Rishabh Agarwal, Pierre-Luc Bacon

The shortcomings of maximum likelihood estimation in the context of model-based reinforcement learning have been highlighted by an increasing number of papers. When the model class is misspecified or has a limited repres…

Model-based Reinforcement Learningreinforcement-learningReinforcement LearningReinforcement Learning (RL)