Neural-iLQR: A Learning-Aided Shooting Method for Trajectory Optimization

TL;DR

Neural-iLQR introduces a learning-aided trajectory optimization method that uses a neural network to model system dynamics, improving performance over traditional iLQR in the presence of model inaccuracies.

Contribution

The paper proposes a novel neural network-based local system model within iLQR, enabling iterative refinement without prior system knowledge.

Findings

Outperforms conventional iLQR with inaccurate models

Effective in complex control tasks

Reduces reliance on precise system models

Abstract

Iterative linear quadratic regulator (iLQR) has gained wide popularity in addressing trajectory optimization problems with nonlinear system models. However, as a model-based shooting method, it relies heavily on an accurate system model to update the optimal control actions and the trajectory determined with forward integration, thus becoming vulnerable to inevitable model inaccuracies. Recently, substantial research efforts in learning-based methods for optimal control problems have been progressing significantly in addressing unknown system models, particularly when the system has complex interactions with the environment. Yet a deep neural network is normally required to fit substantial scale of sampling data. In this work, we present Neural-iLQR, a learning-aided shooting method over the unconstrained control space, in which a neural network with a simple structure is used to represent the local system model. In this framework, the trajectory optimization task is achieved with simultaneous refinement of the optimal policy and the neural network iteratively, without relying on the prior knowledge of the system model. Through comprehensive evaluations on two illustrative control tasks, the proposed method is shown to outperform the conventional iLQR significantly in the presence of inaccuracies in system models.