Learning Models of Model Predictive Controllers using Gradient Data

TL;DR

This paper introduces a method for learning Model Predictive Controllers by leveraging gradient information from differentiable MPC solvers, enabling the approximation of complex controllers with neural networks for easier implementation.

Contribution

It proposes a gradient-based learning approach for MPC models, specifically applying it to explicit MPC and representing it with neural networks for reduced complexity.

Findings

Gradient information improves controller approximation accuracy.

Neural networks can effectively model piece-wise affine MPC laws.

Proposed input sampling enhances model evaluation and design.

Abstract

This paper investigates controller identification given data from a Model Predictive Controller (MPC) with constraints. We propose an approach for learning MPC that explicitly uses the gradient information in the training process. This is motivated by the observation that recent differentiable convex optimization MPC solvers can provide both the optimal feedback law from the state to control input as well as the corresponding gradient. As a proof of concept, we apply this approach to explicit MPC (eMPC), for which the feedback law is a piece-wise affine function of the state, but the number of pieces grows rapidly with the state dimension. Controller identification can here be used to find an approximate lower complexity functional approximation of the controller. The eMPC is modelled with a Neural Network (NN) with Rectified Linear Units (ReLUs), since such NN can represent any piece-wise affine function. A motivation is to replace on-line solvers with neural networks to implement MPC and to simplify the evaluation of the function in larger input dimensions. We also study experimental design and model evaluation in this framework, and propose a hit and run sampling algorithm for input design. The proposed algorithm are illustrated and numerically evaluated on a second order MPC problem.