Work

The Integration of Explicit MPC and ReLU based Neural Networks

Year: 2020

Type: article

Abstract: Using neural networks to capture complex dynamics of highly nonlinear systems is a promising feature for advanced control applications. Recently it has been shown that ReLU based neural networks can be exactly recast in a mixed-integer linear programming formulation. This reformulation enables the incorporation of deep learning models in model predictive control strategies. To alleviate the computational burden of solving the piecewise linear optimization problem online, multiparametric programming is utilized to obtain the full, offline, explicit solution of the optimal control problem. In this work, a strategy is presented for the integration of deep learning models, specifically neural networks with rectified linear units, and explicit model predictive control. The proposed strategy is demonstrated on the advanced control of the ACUREX solar field.

Source: IFAC-PapersOnLine

DOI: 10.1016/j.ifacol.2020.12.544

Authors: Justin Katz, Iosif Pappas, Styliani Avraamidou, Efstratios N. Pistikopoulos

Institution: Texas A&M University

Language: English

Topic: Advanced Control Systems Optimization

Subfield: Control and Systems Engineering

Field: Engineering

Domain: Physical Sciences

Open Access status: diamond

Funders: National Science Foundation, Energy Institute, Texas A and M University

Awards: INFEWS/T3: Decision Support for Water Stressed FEW Nevus Decisions (DS-WSND), G5777797285, SusChEM: An integrated framework for process design, control and scheduling [PAROC]