Input Convex Neural Networks for Building MPC

TL;DR

This paper adapts Input Convex Neural Networks for building Model Predictive Control, enabling real-time energy-efficient temperature regulation with convex models that balance accuracy and computational feasibility.

Contribution

It introduces structural constraints to extend Input Convex Neural Networks for multistep predictions in building MPC, ensuring convexity and practical applicability.

Findings

MPC with Input Convex Neural Networks maintains comfort constraints.

The approach reduces cooling energy consumption.

Models achieve a balance between accuracy and convexity constraints.

Abstract

Model Predictive Control in buildings can significantly reduce their energy consumption. The cost and effort necessary for creating and maintaining first principle models for buildings make data-driven modelling an attractive alternative in this domain. In MPC the models form the basis for an optimization problem whose solution provides the control signals to be applied to the system. The fact that this optimization problem has to be solved repeatedly in real-time implies restrictions on the learning architectures that can be used. Here, we adapt Input Convex Neural Networks that are generally only convex for one-step predictions, for use in building MPC. We introduce additional constraints to their structure and weights to achieve a convex input-output relationship for multistep ahead predictions. We assess the consequences of the additional constraints for the model accuracy and test the models in a real-life MPC experiment in an apartment in Switzerland. In two five-day cooling experiments, MPC with Input Convex Neural Networks is able to keep room temperatures within comfort constraints while minimizing cooling energy consumption.