Model Predictive Climate Control of Connected and Automated Vehicles for Improved Energy Efficiency

TL;DR

This paper develops a model predictive control approach for automotive air conditioning in connected and automated electric vehicles, achieving up to 9% energy savings by optimizing system operation and utilizing vehicle speed preview.

Contribution

It introduces a control-oriented prediction model for A/C systems in CAVs and formulates a nonlinear model predictive control strategy to reduce energy consumption.

Findings

Energy consumption reduced by up to 9% using NMPC.

The control model enforces physical constraints and maintains cabin temperature.

Utilizing vehicle speed preview enhances energy efficiency.

Abstract

This paper considers an application of model predictive control to automotive air conditioning (A/C) system in future connected and automated vehicles (CAVs) with battery electric or hybrid electric powertrains. A control-oriented prediction model for A/C system is proposed, identified, and validated against a higher fidelity simulation model (CoolSim). Based on the developed prediction model, a nonlinear model predictive control (NMPC) problem is formulated and solved online to minimize the energy consumption of the A/C system. Simulation results illustrate the desirable characteristics of the proposed NMPC solution such as being able to enforce physical constraints of the A/C system and maintain cabin temperature within a specified range. Moreover, it is shown that by utilizing the vehicle speed preview and through coordinated adjustment of the cabin temperature constraints, energy efficiency improvements of up to 9% can be achieved.