Model Predictive Control for Battery Thermal Management System

This thesis work aims to determine the effectiveness of Model Predictive Control (MPC) in increasing the vehicle autonomy as well as the battery life in the context of Battery Thermal Management (BTM) System in electric and hybrid vehicles. BTM is, in fact, an important problem to deal with, since it affects vehicle autonomy and the Battery State of health (SOH). In this work, Adaptive Model Predictive Control (AMPC) has been chosen among the possible MPC controllers. The scheme which implements the Vapor Compression Refrigeration Cycle (VCR) and the thermal management system as whole has been developed in Simscape environment by Mathworks company, which has provided it as an example (“Electric Vehicle Thermal Management”). The original compressor control, which was a PID, has been replaced by an AMPC developed in the Matlab and Simulink’s environment. The internal prediction model of the AMPC has been implement using a state space representation of the First Thermodynamic Principle referred to the battery control volume. In this State Space representation, the chosen states have been the battery temperature and the battery State of Charge (SOC), whereas the Manipulated Variable has been the electric power demand of the compressor to be controlled. Both PID and AMPC strategy have been tested and simulated in Matlab and Simulink Environment. The results of this work are shown and discussed.