Centralized Model Predictive Control Strategy for Thermal Comfort and Residential Energy Management

TL;DR

This paper presents a centralized model predictive control approach for residential energy management that optimizes comfort and reduces energy costs by integrating PV, battery storage, and grid interactions.

Contribution

It introduces a novel centralized MPC framework that simultaneously manages heating and energy flow in a residential setup with PV and storage, improving efficiency.

Findings

13.5% reduction in energy cost compared to rule-based strategies

Approximately 31% savings from solar and battery contributions

Effective coordination of heating and energy flow control

Abstract

A novel centralized model predictive control (MPC) is proposed for comfort and energy management in a residential building. The residential setup used here is equipped with a photovoltaic (PV) solar system and a stationary home battery unit. An air-to-air multi-split heat pump (HP) is used as the primary heating system. The electric baseboard (BB) unit in each zone is used as a secondary system. The MPC is simultaneously responsible for controlling the heating inputs of the HP and BB units for comfort management, as well as for the control of energy flow between the PV, the home battery and the bidirectional grid system. Variable Time-of-Use (ToU) rates are considered for the energy cost calculation and Feed-in-Tariff (FiT) is considered for selling energy to the grid. A 13.5% reduction in the energy cost is achieved with the centralized MPC as compared to a rule based energy management strategy. The solar energy generation and battery storage contribute to approximately 31% saving.