A modeling framework for integrating model predictive control into building design optimization

In response to the growing demand for sustainable and energy-efficient buildings, this study presents a novel simulation framework that integrates model predictive control (MPC) within building design optimization (BDO) to identify envelope designs that enhance energy performance, flexibility, and sustainability. The proposed framework utilizes a modular approach that couples the white-box building model in EnergyPlus with a control-oriented gray-box model identified in MATLAB, enabling economic MPC implementation for HVAC through co-simulation. A genetic algorithm (GA) is employed to optimize building envelope parameters (such as insulation and thermal mass) for maximizing cost savings, using a 20-year life-cycle cost analysis to guide optimal design decisions. Applied to an office building in Brussels, the results show that combining MPC with BDO (MPC + GA) achieves significant annual energy savings (23.7 %), cost reductions (29.4 %), CO₂ emission decreases (27.8 %), and life-cycle net savings of 17.9 €/m2, surpassing the individual MPC or GA cases. Additionally, the MPC + GA configuration requires less than half the thermal mass area of the GA case, delivering superior energy savings and greater energy flexibility by reducing peak power demands. This adaptable framework offers practical guidance for building design professionals and policymakers aiming to create resilient, cost-effective, and sustainable buildings across diverse climates and building types.