Data-Driven Operational Control of Ground Source Heat Pump Systems: An Integrated Review of Modeling, Control, Optimization, and Energy System Integration

Ground source heat pump (GSHP) systems play a pivotal role in the decarbonization of heating, ventilation, and air conditioning systems. However, their long-term performance depends heavily on operational control strategies capable of addressing system dynamics, varying load conditions, and the increasing integration with renewable and distributed energy resources. Data-driven methods have emerged as promising approaches for improving the operational flexibility, efficiency, and sustainability of GSHP systems. Existing review studies mainly focus on predictive modeling and provide limited systematic analysis of the interactions among predictive modeling, control, and optimization, particularly in hybrid and networked energy systems. This review shows recent advances in data-driven approaches for GSHP operational control. The reviewed studies demonstrate that data-driven control strategies can improve operational performance and enable real-time decision-making under dynamic conditions, particularly when predictive models are integrated with first-principles models and optimization algorithms. Beyond conventional standalone applications, this review highlights the transition toward hybrid GSHP configurations and integrated energy systems, including fifth-generation district heating and cooling networks. Within these systems, operational objectives shift from maintaining local ground thermal balance and maximizing equipment efficiency to achieving coordinated energy exchange, load flexibility, economic performance, and carbon-aware network operation. Despite substantial methodological advances, challenges related to data quality, boundary uncertainty, model transferability, and computational scalability continue to limit large-scale deployment of data-driven control strategies. Finally, future directions for developing intelligent, interpretable, and scalable data-driven control frameworks for next-generation low-carbon buildings and district energy systems are outlined.