Machine learning-based digital district heating/cooling with renewable integrations and advanced low-carbon transition

Intermittent power production with hybrid storages, dynamic grids' interactions for synergistic complementation, advanced energy management, optimal design and robust operation are critical approaches to realise smart district energy systems. Inter-city energy migration framework with energy flexibility can improve efficiency and enhance resilience in response to fluctuations in power supply and demand. However, limited studies focused on up-to-date technology advances and artificial intelligence-assisted control for district energy systems. This study comprehensively reviewed district heating and cooling networks with diversified grids' interactions, smart energy management and control strategy through multi-disciplinary approaches. An inter-city transportation-based energy migration framework was proposed for district energy sharing and regional energy balance. Technical feasibility and prospects of machine learning methods on energy planning and optimisation have also been demonstrated in terms of demand prediction, energy dispatch, surrogate model development for uncertainty analysis and optimisation, geometrical and operating parameter design. A district energy network was formulated, involving on-site renewable generations, waste heat recovery from centralized power plants, multi-diversified energy storages, advanced energy conversions for distributed renewable energy sharing. Several technical challenges were identified as avenues for future research, including benchmarks for selection of most suitable energy storages considering intrinsic differences and local conditions (e.g., climate and geographical conditions), energy congestions between renewables and hybrid grids, optimisations with advanced algorithms, and multi-criteria decision-making to promote willingness and readiness for stakeholders’ participations.