The growing need to reduce energy consumption and greenhouse gas emissions is driving the search for more efficient heating solutions in buildings. Hybrid heating systems, which combine air-to-water heat pumps (AWHP) with traditional gas boilers, are a common solution after refurbishment investments. However, managing these systems effectively, particularly when integrated with photovoltaic (PV) panels and battery energy storage systems (BESS), remains a complex task. For instance, heat pumps perform poorly in very cold conditions, making boilers a more efficient option; however, it might be advantageous to use it to increase electricity self-consumption. Optimal management depends on multiple factors, including future forecast data. In this paper, a daily optimization program is developed by means of a brute-force approach using forecast data. The core innovation of this paper is the use of an artificial neural network (ANN) that, trained on predictive optimization results, can determine the optimal solution in real-time without the need for future forecasts. The ANN achieved a 99.16% accuracy in new scenarios, successfully optimizing costs, CO2 emissions, and primary energy use. Results indicate up to 19% cost savings in colder cities, a 12% reduction in CO2 emissions, and a 3% decrease in primary energy consumption. This approach holds significant potential for enhancing the integration of renewable energy sources, contributing to long-term sustainability goals.

中文翻译：

---

混合热泵的最优运行策略 - 光伏和电池储能锅炉系统


对减少能源消耗和温室气体排放的需求不断增长，这推动了对建筑物中更高效供暖解决方案的探索。混合供暖系统将空气-水热泵 （AWHP） 与传统燃气锅炉相结合，是翻新投资后的常见解决方案。然而，有效管理这些系统，尤其是与光伏 （PV） 面板和电池储能系统 （BESS） 集成时，仍然是一项复杂的任务。例如，热泵在非常寒冷的条件下性能不佳，使锅炉成为更高效的选择;但是，使用它来增加电力自用可能是有利的。最佳管理取决于多种因素，包括未来预测数据。在本文中，通过使用预测数据的蛮力方法开发了一个每日优化程序。本文的核心创新是使用人工神经网络 （ANN），该网络在预测优化结果上进行训练，可以实时确定最佳解决方案，而无需进行未来预测。ANN 在新场景中实现了 99.16% 的准确率，成功优化了成本、二氧化碳排放和一次能源使用。结果表明，在较冷的城市中，成本节省高达 19%，二氧化碳排放量减少 12%，一次能源消耗减少 3%。这种方法在加强可再生能源的整合方面具有巨大潜力，有助于实现长期可持续发展目标。