ネットゼロ建築に向けて：人工知能、エネルギー効率、再生可能エネルギーシステムのレビュー

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:AIと建築エネルギー最適化の研究動向を俯瞰し、BEPG解消に向けた技術的課題と機会を把握できる。

🏢実務担当者:建物の省エネ・再エネ統合戦略を立案する際に、AI適用の効果と課題を理解するための参考になる。

🏛政策担当者:建築分野の脱炭素政策やZEB普及策を検討する際に、AI技術の可能性と限界を考慮する材料となる。

The building sector is one of the largest contributors to global energy demand and carbon emissions, making the transition to net-zero buildings (NZBs) a critical component of climate change mitigation strategies. However, the persistent building energy performance gap (BEPG), defined as the discrepancy between predicted and actual energy consumption, continues to hinder the achievement of net-zero operational performance. Accordingly, this review examines the role of artificial intelligence (AI) in enabling NZBs through the integration of energy-efficient building systems, renewable energy technologies, and intelligent operational control. A comprehensive review of the literature published between 2018 and 2025 was conducted, focusing on three complementary domains: heating, ventilation, and air conditioning (HVAC) system efficiency as the demand-side pillar, renewable energy integration as the supply-side pillar, and AI as the enabling layer connecting both domains. Synthesis of the reviewed literature reveals that demand-side HVAC technologies achieve energy savings ranging from 20% to 67%, while supply-side renewable energy integration increases photovoltaic (PV) self-consumption by 11–13%. Furthermore, AI-driven optimization, particularly through reinforcement learning (22.3% ± 8.4% energy savings) and digital twins (up to 70% renewable energy utilization), substantially enhances building performance within integrated energy management frameworks. The reviewed studies further demonstrate that AI techniques, including machine learning, deep learning, reinforcement learning, and digital twins, enable accurate energy forecasting (R2 > 0.90), intelligent operational control, and effective coordination of integrated PV–battery energy storage system–electric vehicle systems, improving building energy flexibility and reducing grid fluctuations by up to 12.78%. Despite these advances, challenges related to data quality, interoperability, model explainability, cybersecurity, and limited large-scale real-world validation remain significant barriers to widespread adoption. Overall, the evidence indicates that AI serves as a key enabler for reducing the BEPG and improving the reliability, resilience, and operational efficiency of NZBs, thereby supporting the transition toward intelligent, low-carbon built environments.

• openalex https://doi.org/10.3390/app16168111first seen 2026-09-05 05:05:11
• scopus https://api.elsevier.com/content/abstract/scopus_id/105048967323first seen 2026-09-14 05:30:21 · last seen 2026-09-19 05:58:57