国能日新：聚焦“AI+新能源” 大模型助力电力交易收益提升

Core Viewpoint - The integration of artificial intelligence (AI) with the energy sector is accelerating, driven by the need for precise forecasting and intelligent trading strategies in the context of a new power system and high penetration of renewable energy sources [1][7]. Group 1: AI and Renewable Energy Integration - The implementation of AI models is essential for enhancing power forecasting accuracy and optimizing trading strategies in the energy market [1][7]. - The company, Guoneng Rixin, is focusing on deepening the application of AI models in scenarios such as power forecasting, electricity trading, and virtual power plants to maintain a leading position in the renewable energy information service market [1][7]. Group 2: Technological Advancements - The company has developed the "Kuangming" renewable energy model, which has undergone two technical iterations this year, improving the overall accuracy of renewable energy forecasting by 1%-1.5% after the upgrade to version 2.0 in May [2][3]. - The latest version 3.0 of the model incorporates a stable and efficient computing architecture, allowing for pricing every 15 minutes, which significantly enhances revenue per kilowatt-hour for renewable energy companies [3]. Group 3: Revenue Growth and Market Position - In the first half of the year, the company's revenue from renewable energy power forecasting products reached 205 million yuan, a year-on-year increase of 55.14%, driven by the growth in installed renewable capacity and new demand from distributed photovoltaic customers [4]. - Despite the revenue growth, the gross margin for this business segment declined due to changes in the revenue structure, with expectations for recovery as software service fees are recognized [4]. Group 4: Business Strategies and Innovations - The company employs three main strategies to enhance customer retention and renewal rates: optimizing algorithms and AI model applications to improve forecasting accuracy, establishing a nationwide 24/7 operational service system, and adapting products to meet provincial grid assessment rule changes [4]. - The company has also ventured into innovative businesses such as energy storage management, electricity trading, virtual power plants, and microgrid energy management, generating 16 million yuan in revenue in the first half of the year [4]. Group 5: Future Directions - The company plans to further enhance its market share in power forecasting, particularly in the distributed market, while reducing hardware and service costs through standardized service processes [7]. - Continued focus on the iterative development of the "Kuangming" model and the integration of AI technology in deep learning and renewable energy business scenarios is expected to prepare the company for comprehensive marketization of electricity trading [7].