以下为研究报告摘要： 电力设备板块：25H1业绩稳健。2025H1电网板块合计实现营收1740.49亿元，同比增长 10%；实现归母净利润139.98亿元，同比增长10%。25Q2板块保持增长趋势，收入实现 1045.20亿元，同比增长12%；归母净利润实现90.71亿元，同比增长8%。25H1板块毛利率为 25.57%，同比下降1.52pcts；净利率为7.33%，同比上升0.72pct，25H1盈利能力有所承压判 断主要系大宗材料价格波动小幅影响。 人形机器人：25H1营收和利润均实现双位数同增。2025H1板块合计实现营收1122.08亿 元，同比增长11.94%；实现归母净利润119.83亿元，同比增长21.23%。单季度来看，25Q2 实现营收599.73亿元，同比增长11.44%；实现归母净利润66.21亿元，同比增长16.71%。 风险提示：政策不达预期、电网投资不及预期、行业竞争加剧致价格超预期下降、资产 减值风险、原材料价格波动风险、新技术不及预期、海外经营风险等。（民生证券 邓永康, 王一如,朱碧野,李孝鹏,李佳,林誉韬,席子屹,赵丹,许浚哲,郝元斌） 民生证券近日发布电新行业2025年半 ...

Core Viewpoint - The implementation opinion released by the National Development and Reform Commission and the National Energy Administration aims to establish a preliminary integration innovation system between energy and artificial intelligence by 2027, enhancing the collaboration between computing power and electricity, and achieving significant breakthroughs in core technologies empowered by AI in the energy sector [1][3]. Group 1: Overall Requirements - The initiative is guided by Xi Jinping's thoughts and aims to deepen the integration of AI with the real economy, focusing on expanding application scenarios in the energy sector and enhancing the technological level of AI innovation [2]. - The goal is to improve the safety, reliability, and efficiency of energy systems, ensuring stable energy supply and supporting green and low-carbon transitions [2]. Group 2: Goals by 2027 - By 2027, the integration innovation system will be established, with significant breakthroughs in AI technologies applied to energy, including the deployment of over five specialized large models in various energy sectors [3]. - The initiative aims to identify over ten replicable and competitive demonstration projects and explore hundreds of typical application scenarios, fostering a batch of AI technology application R&D platforms in the energy sector [3]. Group 3: Application Scenarios - AI applications in the power grid will focus on safety, efficiency, and renewable energy consumption, including intelligent forecasting, diagnostic analysis, and planning [5][7]. - AI will also enhance new energy applications, such as virtual power plants and distributed energy storage, improving load-side optimization and dynamic response capabilities [9][10]. Group 4: Key Technology Supply - The initiative emphasizes addressing technical bottlenecks in the energy sector, including data isolation, fragmented computing power, and algorithm opacity, by promoting common key technology breakthroughs [31]. - It aims to establish high-quality data sets and enhance the integration of computing power and electricity, ensuring a robust foundation for AI applications in energy [31][32]. Group 5: Implementation and Support - Local energy authorities and relevant enterprises are encouraged to establish mechanisms to promote the development of AI in the energy sector, ensuring safety and innovation [33]. - The initiative will support pilot demonstrations of AI applications in energy, encouraging cross-sector collaboration and the establishment of innovation alliances [34][35].

Core Viewpoint - The implementation opinions released by the National Development and Reform Commission and the National Energy Administration aim to promote the integration of artificial intelligence (AI) with the energy sector, targeting significant advancements and applications by 2027 and 2030 [1][2][3]. Group 1: Overall Requirements - The initiative is guided by Xi Jinping's thoughts and aims to enhance the integration of AI with the energy sector, focusing on application scenarios and improving innovation levels in AI technologies [2][3]. - The goal is to ensure energy security, stability, and a green low-carbon transition while fostering new productive forces for the new energy system [2][3]. Group 2: Goals by 2027 - By 2027, a preliminary integration system of energy and AI will be established, with significant breakthroughs in core technologies and broader applications [3][4]. - The plan includes the application of over five specialized large models in various energy sectors, the identification of more than ten replicable and competitive demonstration projects, and the exploration of a hundred typical application scenarios [3][4]. Group 3: Goals by 2030 - By 2030, AI technologies in the energy sector are expected to reach a world-leading level, with improved collaborative mechanisms between computing power and electricity [4]. - The focus will be on achieving breakthroughs in intelligent control of electricity, intelligent exploration of energy resources, and intelligent prediction of new energy [4]. Group 4: Accelerating Application Scenarios - The integration of AI in the power grid will enhance safety, efficiency, and the management of electricity supply and demand [5][6]. - AI will also be applied in new energy businesses, such as virtual power plants and distributed energy storage, to optimize load control and enhance energy efficiency [7][8]. Group 5: Key Technology Supply - The initiative addresses technical bottlenecks in the energy sector, including data isolation and high energy consumption of algorithms, by promoting the development of common key technologies [23][24]. - Emphasis will be placed on building high-quality data sets, enhancing computational support, and improving model capabilities [23][24]. Group 6: Support Measures - The plan includes establishing a robust organizational framework to implement AI in the energy sector, promoting collaborative innovation, and enhancing standardization efforts [25][26]. - Pilot demonstrations will be organized to showcase replicable and scalable AI applications in the energy sector [26].

Core Viewpoint - The implementation opinion aims to promote the integration of artificial intelligence (AI) and the energy sector, establishing a framework for high-quality development by 2027 and achieving world-leading levels by 2030 [1][10][12]. Group 1: Implementation Goals - By 2027, the initial framework for the integration of energy and AI will be established, focusing on the deep application of over five professional large models in various energy sectors such as power grids, generation, coal, and oil and gas [1][12]. - The plan includes identifying over ten replicable and competitive demonstration projects and exploring a hundred typical application scenarios [1][4][12]. - By 2030, the goal is to achieve systematic breakthroughs in AI-specific technologies and applications within the energy sector, enhancing safety, green transformation, and efficiency [5][13]. Group 2: Key Tasks - The implementation opinion outlines several key tasks, including empowering various energy scenarios with AI, focusing on coal, electricity, oil, and gas [6][7]. - It emphasizes the need for a comprehensive approach to AI applications across eight major scenarios, including smart grid, new energy, and nuclear power [7][8]. - A total of 37 key tasks have been identified, with specific applications in oil and gas, coal, electricity, and renewable energy [7][8]. Group 3: Technical Support - The opinion highlights the importance of strengthening the foundational technologies for AI applications in the energy sector, focusing on data, computing power, and algorithms [8][32]. - It calls for the establishment of high-quality data sets and a collaborative development mechanism for computing power and electricity [32][33]. - The need for enhancing model capabilities and addressing issues related to data security and algorithm transparency is also emphasized [32][33]. Group 4: Implementation Measures - The document stresses the importance of organizational implementation, encouraging local energy authorities and enterprises to establish mechanisms for promoting AI in the energy sector [34][35]. - It advocates for collaborative innovation among enterprises, research institutions, and universities to build a robust ecosystem for AI and energy integration [34][35]. - The need for pilot demonstrations and the selection of replicable scenarios for AI applications in the energy sector is also highlighted [35][36].

Core Viewpoint - The National Development and Reform Commission and the National Energy Administration have issued implementation opinions to promote the integration of artificial intelligence (AI) with the energy sector, aiming for world-leading AI technologies and applications in energy by 2030 [1][2][5]. Group 1: Overall Requirements - The initiative is guided by Xi Jinping's thoughts and aims to enhance the integration of AI with the energy sector, focusing on innovative applications and improving the safety and efficiency of energy systems [3][4]. - By 2027, a preliminary innovation system for AI and energy integration will be established, with significant breakthroughs in core technologies and widespread applications [4]. Group 2: Accelerating Energy Application Scenarios - AI will be applied in various energy sectors, including smart grid management, energy resource exploration, and new energy forecasting, enhancing operational efficiency and safety [6][8]. - Specific applications include intelligent power supply forecasting, grid diagnostics, and emergency response systems [7][8]. Group 3: Key Technology Supply - The focus is on overcoming technical bottlenecks in the energy sector, such as data isolation and high energy consumption of computing [26]. - Efforts will be made to build high-quality data sets and enhance computational support for AI applications in energy [26]. Group 4: Safeguard Measures - Local energy authorities and relevant enterprises are encouraged to establish mechanisms to promote the development of AI in the energy sector [28]. - Collaboration between enterprises, research institutions, and universities is emphasized to foster innovation and talent development in AI and energy [30].

Core Viewpoint - The article discusses the implementation opinions on promoting "Artificial Intelligence + Energy" for high-quality development, emphasizing the integration of AI technologies into various energy sectors to enhance efficiency, safety, and sustainability by 2027 and beyond [3][9][10]. Group 1: Overall Requirements - The initiative aims to deepen the integration of AI with the energy sector, focusing on enhancing innovation and application technology levels, and ensuring the safety and reliability of energy systems [10][11]. - By 2027, a preliminary integration system of energy and AI is expected to be established, with significant breakthroughs in core technologies and widespread applications [11][12]. Group 2: Accelerating Energy Application Scenarios - AI will be applied across various energy sectors, including power grids, new energy, and traditional energy sources, to optimize operations and enhance flexibility [4][17]. - Specific applications include virtual power plants, distributed energy storage, and intelligent microgrids, aimed at improving load control and dynamic response capabilities [4][19]. Group 3: Key Technology Supply - The focus is on addressing technical bottlenecks in the energy sector, such as data silos and high energy consumption in computing, by developing common key technologies [40]. - Emphasis is placed on building high-quality data sets, enhancing computational support, and improving model capabilities to meet the specific needs of the energy sector [40][41]. Group 4: Implementation Measures - The article outlines measures for effective implementation, including strengthening organizational frameworks, promoting collaborative innovation, and establishing standards for AI applications in the energy sector [42][43]. - Pilot demonstrations will be organized to showcase replicable and scalable AI applications in energy, encouraging cross-sector collaboration [44].

固态电池先发优势确立，反内卷驱动风光储行情 20250907 摘要 固态电池技术迭代逻辑清晰，预计 2025-2026 年成为重要投资方向， 头部锂电池企业已启动中试线，中期审核及设备招标进展值得关注，推 动近期行情。 选择固态电池投资标的需评估增量环节（如氯化锂、干法电极、金属负 极），关注下游锂电企业反馈，考察主业传统材料供应及盈利修复潜力 （如天赐材料、璞泰来）。 反内卷推动光伏行业供给侧调整，并向储能、新能源车、风电外溢。储 能市场需求良好，供需改善驱动涨价预期，亿纬锂能、欣旺达、宁德时 代等产能接近满产。 8 月国内新能源汽车批发量同比增长 24%，储能系统招标量达 70 吉瓦 时，美国储能需求强劲。供需关系改善，天赐材料 6 伏 03 锂产能利用 率达 80%，有望涨价。 储能市场受益于强劲需求和涨价预期，工商业储能和户用储能领域值得 关注，德业股份业绩优秀且估值较低，印尼混用光储项目具期权空间。 Q&A 当前电新板块的核心观点是什么？ 当前电新板块的核心观点主要集中在固态电池、反内卷和 i 电源三个方向。固 态电池方面，先导智能宣布其在全国固态供应链条中已完全打通，这标志着设 备端的重要进展。此 ...

Core Viewpoint - The energy sector is crucial for economic development and national security, with digitalization and intelligence reshaping the global energy landscape [1] Group 1: New Energy Systems - The integration of digital technology with new power systems is essential for creating a secure, green, and economically viable energy framework in China [1] - Energy activities account for 80% of China's carbon emissions, with electricity generation contributing 40% of that [1] Group 2: Nuclear Energy - Nuclear energy is recognized globally as a near-zero carbon clean energy source, supporting high-energy industries in carbon reduction and contributing to the new energy system [3] - The high energy density of nuclear fuel makes it suitable for various applications, including those in challenging environments [3] Group 3: Smart Grid and Digitalization - The State Grid Corporation emphasizes the importance of digitalization in enhancing the energy transition, implementing various applications to support the new power system [4] - Initiatives include the development of a comprehensive digital grid and AI applications to improve service delivery [4] Group 4: Hydrogen Energy - Hydrogen is viewed as the ultimate energy source of the 21st century, with over 60 countries developing hydrogen strategies, and China's hydrogen industry expected to exceed one trillion by 2025 [6] - China Petrochemical Corporation is actively developing hydrogen infrastructure in Liaoning, including the establishment of hydrogen refueling stations [6] Group 5: Oil and Chemical Industry Transition - The oil and chemical industry in China is transitioning from primary chemical production to fine chemical manufacturing, with fossil fuels remaining dominant for the foreseeable future [8] - The industry aims for high-quality development through technological innovation and green low-carbon pathways during the 14th Five-Year Plan [8] Group 6: Future Trends in Energy Transition - The energy transition in China is expected to showcase multi-energy collaboration, diverse application scenarios, and breakthroughs in disruptive technologies [11] - The discussions at the event highlighted the need for a systematic approach to digital and intelligent transformation in the energy sector [13]

Core Viewpoint - The "vehicle-grid interaction" initiative in Guangzhou allows electric vehicle owners to sell stored energy back to the grid during peak hours, providing economic benefits and alleviating grid pressure [2][3]. Group 1: Vehicle-Grid Interaction - The pilot program enables electric vehicle owners to charge their vehicles during off-peak hours and discharge energy back to the grid during peak hours, resulting in financial gains for users [2]. - The program is particularly beneficial for two groups: commuters who can discharge during peak parking hours and those who primarily use their vehicles on weekends [5]. - The initiative is expected to enhance grid stability by utilizing the stored energy in electric vehicles, which can help in peak shaving and valley filling [5]. Group 2: Battery and Energy Storage Potential - The lithium iron phosphate batteries used by GAC maintain over 80% capacity after 3,000 charge-discharge cycles, indicating untapped potential even at the end of their lifecycle [2]. - With approximately 3 million electric vehicles in Guangdong, the total vehicle-mounted energy storage capacity is estimated to be around 200 million kilowatt-hours [2]. - The average electric vehicle can store enough energy to power ten 1-horsepower air conditioners for six hours, showcasing the significant energy storage potential of electric vehicles [5]. Group 3: Regulatory and Market Developments - The Guangdong Provincial Development and Reform Commission has set a feed-in tariff for electric vehicles discharging to the grid, with peak prices reaching approximately 0.96 yuan per kilowatt-hour during summer months [3]. - The Guangdong grid company plans to establish a "vehicle-grid interaction" alliance, aiming to involve major automakers and operators, with expectations of over 600,000 electric vehicles participating in grid regulation by 2028 [6].

Summary of Conference Call Records Industry Overview - The conference call discusses the **AIDC (Artificial Intelligence Data Center)** and **power equipment** sectors, highlighting the strong growth potential driven by AI applications and cloud service providers' capital expenditures [1][6][9]. Key Points and Arguments AIDC Market Growth - There is a significant increase in demand for AI applications, with domestic enterprise-level large model token consumption rising over **300 times** since last year, reaching **10.2 trillion tokens** daily, accounting for over **one-third** of total consumption [5]. - The global AIDC power equipment market is projected to reach **$29 billion** (approximately **200 billion RMB**) by the end of 2025, with the Chinese market estimated at **44 billion RMB** [1][9]. Cloud Providers' Capital Expenditure - Major cloud providers like **Amazon, Meta, and Google** reported a year-on-year capital expenditure increase of **double to triple digits**, with a **30%** quarter-on-quarter growth in Q2 2025 [6]. - Domestic cloud providers, particularly **Alibaba**, showed a **200%** year-on-year growth in capital expenditure, significantly outpacing **Tencent** and **Baidu** [7]. PSU (Power Supply Unit) Trends - The PSU sector is experiencing a shift towards higher power density and direct current (DC) systems, with **5.5 kW** high-power products expected to become mainstream [12]. - The input voltage for cabinets is anticipated to transition from over **200 volts** to **±400 or 800 volts**, indicating a significant change in PSU design [12]. Investment Recommendations - The AIDC sector is expected to maintain high growth, but some stocks are entering a short-term overvaluation phase. Investors are advised to focus on companies with strong mid-year performance and reasonable valuations [2]. - The power grid sector is characterized by limited growth but is suitable for defensive investment, particularly in low-valuation stocks that align with dividend characteristics [3][4]. HVDC (High Voltage Direct Current) Trends - HVDC pricing is expected to follow a trend of high initial prices followed by declines as production scales up. Current domestic prices remain stable, while overseas prices may initially be higher due to demand for high voltage levels [17]. - The industry is moving towards higher voltage levels, with a shift from **240 volts** to **±400 or 800 volts**, and a centralized power supply model is emerging [15]. Future Market Potential - By 2026, global cloud providers' total capital expenditure could double to over **4 trillion RMB**, indicating substantial market potential. Assuming a **20%** net profit margin, this could yield a profit space of **800 billion RMB** [10]. - The domestic IDC market is currently valued at around **400-500 billion RMB**, with expectations for continued high growth in the coming years [11]. Competitive Landscape - In the power distribution and transformer sectors, domestic manufacturers are gaining an edge due to advancements in DC and solid-state technologies, with companies like **Jinpan** leading the way in prototype demonstrations [19][20]. Overall Market Sentiment - The overall sentiment in the power grid sector is cautious, with limited growth expected in 2025, but it remains a defensive investment choice amid market volatility [23][24]. This summary encapsulates the key insights and projections from the conference call, providing a comprehensive overview of the AIDC and power equipment sectors.