Issue Ebookcover storyIssue CoverIssue Contents
List of IssuesDocument Types
- Abstract：ObjectivesThe construction and application of new-type power systems driven by China’s “dual carbon” goals have demonstrated unprecedented complex dynamics characteristics, featuring a high-proportion integration of renewable energy and power electronic devices, posing new challenges to the safe and stable operation of power systems. Intelligent computing in power systems leverages new-generation artificial intelligence technologies—especially large models and pre-training techniques—to achieve the integration of physical models, data-driven models, and knowledge models, thereby developing a novel calculation and analysis method for power systems. This study reviews and discusses the applications of intelligent computing methods and technologies in the analysis, optimization, operation, scheduling, and control of new-type power systems.MethodsFirst, the concept and key technologies of intelligent computing in power systems are introduced. Technical roadmaps are discussed through case studies on intelligent processing in key areas such as time-series prediction, security region optimization, and multi-energy microgrid collaborative optimization. Finally, the prospects of potential applications are explored.ConclusionsIntelligent algorithms outperform traditional methods across multiple evaluation indicators, significantly enhancing prediction accuracy and control efficiency. The results validate the potential of intelligent computing in calculation and analysis for power systems, providing an effective intelligent pathway for new-type power systems and holding substantial significance for the development of intelligent power systems.Keywords：intelligent computing;new-type power system;renewable energy;large model;artificial intelligence (AI);power grid operation and control;optimal dispatch;intelligent evaluation252|5426|3<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
- Abstract：ObjectivesWith the large-scale integration of renewable energy, new-type power systems require greater flexibility and higher prediction accuracy for prediction technology. Traditional prediction methods have limitations in handling dynamic and complex scenarios, highlighting the need for prediction technologies tailored to these systems. Large language models (LLMs), as generative artificial intelligence technologies, have capabilities in multimodal data integrating, few-shot learning, and multitask handling, enabling more intelligent and precise solutions for the prediction of power systems. Therefore, this study focuses on analyzing the current applications and advantages of LLMs in power system prediction.MethodsFirst, the fundamental architecture, training methods, and current application status of LLMs are discussed. Then, their principles and implementations in prediction are explained, with emphasis on advantages and prospects in load prediction, renewable generation prediction, and electricity price prediction. Finally, challenges in LLM-based prediction applications are analyzed from three aspects: data quality management, privacy protection, and computational resources, and feasible solutions are proposed.ConclusionsThrough comparative analysis of various forecasting tasks, LLMs demonstrate superior capabilities in few-shot learning and multimodal data processing compared to traditional methods, making them more adaptable to complex and variable prediction scenarios. Effective application of LLMs can provide innovative solutions for electricity market prediction.Keywords：large language model (LLM);artificial intelligence (AI);new-type power system;load prediction;electricity price prediction;renewable energy output prediction;privacy protection;data training1353|1015|4<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
- Abstract：ObjectivesTo address the challenges faced by new types of nuclear power systems in ensuring energy supply, promoting clean energy transition, and achieving “dual carbon” goals, such as structural diversification, load uncertainty, and data complexity, and to solve problems associated with the application of large language models in the nuclear power domain, such as knowledge limitations, hallucinations, and high reasoning costs, the study explores the application potential of combining large language models with knowledge graphs, especially graph retrieval-augmented generation (GRAG) technology. This combination aims to build more intelligent, efficient, and reliable information processing capabilities for nuclear power systems.MethodsThe research status of large language models and knowledge graphs in the nuclear power domain, their respective advantages and disadvantages, and the complementarity of their integration are analyzed. The advantages of GRAG technology over traditional retrieval-augmented generation (RAG) are highlighted, along with its specific applications in scenarios such as nuclear power risk assessment, intelligent question-answering assistance, knowledge management and decision support, and fault diagnosis and prediction. Furthermore, the technical pathway for introducing and fine-tuning large language models, constructing domain-specific knowledge graphs, and implementing GRAG enhancement is outlined. Finally, an outlook on future research is provided, covering areas such as knowledge graph construction under heterogeneous data, cognitive reasoning and decision-making of large language models, and the controllability of human-computer interaction.ConclusionsGRAG technology combined with knowledge graphs can effectively alleviate the knowledge limitations and hallucination problems of large language models in specialized domains, enhancing the interpretability and reliability of the generated content. The research findings can provide references for the future optimization of knowledge graph construction in the nuclear power domain, enhancing the capabilities of large language models in complex reasoning tasks, and developing artificial intelligence agents for efficient interaction with experts in the nuclear power field.Keywords：dual carbon;nuclear power;smart grid;artificial intelligence (AI);deep learning;large language model;knowledge graph;retrieval-augmented generation259|389|3<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
- Abstract：ObjectivesWith the rising penetration of renewable energy, it is imperative to conduct power and energy balance analysis, design planning schemes, and evaluate market mechanisms through detailed time-series operational simulation. However, due to the randomness and volatility of renewable energy and the continuous expansion of power grid, operational simulation faces challenges in balancing computational speed and accuracy. Artificial intelligence (AI), with its exceptional abilities in representation, generalization, and self-learning, provides new solutions to the current challenges. Therefore, the current status and necessity of the application of AI in accelerating the operational simulation of power systems are systematically analyzed, and the future development is prospected.MethodsFirst, from a mathematical perspective, the concepts of acceleration methods are classified into scenario aggregation, unit aggregation, constraint reduction, and algorithm acceleration. Second, the necessity of applying AI across different aspects is thoroughly analyzed, addressing the crucial question of “why AI is needed”. Then, current AI applications in accelerating operational simulation and their advantages are systematically summarized. Finally, recommended AI application scenarios in power systems and future prospects are presented.ConclusionsAI can provide effective support for accelerating operational simulation from multiple aspects. It shows particular strengths in handling non-linear correlations, substituting expert experience, and characterizing fuzzy rules.Keywords：new-type power system;artificial intelligence (AI);operational simulation;unit aggregation;scenario clustering;redundant constraint deletion;algorithm acceleration;machine learning212|1028|2<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
- Abstract：ObjectivesAs an essential sustainable energy technology, renewable energy-powered water electrolysis for hydrogen production has attracted widespread attention due to its advantages in environmental protection and low carbon emissions. However, conventional water electrolysis technologies for hydrogen production face challenges in terms of efficiency and cost, the rapid development of artificial intelligence (AI) provides an effective way to solve the difficult problems of hydrogen production technology through electrolysis of water. To address this, this study aims to explore the key applications and development prospects of AI for optimizing the efficiency and economic performance of water electrolysis systems for hydrogen production.MethodsCommon AI tools such as MATLAB, Python, and SimuNPS are employed for algorithm development, deep learning model training, and multi-physics simulation in water electrolysis systems for hydrogen production. By integrating AI technologies, applications such as output prediction, system capacity optimization and scheduling, and fault diagnosis are implemented to improve system performance and stability. A comparative analysis of performance of different AI models in various real-world scenarios is conducted to explore their specific roles and implementation methods in enhancing system performance and controllability.ConclusionsAI technology offers new avenues for enhancing the efficiency and intelligent scheduling of renewable energy-powered water electrolysis hydrogen production systems. Future research should focus on the application of AI in output forecasting, scheduling optimization, and fault diagnosis, promoting deep integration between AI and system operation. Moreover, innovative applications of AI in intelligent monitoring, automatic control, and multi-source coordination should be explored to provide strong support for the development of efficient, stable, and low-carbon hydrogen energy systems.Keywords：renewable energy;water electrolysis for hydrogen production;artificial intelligence(AI);deep learning;alkaline electrolyzer;proton exchange membrane electrolyzer;fault diagnosis256|286|2<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
Research on Large-Signal Stability Assessment Methods of DC Microgrids Based on Artificial Intelligence AI Introduction
Abstract：ObjectivesDC microgrids are prone to the issue of large-signal stability due to the low inertia and constant power load characteristics. Traditional model-based methods involve complex calculations and are difficult to solve. To address these issues, this study investigates intelligent analysis methods for large-signal stability of DC microgrids.MethodsCommon artificial intelligence (AI) classifiers are selected to analyze the stability of DC microgrids. A comparative analysis is conducted on three types of common AI technologies (covering six methods)-deep learning, support vector machine, and decision trees-for large-signal stability assessment in a specific DC microgrid case study.ResultsComparative analysis based on specific examples shows that in the large-signal stability assessment of DC microgrids, long short-term memory (LSTM) networks outperform other methods in terms of overall performance (accuracy, real-time capability, and adaptability).ConclusionsThe LSTM network classifier shows high compatibility with the state-space equations of DC microgrids, making it more suitable than traditional machine learning classifiers for large-signal stability analysis of DC microgrids. Meanwhile, ensuring the performance of the classifier requires appropriate selection of parameter values.Keywords：DC microgrids;distributed energy resources;renewable energy;energy storage;artificial intelligence(AI);machine learning;deep learning;long short-term memory network102|538|3<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29Review of Application on Optimization Strategies for New-Type Power System Based on Reinforcement Learning AI Introduction
Abstract：ObjectivesAs power systems evolve toward higher levels of intelligence and automation, reinforcement learning (RL), a key technology in artificial intelligence, shows great potential in the intelligent development of the power sector. Enhancing research methods for RL applications is crucial for fully exploring its potential in power system operation, control, and optimization. Therefore, the performance of RL in practical electrical applications is analyzed, and the possible research directions in the future are prospected, so as to provide assistance for the intelligent transformation of power systems.MethodsThis study provides a systematic review of RL applications across diverse fields of electrical engineering. It systematically introduces the fundamental principles and landmark algorithms of RL, detailing how these algorithms are applied to address practical problems in new-type power system. The study categorizes mainstream RL algorithms in current research and analyzes the advantages and disadvantages of structural improvements made to these algorithms.ResultsCompared to traditional algorithms, RL significantly enhances the intelligence level of new-type power system. It achieves remarkable success in various application scenarios, particularly in addressing system complexity and uncertainty. However, despite many successful cases, several urgent issues still exist in this sector, such as high computational costs, long training times, and limited generalization abilities.ConclusionsReinforcement learning provides novel solutions for the intelligent development of new-type power system. However, achieving large-scale application still needs to overcome a series of technical and practical challenges. This study provides references and insights for researchers and practitioners in electrical engineering.Keywords：new-type power system;reinforcement learning (RL);deep reinforcement learning (DRL);intelligent grid;strategy optimization;energy management;situational awareness;optimized scheduling;artificial intelligence (AI)393|4009|10<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29Location Detection of False Data Injection Attacks in Power Grid Based on Improved Deep Extreme Learning Machine AI Introduction
Abstract：ObjectivesPower systems are facing threats of false data injection attacks. Existing detection methods for false data injection attacks have the problems of insufficient feature learning ability and slow detection speed, making it difficult to locate false data injection attacks rapidly and accurately. Therefore, this study proposes a method for locating false data injection attacks in power grid based on deep extreme learning machine optimized by Levy flying sparrow search algorithm.MethodsThe proposed method uses a deep extreme learning machine as the feature extraction algorithm and basic classifier to achieve rapid and accurate attack location. At the same time, a Levy flying sparrow search algorithm with strong local search ability is employed to optimize the initial weight and bias to further improve the location detection accuracy of the method.ResultsExtensive simulation analyses are conducted on IEEE-14 and IEEE-57 bus power systems. The proposed method achieves a detection accuracy rate of over 94%.ConclusionsCompared with other detection methods, the proposed method demonstrates better detection accuracy and enables faster location detection of false data injection attacks.Keywords：power grid;artifical intelligence(AI);power system;false data injection attack;deep extreme learning machine;Levy flying;sparrow search algorithm;location detection;feature extraction49|450|4<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29- Abstract：ObjectivesTraditional power grid inspection methods suffer from high labor intensity and low efficiency. Taking Shandong Golden Power Grid as the research object, this study proposes an inspection algorithm using unmanned aerial vehicle (UAV) based on lightweight deep learning network YOLOv5-Mv3 for detecting grid insulators and foreign objects.MethodsFirstly, a dataset is constructed using images captured by UAVs during power grid inspection and is trained. Then, for the grid insulators and foreign objects, Mobilenetv3 is used to replace CSPDarknet53 as the feature extraction network in order to lighten the YOLOv5-Mv3 model, reducing parameters and computational cost while maintaining accuracy and enabling real-time detection.ResultsThe proposed detection algorithm achieves a mean Average Precision of 84.7% and 56.6 frames per second. Compared to Faster RCNN, SSD, and YOLOv4 models, the improved YOLOv5-Mv3 demonstrates higher detection accuracy and faster performance.ConclusionsThe proposed algorithm improves the efficiency of UAV-based power grid inspection and achieves lightweight and high-efficiency effect, fully meeting the requirements for intelligent power grid inspection.Keywords：power system;artifical intelligence (AI);power grid;unmanned aerial vehicle;target detection;transmission line;power inspection;machine vision;image processing166|2079|6<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
State Estimation and Fault Diagnosis of Proton Exchange Membrane Fuel Cells Based on Artificial Intelligence AI Introduction
Abstract：ObjectivesThe proton exchange membrane fuel cell (PEMFC), as a highly promising clean energy technology, has attracted much attention in the field of energy conversion. However, the high complexity and operational uncertainties of PEMFC systems pose significant challenges to state estimation and fault diagnosis, seriously affecting system reliability and safety. To effectively address these challenges, the application strategies and effectiveness of artificial intelligence (AI) technology in PEMFC state estimation and fault diagnosis are studied.MethodsCurrent research progress on PEMFC state estimation and fault diagnosis is analyzed. In the field of state estimation, the nonlinear model characteristics of PEMFC are analyzed, AI-based state estimation technologies are introduced, and the application principles and advantages of different algorithms for PEMFC state estimation are analyzed. In the field of fault diagnosis, common fault types of PEMFC are summarized, their fault manifestations and internal causes are analyzed, and AI-based fault diagnosis technologies are introduced. Finally, the future prospects for AI-based PEMFC state estimation and fault diagnosis technologies are discussed.ConclusionsWith its powerful data processing and pattern recognition capabilities, AI technology can accurately estimate the state of PEMFC and effectively diagnose potential system faults, thereby significantly improving the the operational efficiency and stability of PEMFC systems and enhancing their reliability and safety. Future research can focus on areas such as AI algorithm innovation, optimization of state estimation and fault diagnosis, intelligent system development, and collaboration with other technologies.Keywords：clean energy;hydrogen energy;artificial intelligence (AI);proton exchange membrane fuel cell (PEMFC);state estimation;fault diagnosis;deep learning142|2194|1<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
Application of AI in New Power System
- Abstract：ObjectivesAs clean energy sources and chemical raw materials, green hydrogen, green ammonia, and green methanol play an important role in achieving the “dual carbon” goal. With the increasing demand for clean energy, wind-solar-hydrogen-ammonia-methanol integrated industry has emerged as an important new field of interest. However, the industry in China is still in its early stages of development, with numerous challenges in various links of the industry chain. Therefore, it is necessary to explore the technologies related to wind-solar-hydrogen-ammonia-methanol integration and to analyze the future direction of its development.MethodsFirstly, the technological routes of producing green hydrogen, synthesizing green ammonia, and synthesizing green methanol through water electrolysis are introduced, along with an overview of technologies addressing the instability of wind and solar power. Next, the economy of the wind-solar-hydrogen-ammonia-methanol integrated industry are analyzed, highlighting the implementation of key projects and related policies in recent years. Additionally, the electricity and water consumption for producing green ammonia and green methanol are analyzed and estimated, and the production costs of green hydrogen, green ammonia, and green methanol are predicted. Based on this, suggestions for industry development are proposed, including technological innovation, industry chain coordination, and policy support. Finally, the future development directions of the wind-solar-hydrogen-ammonia-methanol integrated industry are analyzed.ConclusionsIn the future, the integrated development of wind-solar-hydrogen-ammonia-methanol will show a trend of technological integration, diversified applications, regional coordination, and cost-effectiveness, making it one of the core pathways to achieve the “dual carbon” goals and transform the energy system.Keywords：dual carbon;wind power;photovoltaic power generation;green hydrogen production;green ammonia;green methanol;zero carbon emissions;industry chain coordination522|2906|1<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
- Abstract：ObjectivesArundo donax L., as a high-yield energy crop suitable for intensive cultivation, shows great potential for large-scale replacement of coal in power generation and CO2 emission reduction. Under China’s “dual-carbon” goals and the requirements of low-carbon transformation for coal-fired power units, vigorously developing clean and efficient combustion technologies using Arundo donax L. as an energy crop offers an important low-carbon energy pathway well-suited to China’s national conditions. Understanding the current research status of Arundo donax L. combustion is crucial for guiding the utilization of its combustion in China. To this end, the research progress on the combustion and utilization of Arundo arundo is reviewed, aiming to provide reference for the large-scale utilization of biomass energy.MethodsThe study conducts a comprehensive investigation into the utilization of Arundo donax L. combustion through literature research methodology, systematically summarizing and analyzing existing research findings. It summarizes the fuel characteristics of Arundo donax L., pre-treatment methods and performance, the characteristics of combustion and pollutant emissions, and industrial-scale tests of Arundo donax L. combustion. Additionally, the study discusses the environmental effects of its cultivation.ConclusionsArundo donax L. shows significant potential for large-scale replacement of coal in heating and power generation. Given the requirements of low-carbon transformation for coal-fired power units in China, co-firing Arundo donax L. in power plants shows great potential.Keywords：dual-carbon;low-carbon;biomass;renewable energy;Arundo donax L.;combustion;emission reduction;coal-fired power generation234|581|2<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
Research on Shadow Mitigation Strategy for Sloped Photovoltaic Arrays Based on Environmental Parameter Model AI Introduction
Abstract：ObjectivesThe anti-shadow strategy plays a vital role in the large-scale application of photovoltaic systems. In order to cope with the problem of power generation loss caused by shadow shading between photovoltaic arrays in areas with low slope, it is necessary to optimize the daily tracking strategy of photovoltaic modules.MethodsTaking a photovoltaic power station in Ningxia as an example, the dynamic shadow shading calculation model and irradiance model were established based on terrain and environmental factors. An anti-shadow flat single-axis solar tracking strategy was proposed. The feasibility and effectiveness of the proposed solution were tested using a simulation model constructed by MATLAB software. Finally,a typical sunny day was selected for field comparison and verification in the selected area.ResultsUnder low-slope terrain conditions, compared with the photovoltaic array using the original tracking strategy, the photovoltaic array with the proposed anti-shadow and automatic tracking strategy can reduce shadow occlusion by about 12%. Additionally,the overall electricity output efficiency of the system can be improved by about 10%.ConclusionsThe proposed strategy is optimized from the perspective of terrain and environmental factors, which effectively solves the applicability problem of the anti-shadow tracking system, broadens the location area of photovoltaic power station, and provides reliable technical support for it, which is of great significance in the large-scale application of photovoltaic systems.Keywords：solar energy;photovoltaic power station;gently sloping terrain;shadow shading;tracking strategy;flat single-axis systems;power generation benefits63|207|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29Modeling Simulation and Experimental Verification of Focal Length Optimization in Linear Fresnel Collector AI Introduction
Abstract：ObjectivesIn order to improve the concentrated light interception rate of linear Fresnel collector (LFC), increase the error control threshold of the tracking system, and enhance the system’s photothermal conversion efficiency, it is necessary to study the variation patterns of light spot width on the primary reflector. Therefore, the reflected light spot width of the LFC micro-arc mirror is modeled, simulated and experimentally studied.MethodsBased on the structural characteristics of the LFC, a light concentrating model of the micro-arc primary reflector is developed. Through MATLAB simulation analysis, the variation patterns of light spot width under the boundary conditions of the micro-arc primary reflector are obtained. A light spot test platform is developed, which simulates solar incidence angles of different seasons and times of the day by changing the angle of the test platform and adjusting the position of the receiver board, to verify the accuracy and effectiveness of the simulation results.ResultsDue to the presence of the solar angle, the reflected light spot width increases by 75.8% compared to the parallel light condition. The experimental results of rotating test platform show that the variation patterns of the measured light spot width are consistent with the simulation results. Under the influence of the solar angle, the maximum tracking error control margin of the LFC is 0.173° in winter and 0.207° in summer.ConclusionsThe tracking control accuracy of the LFC system in winter should be 16.4% higher than that in summer. Therefore, the focal length of the LFC micro-arc reflector should be designed based on the tracking control margin of the light spot width in winter.Keywords：solar energy;photothermal power generation;heat collection system;linear Fresnel;concentrating heat collection;reflected light spot width;concentrated light interception rate;micro-arc mirror80|119|0<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
New Energy
Study on Dredging Strategy and Electricity Price Calculation of Pumped Storage Power Station in Xinjiang AI Introduction
Abstract：ObjectivesUnder the current electricity price mechanism, it is difficult for Xinjiang pumped storage power station to fully recover the cost through the arbitrage of peak-valley price difference and participation in electric auxiliary services. Therefore, the capacity and electricity charge dredging mechanism of pumped storage power station is studied.MethodsBased on the calculation of the electricity cost of Xinjiang pumped storage capacity, the problems existing in the cost dredging are analyzed. The dredging path of “market-oriented + transmission and distribution electricity price” is proposed for Xinjiang pumped storage power station. That is, the cost is recovered by transmission and distribution price before the maturity of power market construction, and the cost is recovered by marketization after the maturity of power market construction.ResultsCompared with the pure transmission and distribution electricity price recovery, the “market-oriented + transmission and distribution electricity price” method has higher feasibility in dredging capacity electricity price and the increase of transmission and distribution electricity price decrease by more than 90%, which is a feasible model to calm the excessive rise of transmission and distribution electricity price.ConclusionsAs a supplementary scheme for the operation cost recovery of pumped storage power stations, the research results will provide a broad market trading space for the pumped storage power stations.Keywords：pumping and storage power station;cost recovery;cost management;power market;electricity price measurement190|172|1<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29Investment Decision and Economic Analysis of Distributed Photovoltaic Power Stations AI Introduction
Abstract：ObjectivesWith the continuous advancement of China’s photovoltaic (PV) industry, the industry is transitioning from being heavily reliant on government policy subsidies to achieving grid parity, which places higher demands on the economic benefits of PV power stations. Consequently, a more comprehensive and detailed study on the investment decision and economic evaluation of photovoltaic power stations is conducted.MethodsInvestment decision guidance for a 4.45 MW distributed PV power station project is provided from multiple aspects, including site solar resource analysis, solar module selection, and inverter selection. By calculating the construction investment and revenue, economic analysis of the project is conducted, and the effects of construction investment, operating costs, power generation price, and power generation output on the project’s economics are studied.ResultsThe economic analysis reveals that the project achieves an after-tax financial internal rate of return (FIRR) of 7.43%, with an after-tax investment payback period of 11.39 a. For equity capital, the FIRR is 11.43%, and the payback period is 10.38 a.ConclusionsThe project offers favorable economic benefits and the project economics is most sensitive to changes in electricity prices. Furthermore, the project demonstrates strong resilience to risks. The research results can provide theoretical reference and engineering guidance for the research and application of PV projects.Keywords：photovoltaic power generation;distributed photovoltaic;cost reduction and efficiency improvement;investment cost;investment decision;economic analysis;sensitivity analysis;internal rate of return415|1156|4<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
New Power System
- Abstract：ObjectivesThe supercritical CO2 Brayton cycle has the advantages of high efficiency, compact structure, enormous power generation potential, and strong scalability. CO2-based mixtures as working fluids can change the critical point and improve cycle performance, which has been the focus in recent years. It is of great significance to review its research progress for basic theoretical research and engineering application.MethodsA review is conducted on the research progress on the application of CO2-based mixtures in the supercritical Brayton cycle. The types of commonly explored mixed working fluids are summarized. The research on cycle components and structure layout is discussed, and the commonly used cycle performance and operating conditions are analyzed.ConclusionsThe application of CO2-based mixtures can increase or decrease the critical temperature of CO2, but the predicted physical properties range is limited, and experimental data is lacking. The recompression Brayton cycle is a commonly investigated cycle layout, and thermodynamic performance and design conditions are widely focused. It is recommended to strengthen and deepen the theoretical and experimental research on the thermodynamic properties of mixed working fluids, conduct reasonable verification on the compatibility of component materials, and propose the novel cycle structure layout for the different thermodynamic properties of mixed working fluids. Further exploration is needed for the analysis of comprehensive performance and dynamic characteristics of the CO2 mixtures supercritical Brayton cycle.Keywords：supercritical;Brayton cycle;carbon dioxide;mixed fluids;thermodynamic property;cycle structure;cycle performance574|1163|1<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
NOx Prediction for Deep Peaking Regulation of Circulating Fluidized Bed Units Based on Data-Driven AI Introduction
Abstract：ObjectivesUnder the deep peaking regulation operation of circulating fluidized bed (CFB) unit, NOx emission will fluctuate significantly, which will have a serious impact on the stability and economy of the denitrification control system. Therefore, a data-driven NOx prediction model based on the deep peaking regulation of CFB unit is proposed.MethodsThe generation mechanism and influencing factors of NOx are analyzed in depth. Combined with the field operation data of a 300 MW CFB unit, the whale optimization algorithm (WOA) is used to optimize the parameters of the long short-term memory (LSTM) neural network, and the WOA-LSTM neural network model is established.ResultsThe model realizes the online prediction of NOx emission mass concentration of CFB unit under deep peaking regulation. Compared with other neural network models, the proposed model has better prediction results. The mean absolute error of the model is 2.49 mg/m3, the mean absolute percentage error reaches 6.8%, and the model correlation coefficient reaches 0.99.ConclusionsThe model can accurately reflect the NOx emission trend under the wide range of variable load operation of the unit. The results of the study are a good guide for the operation of the automatic control system of denitrification in the field.Keywords：coal-fired power unit;circulating fluidized bed (CFB);NOx emission;deep peaking regulation;artificial intelligence (AI);whale optimization algorithm (WOA);long short-term memory (LSTM) neural network135|42|3<HTML><L-PDF> <WORD> <Meta-XML>Updated：2026-06-29
Power Generation and Environmental Protection
