[Objectives]To enhance the operational flexibility of carbon capture coal-fired power plants and thus support the power grid in integrating renewable energy generation,this study proposes an integrated model of compressed carbon dioxide energy storage and carbon capture coal-fired cogeneration system.[Methods]The study establishes a performance analysis model of a carbon capture coal-fired cogeneration system integrated with compressed carbon dioxide energy storage.The system thermal performance and sensitivity of key parameters are investigated using exergy analysis method.[Results]The research shows that compared with the original car by integrating the compressed carbon dioxide energy storage system,the output power of the carbon capture coal-fired power generation system decreases by 22.59％under the charging condition and increases by 6.19％under the discharging condition.The system power efficiency(SPE)of the compressed carbon dioxide energy storage system reaches 86.32％,the exergy efficiency is 64.82％,and the energy storage density reaches 46.60 kW · h/m3.In addition,the integrated system can effectively recover the waste heat from the carbon capture process to supply heat users,with waste heat recovery rates of 24.9％and 26.1％under the charging and discharging conditions,respectively.[Conclusions]The proposed system enhances the operational flexibility of power generation.Key measures to enhance system performance include improving the expander's isentropic efficiency,reducing the heat transfer terminal temperature difference,and improving heat transfer efficiency.
ObjectivesTo address challenges of short-term electricity price prediction arising from high penetration of renewable energy into the grid, this study proposes a multi-modal, multi-task short-term electricity price prediction model, EP-VLM, based on a vision language model (VLM).MethodsThe model integrates three modules: a patch memory augmenter, a frequency-domain convolutional encoder, and a structured text embedder. These modules process temporal, visual, and textual multi-modal features, respectively, and convert them into tokens that can be embedded into the VLM. By extracting multi-modal fused features, the model can comprehensively capture the complex factors influencing electricity prices.ResultsExperiments based on real data from a provincial electricity market in northern China demonstrate that EP-VLM significantly outperforms existing time-series prediction models across different prediction horizons and exhibits strong few-shot learning capability.ConclusionsThe model effectively improves electricity price prediction precision, validating its accuracy and robustness in ultra-short-term and day-ahead electricity price prediction.
ObjectivesWith the continuous growth of the proportion of renewable energy in the energy structure, the application of electricity-based hydrogen production coupled with fuel cell power generation technology in the electricity-hydrogen-heat integrated energy system has received widespread attention from the industry. These systems are gradually evolving from single independent systems to clustered ones, which makes comprehensive control technology play a crucial role in them. To this end, the research progress of coordinated control technology for electricity-hydrogen coupling system clusters is systematically analyzed.MethodsThree main modeling methods—mechanistic modeling, empirical modeling, and semi-empirical modeling—are comprehensively reviewed in terms of their roles in optimizing the electrolyzer performance. An in-depth analysis is conducted on the response capabilities of three control strategies, namely chain allocation strategy, average allocation strategy, and rotation control strategy, under fluctuating operating conditions. The application potential of integrated energy systems, including electricity-hydrogen coupling systems in fields such as transportation, construction, and industrial parks, is explored, with particular emphasis on their critical role in power production and supply. Finally, an economic analysis is conducted to evaluate the cost-effectiveness of different models in practical applications, in order to provide a theoretical basis and decision support for the optimization design and efficient operation of electricity-hydrogen-heat integrated energy systems.ConclusionsIn the future, the integrated application of artificial intelligence and machine learning technology in the electricity-based hydrogen production system is expected to further improve modeling accuracy, optimize operation strategies, enhance system stability and intelligence level, thereby promoting the development of electricity-based hydrogen production technology towards high efficiency and sustainability.
ObjectivesPhotovoltaic power generation is characterized by randomness, intermittency, and volatility. Its large-scale grid integration may impact power systems, thereby posing serious challenges to grid operation and scheduling. Therefore, developing a high-precision photovoltaic power prediction model and accurately characterizing the fluctuation patterns of photovoltaic power under varying meteorological conditions are of great significance for ensuring the stable operation of the power grid. To this end, this study proposes a photovoltaic power prediction method based on an intelligent optimization algorithm and a combined prediction model.MethodsFirst, a residual network (ResNet)-LSTM-Dropout combined photovoltaic power prediction model is constructed with the long short-term memory network (LSTM) as the core model. In this model, the ResNet is used to extract deep nonlinear features from complex meteorological data, LSTM is employed to learn the temporal variation patterns of photovoltaic power sequences, and the Dropout layer is introduced to reduce the risk of overfitting during model training on complex samples, thereby improving the generalization performance of the model. Second, to address the problems that the rime optimization algorithm (RIME) easily falls into local optima and has slow convergence speed, the cosine strategy is introduced to improve local exploration capability, the role strategy is adopted to enhance global search capability, and the Cauchy mutation strategy is incorporated to avoid premature convergence. Accordingly, an improved rime optimization algorithm (IRIME) is proposed. Finally, IRIME is used to optimize the key hyperparameters of the ResNet-LSTM-Dropout combined model, and the optimized model is applied to short-term photovoltaic power prediction.ResultsThe effectiveness of the proposed model is verified using measured data from a photovoltaic power station in Northwest China. The experimental results show that, under different weather conditions, the IRIME-ResNet-LSTM-Dropout model outperforms other comparison models in prediction performance, with more significant advantages under complex weather conditions such as cloudy and rainy days.ConclusionsThe proposed method effectively improves the accuracy of photovoltaic power prediction, providing important theoretical support for ensuring safe and stable grid operation and optimizing coordinated planning in power systems.
ObjectivesThe multimodal fault signals such as vibration anomalies with nonlinear and multi-scale characteristics may be generated during the operation of hydroturbine units. It is difficult for traditional time-series models to capture long-range dependencies, and the single-modal analysis method cannot integrate heterogeneous data features effectively. In response to the above issues, this study proposes a multimodal heterogeneous graph hybrid feature extraction model consisting of Transformer-Gramian angular summation fields (GASF)-recurrence plot (RP)-two-dimensional (2D)-gated recurrent unit (GRU), aiming to improve the reliability and generalization ability of fault diagnosis.MethodsFirstly, the time-series data of hydroturbine units are converted into 2D images, GASF and RP methods are used to extract spatial features of time-series data, and Transformer model is constructed. Meanwhile, GRU is used to capture dynamic time-series features, and multi-modal feature fusion is used to combine temporal features, image spatial features, and heterogeneous image features. Thus, the accuracy and robustness of fault identification are significantly improved.ResultsThe proposed method shows higher accuracy and stronger generalization ability in the fault diagnosis task of hydroturbine units, and the diagnosis accuracy reaches 100% on multiple test sets.ConclusionsThe proposed method can effectively fuse time-series data and image features, significantly enhance the model's ability to recognize nonlinear fault modes, and accurately capture the abnormal state of the device.
