Journal of Hydroelectric Engineering

水力发电学报

Office Online

Included Databases

The journal is included in the following databases:
(i)Scopus,
(ii)The Chinese Science Citation Database Source Journals,
(iii)The Chinese Science Journal Abstract Database,
(iv)The Chinese Science and Technology Papers Statistics and Analysis Database,
(v)China Newspaper Subscription Guide Information Database,
(vi)Chinese Academic Journals (CD).


2026 Vol. 45, No. 6 Published: 2026-06-25

	1	Optimizing short-term operation of pumped storage stations considering power grid residual load forecasting
		CHENG Ming, ZHOU Yanlai, WEI Yilong, YANG Xu
		DOI: 10.11660/slfdxb.20260601
		Large-scale integration of wind power and solar power has led to highly fluctuating residual loads on the power systems, posing a great challenge to the short-term operation of pumped storage stations. This study focuses on a real case of the Heimifeng pumped storage station that connects to the Hunan power grid in Central China. Based on previous studies of residual load forecasting, we develop a short-term model for optimal operation of the station, incorporating multi-period operational constraints. This model considers dual objectives－maximizing power generation benefit and minimizing residual load fluctuations; it is equipped with an improved genetic algorithm for efficient solution, applicable to the cases of complicated constraints－including operational mode transitions and reservoir water balance. Results demonstrate that with the forecasted typical residual load sequences used as input, the optimized operation achieves an average daily power generation benefit of 1.5254 million Yuan, or an increase of 36% relative to the existing real operation, while reducing residual load fluctuations by 34%. This study is practically useful for economical operation of a pumped storage station, and helps enhance grid flexibility under high penetration of renewable energy integration.
		2026 Vol. 45 (6): 1-11 [Abstract] ( 79 ) PDF (2975 KB) ( 67 )

	12	Study on flow behaviors and energy loss mechanisms in S-characteristic zone of pumped storage units
		LI Chunjie, ZHOU Jinpeng, FU Yunheng, CHEN Yong, FU Xiaolong
		DOI: 10.11660/slfdxb.20260602
		To investigate the flow behaviors and energy loss mechanisms of a pumped storage unit operating in the S-characteristic zone, this study focuses on a model pump-turbine with a rated head of 451 m. Unsteady flows across the full flow passage are simulated numerically for the operating conditions in turbine, runaway, and braking modes; Distribution patterns and dominant mechanisms of energy losses are revealed by integrating an analysis of internal flow and the entropy production calculated using an improved near-wall loss function. The results indicate that near-wall entropy production accounts for 21.7% to 26.8% of the total, obviously a significant contribution; the major production, however, comes from the mainstream. The runner section contributes most of the total production, or in the 45% - 51% range in the three typical modes, establishing it as the core region for energy dissipation and hydraulic excitation in S-zone operation. Key flow structures responsible for flow instability and increased energy loss include runner inlet backflows, blade channel vortices, and draft tube vortices. This study lays a basis for future studies of stable S-zone operation and hydraulic optimization of pumped storage units.
		2026 Vol. 45 (6): 12-22 [Abstract] ( 44 ) PDF (5747 KB) ( 46 )

	23	Reinforcement learning approach for speed control of pump turbines
		XIAO Wensheng, HE Jia, ZHAO Zhonggai, LANG Yandong, CHEN Jinbao
		DOI: 10.11660/slfdxb.20260603
		Hydropower serves as a critical renewable energy source that is often used for essential peak-shaving and frequency regulation for power grids; the agility of hydropower units’ speed-governing system in response to load fluctuations directly impacts power quality and grid stability. However, the generating units usually operate across diverse conditions and suffer from severe nonlinearities, posing a huge challenge to the conventional method of proportional-integral-derivative (PID) control. To enhance system control performance and robustness, this paper describes a new intelligent control strategy that is based on the Soft Actor-Critic (SAC) reinforcement learning algorithm. By using the strategy and a nonlinear pump-turbine governing system, first a framework is constructed to train a network of Nonlinear Autoregressive with Exogenous Input Long Short-Term Memory (NARX-LSTM) as a surrogate model. It uses one module for LSTM-based error-correction to raise model accuracy. Then, this error-corrected NARX-LSTM environment is leveraged for iterative training of the SAC agent. Simulation results demonstrate that the new method outperforms traditional PID control in response speed and overshoot suppression across multiple operating points. And, the strategy exhibits superior resilience to operational transitions with minimal fluctuations. This study has verified the efficacy of reinforcement learning in achieving a complicated industrial control, and a promising new paradigm for hydropower speed regulation.
		2026 Vol. 45 (6): 23-36 [Abstract] ( 53 ) PDF (2306 KB) ( 21 )

	37	Intelligent earth-rock dams
		LIU Changyang, MA Rui, LI Qingbin, HU Yu, LIAO Jingxia, HAN Zhi, SUN Zhengdong, ZHANG Fengqiang
		DOI: 10.11660/slfdxb.20260604
		China has stepped up on a new stage of high-quality development in the national water network. As a key node in this network system, earth-rock dams have started an iterative upgrading trend toward an intelligent structure in the water conservancy industry. However, a variety of crucial problems exist in this development stage, such as incomplete theories, scattered technical systems, and lack of evaluation standards. This paper presents a comprehensive study focusing on four aspects－definition and connotation, technical system, evaluation standard, and application cases. First, we clarify the four major functional connotations for an intelligent earth-rock dam－appropriate sensing, realistic analysis, intelligent decision-making, and automatic control. Then, we define the technical composition of an intelligent earth-rock dam, and demonstrate key technologies and how to implement them－digital and intelligent infrastructure, fusion-intelligence model, intelligent decision model, and automatic control system. On this basis, we suggest a grade evaluating standard of four categories and twelve grades for the intelligence degrees of earth-rock dams. Finally, we apply this standard to typical pilot projects－two intelligent earth-rock dams, i.e. the Qingshan reservoir and the Tankeng hydropower station. This verifies the applicability of the standard and the technical system, which would guide and promote the design and construction of intelligent earth-rock dams.
		2026 Vol. 45 (6): 37-51 [Abstract] ( 75 ) PDF (2512 KB) ( 80 )

	52	Research on improved factor model and prediction method for dam deformation in pumped storage power stations
		SONG Jintao, XIE Jinhua, XU Zengguang, QIN Yuan, CHENG Lin, MA Chunhui
		DOI: 10.11660/slfdxb.20260605
		China has planned and built many pumped-storage hydropower stations which are characterized by rapid and large-amplitude water-level fluctuations. Developing dam deformation factor models and prediction methods that can describe these operating characteristics is crucial for dam safety assessment. Based on traditional models, this study develops an improved deformation factor model that is equipped with an additional factor representing the water-level change rate, to address the limitation of conventional factor models caused by lack of this rate factor. To mitigate redundancy in high-dimensional factors, we work out a new hybrid framework that is used to integrate kernel principal component analysis and a deep autoregressive neural network, thereby enhancing predictive performance through jointly applying factor dimensionality reduction and deep-learning-based modeling. Application to a case study of the dam for a pumped-storage station shows this method achieves an accuracy significantly higher than that of the conventional models by adding the water-level change rate factor. And, it reduces the mean absolute error by roughly 36.27% on average relative to the benchmark models, demonstrating a significant improvement in model prediction, as a new approach to dam deformation analysis and safety evaluation.
		2026 Vol. 45 (6): 52-63 [Abstract] ( 41 ) PDF (3357 KB) ( 32 )

	64	Seismic response of earth-rock dams considering varying thickness of reservoir bottom sediment layers
		CHEN Denghong, WU Yu, YUE Meng
		DOI: 10.11660/slfdxb.20260606
		Based on the theory of fluid-structure interaction, this paper develops a dynamic simulation method that accounts for the compressibility of reservoir water and the effect of bottom sediment layers. It models water as an acoustic medium, and uses boundary impedance conditions to simulate the absorption and reflection effects of sediment layers on wave energy. Accordingly, we can transform the sediment layer problem into one-dimensional wave propagation through the foundation depth, and analytically derive the corresponding boundary parameters. Validation shows the method effectively captures the dynamic response of sediment layers with varying thicknesses. In a specific case study, we construct a three-dimensional finite element model for dam–reservoir water–sediment–foundation system of a real asphalt-concrete core rockfill dam, and simulate its dynamic behaviors at different sediment layer thicknesses. The results indicate that sediment thickness has a minor influence on dam displacement, but imposes a significant effect on the dams￠ hydrodynamic pressure, acceleration, and stress responses. Sediment layers produce suppression on the hydrodynamic pressure, and the suppression is most pronounced in the case of the sediment thickness ranges of 0.1H - 0.2H (H is the dam height). This study helps to understand the bottom sediment effect on the dynamic responses of dam–reservoir water–foundation system.
		2026 Vol. 45 (6): 64-76 [Abstract] ( 46 ) PDF (7155 KB) ( 21 )

	77	Variation trends in hourly-scale extreme precipitation in Beijing and its responses to urbanization
		YANG Puxin, LU Yajing, LI Yongkun, YANG Simin, YIN Yue, ZHANG Xumin, LI Xin, LIU Chenyang
		DOI: 10.11660/slfdxb.20260607
		This paper presents a systematical analysis on the trends in the hourly-scale extreme precipitation in Beijing from 2000 to 2023 under the influence of urban heat islands and urbanization, based on the observation data of high-resolution precipitation, surface temperature, and impervious surface coverage. To achieve a comprehensive analysis, we use multiple statistical and spatial analysis methods－including trend testing, change point detection, L-moment analysis, cluster analysis, and space-time cube and emerging hot spot analysis. The results indicate a significant overall increase in extreme precipitation. The urbanizing process underwent a structural shift in 2011, leading to an elevated risk of extreme precipitation. High-value precipitation areas are primarily distributed in the city’s southwestern part, while the regions with strong variability and extreme intensity are concentrated in the transitional zones between the urban core and the western and northern mountainous areas. The spatiotemporal patterns of urban heat island are characterized by a stable core and active periphery, closely related to that of the extreme precipitation. Combination of ongoing urbanization and intensifying heat island effects has intensified hourly-scale extreme precipitation in Beijing, increasing spatial heterogeneity and complexity in associated risks.
		2026 Vol. 45 (6): 77-89 [Abstract] ( 36 ) PDF (4004 KB) ( 49 )

	90	Precipitation nowcasting model based on spatiotemporal characteristics and error correction
		LI Leijing, LI Jianzhu, WU Chaowen, YAN Fengxiang, LI Yuanyuan, WANG Yongtao
		DOI: 10.11660/slfdxb.20260608
		To enhance the performance of precipitation intensity forecasting and spatiotemporal distribution pattern forecasting, this paper describes a deep learning model of error-corrected spatiotemporal Rain-Net (ECST-RainNet). With radar echoes and quantitative precipitation estimation sequences as dual-channel inputs, this new model adopts convolutional neural networks to extract spatial features and a spatiotemporal long-short term memory network to capture temporal characteristics, incorporating an error correcting module to reduce systematic bias. We verify its performance and forecasting of typical precipitation events against the rain gauge measurements from the Liulin experimental watershed in Hebei Province. The results show that in comparison with the rainfall measured at these rain gauges, its quantitative precipitation estimation achieves a correlation coefficient of 0.67 and a root mean square error of 4.77 mm/h, substantially outperforming the dynamic radar reflectivity factor (Z) - surface precipitation intensity (R) relationship method. For one-hour lead-time precipitation forecasts, it generates an estimation error lower than that of the machine learning or deep learning model that uses radar echoes as a single data source. Meanwhile, it improves the forecasting accuracy of the distribution patterns and the precipitation center of cumulative areal rainfalls, showing the performance improved by integrating multiple data sources and its significance in radar precipitation estimation and nowcasting.
		2026 Vol. 45 (6): 90-99 [Abstract] ( 44 ) PDF (3749 KB) ( 36 )

	100	Study on high performance computing of waterlogging process in central urban area of Ordos
		CHEN Tong, SUN Jian, ZHANG Zihao, LIN Binliang
		DOI: 10.11660/slfdxb.20260609
		Waterlogging disasters have occurred frequently in Ordos in recent years, but few previous studies conducted a high-resolution simulation of the waterlogging process in its urban area. Usually, traditional hydrodynamic models struggle to meet such real-time forecasting needs due to large computational demands. This study develops a numerical simulation approach for urban waterlogging that balances simulation accuracy and computational efficiency, aimed at a practical tool efficient and reliable for urban flood prevention and disaster mitigation. We construct a high-performance hydrodynamic model of urban waterlogging in the study area－the central urban area of Kangbashi District, Ordos City－and generate a meter level elevation mesh through multi-source data fusion; Then we simulate and examine its waterlogging process under the design rainstorms of different return periods. The parallel computing performance of this model is tested on the national supercomputing platform. Results show that it achieves waterlogging simulations with a satisfactory accuracy, identifies the waterlogging prone points, and reconstructs the surface runoff process. Its thousand-CPU core calculations are a thousand times faster than serial calculation, so that simulating a waterlogging process of several hours long takes just a few minutes of wall clock time. The findings would help implement the real-time warning and emergency management of urban waterlogging and promote parallel computing of large CPU-cost hydrodynamic models.
		2026 Vol. 45 (6): 100-111 [Abstract] ( 45 ) PDF (7470 KB) ( 24 )

	112	Acoustic fault diagnosis of hydraulic turbines based on fusion of convolutional attention and WGAN-AE
		WANG Yu, HUA Tianlin, SHI Min
		DOI: 10.11660/slfdxb.20260610
		To address the scarcity of labeled data, weak early fault signals, and severe background noise interference in the acoustic diagnosis of hydraulic turbine flow channels, An unsupervised acoustic fault diagnosis method based on the Convolutional Block Attention Module (CBAM) and Generative Adversarial Networks (GAN) was proposed. This method constructs a deep diagnostic model (CBAM-WGAN-AE) that integrates a Wasserstein Generative Adversarial Network (WGAN) with an Autoencoder (AE). It is trained using exclusively acoustic signal data from normal operating conditions to learn the deep feature distribution, and leverages the CBAM to enhance sensitivity to key fault features while suppressing irrelevant noises. Additionally, an anomaly detection mechanism based on the K-Nearest Neighbors (KNN) algorithm was introduced and the turbine’s abnormal states were detected by calculating the nearest neighbor distance deviations of test samples from the corresponding normal samples in the feature space. Through validating against the fault experimental data from a model hydraulic turbine, The new model improves the Area Under Curve (AUC) up to 91.71%, outperforming previous diagnostic models reported in both overall accuracy and the ability to identify weak fault features.
		2026 Vol. 45 (6): 112-124 [Abstract] ( 56 ) PDF (4785 KB) ( 24 )