Article
Geosciences, Multidisciplinary
Yuxue Guo, Xinting Yu, Yue-Ping Xu, Hao Chen, Haiting Gu, Jingkai Xie
Summary: This study developed an AI-based management methodology integrating multi-step streamflow forecasts and multi-objective reservoir operation optimization for water resource allocation. The study found that both GRU and LSTM performed equally well on streamflow forecasts, with GRU potentially being the preferred method due to its simpler structure and less modeling time. Higher forecast performance could lead to improved reservoir operation, while uncertain forecasts were more valuable than deterministic forecasts.
HYDROLOGY AND EARTH SYSTEM SCIENCES
(2021)
Article
Environmental Sciences
Yang Song, Chunqi Shen, Ying Wang
Summary: By coupling the non-dominated sorting genetic algorithm-II (NSGA-II) model and the General Lake Model-Aquatic EcoDynamics library (GLM-AED) model, reservoir operation strategies (ROSs) can effectively control algal blooms. In the case of Zipingpu Reservoir, the peak of outflow discharge can be reduced by 19%, total power generation can be increased by 8%, and the peak of chlorophyll a concentration can be decreased by 36% compared to the original reservoir operation. Balancing the objectives of algal bloom control, flood prevention, and power generation is crucial in reservoir operations.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2023)
Article
Engineering, Civil
Chengxin Luo, Wei Ding, Chi Zhang, Xuan Yang
Summary: To effectively mitigate droughts, multiple hydrological forecasts are needed. This study proposes a novel Model Predictive Control (MPC) that integrates streamflow forecast, regime state forecast, and annual streamflow volume state forecast. By incorporating these forecasts, significant performance gains can be achieved in drought mitigation. However, forecast value is influenced by forecast uncertainty and other factors.
JOURNAL OF HYDROLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Chunliang Zhao, Yuanyuan Hao, Dunwei Gong, Junwei Du, Shujun Zhang, Zhong Li
Summary: This paper presents a general method incorporating transfer learning for multi-scenario multi-objective optimization problems (MSMOPs). It develops a multi-scenario ensemble framework that transfers knowledge between scenarios to combine arbitrary multi-objective evolutionary algorithms. An adaptive decomposition-based multi-objective evolutionary algorithm with bi-layer selection (EADaBS) is proposed and embedded within the framework as a base learner. Experimental results demonstrate the effectiveness and efficiency of the proposed algorithms, outperforming existing state-of-the-art algorithms.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Economics
Tobias Fissler, Yannick Hoga
Summary: Systemic risk measures like CoVaR, CoES, and MES are widely used in finance, macroeconomics, and regulatory bodies, yet they lack the ability to be observed and identified, making forecast comparison and validation impossible. To solve this problem, we introduce the concept of multi-objective elicitability using bivariate scores with a lexicographic order. We propose Diebold-Mariano type tests with suitable bivariate scores to compare systemic risk forecasts and illustrate the test decisions using an easy-to-apply traffic-light approach. Applying this approach to DAX 30 and S&P 500 returns, we provide recommendations for regulators.
JOURNAL OF BUSINESS & ECONOMIC STATISTICS
(2023)
Article
Mathematics, Applied
Parnian Hosseini, Nathan L. Gibson, Duan Chen, Arturo S. Leon
Summary: This study explores optimal control under input uncertainties that cannot be accurately quantified, proposing a framework that incorporates decision flexibility as an additional objective to balance between flexibility and achieving operational objectives, utilizing uncertainty quantification techniques to compute expected values of objectives. A clear trade-off is identified between the amount of flexibility for decision variables and the expected values of other objectives.
INTERNATIONAL JOURNAL OF COMPUTER MATHEMATICS
(2021)
Article
Environmental Sciences
Jinshu Li, Wei Zhang, William W-G Yeh
Summary: This study presents a multi-objective, multi-stage stochastic programming model for reservoir management and operation, utilizing utility theory to select the best compromise solution. The model successfully produces the optimal water release policy under different hydrological scenarios, considering both inflow uncertainty and the tradeoff between conflicting objectives.
WATER RESOURCES RESEARCH
(2021)
Article
Engineering, Civil
Jiqing Li, Jing Huang, Pengteng Liang, Jay R. Lund
Summary: Through fuzzy representation and multi-objective optimization model, the study achieved a reasonable determination of environmental flow, which can guide reservoir discharge; Environmental flow as an optimization objective rather than a constraint is conducive to formulating environmentally friendly reservoir operation schemes; Multi-objective risk analysis can avoid the one-sidedness of single-objective risk analysis and provide more basis for reservoir management.
WATER RESOURCES MANAGEMENT
(2021)
Article
Environmental Sciences
Icen Yoosefdoost, Milad Basirifard, Jose Alvarez-Garcia
Summary: This study examined three different policies to improve the performance of reservoirs, finding that the Hedging Rule (HR) significantly outperformed the Standard Operation Policy (SOP) in enhancing reservoir performance indexes, and the Multi-Objective Optimization method (MOO) also yielded significant improvements in reservoir performance indexes.
Article
Computer Science, Artificial Intelligence
Dong Liu, Tao Bai, Mingjiang Deng, Qiang Huang, Xiaoting Wei, Jin Liu
Summary: This study proposes a novel parallel approximate evaluation-based model (PAEM) to enhance the search ability and shorten the calculation time in reservoir operation optimization, while accurately controlling approximation errors. By combining PAEM with LSTM, the PAEM-LSTM model is developed for fast formulation of operating rules. Results show that PAEM provides significantly better Pareto-optimal solutions at a faster speed compared to other algorithms, while maintaining extremely low approximation errors. PAEM recommends a small population size and large mutation size, and its efficiency increases with the scale of the reservoir group. The PAEM-LSTM model can increase hydropower generation and reduce ecological water shortage compared to conventional operating rules.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Water Resources
Thatkiat Meema, Yasuto Tachikawa, Yutaka Ichikawa, Kazuaki Yorozu
Summary: This study focused on forecasting river flows and optimizing dam release in the Nan River Basin in Thailand using a distributed hydrological model with ensemble weather forecasting. The research found that utilizing ensemble forecasts with dynamic programming led to more efficient real-time reservoir optimization compared to using historical data. The results provide valuable insights for water resource management in the region and can potentially be applied to other basins as well.
JOURNAL OF HYDROLOGY-REGIONAL STUDIES
(2021)
Article
Environmental Sciences
Yanjun Kong, Yadong Mei, Xianxun Wang, Yue Ben
Summary: This study introduces a clustering-based decision-making method for multi-objective reservoir operation optimization. By innovatively clustering reservoir operation processes and analyzing trade-off surfaces, a compromise solution is selected. The method was applied to the Three Gorges cascade reservoirs system and proved to effectively distinguish operation processes, reducing the selection range to a set containing relatively few solutions.
Article
Engineering, Civil
Wenting Jin, Yimin Wang, Jianxia Chang, Xuebin Wang, Chen Niu, Yu Wang, Shaoming Peng
Summary: This paper proposes a novel framework to balance multiple objectives in reservoir operation and optimize the model to enhance the satisfaction of multiple requirements.
JOURNAL OF HYDROLOGY
(2021)
Article
Engineering, Multidisciplinary
Jingsi Huang, Xiangyu Wu, Zhijie Zheng, Yuansheng Huang, Wei Li
Summary: This article studies the optimal operation of a combined cascade reservoir and hydrogen system to improve the utilization of water resources and the system's economic benefits. The research shows that this combined system can effectively improve overall economic benefits and reservoir operation efficiency.
IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS
(2022)
Article
Engineering, Civil
Aadhityaa Mohanavelu, Bankaru-Swamy Soundharajan, Ozgur Kisi
Summary: This study compares six different modeling techniques to develop optimal reservoir operation solutions for competing objectives of irrigation and flood control. The results indicate that Deterministic Dynamic Programming achieves the best performance, and the current reservoir operation exhibits high vulnerability and low resilience.
WATER RESOURCES MANAGEMENT
(2022)
Article
Engineering, Civil
Liting Zhou, Pan Liu, Ziling Gui, Xiaojing Zhang, Weibo Liu, Lei Cheng, Jun Xia
Summary: This study focuses on the identification of time-varying parameters in hydrological modeling and explores the relationship between model structural deficiencies and parameters. It proposes a method to improve model structures by considering the variations of parameters over time.
JOURNAL OF HYDROLOGY
(2022)
Article
Engineering, Civil
Xiao Li, Pan Liu, Bo Ming, Kangdi Huang, Weifeng Xu, Yan Wen
Summary: This paper proposes a method to jointly derive forward contracts and operating rules, which helps hydropower producers gain more revenue in long-term operations. Using the Qing River cascade reservoir system in China's Hubei Province as a case study, the two-layer nested approach is used to determine optimal contract targets and operation trajectories, and a simulation-based optimization approach is employed to refine the operating rules and contracts.
JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT
(2022)
Article
Green & Sustainable Science & Technology
Xiao Li, Pan Liu, Yibo Wang, Zhikai Yang, Yu Gong, Rihui An, Kangdi Huang, Yan Wen
Summary: This study focuses on deriving operating rule curves to maximize revenue by integrating reservoir operations and a spot market. The derived rule curves effectively increase hydropower generation and revenue, reduce spill water, and expand the operational zone.
Article
Thermodynamics
Yu Gong, Pan Liu, Bo Ming, Maoyuan Feng, Kangdi Huang, Yibo Wang
Summary: This study identifies the functional form of operating rules for hydro-photovoltaic hybrid power systems using mathematical derivation, providing guidance for improving operational efficiency.
Article
Environmental Sciences
Xinran Luo, Pan Liu, Qianjin Dong, Yanjun Zhang, Kang Xie, Dongyang Han
Summary: This study investigates the use of LSTM-based preprocessing and postprocessing techniques in a hydrological model. The results show that LSTM-based bias correctors are effective and the integrated model performs better than the LSTM-only model when trained with limited data.
JOURNAL OF FLOOD RISK MANAGEMENT
(2023)
Article
Thermodynamics
Kangdi Huang, Pan Liu, Jong-Suk Kim, Weifeng Xu, Yu Gong, Qian Cheng, Yong Zhou
Summary: In this study, a three-stage model is proposed to formulate the daily generation scheduling of a wind-solar-hydro complementary system (WSHCS), addressing the significant difference between non-adjustable and adjustable periods for hydropower. The case study results show that the proposed three-stage model can increase the energy production and reduce the power curtailment rate of the WSHCS compared to the two-stage model. Furthermore, a three-layer nested approach is applied to simplify the high-dimensional three-stage model and improve computational efficiency.
Article
Environmental Sciences
Dongyang Han, Pan Liu, Kang Xie, He Li, Qian Xia, Qian Cheng, Yibo Wang, Zhikai Yang, Yanjun Zhang, Jun Xia
Summary: This study proposes an artificial intelligence-based LSTM model with attention mechanisms to forecast long-term runoff. The model improves accuracy by automatically assigning weights to key factors and analyzes their impacts on runoff.
ENVIRONMENTAL RESEARCH LETTERS
(2023)
Article
Environmental Sciences
Yibo Wang, Pan Liu, Dimitri Solomatine, Liping Li, Chen Wu, Dongyang Han, Xiaojing Zhang, Zhikai Yang, Sheng Yang
Summary: Aquatic community dynamics are greatly influenced by flow regime and water quality conditions, but these factors have rarely been integrated into existing ecological models. This study proposes a new niche-based metacommunity dynamics model that successfully simulates the coevolution processes of multiple populations in the mid-lower Han River, China. The study finds that different populations have varying responses to flow regime and water quality conditions.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2023)
Article
Engineering, Civil
Hang Su, Lei Cheng, Yue Wu, Shujing Qin, Pan Liu, Quan Zhang, Hujie Cheng, Yuran Li
Summary: This study investigates the dynamics of dissolved organic carbon (DOC) in a flash flood catchment during flood seasons. The results show three phases of DOC concentration and discharge: slow increase, rapid increase, and flat phases during inter-storm, moderate storm, and extreme storm periods, respectively. Hysteresis analysis reveals a shift in the clockwise pattern during extreme storms, indicating a transition from transport-limited to source-limited DOC export. Neglecting this shift can lead to significant overestimation of DOC yield in flash flood catchments.
JOURNAL OF HYDROLOGY
(2023)
Article
Thermodynamics
Qian Cheng, Pan Liu, Qian Xia, Lei Cheng, Bo Ming, Wei Zhang, Weifeng Xu, Yalian Zheng, Dongyang Han, Jun Xia
Summary: This study proposes an analytical method based on daily hydropower and PV power to evaluate the power curtailment rate of HPESs under future climate variability. The effectiveness of the method is verified through short-term and long-term operation models, and the accuracy of the PV curtailment function is validated through a case study. Results show that future hydropower and PV power will increase with substantial variances, while the PV curtailment rate exhibits an overall increase in the near future and a larger increase in the far future.
Article
Environmental Sciences
Xinran Luo, Pan Liu, Qian Xia, Qian Cheng, Weibo Liu, Yiyi Mai, Chutian Zhou, Yalian Zheng, Dianchang Wang
Summary: Quantifying the uncertainty of stormwater inflow is critical for improving the resilience of urban drainage systems. However, the computational complexity and time consumption have hindered the implementation of uncertainty-addressing methods for real-time control of these systems. To address this issue, this study developed a machine learning-based surrogate model that reduces computational complexity while maintaining high-fidelity descriptions of drainage dynamics. By using stormwater inflow and controls as inputs and system overflow as the output, the surrogate model can quickly evaluate system performance, making stochastic optimization feasible. The proposed real-time control strategy, which combines the surrogate model with stochastic model predictive control, aims to minimize expected overflow under various stochastic stormwater inflow scenarios. An ensemble of stormwater inflow scenarios is generated by assuming forecast errors follow normal distributions, and representative scenarios with their probabilities are selected to downsize the ensemble. The results of applying the proposed control strategy to a combined urban drainage system in China are promising, showing improved system resilience to uncertainty compared to classical deterministic model predictive control.
JOURNAL OF ENVIRONMENTAL MANAGEMENT
(2023)
Article
Engineering, Civil
Qian Cheng, Pan Liu, Lei Cheng, Bo Ming, Zhikai Yang, Xinran Luo, Weifeng Xu, Lanqiang Gong, Yang Zhang
Summary: Hydro-wind-photovoltaic (PV) complementary power systems (HWPCSs) provide a promising solution for integrating intermittent wind and PV power. However, the evaluation of HWPCSs under climate change often neglects short-term features, resulting in potential misestimations of climate change impacts. This study proposes a framework to quantify this misestimation by extracting short-term features and incorporating them into a long-term model. The results validate the importance of considering short-term features in accurately evaluating long-term HWPCS performance under climate change.
JOURNAL OF HYDROLOGY
(2023)
Article
Environmental Sciences
Zhenhui Wu, Dedi Liu, Yadong Mei, Shenglian Guo, Lihua Xiong, Pan Liu, Jie Chen, Jiabo Yin, Yujie Zeng
Summary: This paper proposes a novel simulation method for the WHE nexus system based on a nonlinear dynamic resilience model, and the results show that it outperforms linear models in simulating and predicting the resilience of the system, thus providing better guidance for resilience management in water resource systems.
WATER RESOURCES RESEARCH
(2023)
Article
Green & Sustainable Science & Technology
Shitian Xu, Pan Liu, Xiao Li, Qian Cheng, Zheyuan Liu
Summary: In order to integrate large-scale variable renewables into the power grid, researchers have established an operating model for the hydro-wind-PV system and derived operating rules considering long-term electricity prices. The case study of Jinping I HWPES in China shows that the operating rules significantly depend on electricity prices, and the proposed method improves power generation and benefits.
Article
Geosciences, Multidisciplinary
Yujie Zeng, Dedi Liu, Shenglian Guo, Lihua Xiong, Pan Liu, Jiabo Yin, Zhenhui Wu
Summary: This study proposes a new approach to model the water-energy-food nexus by incorporating human sensitivity and reservoir operation. The results indicate that environmental awareness can reduce resource shortages and improve the sustainability of the system.
HYDROLOGY AND EARTH SYSTEM SCIENCES
(2022)