PDF(3472 KB)
Improved Closed-Loop Predict-and-Optimize Intertwined Framework-Based Two-Stage Dispatch for Hydro-Wind- Photovoltaic-Involved Power Systems
LIU Jichun, XIAO Yujin, QIU Gao, TANG Lun, SUN Yi, LI Linghao
Electric Power Construction ›› 2025, Vol. 46 ›› Issue (12) : 143-158.
PDF(3472 KB)
PDF(3472 KB)
Improved Closed-Loop Predict-and-Optimize Intertwined Framework-Based Two-Stage Dispatch for Hydro-Wind- Photovoltaic-Involved Power Systems
[Objective] The high randomness of hydro，wind，and photovoltaic powers exacerbates the problem of achieving a trade-off between the computational complexity and operational economy of the traditional open-loop predict-then-optimize （OPO） dispatch. [Methods] This limitation is addressed by proposing a two-stage dispatch method based on an improved closed-loop predict-and-optimize intertwined framework （CPO） for hydro-wind-photovoltaic-involved power systems. First，a two-stage dispatch model for hydro-wind-photovoltaic power systems involving series，parallel，and hybrid-connected hydropower groups is constructed. Next，to train an economy-oriented prediction model for inflow and renewable energy generation，a loss function is established by considering the absolute deviation between the system cost calculated from the ground truths and the predictions of inflow and renewable energies. Finally，the variance，Bollinger bands，and autocorrelation function are combined to quantify the fluctuation intensities of renewable energy generation and hydropower inflow such that a hybrid regularization strategy involving elastic net regression is constructed to balance the training complexity and performance of the CPO under multiple uncertainties. [Results] MATLAB simulation results show that during the typical wet，dry，and normal months，the monthly average actual system cost obtained using the improved CPO method is reduced by 0.74%，0.57%，and 0.66%，respectively，compared with that obtained using the traditional OPO method； this verifies the effectiveness of the proposed method for improving the economic efficiency of power dispatch. [Conclusions] The improved CPO method proposed in this study significantly reduces the actual system cost and optimizes the economic efficiency of dispatch when the overall prediction accuracy of hydropower inflow，wind power，and photovoltaic power decreases slightly and the accuracy increases in certain periods. Moreover，in scenarios with high degrees of uncertainty，the effect of this method on improving economic efficiency is even more prominent.
hydropower plant group / inflow prediction / unit commitment / closed-loop predict-and-optimize intertwined framework / elastic net regression
| [1] |
陈国平, 董昱, 梁志峰. 能源转型中的中国特色新能源高质量发展分析与思考[J]. 中国电机工程学报, 2020, 40(17): 5493-5506.
|
| [2] |
李建林, 丁子洋, 游洪灏, 等. 构网型储能支撑新型电力系统稳定运行研究[J]. 高压电器, 2023, 59(7): 1-11.
|
| [3] |
赵毅, 陈雨婷, 孙文瑶. 多微网配电系统区间调度方法与市场交易策略研究[J]. 电网技术, 2022, 46(1): 47-56.
|
| [4] |
李伟, 韩瑞迪, 孙晨家, 等. 基于用电偏好的可平移负荷参与需求响应最优激励合同与激励策略[J]. 中国电机工程学报, 2021, 41(S1): 185-193.
|
| [5] |
郭红霞, 陈凌轩, 张启, 等. 电力电量平衡视角下新型电力系统极端场景研究及应对综述[J]. 电网技术, 2024, 48(10): 3975-3994.
|
| [6] |
黄河, 王建学, 肖云鹏, 等. 新型电力系统电力电量平衡分析关键技术与研究框架[J]. 电力建设, 2024, 45(9): 1-12.
随着新型电力系统加速构建,可再生能源发电装机与电量占比不断攀升。如何在源荷双重强不确定性下实现准确、高效、实用的电力电量平衡分析,是新型电力系统调控与规划中的基础与关键问题。通过对电力电量平衡分析在研究对象、分析模型与方法、评价指标与应用等方面的综述,总结了现有研究在平衡状态刻画和分析计算方面的不足。针对“强随机性与策略性下的电力电量平衡状态刻画”与“集中调控-协调自治平衡机制下的电力电量平衡分析与计算”两大关键科学问题,首先提出了涵盖“平衡边界生成-平衡分析计算-平衡状态评估-平衡预警-平衡能力提升”的新型电力系统电力电量平衡分析研究内容体系,接着阐述随机性平衡场景生成、多周期协调解耦的平衡分析、平衡状态评估指标体系设计、多时空分级平衡能力预警和新型平衡协作机制的研究思路与技术方案。最后总结并展望了新型电力系统电力电量平衡分析研究中需要重点突破的关键技术。
With the accelerated construction of new types of power systems, the proportion of integrated renewable energy sources has increased rapidly. Achieving an accurate, efficient, and practical power balance analysis under strong uncertainty of the source and load is a basic theme in the dispatching, controlling, and planning of new power systems. Hence, the shortcomings of current research in the description and calculation of the balance state are reviewed by summarizing the research objects, analytical models and methods, evaluation indices, and applications of power and quantity balance analyses. Based on the two key scientific issues of “electric power and energy balance state description under strong randomness and strategy” and “electric power and energy balance analysis and calculation under the centralized regulation-coordination autonomous balancing mechanism,” a research theoretical system for electric power and energy balance is proposed. First, this system covers “balance boundary generation-balance analysis and calculation-balance state assessment- balance early warning-balance capacity improvement.” Then, research ideas and technical schemes are proposed, including random balanced scenario generation, multi-period coordination decoupling balanced analysis, balanced state evaluation index system design, multi-spatial-temporal hierarchical balanced early warning, and balanced cooperation mechanism construction. Finally, some key technologies are worthy of consideration in further research on electric power and energy balance analyses in new types of power systems. |
| [7] |
陈皇森, 石立宝. 考虑多风电场出力预测误差分布特征的随机机组组合[J]. 电网技术, 2023, 47(12): 5026-5040.
|
| [8] |
|
| [9] |
朱继忠, 董瀚江, 李盛林, 等. 基于分布式新能源集群的微能源网优化调度综述[J]. 中国电机工程学报, 2024, 44(20): 7952-7970.
|
| [10] |
李嘉森, 王进, 杨蒙, 等. 基于随机优化的虚拟电厂热电联合经济优化调度[J]. 太阳能学报, 2023, 44(9): 57-65.
针对三北地区现有能源结构调节能力不足而导致的弃风问题,将风电场、光热电站、火电机组和热电联产机组聚合为虚拟电厂。采用随机优化处理风光不确定性问题,通过拉丁超立方抽样生成大量随机风光场景,并在充分考虑风光相关性和分布随机特性的基础上,利用Kantorovich距离削减与K-均值聚类算法对随机场景进行降维优化,获得风电、太阳直接辐照度典型预测场景。结合光热电站的灵活性与供能惯性,构建含光热虚拟电厂热电联合优化调度模型,并建立系统总运行成本最小的目标函数。最后在算例部分验证所提随机优化方法在计算效率、预测精度和处理风光随机问题的优越性;对不同运行模式下的目标函数进行求解,验证所提出的优化调度策略能够在满足系统运行经济性的同时实现风电的最大消纳。
Aiming at the problem of wind curtailment caused by the energy structure lacked the adjustment ability in the three north area, this paper aggregated wind farm, concentrating solar power plant（CSPP）, thermal power units and combined heat and power（CHP） plant into virtual power plant（VPP）. Using stochastic optimization to deal with the uncertainty of wind-solar, Latin hypercube sampling （LHS） was used to generated a large number of random scenes, and based on considering the random characteristics and correlation of wind-solar distribution fully,Kantorovich distance reduction and <em>K</em>-means clustering algorithm were used to optimized and reduced the dimension of random scenes, for obtaining typical prediction wind-solar scenes. Combined with the flexibility and energy supply inertia of CSPP, the optimal dispatching model of the VPP contained photothermal was constructed, and the objective function of minimizing the total operation cost of the system was established. Finally, an example was given to verify the superiority of the proposed stochastic optimization method in computational efficiency and prediction accuracy; The objective functions under different operation scenarios were solved to verify that the optimal dispatching model could improve the wind power consumption capacity while reducing the system operation cost effectively.
|
| [11] |
王海伦, 丁一凡, 李杨, 等. 计及混合式抽水蓄能改造的清洁微网分布鲁棒容量优化配置[J]. 电工技术学报, 2025, 40(7): 2112-2126.
|
| [12] |
刘亚鑫, 蔺红, 马越, 等. 考虑分时碳价的虚拟电厂分布鲁棒竞标方法[J]. 电测与仪表, 2024, 61(8): 12-19.
|
| [13] |
张亚超, 刘开培, 秦亮, 等. 基于聚类经验模态分解-样本熵和优化极限学习机的风电功率多步区间预测[J]. 电网技术, 2016, 40(7): 2045-2051.
|
| [14] |
黎敏, 林湘宁, 张哲原, 等. 超短期光伏出力区间预测算法及其应用[J]. 电力系统自动化, 2019, 43(3): 10-16.
|
| [15] |
何常新, 彭旭, 方福东, 等. 基于LASSO回归和QRLSTM的来水预测方法研究[J]. 人民长江, 2024, 55(11): 138-145, 165.
|
| [16] |
黄晟, 潘丽君, 屈尹鹏, 等. 基于改进深度Q网络的无预测风电场日前拓扑优化[J]. 电力系统自动化, 2025, 49(2): 122-132.
|
| [17] |
|
| [18] |
|
| [19] |
|
| [20] |
. The potential benefits of seasonal streamflow forecasts for the hydropower sector have been evaluated for several basins across the world but with contrasting conclusions on the expected benefits. This raises the prospect of a complex relationship between reservoir characteristics, forecast skill, and value. Here, we unfold the nature of this relationship by studying time series of simulated power production for 735 headwater dams worldwide. The time series are generated by running a detailed dam model over the period 1958–2000 with three operating schemes: basic control rules, perfect forecast-informed operations, and realistic forecast-informed operations. The realistic forecasts are issued by tailored statistical prediction models – based on lagged global and local hydroclimatic variables – predicting seasonal monthly dam inflows. As expected, results show that most dams (94 %) could benefit from perfect forecasts. Yet, the benefits for each dam vary greatly and are primarily controlled by the time-to-fill value and the ratio between reservoir depth and hydraulic head. When realistic forecasts are adopted, 25 % of dams demonstrate improvements with respect to basic control rules. In this case, the likelihood of observing improvements is controlled not only by design specifications but also by forecast skill. We conclude our analysis by identifying two groups of dams of particular interest: dams that fall in regions expressing strong forecast accuracy and having the potential to reap benefits from forecast-informed operations and dams with a strong potential to benefit from forecast-informed operations but falling in regions lacking forecast accuracy. Overall, these results represent a first qualitative step toward informing site-specific hydropower studies.
|
| [21] |
|
| [22] |
|
| [23] |
Many real-world analytics problems involve two significant challenges: prediction and optimization. Because of the typically complex nature of each challenge, the standard paradigm is predict-then-optimize. By and large, machine learning tools are intended to minimize prediction error and do not account for how the predictions will be used in the downstream optimization problem. In contrast, we propose a new and very general framework, called Smart “Predict, then Optimize” (SPO), which directly leverages the optimization problem structure—that is, its objective and constraints—for designing better prediction models. A key component of our framework is the SPO loss function, which measures the decision error induced by a prediction. Training a prediction model with respect to the SPO loss is computationally challenging, and, thus, we derive, using duality theory, a convex surrogate loss function, which we call the SPO+ loss. Most importantly, we prove that the SPO+ loss is statistically consistent with respect to the SPO loss under mild conditions. Our SPO+ loss function can tractably handle any polyhedral, convex, or even mixed-integer optimization problem with a linear objective. Numerical experiments on shortest-path and portfolio-optimization problems show that the SPO framework can lead to significant improvement under the predict-then-optimize paradigm, in particular, when the prediction model being trained is misspecified. We find that linear models trained using SPO+ loss tend to dominate random-forest algorithms, even when the ground truth is highly nonlinear.
|
| [24] |
|
| [25] |
赵长飞. 新能源电力系统概率预测-决策一体化研究[D]. 杭州: 浙江大学, 2022.
|
| [26] |
|
| [27] |
|
| [28] |
|
| [29] |
|
| [30] |
|
| [31] |
加鹤萍, 郭宇辰, 马乾鑫, 等. 基于机器学习方法的现货电价预测研究综述[J]. 电力建设, 2025, 46(2): 160-179.
|
| [32] |
|
| [33] |
|
| [34] |
|
| [35] |
侯慧, 甘铭, 吴细秀, 等. 考虑移动氢能存储的港口多能微网两阶段分布鲁棒优化调度[J]. 中国电机工程学报, 2024, 44(8): 3078-3093.
|
| [36] |
杜刚, 赵冬梅, 刘鑫. 计及风电不确定性优化调度研究综述[J]. 中国电机工程学报, 2023, 43(7): 2608-2627.
|
/
| 〈 |
|
〉 |
