PDF(2966 KB)
PDF(2966 KB)
PDF(2966 KB)
基于深度强化学习的农村风光水储微电网容量配置研究
Capacity Configuration of Rural Wind-PV-Hydro-Storage Microgrids Based on Deep Reinforcement Learning
针对农村地区基础设施薄弱、电压稳定性差、可再生能源利用效率低等现实问题，构建了一种兼顾电压稳定性与经济性的分布式电源容量与选址优化模型。该模型以提高电压稳定性为目标之一，以降低分布式电源接入配电网造成的电能质量影响；同时，以分布式电源全生命周期的平准化度电成本（levelized cost of energy，LCOE）为另一优化目标，综合考虑投资、运维成本及发电能力，以提高系统经济性。采用双重深度Q网络（double deep Q-network，DDQN）算法求解该模型，制定兼顾电压稳定与经济性的最优方案。最后，以改进后的IEEE 33节点系统进行仿真验证，结果表明：采用该模型的改造方案能有效提高农村配电网电压稳定性并降低系统改造成本；通过与深度Q网络（deep Q-network， DQN）算法、非支配排序遗传算法（nondominated sorting genetic algorithm II ， NSGA-II）、多目标粒子群（multi-objective particle swarm optimization， MOPSO）算法进行对比分析，进一步验证了所提方法的优越性，体现出强化学习算法在解决复杂优化问题时的高效性与灵活性。
To address weak infrastructure， poor voltage stability， and low renewable-energy utilization in rural areas， this paper proposes a siting-and-sizing model for distributed generation （DG） that simultaneously optimizes voltage quality and economic performance. One objective aims to minimize voltage deviations caused by DG integration， thereby enhancing distribution-network power quality； the other seeks to minimize the levelized cost of energy （LCOE） over the full life cycle of the DG portfolio， accounting for investment， operation and maintenance expenses， and energy yield. The model is solved with a double deep Q-network （DDQN）， yielding a configuration that balances voltage stability and cost. Simulation on a modified IEEE 33-bus rural feeder shows that the DDQN-based scheme markedly improves voltage profiles while reducing upgrade costs. Furthermore， comparative analyses with the deep Q-network （DQN）， non-dominated sorting genetic algorithm II （NSGA-II）， and multi-objective particle swarm optimization （MOPSO） methods verify the superiority of the proposed approach， highlighting the efficiency， adaptability， and robustness of reinforcement learning for complex energy-system optimization.
配电网 / 可再生能源微电网 / 强化学习 / 平准化度电成本（LCOE） / 容量配置
distribution networks / renewable energy microgrids / reinforcement learning / levelized cost of energy （LCOE） / capacity allocation
| [1] |
国家发展和改革委员会. 关于实施农村电网巩固提升工程的指导意见[EB/OL].(2023-07-04)[2025-10-16]. https://www.gov.cn/zhengce/zhengceku/202307/content_6891875.htm.
|
| [2] |
宋美琪, 李洪全, 肖赫, 等. 面向高比例新能源接入主动配电网的多层模型预测电压控制[J]. 分布式能源, 2025, 10(1): 53-61.
|
| [3] |
方圆, 陈洁, 田小壮, 等. 孤岛模式下风-氢互补微电网的容量经济优化[J]. 计算机仿真, 2020, 37(2): 110-114.
|
| [4] |
周京华, 翁志鹏, 宋晓通. 兼顾可靠性与经济性的孤岛型光储微电网容量配置方法[J]. 电力系统自动化, 2021, 45(8): 166-174.
|
| [5] |
王世震, 窦迅, 王俊. 考虑微电网群和系统经济运行的配电网储能优化配置[J]. 现代电力, 2022, 39(1): 88-94.
|
| [6] |
韩莹, 于三川, 李荦一, 等. 计及阶梯式碳交易的风光氢储微电网低碳经济配置方法[J]. 高电压技术, 2022, 48(7): 2523-2533.
|
| [7] |
李争, 罗晓瑞, 徐若思, 等. 风光-氢储微电网系统多目标容量优化配置[J]. 热能动力工程, 2023, 38(4): 131-138.
|
| [8] |
屈克庆, 乔敬茂, 毛玲, 等. 共享储能电站优化配置及选址评价方法[J]. 现代电力, 2025, 42(3):614-622.
|
| [9] |
张莲, 赵梦琪, 廖宗毅, 等. 计及多因素聚合储能寿命的微电网容量优化配置[J]. 重庆理工大学学报(自然科学), 2023, 37(1): 196-203.
|
| [10] |
张世旭, 李姚旺, 刘伟生, 等. 面向微电网群的云储能经济-低碳-可靠多目标优化配置方法[J]. 电力系统自动化, 2024, 48(1): 21-30.
|
| [11] |
辛曦, 欧阳森, 黄祎, 等. 考虑储能容量衰减的多保供电型微网最优经济配置及可靠性评估[J]. 电力建设, 2024, 45(10): 100-113.
|
| [12] |
孙建梅, 陈璐. 基于LCOE的分布式光伏发电并网效益分析[J]. 中国电力, 2018, 51(3): 88-93.
|
| [13] |
昌敦虎, 田川, 张泽阳, 等. 基于LCOE模型的光伏发电经济效益分析:以宜昌农村地区光伏扶贫电站项目为例[J]. 环境科学研究, 2020, 33(10): 2412-2420.
|
| [14] |
赵振宇, 张玉洁. 光储项目成本效益模型及平价上网预测研究[J]. 太阳能学报, 2023, 44(7): 214-220.
为科学、准确地确定光储项目的经济性和平准化度电成本（LCOE）,考虑技术进步、政策调控等因素的影响建立光储项目成本效益模型、LCOE模型以及系统动力学仿真模型。以河北省光储开发项目为案例,分别在有无补贴的情况下应用模型,对光储项目LCOE、内部收益率和综合效益等指标进行仿真分析。选取青海省、云南省光储开发项目做对比研究,测算出光储项目在上述3省中实现平价上网的不同临界点。所建模型可用于分析光储项目成本效益、测算LCOE值、预测光储发电实现平价上网的初始年份。
It is especially urgent to calculate the cost and benefit of photovoltaic energy storage power project accurately. In order to scientifically and accurately determine the economic and levelized cost of energy（LCOE） of photovoltaic energy storage power project, in this paper, the cost benefit model, LCOE model and system dynamics simulation model of photovoltaic energy storage power project are established considering the influence of technological progress and policy regulation. Taking the photovoltaic energy storage power project in Hebei province as an example, the LCOE, internal rate of return and comprehensive benefit of the photovoltaic energy storage power project are simulated and analyzed by using the model with or without subsidies. It also selects the photovoltaic energy storage power projects in Qinghai province and Yunnan province for comparative study, and calculates the different critical points for photovoltaic energy storage power projects to achieve grid parity in the above three provinces. The model can be used to analyze the cost benefit of photovoltaic energy storage power project, to measure LCOE, and to predict the initial year when photovoltaic energy storage power project will realize grid parity in the future.
|
| [15] |
张有兵, 林一航, 黄冠弘, 等. 深度强化学习在微电网系统调控中的应用综述[J]. 电网技术, 2023, 47(7): 2774-2788.
|
| [16] |
陈维江, 靳晓凌, 吴鸣, 等. 双碳目标下我国配电网形态快速演进的思考[J]. 中国电机工程学报, 2024, 44(17): 6811-6818.
|
| [17] |
郭一帆, 欧阳森, 张晋铭, 等. 含小水电配电网重要负荷评估及其水光储优化配置方法[J]. 广东电力, 2024, 37(5):32-42.
|
| [18] |
李鑫伟, 陈彬剑, 于明志, 等. 基于多目标优化的多能互补冷热电联产系统运行优化研究[J]. 热力发电, 2024, 53(7): 73-81.
|
| [19] |
智筠贻, 凌浩恕, 吴昊, 等. 风光储多能互补能源系统容量配置优化[J]. 储能科学与技术, 2024, 13(11): 3874-3888.
风光储多能互补能源系统可充分利用可再生能源提高供能的经济性和环保性。本文提出了一种风光储多能互补能源系统，建立了系统的能量模型；综合考虑系统运行的经济性和环保性，提出了系统综合成本和碳排放量最低的目标；开发了改进型非支配遗传算法求解仿真模型，得到了多目标问题的帕累托最优解集，并通过逼近理想解排序法获得了系统的最优容量配置运行方案；利用线性规划软件CPLEX求解器开展了系统的运行调度优化，验证了该系统框架和优化调度模型的有效性和正确性。研究结果表明，本文所提出的风光储多能互补能源系统容量配置优化方法有效提高了可再生能源利用率，实现了经济成本和碳排放量最低，提高了系统的经济性和环保性。本文为可再生能源系统实现持续稳定可靠的供能和园区的低碳化转型提供了参考。
The multi-energy complementary system integrating wind, solar, and energy storage technologies optimizes the use of renewable energy resources, enhancing both economic and environmental benefits. This study proposes a multi-energy complementary system model that incorporates wind, solar, and energy storage. The objective is to minimize the system's overall cost and carbon emissions, addressing both economic and environmental concerns. An improved non-dominated genetic algorithm is developed to obtain the Pareto optimal solution set for the multi-objective optimization problem. The optimal capacity configuration and operation scheme are determined using the technique for order preference by similarity to ideal solution. The system's operation scheduling is optimized using the CPLEX solver, a linear programming software, to validate the effectiveness and accuracy of the proposed system framework and scheduling model. Results demonstrate that the proposed optimization method significantly enhances renewable energy utilization, minimizes economic costs and carbon emissions, and improves the system's economic and environmental performance. This research offers valuable insights for the sustainable, stable, and reliable energy supply of renewable energy systems and supports the low-carbon transition of industrial parks. |
| [20] |
谢雨龙, 罗逸飏, 李智威, 等. 考虑微网新能源经济消纳的共享储能优化配置[J]. 高电压技术, 2022, 48(11): 4403-4413.
|
| [21] |
刘健辰, 刘山林. 基于二阶锥松弛和Big-M法的配电网分布式电源优化配置[J]. 电网技术, 2018, 42(8): 2604-2611.
|
| [22] |
高冠中, 姚建国, 严嘉豪, 等. 基于多智能体深度强化学习的配-微网协同优化调度研究[J]. 智慧电力, 2024, 52(9): 80-87.
|
| [23] |
杜明, 任建国, 张清杨. 基于双Q学习算法的安全容量的优化[J]. 现代电子技术, 2024, 47(11): 181-186.
|
/
| 〈 |
|
〉 |
清华大学出版社期刊中心
