| 引用本文: | 高 波,刘 川,韩 建,等.基于PSO-ELM的可植入UPQC的“源-网-荷-储”系统最优控制策略[J].电力系统保护与控制,2025,53(02):62-72.[点击复制] |
| GAO Bo,LIU Chuan,HAN Jian,et al.Optimum control for UPQC-embedded source-network-load-storage system using PSO-ELM[J].Power System Protection and Control,2025,53(02):62-72[点击复制] |
|
| 摘要: |
| 针对传统“源-网-荷-储”(source network load storage, SNLS)系统的可再生能源渗透率低及电能质量差等问题，提出了一种可植入统一电能质量调节器(unified power quality conditioner, UPQC)的SNLS系统最优控制方案。该方案通过基于粒子群优化(particle swarm optimization, PSO)的极限学习机(extreme learning machine, ELM)方法实现。在多目标优化运行方案中：第一个优化目标为最大化光伏阵列发电量；第二、三个优化目标分别为最小化负荷电压偏差和最大化网侧功率因数；第四个优化目标则为最大化变换器的利用率。由于多目标优化问题不易实时求解，提出了一种基于优化目标优先权顺序的分层优化思想，将多目标优化问题简化为若干个单目标优化问题。然后，通过将求解的所有最优解集训练为PSO-ELM代理模型，以实现所提策略的快速精确执行。最后，通过仿真验证了所提方法的有效性。算例表明所提策略可提升可再生能源的消纳率与系统变换器的利用率，并优化电能质量。 |
| 关键词: 统一电能质量调节器 “源-网-荷-储”系统 光伏 PSO-ELM |
| DOI：10.19783/j.cnki.pspc.240749 |
| 投稿时间：2024-06-17修订日期：2024-10-22 |
| 基金项目:国家自然科学基金项目资助(52407101，52467006，52367015)；江西省自然科学基金项目资助(20232BAB214061，2023BAB214065) |
|
| Optimum control for UPQC-embedded source-network-load-storage system using PSO-ELM |
| GAO Bo,LIU Chuan,HAN Jian,LI Zewen,WEI Baoquan |
| (East China Jiaotong University, Nanchang 330000, China) |
| Abstract: |
| To address the issues of low renewable energy penetration and poor power quality in traditional source-network-load-storage (SNLS) systems, an optimum control scheme for an SNLS system embedded with a unified power quality conditioner (UPQC) is presented. The proposed scheme is implemented using the particle swarm optimization (PSO) based extreme learning machine (ELM) algorithm. In the multi-objective optimization operation scheme: the first optimization objective is to maximize the power generation of photovoltaic (PV) arrays; the second and third optimization objectives are to minimize the load voltage deviation and maximize the network side power factor, respectively; and the fourth optimization objective is to maximize the utilization rate of the converter. Since the multi-objective optimization problems are difficult to solve in real-time, a hierarchical optimization approach based on the priority order of optimization objectives is presented to simplify the multi-objective optimization problem into several single-objective ones. Then, by training all the optimum solution sets obtained as an PSO-ELM surrogate model, the proposed strategy can be executed quickly and accurately. Finally, the effectiveness of the proposed scheme is verified through simulations. The case studies show that the proposed strategy can improve the absorption rate of renewable energy and the utilization rate of converters, and optimize power quality. |
| Key words: unified power quality conditioner (UPQC) source-network-load-storage (SNLS) photovoltaic (PV) particle swarm optimization-based extreme learning machine (PSO-ELM) |
