• CSCD核心库收录期刊
  • 中文核心期刊
  • 中国科技核心期刊

电力建设 ›› 2015, Vol. 36 ›› Issue (11): 10-16.doi: 10.3969/j.issn.1000-7229.2015.11.002

• 配电网规划专栏 • 上一篇    下一篇

基于随机机会约束规划的有源配电网多目标规划

屈高强1,李荣2,董晓晶1,康健3,党东升1,刘洪2   

  1. 1.国网宁夏电力公司经济技术研究院,银川市 750011;2.天津大学电气与自动化工程学院,天津市 300072;3.国网宁夏电力公司,银川市 750010
  • 出版日期:2015-11-01
  • 作者简介:屈高强(1983),男,本科,工程师,主要从事电力系统规划及分析研究工作; 李荣(1991),男,硕士研究生,主要研究主动配电网规划; 董晓晶(1980),男,本科,工程师,主要从事电力系统及电网规划工作; 康健(1979),男,本科,工程师,主要从事配电网规划及新能源接入管理工作; 党东升(1982),男,本科,工程师,从事继电保护工作; 刘洪(1979),男,博士,副教授,主要从事城市电网规划、评估等方面的研究和应用工作。
  • 基金资助:

    国网宁夏电力公司科技项目(5229JY1307G6)

Multiple-Objective Planning of Active Power Distribution Network Base0d on Random Chance Constrained Programming

QU Gaoqiang1, LI Rong2, DONG Xiaojing1, KANG Jian3, DANG Dongsheng1, LIU Hong2   

  1. 1.    State Grid Ningxia Electric Power Science Research Institute, Yinchuan 750011, China;2. School of Electrical Engineering and Automaton, Tianjin University, Tianjin 300072, China;3. State Grid Ningxia Electric Power Corporation, Yinchuan 750010, China
  • Online:2015-11-01

摘要:

含分布式新能源配电网规划均采用被动、保守接入分布式新能源的规划方法,固然保证了配电网安全,但并没有反映分布式新能源的出力特征,因而造成不必要的配电网建设投资。为了解决这一问题,采用基于随机机会约束规划的有源配电网规划方法,将有源配电网规划中必须满足的硬性约束条件转变为较高置信度的软约束形式,同时,在模型中引入反映经济效益的投资成本、网络损耗以及反映配电网供电安全性的电压偏移度这3个目标函数作为优化对象,形成了有源配电网规划的多目标随机机会约束规划模型。采用结合量子法改进的非支配排序多目标优化遗传算法(non-dominated sorting genetic algorithm2, NSGA-2)求解获得非劣解帕累托前沿,在此基础上,运用逼近理想解排序法(technique for order preference by similarity to ideal solution,TOPSIS)对非劣解排序得到最优方案。最后,以57节点的配电网网络为算例,验证了方法的可行性、有效性。

关键词: 随机机会约束, 分布式电源, 时序特性, 有源配电网规划, 改进NSGA-2

Abstract:

Most of the typical distribution network programmings with distributed new energy utilize passive, conservative programming method to access distributed new energy, which indeed secure the safe of distribution network, but cannot reflect the output characteristics of distributed new energy and cause unnecessary investment on distribution network construction. To solve this problem, this paper proposed active power distribution network programming method based on random chance constrained programming. Firstly, the hard constraint conditions in active power distribution network programming were transformed into soft ones with higher confidence level. Meanwhile, three independent objective functions including the investment cost reflecting the economic benefit, the power loss and the voltage deviation degree reflecting the distribution network power supply security were set to form the multiple-objective active power distribution network planning model based on random chance constrained programming. Then, the model was solved to obtain non-inferior solution Pareto frontier by the improved NSGA-2 (non-dominated sorting genetic algorithm2) combined with the quantum method. On this basis, the TOPSIS (technique for order preference by similarity to ideal solution) was used to sort the non-inferior solution, in order to obtain the optimal solution. Finally, a distribution network with 57 nodes was used as example to verify the feasibility and availability of the proposed method.

Key words: random chance constraint, distributed generation, time sequential characteristics, active power distribution network planning, improved NSGA-2

中图分类号: