计算机工程与应用 ›› 2016, Vol. 52 ›› Issue (16): 251-256.

• 工程与应用 • 上一篇    下一篇

基于改进遗传算法的分布式电源并网优化配置

包广清1,杨国金2,杨  勇3,王晓兰1   

  1. 1.兰州理工大学 电气工程与信息工程学院,兰州 730050
    2.国网青海省电力公司经济技术研究院,西宁 810008
    3.甘肃省电力科学研究院,兰州 730050
  • 出版日期:2016-08-15 发布日期:2016-08-12

Improved Genetic Algorithm-based optimal planning for grid-connected distributed generation

BAO Guangqing1, YANG Guojin2, YANG Yong3, WANG Xiaolan1   

  1. 1.College of Electrical Engineering & Automation Engineering, Lanzhou University of Technology, Lanzhou 730050, China
    2.Economic Research Institute of Qinghai Province, Xining 810008, China
    3.Electric Power Research Institute of Gansu Province, Lanzhou 730050, China
  • Online:2016-08-15 Published:2016-08-12

摘要: 随着国家电网对分布式电源并网市场的开放,将分布式电源集成到现有配电系统是今后电力系统的发展趋势。以配电网网损和节点电压偏移最小化为优化目标,考虑支路电流约束、分布式发电单元容量和总接入容量等约束条件,构建大规模分布式电源并网优化配置模型。并提出基于均匀设计的改进遗传算法进行寻优计算,避免了遗传算子的盲目试凑,可以较好地兼顾多目标优化Pareto解集的多样性与快速性,有效提高优化精度。算例对比分析结果表明,通过对分布式电源接入配电网的合理优化配置,可以有效降低系统网损,提高配电网电压的稳定性。

关键词: 分布式发电, 系统网损, 节点电压偏移, 多目标优化, 遗传算法

Abstract: With the opening of the national grid for grid-connected Distributed Generation(DG) market, the DG is integrated into existing distribution systems becomes the future development trend of electric power system. To minimize the distribution network loss and node voltage deviation, the optimal location and capability configuration model for grid-connected DG is built, taking into account the branch current constraints, distributed generation units and total access capacity constraints. Then the improved Genetic Algorithms(GA) are applied to the optimization calculation. The genetic operators are set by uniform design to avoid the time consuming blind trials, and both the diversity and fastness of Pareto solution set are coordinated well. The results of example analysis show that the obtained configuration scheme of DG can effectively reduce the power loss and voltage deviation, thus the feasibility and effectiveness of the proposed model and algorithm are validated.

Key words: distributed generation, power loss, node voltage deviation;multi-objective optimization; genetic algorithm