文章摘要
李鹏,韩建沛,殷云星,等.电转氢作为灵活性资源的微网容量多目标优化配置[J].电力系统自动化,2019,43(17):28-35. DOI: 10.7500/AEPS20181023006.
LI Peng,HAN Jianpei,YIN Yunxing, et al.Multi-objective Optimal Capacity Configuration of Microgrid with Power to Hydrogen as Flexible Resource[J].Automation of Electric Power Systems,2019,43(17):28-35. DOI: 10.7500/AEPS20181023006.
电转氢作为灵活性资源的微网容量多目标优化配置
Multi-objective Optimal Capacity Configuration of Microgrid with Power to Hydrogen as Flexible Resource
DOI:10.7500/AEPS20181023006
关键词: 微网(微电网)  电转氢  静态灵活性  Tchebycheff法  多目标优化
KeyWords: microgrid  power to hydrogen(P2H)  static flexibility  Tchebycheff approach  multi-objective optimization
上网日期:2019-03-26
基金项目:国家自然科学基金资助项目(51577068);国家高技术研究发展计划(863计划)资助项目(2015AA050104)
作者单位E-mail
李鹏 华北电力大学电气与电子工程学院, 河北省保定市 071003  
韩建沛 华北电力大学电气与电子工程学院, 河北省保定市 071003 ncepu_hjp@outlook.com 
殷云星 华北电力大学电气与电子工程学院, 河北省保定市 071003  
韦巍 浙江大学电气工程学院, 浙江省杭州市 310027  
摘要:
      高比例可再生能源接入情形下,由于没有大电网的支撑,独立型微网的规划设计对灵活性水平提出了更高的要求。为了提高独立型微网接纳可再生能源的能力,将电转氢作为灵活性资源,以微网中风光机组容量和灵活性资源容量为决策变量,建立以微网年投资运行成本最低和静态灵活性水平最高为优化目标的多目标优化模型。针对风光及负荷长时间尺度的不确定性,采用X-means聚类得到风速、光照强度和负荷情况的典型场景;针对所建立的混合整数多目标规划模型,采用Tchebycheff法将多目标模型转化为多个单目标模型进行求解;针对求解一系列单目标问题得到的Pareto非劣解集,基于模糊熵权法和模糊隶属度构建排序指标,选择排序函数值最高的非劣解作为最优解。最后,基于MATLAB仿真对所提优化配置方法的正确性和合理性进行了验证。
Abstract:
      With the high proportion of renewable energy sources, the planning and design of the stand-alone microgrid put higher requirements on the level of flexibility because of no support from the grid. In order to enhance the ability of stand-alone microgrid to accommodate renewable energy, power to hydrogen(P2H)is employed as a flexible resource. Additionally, the capacities of wind turbine and photovoltaic cells, together with the capacities of flexibility resources in microgrid, are taken as decision variables to establish a multi-objective optimization model, which takes the lowest annual investment operation cost and the highest static flexibility level as optimization objectives. Aiming at uncertainties of wind power, photovoltaic intensity and loads during long-time period, the X-means clustering method is introduced to obtain typical scenarios of wind speed, photovoltaic intensity and loads. With regard to the established mixed integer multi-objective programming model, the Tchebycheff approach is adopted to convert the original multi-objective model into multiple single-objective models. For the Pareto non-inferior solution sets obtained by solving a series of single-objective problems, the ranking index is constructed with the utilization of the fuzzy entropy weight method and fuzzy membership degree, and the optimal solution is defined as the non-inferior solution with the highest ranking function value. Finally, the correctness and rationality of the proposed optimal configuration method are verified based on the MATLAB simulation.
查看全文(Free!)   查看附录   查看/发表评论  下载PDF阅读器