中国电机工程学报
中國電機工程學報
중국전궤공정학보
ZHONGGUO DIANJI GONGCHENG XUEBAO
2015年
12期
2919-2926
,共8页
检修计划%风电%鲁棒优化%不确定性集合%保守性
檢脩計劃%風電%魯棒優化%不確定性集閤%保守性
검수계화%풍전%로봉우화%불학정성집합%보수성
maintenance schedule%wind power%robust optimization%uncertainty set%conservatism
大规模风电接入电网给传统电力系统机组检修决策问题带来了新的挑战,如何在历史统计数据不足的情况下考虑风电对机组检修的影响成为难点之一。将鲁棒优化理论引入检修决策领域,构建机组检修鲁棒优化模型,并给出风电不确定集合的构造方法。模型以风电最恶劣情景下系统运行最优为优化目标,保证系统运行安全性,通过调整不确定集合边界,可控制鲁棒优化模型的保守性,实现决策的经济性与安全性平衡。算例对比证明所提模型和方法在计算效率和计算结果中的优势。通过调整保守性参数,鲁棒优化模型可获得比传统随机规划模型更优的结果。
大規模風電接入電網給傳統電力繫統機組檢脩決策問題帶來瞭新的挑戰,如何在歷史統計數據不足的情況下攷慮風電對機組檢脩的影響成為難點之一。將魯棒優化理論引入檢脩決策領域,構建機組檢脩魯棒優化模型,併給齣風電不確定集閤的構造方法。模型以風電最噁劣情景下繫統運行最優為優化目標,保證繫統運行安全性,通過調整不確定集閤邊界,可控製魯棒優化模型的保守性,實現決策的經濟性與安全性平衡。算例對比證明所提模型和方法在計算效率和計算結果中的優勢。通過調整保守性參數,魯棒優化模型可穫得比傳統隨機規劃模型更優的結果。
대규모풍전접입전망급전통전력계통궤조검수결책문제대래료신적도전,여하재역사통계수거불족적정황하고필풍전대궤조검수적영향성위난점지일。장로봉우화이론인입검수결책영역,구건궤조검수로봉우화모형,병급출풍전불학정집합적구조방법。모형이풍전최악렬정경하계통운행최우위우화목표,보증계통운행안전성,통과조정불학정집합변계,가공제로봉우화모형적보수성,실현결책적경제성여안전성평형。산례대비증명소제모형화방법재계산효솔화계산결과중적우세。통과조정보수성삼수,로봉우화모형가획득비전통수궤규화모형경우적결과。
ABSTRACT:The integration of large scale wind power brings great challenges to the classical generator maintenance scheduling in power systems. One of the major difficulties is that how to consider the impact of wind power on the maintenance schedule under the absence of historical statistic data of wind power. The robust optimization theory was applied to the maintenance scheduling problems and a robust optimization model, as well as the method of constructing uncertainty set of wind power, was proposed in this paper. The security of power system was reflected in the objective of the robust optimization model, which was to optimize the operation cost under the worst case of uncertainty factors. The economy of the optimal decision could be achieved by adjusting the boundary of uncertainty set to control the conservatism of the model. The computational efficency and advantage of the model are demonstrated by numerical examples. By adjusting the uncertainty set, the optimal maintenance schedule obtained from the robust model is expected to be better than the one from the stochastic programming model.