电网技术
電網技術
전망기술
POWER SYSTEM TECHNOLOGY
2013年
12期
3461-3466
,共6页
和敬涵%柏丹丹%王小君%叶豪东
和敬涵%柏丹丹%王小君%葉豪東
화경함%백단단%왕소군%협호동
低频减载%单位负荷切除因子%层次分析法%综合代价%粒子群优化算法
低頻減載%單位負荷切除因子%層次分析法%綜閤代價%粒子群優化算法
저빈감재%단위부하절제인자%층차분석법%종합대개%입자군우화산법
under-frequency load-shedding%per unit load shedding factor%analytic hierarchy process%comprehensive cost%particle swarm optimization
引入单位负荷切除因子来定量表征不同切负荷节点的负荷重要性、单位电能停电损失、负载率和负荷频率调节效应系数的差异对切负荷动作效果的影响,基于层次分析法求解单位负荷切除因子的数值大小,并将其作为负荷节点切负荷容量的权值。计及各切负荷点的容量约束、系统切负荷总量的限制和系统稳态频率约束,提出一种低频减载综合代价最优模型,并采用粒子群优化算法求解该多约束优化问题。仿真算例结果表明所提算法相对于传统低频减载算法表现更佳,初步论证了模型和算法的可行性和有效性。
引入單位負荷切除因子來定量錶徵不同切負荷節點的負荷重要性、單位電能停電損失、負載率和負荷頻率調節效應繫數的差異對切負荷動作效果的影響,基于層次分析法求解單位負荷切除因子的數值大小,併將其作為負荷節點切負荷容量的權值。計及各切負荷點的容量約束、繫統切負荷總量的限製和繫統穩態頻率約束,提齣一種低頻減載綜閤代價最優模型,併採用粒子群優化算法求解該多約束優化問題。倣真算例結果錶明所提算法相對于傳統低頻減載算法錶現更佳,初步論證瞭模型和算法的可行性和有效性。
인입단위부하절제인자래정량표정불동절부하절점적부하중요성、단위전능정전손실、부재솔화부하빈솔조절효응계수적차이대절부하동작효과적영향,기우층차분석법구해단위부하절제인자적수치대소,병장기작위부하절점절부하용량적권치。계급각절부하점적용량약속、계통절부하총량적한제화계통은태빈솔약속,제출일충저빈감재종합대개최우모형,병채용입자군우화산법구해해다약속우화문제。방진산례결과표명소제산법상대우전통저빈감재산법표현경가,초보론증료모형화산법적가행성화유효성。
A unit load-shedding factor is led in to quantitatively characterize the impacts of load importance at different load-shedding points, outage cost of per unit electric energy, load rate and the difference between coefficients of load-frequency regulation effects on load-shedding effect. Based on analytical hierarchy process (AHP) the numerical value of per unit load-shedding factor is solved and taken as the weight of load-shedding capacity at load bus. Considering the capacity constraints at load-shedding points, the limitation of total load-shedding amount and the system frequency constraint under steady state, an optimal model of comprehensive cost for under-frequency load-shedding (UFLS) is proposed, and particle swarm optimization algorithm is used to solve the proposed problem with multi-constraints. Simulation results show that relative to traditional UFLS algorithm, better load-shedding effects can be achieved by using the proposed optimal algorithm, thus the feasibility and effectiveness of the proposed model and algorithm are preliminarily demonstrated.