不同时间尺度的风功率建模及LOLP算法研究  

Different time scales of wind power modeling and LOLP algorithm research

作　　者：张紫凡[1] 

机构地区：[1]华南理工大学广州学院,广东广州510800

出　　处：《可再生能源》2014年第11期1658-1664,共7页Renewable Energy Resources

基　　金：电气工程及其自动化重点学科建设基金项目(JY110104)

摘　　要：风功率与时间的紧密相关性,使得风电有别于常规机组,因此,必须建立能够反映风功率时间特性的模型。大规模风电的接入对电力系统的可靠运行带来威胁,采用随机生产模拟以及序列运算等方法可以评估电源的可靠性,LOLP和LOLE则是其中的核心指标。文章利用每10 min风功率的测量数据,建立在时间序列上的风功率多状态模型,并采用随机生产模拟的方法,提出了含风电场的电力系统长期LOLP算法;建立在时间断面上的风功率马尔可夫链模型和状态转移矩阵;采用序列运算的方法,提出了含风电场的电力系统短期LOLP算法;最后,在IEEE-RTS 79模型中实现了上述算法,得到了相关结论。文章的研究对于含风电场的电力系统电源规划、运行规划以及可靠性评估等具有重要意义。Close correlation between wind power and time makes wind power generator different from conventional unit, therefore, wind power model, which can reflect the time characteristic, must be established. With large scale wind power integration, reliable operation of electric power system is under threat. By adopting the method of stochastic simulation and production sequence al- gorithm, the power supply reliability can be evaluated, LOLP and LOLE are core indicators. In this paper, wind power state model of time sequence is set up by using every 10 minutes wind power measurement data, and the probabilistic production simulation method is used to propose a long term LOLP algorithm for power systems with wind farms. Markov chain model and state transition matrix are built based on time section. The method of serial operation is adopted to put forward a short-term LOLP algorithm for power systems which contain wind farms. Finally, the algorithm is achieved in IEEE-RTS 79 model, and relevant conclusions are gotten. The research has great signif- icance for power planning, operational planning and reliability evaluation of power systems which contain wind farms.

关 键 词：风功率模型 LOLP 多态模型 马尔可夫链 随机生产模拟 序列运算 

分 类 号：TK81[动力工程及工程热物理—流体机械及工程]