大电网可靠性的序贯和非序贯蒙特卡洛仿真的收敛性分析及比较  

Convergence Analysis and Comparison of Sequential and Nonsequential Monte-Carlo Simulation for Bulk Power System Reliability Assessment

作　　者：赵渊[1] 周家启[1] 刘志宏[2] 

机构地区：[1]重庆大学输配电装备及系统安全与新技术国家重点实验室,重庆400044 [2]重庆市电力公司杨家坪供电局,重庆400050

出　　处：《电工技术学报》2009年第11期127-133,共7页Transactions of China Electrotechnical Society

基　　金：国家自然科学基金(50607021);国家重点基础研究发展计划项目(973项目)(2004CB217908)资助项目

摘　　要：深入研究了大电网可靠性评估的序贯和非序贯蒙特卡洛仿真方法的收敛特性,分析了两种方法的计算精度(方差系数)和样本容量(仿真总年数或状态抽样总数)的概率不确定性关系,基于同分布的中心极限定理推导了计算精度和样本容量的置信区间公式,得出结论:①序贯蒙特卡洛仿真的收敛性取决于可靠性指标的标准方差与其期望值的比值,该比值越小,则收敛越快;②非序贯蒙特卡洛仿真的收敛性取决于失负荷概率(LOLP)期望值的大小,该值越大,则收敛越快。通过对RBTS、IEEE-RTS 79和IEEE-RTS 96可靠性测试系统的评估分析验证了本文结论的正确性。The convergence performance of sequential and non-sequential Monte Carlo simulation for power grid reliability assessment is researched in detail. The probabilistic dependence relationship between computation accuracy （variance coefficient） and sample size （total years simulated or the sampling amount of states） of the two methods is analyzed. Based on the central limit theorem, the confidence interval formulae for computation accuracy and sample size of the two methods are derived, and then the conclusion is obtained：①the convergence of sequential Monte Carlo simulation depends on the ratio of standard variance of reliability index to its expected value, the smaller the ratio, the better the convergence performance; ②the convergence of non-sequential Monte Carlo simulation depends on the expected value of loss of load probability （LOLP）, the larger the value, the faster the convergence. By means of the assessment analysis in RBTS, IEEE-RTS 79 and IEEE-RT S96 test systems, the correctness of the proposed method is verified.

关 键 词：可靠性评估 蒙特卡洛仿真 收敛性 置信区间 

分 类 号：TM732[电气工程—电力系统及自动化]