基于蒙特卡洛法和最小二乘支持向量机的复杂电力系统可靠性评估  

Reliability Evaluation of Complex Power System Based on Monte Carlo Simulation and Least Squares Support Vector Machine

作　　者：王景辰[1] 李孝全[1] 杨洋[2] 林茂[1] 

机构地区：[1]空军工程大学防空反导学院,西安710051 [2]长安大学信息工程学院,西安710064

出　　处：《华东电力》2013年第5期1001-1004,共4页East China Electric Power

摘　　要：针对复杂电力系统可靠性评估效率低和计算时间长等问题,提出了一种将蒙特卡洛法(MCS)和最小二乘支持向量机(LSSVM)相结合的方法。该方法采用非序贯蒙特卡洛法抽取系统状态,利用LSSVM分类模型对抽取的系统状态进行分类,对分类后的故障状态进行可靠性指标计算,从而避免了计算正常运行状态,节省了评估时间。将该方法应用于可靠性测试系统IEEE-RTS-79中,仿真结果表明该方法在保证评估精度的基础上,大大提高了计算速度。This paper combines Monte Carlo simulation with the least squares support vector machine （ LSSVM）, to re- duce the inefficiency and long computational time in the reliability evaluation of complex power system. LSSVM is used to classify the power system states sampled by the Monte Carlo simulation. These classified failure states are then calculated to obtain reliability indices. As a result, the computing time for normal operation state is eliminated, reduc- ing the assessment time. The proposed hybrid method has been applied to the IEEE Reliability Test System （ IEEE- RTS-79） , and simulation results demonstrate the proposed MCS-LSSVM based hybrid method has an excellent per- formance in both efficiency and computational time in evaluating the complex power system reliability.

关 键 词：电力系统 可靠性评估 蒙特卡洛法 最小二乘支持向量机 K-平均聚类 

分 类 号：TM712[电气工程—电力系统及自动化]