采用SOM的发输电系统可靠性综合评估快速算法  

A fast method for comprehensive reliability assessment of power generation and transmission system using SOM

作　　者：张瑞华[1] 宋云亭[2] 陈曦 

机构地区：[1]中国科学院电工研究所,北京100080 [2]中国电力科学研究院,北京100085 [3]深圳市许继昌达电网控制设备有限公司,广东深圳518040

出　　处：《中国电力》2005年第6期12-16,共5页Electric Power

摘　　要：针对目前发输电系统可靠性综合评估算法中蒙特卡罗仿真效率不高的问题,提出了一种利用模糊自组织映射(SOM)神经网络进行状态筛选的可靠性快速评估算法。SOM神经网络具有拓扑特征保持性质,经过训练后的SOM网络具有模式聚类能力,即不同的运行模式被映射到输出平面的相应位置,因而能对暂态稳定性进行判别。由于SOM具有学习训练时间短的优点,因而特别适于可靠性评估。将模糊SOM网络和序贯蒙特卡罗仿真结合在一起对系统状态进行筛选,首先将明显不失稳的无效系统状态筛掉,大大减少了需要完全评估的系统状态数,从而显著提高了综合评估的效率。通过对IEEE-RTS标准算例系统的计算,结果表明所提算法快速有效,具有良好的应用价值。For overcoming the problem that Monte-Carlo sampling technique normally used in power system probabilistic simulation has low efficiency, this paper proposed a fast method using fuzzy SOM neural network as system states filter to evaluate the reliability of bulk power system for the first time. SOM learns to recognize groups of similar input vectors in such a way that neurons physically near each other in the neuron layer respond to similar input vectors. After training, the weight vectors are self-organized and represent prototypes of the input vectors. Input vectors, which are similar in their original space, share similarity in the map. SOM can be used to label the security of power system states quickly. SOM is especially appropriate to estimate the reliability of power system because it's training time shorter than other neural network. Invalid system states can be filtered by fuzzy SOM neural network, it reduces significantly the number of system states that should be evaluated. Fuzzy SOM neural network combined with sequential Monte-Carlo simulation can greatly reduce the calculation burden of reliability evaluation of bulk power system. A case study of the IEEE-RTS test system is presented to demonstrate the effectiveness and feasibility of this developed algorithm.

关 键 词：SOM网络 发输电系统 可靠性评估 状态筛选 蒙特卡罗仿真 

分 类 号：TM732[电气工程—电力系统及自动化]