基于改进的重要抽样算法的输电系统可靠性评估
发布时间:2018-01-13 04:37
本文关键词:基于改进的重要抽样算法的输电系统可靠性评估 出处:《华北电力大学》2015年硕士论文 论文类型:学位论文
更多相关文章: 输电系统 可靠性评估 蒙特卡洛法 改进的重要抽样法
【摘要】:随着经济社会的发展,各行各业电力用户对供电可靠性的要求越来越高,发生停电故障给人类带来的负面影响也就越来越大,保证安全可靠供电是电力系统的重中之重。面对结构日趋复杂的输电网络,客观、准确、快速地对系统可靠供电能力进行评估,可以全面评价电网可靠性,找到系统的可靠性薄弱环节,为电力系统规划、运行和改进提供数据支持,有着重要的意义。在各种可靠性评估方法中,蒙特卡洛法的应用最为广泛,尤其适用于大型电力系统的可靠性评估计算,因为其抽样次数与系统的规模无关。然而该方法最大的问题在于,要想获得较高精度的可靠性指标需要进行大量的抽样计算,这种计算精度和计算速度之间的矛盾很大程度上限制了可靠性评估在实际运用时的收敛速度。针对蒙特卡洛法的这一问题,学者们提出了各种旨在通过降低方差加快收敛速度的抽样算法,其中被认为最有效的是重要抽样法。该方法通过把系统内元件的强迫停运率扩大k倍,实现对元件失效情况的放大,提高抽样效率。然而该方法通常被用于元件故障率相对较高的发电系统,在现实输电系统中,元件故障率较低、网络可靠性较好,通过迭代求取最优乘子k的算法收敛效果并不明显,限制了该方法在输电网络工程中的应用。本文在对以上问题进行分析的基础上,提出了适用于实际输电系统的改进重要抽样法:在重要抽样的基础上对最优乘子k的取值和修正方式进行改进,将抽样分为预抽样和正式抽样两步,通过对小样本的预抽样确定k的取值,然后正式抽样计算系统的可靠性指标。分别应用常规蒙特卡洛法、重要抽样法和改进的重要抽样法,对可靠性测试系统RBTS和IEEE-RTS的输电部分进行可靠性计算,分析比对计算结果和计算精度,验证了改进算法的准确性和高效性。为了测试改进算法在工程应用中的适用性,以某地区实际电网M为研究对象,分别用上述三种方法计算该电网整体以及各节点的可靠性指标,对比计算结果和精度,证明改进算法在实际工程中应用良好。最后,以该算法计算结果为基础,结合M电网拓扑结构和元件参数特点,找出该地区电网可靠性薄弱的环节,提出改进可靠性的方案,并验证了这些方案的有效性。
[Abstract]:With the development of economy and society, all walks of life power users on power supply reliability of the increasingly high demand, the negative impact of power outage brought to mankind is more and more big, to ensure safe and reliable power supply is the priority among priorities of the power system. In the face of the transmission network, the increasingly complex structure of objective, accurate, fast and reliable to evaluate power supply system, a comprehensive evaluation of reliability of power grid, find the weakness of system reliability for power system planning, operation and provide data support to improve, has important significance. In the evaluation of various methods of reliability, Monte Carlo method is the most widely used, especially suitable for reliability assessment of power system calculation, because the sampling the number and the scale of the system. However, this method is not the biggest problem, in order to obtain the reliability index of high precision required a lot of sampling This calculation, the contradiction between computational accuracy and computational speed greatly limits the reliability evaluation of the convergence speed in practical application. In order to solve the problem of Monte C Lofa, scholars have put forward various aims to reduce the variance of sampling algorithm to speed up the convergence rate, which is considered the most effective is the important sampling method. By this method the the system components of the forced outage rate of expansion of K times, to achieve amplification of component failure, improve the sampling efficiency. However, this method is usually used for power system component failure rate is relatively high, in the real transmission system, component failure rate is low, the network reliability, through the iterative convergence results from optimal multiplier K is not obvious, which restricts the application of this method in transmission network engineering. Based on above analysis, put forward for the actual transmission The improvement of the system of importance sampling method: Based on importance sampling on optimal value K and correction method was improved, the sample is divided into pre sampling and formal sampling two step, to determine the value of K by pre sampling of small samples, then the formal sampling calculation system reliability index using conventional Monte Carlo method respectively. And the importance sampling method and improved important sampling method, the transmission part of the reliability testing system of RBTS and IEEE-RTS for the calculation of reliability, analysis and comparison of calculation results and calculation accuracy, verify the accuracy and efficiency of the improved algorithm. In order to improve the applicability of the algorithm in the engineering application of the test in the practical power system M as the research object the whole power grid were calculated, and the reliability index of each node by the above three methods, comparing the results of calculation and accuracy in practical engineering shows that the improved algorithm should be used. Finally, based on the calculation results of the algorithm, combined with the characteristics of M power network topology and component parameters, we find out the weak links of the power grid in this area, and put forward the scheme of improving reliability, and verify the effectiveness of these schemes.
【学位授予单位】:华北电力大学
【学位级别】:硕士
【学位授予年份】:2015
【分类号】:TM732
【参考文献】
相关期刊论文 前5条
1 薛禹胜;综合防御由偶然故障演化为电力灾难——北美“8·14”大停电的警示[J];电力系统自动化;2003年18期
2 别朝红,王锡凡;蒙特卡洛法在评估电力系统可靠性中的应用[J];电力系统自动化;1997年06期
3 何国锋,谭震宇;采用等分散抽样法的电力系统概率仿真[J];电力自动化设备;2004年07期
4 宋晓通,谭震宇;改进的重要抽样法在电力系统可靠性评估中的应用[J];电网技术;2005年13期
5 别朝红,王建华,王锡凡;一种减小蒙特卡洛模拟方差的新方法[J];中国电力;1999年12期
,本文编号:1417526
本文链接:https://www.wllwen.com/kejilunwen/dianlilw/1417526.html
教材专著