毕业论文

打赏
当前位置: 毕业论文 > 自动化 >

GSA间隙统计算法的电力系统不良数据辨识研究

时间:2021-04-09 22:54来源:毕业论文
利用UCI标准数据库的IRIS数据集对GSA与肘形判据进行了仿真实验,同时论文还利用江苏电网局部实测数据对肘形判据的实用性进行了验证,结果表明:肘形判据具有更好的辨识性能和计算

摘要电力系统的不良数据将直接影响对电力系统的状态估计,进而威胁电力系统安全稳定运行。

本文分析了应用于电力系统不良数据辨识中的传统GSA辨识方法和肘形判据理论。GSA方法基于数理统计原理,可自动确定最佳聚类个数,但却容易陷入局部最优解和受到样本容量大小的影响,使得辨识性能下降。肘形判据理论通过分析待检数据集的聚类离散度与聚类个数间的关系,利用聚类离散度来计算肘形折角值,并以最小的肘形折角值来作为最佳聚类个数的判断依据,可有效克服传统GSA方法的缺点。本文利用UCI标准数据库的IRIS数据集对GSA与肘形判据进行了仿真实验,同时论文还利用江苏电网局部实测数据对肘形判据的实用性进行了验证,结果表明:肘形判据具有更好的辨识性能和计算速度,在电力系统不良数据辨识方面具有独到的优势。65198

毕业论文关键词  间隙统计算法,不良数据辨识,肘形判据,电力系统

毕业设计说明书(论文)外文摘要

Title Bad Data Identification of Power System Research Based on GSA

Abstract

Bad data of power system influences the power system state estimation directly, furthermore the safety and stability of power system.

In this paper, traditional GSA algorithm and elbow judgment theory applied in bad date identification of the power system were studied. GSA method is based on principle of statistics, which can quantify the optimum clusters automatically, but sometimes it could be influenced by the local optimal solution and the sample capacity, which weakens the identification performance. Elbow judgment can analyze the relationship between the degree of clustering dispersion and clustering numbers of data set to be tested, and calculate the elbow angle by the degree of clustering dispersion, and then the optimal clustering number can be obtained by the least elbow angle. IRIS data set of UCI standard database was used to validate the GSA and elbow judgment, and the measured data of Jiangsu power network is also used to validate the practicability of the elbow judgment. The results show that, the elbow judgment has a better identification performance and calculation speed, which bings a great advantage in the bad data identification of power system.

Keywords  gap statistic algorithm, bad data identification,  elbow judgment, power system

目录

毕业设计说明书(论文)中文摘要 2

1 绪论 5

1.1 课题的研究意义 5

1.2 国内外的研究现状 5

1.3 本文的主要研究工作 11

2 传统电力系统不良数据辨识方法 13

2.1 故障数据与不良数据的区别 13

2.2 基于聚类算法的不良数据辨识 14

2.3 基于状态估计的不良数据辨识 22

2.4 本章小结 25

3 间隙统计算法理论 26

3.1 传统间隙统计算法 26

3.2 传统GSA算法的实例仿真 29

3.3本章小结 34

4 基于GSA的肘型判据理论 36

4.1 肘型判据理论 36

4.2 肘型判据实例仿真 40

4.3 本章小结 48

GSA间隙统计算法的电力系统不良数据辨识研究:http://www.751com.cn/zidonghua/lunwen_72710.html
------分隔线----------------------------
推荐内容