基于遗传LM算法的分布式电网异常负荷自动识别方法
Automatic Identification Method for Abnormal Load in Distributed Power Grid Based on Genetic LM Algorithm
朱先茂 1于良 2王超3
作者信息
- 1. 济南莱电新源电力建设有限公司,山东 济南 271100
- 2. 国网莱芜供电公司,山东 济南 271100
- 3. 山东东瑞规划设计研究院有限公司,山东 菏泽 274300
- 折叠
摘要
传统分布式电网异常负荷自动识别方法直接对电网负荷数据实施分类,未进行预处理,识别错误率较高.提出基于遗传LM算法的分布式电网异常负荷自动识别方法.首先,预处理采集到的分布式电网异常负荷数据;其次,通过遗传算法生成初始种群,其中,每个个体对应一个电网负载数据、异常负荷特征的参数组合;然后,利用LM算法的局部优化特性对电网数据参数组合进行局部优化,对遗传算法进行改进,计算其适应度,得到准确的电网负荷数据异常分类结果;最后,进行电力负荷异常值的识别.结果表明,该研究方法识别错误率更低,具有实用性.
Abstract
The traditional automatic identification method for abnormal loads in distributed power grids directly classifies the load data without preprocessing,resulting in a high recognition error rate.Propose a distributed power grid abnormal load automatic identification method based on genetic LM algorithm.Firstly,preprocess the collected abnormal load data of the distributed power grid.Secondly,an initial population is generated through genetic algorithm,where each individual corresponds to a parameter combination of power grid load data and abnormal load characteristics.Then,using the local optimization characteristics of the LM algorithm to optimize the combination of power grid data parameters,improving the genetic algorithm,calculating its fitness,and obtaining accurate abnormal classification results of power grid load data.Finally,identify abnormal values of power load.The results indicate that this research method has a lower recognition error rate and practicality.
关键词
遗传LM算法/分布式电网/异常负荷Key words
genetic LM algorithm/distributed power grid/abnormal load引用本文复制引用
出版年
2024
NETL
NSTL
万方数据