|
| Diagnosis of feeder-transformer connectivity in a distribution network based on voltage first-order difference outlier correlation analysis |
| DOI：10.19783/j.cnki.pspc.240175 |
| Key Words:distribution network feeder-transformer connectivity voltage fluctuation outlier detection |
| Author Name | Affiliation | | CHEN Jinming1,2 | 1. School of Electrical Engineering, Southeast University, Nanjing 210096, China
2. Electric Power Research Institute, State
Grid Jiangsu Electric Power Co., Ltd., Nanjing 211103, China
3. College of Automation & College of Artificial
Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China | | JIANG Wei1 | 1. School of Electrical Engineering, Southeast University, Nanjing 210096, China
2. Electric Power Research Institute, State
Grid Jiangsu Electric Power Co., Ltd., Nanjing 211103, China
3. College of Automation & College of Artificial
Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China | | YUAN Yubo2 | 1. School of Electrical Engineering, Southeast University, Nanjing 210096, China
2. Electric Power Research Institute, State
Grid Jiangsu Electric Power Co., Ltd., Nanjing 211103, China
3. College of Automation & College of Artificial
Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China | | ZENG Fei2 | 1. School of Electrical Engineering, Southeast University, Nanjing 210096, China
2. Electric Power Research Institute, State
Grid Jiangsu Electric Power Co., Ltd., Nanjing 211103, China
3. College of Automation & College of Artificial
Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China | | LU Qingning2 | 1. School of Electrical Engineering, Southeast University, Nanjing 210096, China
2. Electric Power Research Institute, State
Grid Jiangsu Electric Power Co., Ltd., Nanjing 211103, China
3. College of Automation & College of Artificial
Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China | | XU Junjun3 | 1. School of Electrical Engineering, Southeast University, Nanjing 210096, China
2. Electric Power Research Institute, State
Grid Jiangsu Electric Power Co., Ltd., Nanjing 211103, China
3. College of Automation & College of Artificial
Intelligence, Nanjing University of Posts and Telecommunications, Nanjing 210023, China |
|
| Hits: 2691 |
| Download times: 474 |
| Abstract:The voltage correlation calculation and its extension method is an effective means to realize the diagnosis of feeder-transformer relationship in distribution networks. However, due to the complex operating conditions of low-voltage distribution networks, the distribution transformer voltage fluctuates abnormally in some periods, which leads to certain limitations of this type of method. For this reason, this paper expands on the voltage correlation method, utilizes the mechanism of voltage mutation feature diffusion with the trend, and proposes a feeder-transformer relationship diagnosis method based on the first-order difference separation group point correlation analysis of voltage. First, based on the voltage correlation method, the initial screening of the distribution transformers is carried out to derive the credible and abnormal distribution transformer sets. Secondly, a univariate time series streaming method is applied to carry out voltage first-order difference outlier detection, and the bus outlier labeling is realized by correlating with the substation voltage regulation events. Subsequently, the outlier points are evaluated from three dimensions of interpretability, perceptibility and distinguishability to form a sequence of bus weighted outlier points. Finally, a feeder-transformer relationship diagnosis method based on the matching degree of outlier points is proposed to recommend the feeder to which the anomalous distribution transformer belongs through a comprehensive horizontal and vertical comparison. Experimental results show that the feeder-transformer relationship diagnosis method proposed in this paper can effectively overcome the influence of complex working conditions and significantly improve the performance of the diagnosis algorithm. |
| View Full Text View/Add Comment Download reader |
|
|
|