Application of grey theory in pollution prediction on insulator surface in power systems
| Autor: | Zhijin Zhang, Xun Li, Yushen He, Xingliang Jiang, Xinhan Qiao |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Pollution
media_common.quotation_subject General Engineering Building model 020101 civil engineering Insulator (electricity) 02 engineering and technology Wind speed 0201 civil engineering Electric power system 020303 mechanical engineering & transports 0203 mechanical engineering Approximation error Statistics Arc flash Environmental science General Materials Science Salt deposit density media_common |
| Zdroj: | Engineering Failure Analysis. 106:104153 |
| ISSN: | 1350-6307 |
| Popis: | Pollution-induced flashover is critically affecting the safety operation of the power system so as to puts social production at risk. However, few literatures have predicted insulator pollution levels. Therefore, in this paper grey theory is introduced and GM (1, N) method for the prediction of pollution on insulator surface was proposed. Further, a pollution test lasting for 9 months was conducted to build the GM (1, N) model. The test of another 4 months was also conducted to contrast the test results (ΔESDD (Equivalent salt deposit density)) and predicted results. The research results have shown that: GM (1, N) was applicable to predict the pollution on insulators at the cases of poor information with less data for building model. Besides, considering more environmental factors is conducive to improving the prediction accuracy of the GM (1, N) model. However, wind speed that varied very little during the 13 months' nature contamination periods in Chongqing can be neglected, which may improve the accuracy of the model (with its relative error decreasing from 7.9% to 6.3%). Next, the application scenario of the model is proposed. In addition, GM (1, N) method has potential applications in other field, and the prediction accuracy of GM (1, N) model can also be improved by increasing or decreasing the number of influencing variables as well as the number of samples. The proposed GM (1, N) method provides guidance for anti-pollution work in power systems, which promotes the reduction of flashover accidents and maintain social production safety. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect