Study on Power Transformer Faults Based on Neural Network Combined Plant Growth Simulation Algorithm

Author(s): Ying Xia*, Bo Zhou, Mengxiong Lu, Chuanqing Sun.

Journal Name: Recent Patents on Computer Science

Volume 10 , Issue 3 , 2017

Become EABM
Become Reviewer

Graphical Abstract:


Abstract:

Background: In the modern age when electric power is one of the major energies, electrical equipment and power transformer are indispensable. Power transformer is an important electric transmission and transformation equipment in electrical power system. The failure of power transformer may induce long-term power supply interruption and large economic loss. Therefore, the diagnosis and repair of broken-down power transformer are of great urgency. This study developed a patent which could improve fault diagnosis applicability and reduce transformer faults using plant growth simulation algorithm (PSGA) in combination with neural network method.

Methods: First of all, an improved PSGA model was established according to the characteristics and defects of PSGA. Secondly, the improved model was combined with neural network for the analysis of transformer faults; a back-propagation (BP) neural network fault analysis model was then established as well as given a simulation experiment.

Results: Simulation results showed that such a method could accurately diagnose transformer faults. Finally, according to internal faults of transformers, a recognition mathematical model of transformer winding was established and given an experimental simulation. The simulation results indicated that the patent was enough to satisfy the synthetic fault diagnosis of power transformers.

Conclusion: In conclusion, such patent provides an effective theoretical and experimental basis for the diagnosis and fixing of power transformer faults.

Keywords: Plant growth simulation algorithm, back-propagation neural network, power transformer, fault study, neural network, electric transmission.

Rights & PermissionsPrintExport Cite as


Article Details

VOLUME: 10
ISSUE: 3
Year: 2017
Page: [216 - 222]
Pages: 7
DOI: 10.2174/2213275910666170502150006

Article Metrics

PDF: 18
HTML: 2
EPUB: 1
PRC: 1