Investigation on Membership Function Design for Fault Classification based on Fuzzy Logic
Mohnlein N, Schindler J, Jäger J (2018)
Publication Type: Conference contribution, Conference Contribution
Publication year: 2018
Pages Range: 59-64
Conference Proceedings Title: Power and Energy Student Summit
Event location: Kaiserslautern
Abstract
In this paper, a fuzzy logic based fault classification in transmission lines is analyzed. For this, a fuzzy inference system relying only on current measurements from the literature is used. It is able to classify faults correctly in most cases, but still shows weaknesses for extreme line loadings and fault parameters. Besides changing the fuzzy logic operators, mainly the design of the membership functions is thoroughly analyzed. The impact of different line configurations like single and double lines as well as the impact of line transposition is discussed. Value ranges of fuzzy system input parameters are analyzed to optimize membership functions. Triangle and Gaussian type functions are studied and their classification performance is compared to mere boolean logic. Additionally, a universal fuzzy system is designed for optimal classification of all tested fault cases, as opposed to systems specifically adapted to a certain configuration. In total, the paper provides valuable insights into the functioning and design of fuzzy logic based fault classification.
Authors with CRIS profile
How to cite
APA:
Mohnlein, N., Schindler, J., & Jäger, J. (2018). Investigation on Membership Function Design for Fault Classification based on Fuzzy Logic. In Power and Energy Student Summit (pp. 59-64). Kaiserslautern, DE.
MLA:
Mohnlein, Nicholas, Jakob Schindler, and Johann Jäger. "Investigation on Membership Function Design for Fault Classification based on Fuzzy Logic." Proceedings of the IEEE PESS Power and Energy Student Summit, Kaiserslautern 2018. 59-64.
BibTeX: Download