Open AccessDevelopment of Levenberg-Marquardt theoretical approach for electric networks
Author(s) -
Alexey Mikhaylov,
S. I. Tarakanov
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1515/5/052006
Subject(s) - levenberg–marquardt algorithm , artificial neural network , computer science , ranking (information retrieval) , artificial intelligence , measure (data warehouse) , basis (linear algebra) , operator (biology) , machine learning , data mining , mathematics , biochemistry , chemistry , geometry , repressor , transcription factor , gene
The algorithm for artificial neural networks is presented for the optimal distribution of tasks in electric networks in an automatic mode without operator participation. The article presents the artificial neural networks algorithm based on Levenberg-Marquardt approach that implements the specified task, as well as substantiation of its characteristics. It is proposed to use the technology of artificial neural networks (ANN), which on the basis of the developed multi-criteria evaluation electricity system of ARES allows ranking. The ANN architecture with Levenberg-Marquardt algorithm of weights optimization and their efficiency is estimated. As indicators of efficiency, the F-measure and the percentage of correctly made decisions (accuracy) were chosen for optimal network parameters. The obtained ANN was successfully tested.