NEURAL NETWORK CLASSIFIER OF ENERGY FACILITIES OPERATING MODES AND ITS RECOGNITION ABILITY ASSESSMENT AT DIFFERENT NUMBER OF PRECEDENTS

Yuri A. Dementiy, Aleksandr N. Maslov

DOI: 10.47026/1810-1909-2021-3-45-52

Key words

machine learning, neural network, classification of object operation modes, recognizing ability.

Abstract

Classical algorithms of relay protection construction do not use all available information base and therefore cannot provide the highest possible sensitivity with guaranteed selectivity. These algorithms, as a rule, concentrate different information, as a result of which it is partially lost. For example, the resistance relay operates with complex resistance, that is, two real parameters, although two complex variables – voltage and current – are used to calculate the complex resistance. This paper shows the solution to the problem of classification of power line operating modes using a neural network algorithm. The simplest neural network, a perceptron, is a universal classifier, since a convergence theorem has been proved for it, showing that if a classification exists, a perceptron of sufficient complexity is able to describe it. The statistical and geometrical interpretations of various algorithms are discussed. The dependence of the quality of the classifier’s work on the distribution of precedents in the training sample, on which the training is based, as well as on the structure and parameters of the neural network, is shown. The recognition ability of the neural network classifier, i.e. the ability to distinguish short circuits within the protected zone from short circuits outside the protected zone at different number of precedents in the training sample, is evaluated. The limits of applicability of such algorithms to the task of classification of object operation modes in electric power industry are shown and recommendations for their practical application are formulated. The results obtained indicate the need to develop methods for training classifiers that are based on a source of informative precedents in the form of a simulation model of the object.

References

1. Dementiy Yu.A. Aktivnoe obuchenie klassifikatora rezhimov raboty ob”ekta s ispol’zovaniem imitacionnoj modeli [Active training classifier operating modes of an object using a simulation model]. In: Sovremennye tendentsii razvitiya tsifrovykh sistem releinoi zashchity i avtomatiki: materialy nauch.-tekhn. konf. molodykh spetsialistov foruma «RELAVEKSPO-2021» [Proc. of Sci. Conf. «Modern trends in the development of digital relay protection and automation systems»]. Cheboksary, Chuvash State University Publ, 2021, pp. 157–162.
2. Dementiy Yu.A., Maslov A.N., Nikolaev K.P. Neirosetevaya klassifikatsiya rezhimov [Neural classification of modes]. In: Sovremennye tendentsii razvitiya tsifrovykh sistem releinoi zashchity i avtomatiki: materialy nauch.-tekhn. konf. molodykh spetsialistov foruma «RELAVEKSPO-2021» [Proc. of Sci. Conf. «Modern trends in the development of digital relay protection and automation systems»]. Cheboksary, Chuvash State University Publ, 2021, pp. 147–152.
3. Haykin S. Neural Networks: A Comprehensive Foundation Subsequent Edition. New York, Prentice Hall PTR, 1998, 842 p.
4. Kosheev M.I., Slavutskii A.L., Slavutskii L.A. Elementarnyi perseptron kak instrument analiza perekhodnykh protsessov [Elementary perceptron as a tool for the transients analyzing]. Vestnik Chuvashskogo Universiteta, 2020, no. 3, pp. 84–93.
5. Koshcheev M.I., Slavutskii A.L., Slavutskii L.A. Prostyye neyrosetevyye algoritmy dlya volnovogo metoda opredeleniya mesta povrezhdeniya elektroseti [Simple neural network algorithms for the wave method of fault location in power networks]. Vestnik Chuvashskogo Universiteta, 2019, no. 3, pp. 110–118.
6. Kruglov V.V., Borisov V.V. Iskusstvennye neironnye seti. Teoriya i praktika [Neural networks. Theory and practice]. Moscow, Goryachaya liniya Telekom Publ., 2001, 382 p.
7. Martynov M.V. Issledovanie i razrabotka obuchaemykh modulei mikroprotsessornykh zashchit linii elektroperedachi: avtoref. dis. … cand. techn. nauk [Research and development of trainable modules for microprocessor protection of power lines: Abstract of the Cand. Diss.]. Cheboksary, 2014, 24 p.
8. Scikit-learn: Machine Learning in Python, Pedregosaet al., JMLR 12, 2011, pp. 2825–2830.

Information about the authors

Yuri A. Dementiy – Candidate of Technical Sciences, Head of the Group, Relematika LLC, Russia, Cheboksary (dementiy.yu.a@gmail.com).

Aleksandr N. Maslov – Candidate of Technical Sciences, Head of Sector, Relematika LLC, Russia, Cheboksary (maslov_an@relematika.ru).

For citations

Dementiy Yu.A., Maslov A.N. NEURAL NETWORK CLASSIFIER OF ENERGY FACILITIES OPERATING MODES AND ITS RECOGNITION ABILITY ASSESSMENT AT DIFFERENT NUMBER OF PRECEDENTS. Vestnik Chuvashskogo universiteta, 2021, no. 3, pp. 45–52. DOI: 10.47026/1810-1909-2021-3-45-52 (in Russian).

Download the full article