Analysis of the state of high-voltage current transformers based on gradient boosting on decision trees
This paper addresses the problem of instrument current transformers technical state assessment based on machine learning methods. The introductory parts of the paper provide a detailed analysis of modern methods and approaches for technical state assessment of high-voltage power equipment of power plants and substations as well as a review of modern software tools and the latest trends in the given field of study. Within the framework of the study, a comparative analysis of gradient boosting on decision trees and random forest algorithms was carried out for a given mathematical problem formulation. The resulting classification quality metrics of current transformers technical state assessment, Precision and Recall, are estimated to be 87.1% and 83.7%, correspondingly.
Machine Learning Application for the High-Voltage Equipment Lifecycle Forecasting
This paper presents the model of the intelligent lifecycle management system for high-voltage power network equipment. The model is validated by providing technical state estimates for 110 kV oil-filled power transformers based on the technical diagnostic data retrospective. The transformers are operated at the substations of the regional power system. In the framework of the author’s research, the adaptive system for managing and forecasting of the power equipment lifecycle based on machine learning methods was developed, where gradient boosting over decision trees is used as the core mathematical algorithm for solving the above problems: both for classification and regression ones.
The Methodology of High-Voltage Instrument Transformers Technical State Index Assessment
The paper presents a methodology for technical state assessment of high-voltage instrument current and voltage transformers based on the basic requirements, imposed by governmental regulations. The developed methodology includes the analysis of the power equipment structure, allocation of sub-systems, evaluation criteria and a list of essential parameters, technical state assessment methodology, list of the boundary values, formulation of standard recommendations corresponding to different values of technical state index. The presented approach is validated by providing technical state index calculation results along with failure statistics analysis of the regional power system.
The Application of Partial Discharge Monitoring System for Instrument Transformers: Special Issues
The paper describes the results of research and development project, within the framework of which a system of partial discharges monitoring of oil-filled current and voltage transformers was developed and investigated. The paper provides an overview of modern PD monitoring approaches and describes the scientific and technical rationale for choosing the most effective PD registration approach. The effectiveness of the considered approaches for solving the problems of high-voltage equipment technical state assessment is also analyzed and reported.
Electrical energy storage systems for increasing technical and economical characteristics of gas engine power plants
This paper contains actual testing results of electrical energy storage system based on lithium-ion technology working with gas engine generator and altering load power. As is known, gas generators are sensitive to changes in load power, obtained results show the effectiveness of the usage EESS to supply sustainability of generators in high amplitude and frequency altering load power environment. These results achieved through the ability of the developed converter and control strategy to compensate rapid deviations of the load power. In additional, in control system realized others common functionality: active power limitation, voltage and frequency control, reactive power control, power factor control and others.
А. И. Хальясмаа, В. З. Манусов
Опыт реализации комплексной системы диагностики высоковольтного оборудования. Вестник Казанского государственного энергетического университета.2020. Т. 12, № 1 (45). – С. 82–92.
В данном исследованим для реального энергетического объекта для ОРУ-110 кВ разработана и апробирована комплексная методика оценки технического состояния электрооборудования на базе дополнительных методов технического диагностирования состояния оборудования на основе инфракрасного и ультрафиолетового периодического контролей, а также метода измерения частичных разрядов. В результате исследования выявлены дефекты в различных видах оборудования с разной степенью их развития, где каждый из методов в отдельности покрывает только определенную (свою) группу дефектов, которая может быть идентифицирована только в 20-25% случаев с помощью другого метода из перечисленного списка, при этом точность идентификации дефекта при использовании всех трех методов технического диагностирования, как например, было продемонстрировано на трансформаторах тока и напряжения повышается до 90%. А применение ультрафиолетового контроля совместно с тепловизионным контролем позволяет повысить точность идентификации дефектов на 15%.
В. З. Манусов, В. М. Левин, А. И. Хальясмаа, Д. С. Ахьёев.
Применение методов искусственного интеллекта в задачах технической диагностики электрооборудования электрических систем. 2020. 446 с - (Монографии НГТУ)
В монографии рассматриваются методы искусственного интеллекта в задачах управления и оптимизации режимов электрических интеллектуальных сетей (Smart Grid) с альтернативными и возобновляемыми источниками энергии. Введены и обоснованы понятия генерирующих потребителей (холонов) и иерархической структуры в виде холархии для сетей с двусторонним потоком энергии и информации. Применены различные алгоритмы роевого интеллекта и их сравнительный анализ с методом градиентного спуска. Рассмотрены принципы Q-обучения с подкреплением.
Книга может представлять интерес для студентов, магистров и аспирантов высших учебных заведений, а также научных работников и инженеров-электриков проектных и производственных предприятий, в том числе и для широкого круга читателей.
Robust Short-Term Wind Speed Forecasting Using Adaptive Shallow Neural Networks // Problems of the Regional Energetics. 2020. № 3 (47). 68 – 79 pp.
Wind speed forecasting is necessary to integrate wind farms into power systems. This work aims to develop a machine learning model for short-term wind speed forecasting with acceptable accuracy but high robustness and the possibility of automatic online retraining. A shallow multilayer perceptron, trained only on retrospective data on wind speed, is proposed. The most significant results are combining simple neural network architecture with ReLU activation function, Adam training method developed for deep neural networks; and the automatic hy- per-parameters selection using Grid search with open upper bounds. The significance of the obtained results is that shallow neural networks using ReLU, Adam, and Grid search are practically not inferior to adaptive models in terms of tuning speed and the risk of subsequent differences in accuracy be- tween training data and data supplied during operation. At the same time, shallow neural networks make it possible to obtain more accurate forecasts, and due to their small size, they are trained quickly; and retraining can be performed automatically when new data arrives.
Manusov V., Igumnova E., Stanislav E., Nesterenko G., Matrenin P.
Comparison study of wind flow velocity short-term forecasting methods based on adaptive models and neural networks // International Journal of Advanced Science and Technology, 29(8 Special Issue), pp. 2108-2115. 2020. (Scopus Q4).
The article is devoted to the study and comparison analysis of wind flow velocity short-term forecasting methods based on adaptive models and neural networks. The research used data obtained from the Russian island. The adaptive Holt, Brown and Holt-Winters models are compared, and their best parameters are selected for the problem at hand. The paper gives the results of research on the influence of the number of hidden layers, the activation function, and the method of training a multilayer neural network on the accuracy in forecasting. The results showed a significant advantage of a rectified linear unit (ReLU) as an activation function relative to the sigmoid function and hyperbolic tangent, as well as the advantage of the Adam method over classic Gradient Descent. Adaptive methods showed MAPE 16%, and Neural networks showed 10%. Adaptive methods ensure quick tuning a task solution with a low risk of overfitting, so the accuracy between the training data and real-life accuracy will be the same. Artificial neural networks allow to obtain more accurate forecasts but require careful tuning and have the risk of overfitting as a significant accuracy decreasing in case of changing climatic conditions.
P. V. Matrenin, V. Z. Manusov, N. Khasanzoda, D. V. Antonenkov.
Application of swarm intelligence algorithms to energy management of prosumers with wind power plants // International Journal of Electrical and Computer Engineering. Vol. 10. no. 6. 2020. pp. 6172-6179. (Scopus Q2.)
The paper considers the problem of optimal control of a prosumer with a wind power plant in smart grid. It is shown that control can be performed in non-deterministic conditions due to the impossibility of accurate forecasting of the generation from renewable plants. A control model based on a priority queue of logical rules with structural-parametric optimization is applied. The optimization problem is considered from a separate prosumer, not from the entire distributed system. The solution of the optimization problem is performed by three swarm intelligence algorithms. Computational experiments were carried out for models of wind energy systems on Russky Island and Popov Island (Far East). The results obtained showed the high effectiveness of the swarm intelligence algorithms that demonstrated reliable and fast convergence to the global extreme of the optimization problem under different scenarios and parameters of prosumers. Also, we analyzed the influence of accumulator capacity on the variability of prosumers. The variability, in turn, affects the increase of the prosumer benefits from the interaction with the external global power system and neighboring prosumers.
Methodology for Analysing the Technical State and Residual Life of Overhead Transmission Lines
The paper presents an industry-validated technique for analysing the technical state of overhead power transmission lines and predicting their residual life with the ability to predict the schedule of their repairs and maintenance. The authors propose a new approach to overhead lines technical state identification based on the analysis of its components: towers, conductors and insulation nodes. Also, within the framework of the study, an indirect method for assessing the physical wear indicator of overhead power transmission lines has been developed, taking into account the previously carried out replacement of its elements and components. The feasibility of extending the life of the overhead power transmission line components based on residual life assessment methods is justified. The developed methodology is intended for energy utilities as well as engineering supervision organisations conducting overhead power transmission lines surveys. Novel recommendations are formulated to assess the physical wear of the overhead transmission lines.
Industry Experience of Developing Day-Ahead Photovoltaic Plant Forecasting System Based on Machine Learning
This article highlights the industry experience of the development and practical implementation of a short-term photovoltaic forecasting system based on machine learning methods for a real industry-scale photovoltaic power plant implemented in a Russian power system using remote data acquisition. One of the goals of the study is to improve photovoltaic power plants generation forecasting accuracy based on open-source meteorological data, which is provided in regular weather forecasts. The article scrutinizes the application of different machine learning algorithms, including Random Forest regressor, Gradient Boosting Regressor, Linear Regression and Decision Trees regression, to the remotely obtained data. As a result of the application of the aforementioned approaches together with hyperparameters, tuning and pipelining of the algorithms, the optimal structure, parameters and the application sphere of different regressors were identified for various testing samples. The mathematical model developed within the framework of the study gave the opportunity to provide robust photovoltaic energy forecasting results with mean accuracy over 92% for mostly-sunny sample days and over 83% for mostly cloudy days with different types of precipitation.
Improving Accuracy and Generalization Performance of Small-Size Recurrent Neural Networks Applied to Short-Term Load Forecasting
The load forecasting of a coal mining enterprise is a complicated problem due to the irregular technological process of mining. It is necessary to apply models that can distinguish both cyclic components and complex rules in the energy consumption data that reflect the highly volatile technological process. For such tasks, Artificial Neural Networks demonstrate advanced performance. In recent years, the effectiveness of Artificial Neural Networks has been significantly improved thanks to new state-of-the-art architectures, training methods and approaches to reduce overfitting. In this paper, the Recurrent Neural Network architecture with a small-size model was applied to the short-term load forecasting of a coal mining enterprise. A single recurrent model was developed and trained for the entire four-year operational period of the enterprise, with significant changes in the energy consumption pattern during the period. This task was challenging since it required high-level generalization performance from the model. It was shown that the accuracy and generalization properties of small-size recurrent models can be significantly improved by the proper selection of the hyper-parameters and training method. The effectiveness of the proposed approach was validated using a real-case dataset.
от 16 июля 2020 года
Программа распознавания образов технического состояния силовых высоковольтных трансформаторов на основе алгоритмов машинного обученияя
от 31 июля 2020 года
от 16 июля 2020 года
Программа распознавания образов технического состояния высоковольтных измерительных трансформаторов тока и напряжения на основе алгоритмов машинного обучения
от 26 августа 2020 года
Способ диагностики технического состояния высоковольтного оборудования открытых распределительных устройств
от 16 июля 2020 года
Программа для оперативного прогнозирования выработки электрической энергии фотоэлектрическими станциями на основе алгоритмов машинного обучения