Розробка алгоритму для побудови прототипу експертної системи діагностики аварійних ситуацій мережі
Loading...
Date
2021
Journal Title
Journal ISSN
Volume Title
Publisher
НМетАУ, Дніпро
Abstract
UKR: Сучасні електроенергетичні системи відносяться до класу великих людино-машинних систем кібернетичного типу, тому розробка експертної системи діагностики аварійних ситуацій є актуальною. В роботі було розроблено модифікований алгоритм Дейкстри. Мета модифікації полягає в підвищенні ефективності запропонованого рішення, можливість роботи в зваженому графі і виключення помилок під час пошуку альтернативного шляху.
RUS: Используя экспертную систему, можно уменьшить количество информации, которое нужно учитывать диспетчеру и ускорить процесс принятия решения. Также правильно построенная экспертная система позволит уменьшить количество потенциальных ошибок в принятых решениях. В рамках данной работы было разработано модифицированной алгоритм Дейкстри. Модифицированный алгоритм показал большую эффективность по сравнению с другими алгоритмами, которые рассматривались, потому на базе его в дальнейшем реализован прототип экспертной системы.
ENG: Modern electric power systems belong to the class of large human-machine systems of the cybernetic type that is why development of consulting model of diagnostics of emergency situations is actual. In the conditions of operative management it is related the modes of grids to hard limitation of time for a decision-making, and during prognostication of the perspective modes and planning of development of grids - with a vagueness and ambiguousness of initial information, for example, of data about the perspective loading of grid. Using a consulting model, it is possible to decrease information content it needs to take into account that to the controller and accelerate a decision-making process. Also the correctly built consulting model will allow to decrease the amount of potential errors in made decision. The up-diffused electric network after the structure reminds count. An example of a task that needs to be decoupled using an automated system would be the selection of a route for reconnecting the power grid to resume the system. In our view, more effective for this task there will be the use algorithm of round of count. Dijkstree's algorithm is a solution to a problem in general form, which will not always be relevant for a particular case. As part of this work, a modified Dijkstree algorithm was developed. The aim of modification consists in the increase of efficiency of an offer solution, possibility of work in the self-weighted count and exception of errors during the search of alternative way. This result is achieved by memorizing all the connecting nodes and calculating the minimum routes to all vertices, taking into account the lines that are in a state of failure. Findings. The choice of an algorithm to achieve this goal is key. The speed of the program, the correctness of the chosen route, the speed of repair and renovation work depends on the correctly selected algorithm. The modified algorithm showed large efficiency as compared to other algorithms that was examined, that is why the prototype of consulting model is realized in future on the base of him.
RUS: Используя экспертную систему, можно уменьшить количество информации, которое нужно учитывать диспетчеру и ускорить процесс принятия решения. Также правильно построенная экспертная система позволит уменьшить количество потенциальных ошибок в принятых решениях. В рамках данной работы было разработано модифицированной алгоритм Дейкстри. Модифицированный алгоритм показал большую эффективность по сравнению с другими алгоритмами, которые рассматривались, потому на базе его в дальнейшем реализован прототип экспертной системы.
ENG: Modern electric power systems belong to the class of large human-machine systems of the cybernetic type that is why development of consulting model of diagnostics of emergency situations is actual. In the conditions of operative management it is related the modes of grids to hard limitation of time for a decision-making, and during prognostication of the perspective modes and planning of development of grids - with a vagueness and ambiguousness of initial information, for example, of data about the perspective loading of grid. Using a consulting model, it is possible to decrease information content it needs to take into account that to the controller and accelerate a decision-making process. Also the correctly built consulting model will allow to decrease the amount of potential errors in made decision. The up-diffused electric network after the structure reminds count. An example of a task that needs to be decoupled using an automated system would be the selection of a route for reconnecting the power grid to resume the system. In our view, more effective for this task there will be the use algorithm of round of count. Dijkstree's algorithm is a solution to a problem in general form, which will not always be relevant for a particular case. As part of this work, a modified Dijkstree algorithm was developed. The aim of modification consists in the increase of efficiency of an offer solution, possibility of work in the self-weighted count and exception of errors during the search of alternative way. This result is achieved by memorizing all the connecting nodes and calculating the minimum routes to all vertices, taking into account the lines that are in a state of failure. Findings. The choice of an algorithm to achieve this goal is key. The speed of the program, the correctness of the chosen route, the speed of repair and renovation work depends on the correctly selected algorithm. The modified algorithm showed large efficiency as compared to other algorithms that was examined, that is why the prototype of consulting model is realized in future on the base of him.
Description
Keywords
експертна система, модифікований алгоритм, граф станів, електроенергетичні системи, КІТС
Citation
Глоба Н. І., Дмитрієва І. С. Розробка алгоритму для побудови прототипу експертної системи діагностики аварійних ситуацій мережі. Системні технології. Дніпро, 2021. Т. 2. № 133. С. 128–134. DOI: 10.34185/1562-9945-2-133-2021-14.