Filatova Anna Yurievna

Faculty:
Computer Information Technologies and Automation (CITA)

Speciality:
Information Control Systems (ICS)

Theme of master's work: "«Development of the computerized subsystem of analysis and forecasting of natural gas consumption in “Donetskgorgaz” conditions»"

Scientific adviser: the professor Lazdyn Sergey Vladimirovich

Abstract

    Natural gas is today one of the basic used power resources. The epoch of practically unlimited and cheap power resources has ended with disintegration of a planned economy. It was time when their share made in the cost price of production only some percent.

    Economically expedient limit of their consumption within the limits of the developed technologies has come for each separate enterprise. The high cost factor of power resources has caused cardinal change of the attitude for the organization accurate energy calculation in the industry and other power-intensive branches (transport and residential and communal services) last years. It enables profitable decisions acceptances of management to suppliers and gas distribution company by construction models of consumption and use of predicted sizes for branch or the separate consumer.

    The urgency of the analysis and forecasting of demand modes in sections of time is doubtless, as the proved forecasts are the effective tool of forward planning and management in branch. The researches urgency is shown in application of the received results for operative and tactical management.

    Many enterprises of gas branch experts conduct researches in this area. They are such scientists: Barhudarjan I.G., Oganesjan L.A. - application of forecasting mathematical methods, models construction of natural gas consumption; Kovalenko M.V., Махотило K.V.conduct - use of neural networks at forecasting power resources use ; Ronald H. Brown, Bryan M. Marx and Georges of F.Korliss, and also S.Dzhil, Дж. Деферрари etc. Market relations between manufacturers and consumers of power resources in a new fashion put questions of power supply, normalization and increase of efficiency power resources use. The important task of work is statement and the decision of gas supply problems in interrelation with a power supply system, namely: forecasting and demand management for gas under the demand forecast and the analysis of natural gas offer. The control, forecasting and production management of energy resource are important from the point of view of the established limiting values observance for consumers power and used natural gas quantity.

     The purpose of work is modelling, the analysis and forecasting of natural gas demand. It is necessary to solve the following tasks for achievement of an object in view:

• To analyse a condition of the theory and practice of economic parameters time series modelling;
• To analyse known models of forecasting and to reveal lacks of models of the analysis and forecasting of natural gas demand time series;
• To reveal opportunities of adaptation and updating of known analysis and forecasting models, to offer model of natural gas demand forecasting in view of reliability requirements predictable values;
• To choose the most suitable methods of task in view decision of the various time periods;
• To develop architecture of a subsystem and to realize a complex of software which will allow to make a choice of rational variants forecasting models.

    Scientific novelty of work consists in development of theoretical, methodical and practical aspects of natural gaz demand analysis and forecasting which are expressed in following substantive provisions:

1. The approach to natural gas consumption analysis and forecasting which is realized at analysis and forecasting models construction depending on the forecast period is offered;
2. Construction of natural gas consumption model, allowing to consider the maximal number of the factors influencing volume and consumption character is planned.

     Use of statistical methods and the device of neural networks is supposed. Compositions often give the best result as in this case errors of various models compensate each other. Such system is necessary in practice for decision-making management at any enterprise on gas delivery or distribution company (for example, " Donetskgorgaz").

Planned results

    It is possible to carry the following results of developed subsystem work:

• the organization of storage information of natural gas consumption with an opportunity to extract necessary data;
• forecasting model construction with a taking into account maximal number of the factors influencing on natural gas consumption;
• reception as much as possible approached to valid the expected consumption, group during the specified time period which can be used at acceptance of administrative decisions.

Conclusions

    Thus, the problem of the analysis and consumption of natural gas forecasting, as one of the basic power resources has been presented in the given abstract. In work The existing methods and system of the analysis and forecasting review have been adduced in this abstract, statement of task for the further research is chosen and described.

    The further researches will be made in view of use of modern data analysis means and the specific approach to forecasting. Approach feature to forecast construction is that it is supposed to use depending from the forecasting period various methods of the task decision and their combinations with the purpose to increase forecast accuracy.

References

1. Панкратов В.С., Берман Р.Л.
   Разработка и эксплуатация АСУ газотранспортными системами. - Москва, 1982.
2. Бархударян И.Г., Оганесян Л.А.
   Современный уровень прогнозирования режимов газопотребления и их моделирование. - Ереван,1981
3. Айвазян С.А. и др.
   Прикладная статистика: основы моделирования и первичной обработки данных. - Москва, 1983.
4. Тарасенко В., Филиппов В., Сенюков В., Миденко М.
   «Система управления газовым хозяйством региона»
   Журнал «Системная интеграция. Нефтегазовая промышленность», №2, 2000г.
5. Єріна А.М.
   Статистичне моделювання та прогнозування: Навч.посібник. Київ, 2001.
6. http://www.forecsys.ru/technologies.phphttp://www.forecsys.ru Библиотека средств прогнозирования
7. http://www.statsoft.ru
   STATISTICA, анализ данных, статистический анализ, статистический анализ данных, визуализация, прогнозирование, моделирование, сегментация, исследование, аналитические решения.
8. Андрей Суслов
   Инструменты и технологии аналитической обработки данных. Обзор современных решений в области анализа данных. - "Сетевой журнал" №12,2001.
9. С.Д. Коровкин, И.А. Левенец, И.Д. Ратманова, В.А. Старых, Л.В. Щавелёв.
   Решение проблемы комплексного оперативного анализа информации хранилищ данных – 1998.
10. Коваленко М.В., Махотило К.В.
    "Нейросетевая модель прогнозирования потребления газа в жилищно-бытовом секторе"
    Вести. Нац. техн. ун-та "ХПИ", выпуск №12, 2002г.
11. Дюк В., Самойленко А.
    Конспект книги "Data Mining: учебный курс". - СПб. Изд. "Питер", 2001.
12. http://www.w3.org
    Анализ данных, нейронные сети, прогнозирование. Прогнозирование потребления электроэнергии
13. http://www.bi-grouplabs.ru
    Анализ данных, нейронные сети, прогнозирование.
14. http://www.neuroproject.ru/
    Методы прогнозирования
15. Ивахненко А.Г.
    Долгосрочное прогнозирование и управление сложными системами. - М., «Техника», 1975.
16. Грешилов А.А., Стакун В.А., Стакун А.А.
    Математические методы построения прогнозов. – М.: Радио и связь,1997.
17. http://www.gmdh.net/gmdh.htm
    Метод МГУА для социально экономического прогнозирования, математического моделирования, статистического анализа данных, аналитической оценки систем и программирования.
18. http://www.interface.ru
    Обзор современных решений в области анализа данных
19. Кречетов, Иванов.
    Продукты для интеллектуального анализа данных //ComputerWeek, Москва 1997 №14-15.

Note: Given abstract is not the final version of the author's master’s works abstract. Ending of work is planned to the end of December, 2007. You can address for the final version of the dissertation to the author: nuwa1@list.ru.

Начало