One of the important problems of nowadays humanity is the accurate weather forecast.
Often, the weather forecast is carried out in order to take into account possible adverse and
devastating effects of human activity and it is usually performed with a certain degree of confidence. In order to store initial data of observed weather indexes it is necessary to create one large or several relatively smaller, distributed on a territorial or other basis databases to collect, process information and proper implementation of the forecast.
Generally, it’s allows to react to forecast in time and deal in an appropriate way.
Using of meteorological information gives especially noticeable economic effect in such an areas as aviation,
building industry, fishing, power engineering, shipping and agriculture.
From a scientific point of view, the prediction of weather is one of the most difficult tasks of aerophysics. Improvement of computer equipment allows us to implement mathematical approaches and methods of the atmospheric researches.
Each of them allows to predict weather conditions in a varying degree.
By this means, the theme and the results of research of weather parameters dynamic prediction are determined by the solution of the observed meteorological parameters prediction problem.
The purpose of master's work is to develop the final mathematical model to predict the dynamics
of meteorological parameters in the form of a software product that includes a number of modern methods for pretreatment and forecasting of meteorological data.
The object of the study are time series of observed meteorological parameters: temperature, humidity, pressure, wind speed.
Subjects of research are the methods of preliminary preparation and forecasting data, quality and range of the forecast results.
So, master's work aims are:
1. Examine existing methods of pretreatment and forecasting time series, and introduce them into existing prognostic complex;
2. Perform a review of existing methods of forecasting;
3. Develop the finite mathematical model of the meteorological parameters dynamics;
4. To test the mathematical models;
5. Implement the proposed program model;
6. Provide an information protection program of the mathematical model.
Scientific significance of this work is using new approaches to solve the forecasting problem of the meteorological series dynamics, in particular the use of:
- Interpolation methods in the first phase of work for the initial training time series;
- Methods to determine the smallest dimension of the model, such as the method of false neighbors, "Caterpillar", which ensures the uniqueness of prediction;
- Eglays method to form the predictive models.
The results of the implementation approaches used in this work are planned to compare to the results obtained by implementation the prediction using a predictive complex.
The influence of these phenomena of outcomes prediction is quite significant, as a result based on known methods the theses of practical value of master's work are formulated.
The methods of processing time series were developed and proposed, the proposed methods of forecasting are aimed ultimately to improve the quality of the forecast. The practical value consists in solving the problem, that has a practical focus, namely: to increase the accuracy of the forecast.
1. Searching for presence of voids and filling (in the case of detection).
2. Analysis, identification and correction of anomalous values.
3. The final view, antialiasing.
In order to solve this problem different methods of assessment are used: false neighbors, the main components, Grassberger - Prokachchia, a method well-adjusted basis. We use the method of false neighbors to determine the smallest dimension of a model which provides an unequivocal prediction. The method of the "Caterpillar" is a procedure for transforming one-dimensional time series into a multi-dimensional.
At this stage, the forming of predictive models are done, using the Eglays method (for getting an evolution operator), the utility value of which is needed to synthesize the equation of regression.
1. Suitability of the model. The conclusion about the suitability of the model to predict values is based on the previous (paragraph 2 of Phase 4). The mean square error calculates, mean square error as well as minimum and maximum are determined. If the error is within acceptable limits, it forwards to step 4 step 4, else - goes to the second stage.
2. Use of the model. The dynamics of observed meteorological parameter’s time series (calculated according to the operator of evolution) with a predetermined lead time (number of time periods which are predicting) are forecasting.
In common implemenlation, developed finite model works such as follows:
The development of software that implements a finite mathematical model of the meteorological parameters dynamics in the form of additional functionality for developed prognostic complex, to increase the reliability of the forecast. Series of computational experiments conduct to compare the developed model to the models present in the prognostic complex.
Results of the first phase of the work were presented at the II International Scientific Conference of Students and Young Scientists "Information control systems and computer monitoring in 2011" April 12-14, 2011, and published in a composite book (p. 119-123) [12].
The task of additions and improvements developed predictive complex was adopted as the main task of developing the finite mathematical model of the meteorological parameters dynamics, as well as adapting the model developed for laboratory studies. An analysis of the literature about methods of pretreatment and time series prediction was done as a result. The first phase of the mathematical model was implemented, as well as a series of computational experiments on the preliminary time series preparation for further work with them. The results of the paper were presented at the II International Conference of Students and Young Scientists (material: Volume 1, p.119-123) [12]. The proposed model has the ability to predict weather parameters, using the methods of model equations reconstruction based on the analysis of time series for numerical weather prediction to improve the quality and lead time to forecast.
The work is in development for the day of writing this auto summary. Total willingness of work - in December 2011. The paper and relevant materials can be obtained either form author or its scientific supervisor after that date.
1. Эглайс В.О. Аппроксимация табличных данных многомерным уравнением регрессии. –
Вопросы динамики и прочности: Рига, 1981, Вып. 39. – с.120-125.
2. Википедия: свободная энциклопедия. [Электронный ресурс]:
http://ru.wikipedia.org/wiki/%D0%9C%D0%B5%D1%82%D0%BE%D0%B4_%D0%B3%D0%BB%D0%B0%D0%B2%D0%BD%D1%8B%D1%85%D0%BA%D0%BE%D0%BC%D0%BF%D0%BE%D0%BD%D0%B5%D0%BD%D1%82 (15.05.2011).
3. Татаренко С.И. Методы и модели анализа временных рядов: Метод. указания к лабораторным работам. Тамбов: Тамбовский государственный технический университет, 2008. – 19с.
4. Википедия: свободная энциклопедия. [Электронный ресурс]:
http://ru.wikipedia.org/wiki/Линейная_интерполяция (15.05.2011).
5. Гриценко А.В. Реконструкция уравнений и прогнозирование метеопараметров по их временным рядам. – Донецк, ДонНТУ, 2010. – 149с.
6. Сивяков А.С. Построение прогностического комплекса и внедрение его в электронную сеть университета. – Донецк, ДонНТУ, 2010. [Электронный ресурс]:
http://www.masters.donntu.ru/2010/fknt/sivyakov/diss/index.htm (28.04.2011).
7. Онлайн Энциклопедия Кругосвет ПРОЯВЛЕНИЯ СОЛНЕЧНОЙ АКТИВНОСТИ НА ЗЕМЛЕ. [Электронный ресурс]:
http://www.krugosvet.ru/enc/nauka_i_tehnika/astronomiya/PROYAVLENIYA_SOLNECHNO_AKTIVNOSTI_NA_ZEMLE.html?page=0,1 (10.05.2011).
8. Комиссия по атмосферным наукам. Тринадцатая сессия ОСЛО, НОРВЕГИЯ, 12-20 февраля 2002г. КАН-XIII/PINK 8 (19.02.2002 г.). [Электронный ресурс]:
http://meteoinfo.ru/climate-WMO (16.05.2011).
9. Takens F. Detecting strange attractors in turbulence // Lec. NotesMath., 1981. V. 898. 366-381.
10. Орлов А.И. Статистические методы прогнозирования. Источник: Международная Академия исследований будущего (IFRA). Российское отделение – Академия прогнозирования. [Электронный ресурс]:
http://www.maib.ru/prognostication/methodsandmodels/methodsandmodels_45.html (14.05.2011).
11. Википедия: свободная энциклопедия. [Электронный ресурс]:
http://ru.wikipedia.org/wiki/%D0%9F%D1%80%D0%BE%D0%B3%D0%BD%D0%BE%D0%B7 (14.05.2011).
12. Носов С.С., Беловодский В.Н. Предварительная обработка временных метеорядов: методы, эксперименты, результаты. Источник: Информационные управляющие системы и компьютерный мониторинг – 2011 / Материалы II всеукраинской научно-технической конференции студентов, аспирантов и молодых учёных. – Донецк, ДонНТУ – 2011, с. 119-123.
13. Сивяков А.С., Трофименко Е.С., Беловодский В.Н. Регрессионное уравнение Эглайса: метод, реализация, эксперимент. Источник: Компьютерный мониторинг и информационные технологии 2009 / Материалы V международной научной конференции студентов, аспирантов и молодых ученых. – Донецк, ДонНТУ – 2009. [Электронный ресурс]:
http://www.uran.donetsk.ua/~masters/2010/fknt/sivyakov/library/article4.htm (02.05.2011).
14. Климова Е.А. Разработка дифференциальной математической модели эволюции метеопараметров и решение задач прогноза на ее основе. – Донецк, ДонНТУ, 2011. [Электронный ресурс]:
http://www.masters.donntu.ru/2011/fknt/klimova/diss/index.htm (на момент выполнения автореферата материалы автора не индексируются, однако переданы в электронном виде).