Article

Article title CHOICE OF PARAMETERS FOR ADJUSTMENT MODELS OF A SMOOTHING TIME NUMBER FOR SHORT-TERM FORECASTING OF PHYSICAL SIZE
Authors S.I. Klevtsov
Section SECTION II. MATHEMATICAL MODELS AND METHODS
Month, Year 05, 2011 @en
Index UDC 621.302
DOI
Abstract The scheme of forecasting of physical size by means of adaptive polynomial models of a time series of the first order is developed. Influence of adjusting parameters for model of a time number on a forecasting error is estimated. The general rules of a choice of their values are defined. The analysis shows, that for maintenance of good accuracy of forecasting by means of time numbers it is necessary to choose the constant of smoothing corresponding to dynamics of predicted process. For elimination or the maximum reduction of an adjusting site for adaptation of a time number and expansion of a site for forecasting it is necessary to set precisely initial values of approximating factors for representation of initial function.

Download PDF

Keywords Time series; model; forecasting; microcontroller; physical quantity.
References 1. Лукашин Ю.П. Адаптивные методы краткосрочного прогнозирования временных рядов. – М.: Финансы и статистика, 2003. – 416 с.
2. Бокс Дж., Дженкинс Г. Анализ временных рядов, прогноз и управление. – М.: Мир, 1974. – 197 с.

Comments are closed.