Forecasting Algorithm based on Temperature Errors Pre-Diction Using Kalman Filter for Building Management Sys-Tem

Latvian Journal of Physics and Technical Sciences 2021
Nikolajs Bogdanovs, Romualds Beļinskis, Vadims Bistrovs, Ernests Pētersons, Aleksandrs Ipatovs

This work offers a new method of collection and processing of meteorological data of meteorological service based on observations and correction of numerical weather forecast errors by using a new prediction algorithm. This algorithm vastly increases the accuracy of the short-term forecast of outdoor air temperature which inherent uncertainty due to the stochastic nature of atmospheric processes. Processing of temperature data using Kalman Filter provides the decrease in predicted temperature errors. The main setup methods of Kalman Filter have been examined. The article also describes practical use and implementation of accuracy improving algorithm of predicted temperature by using Python.

Keywords
Kalman Filter, algorithm, forecast, prediction, temperature
DOI
10.2478/lpts-2021-0038
Hyperlink
https://www.sciendo.com/article/10.2478/lpts-2021-0038

Bogdanovs, N., Beļinskis, R., Bistrovs, V., Pētersons, E., Ipatovs, A. Forecasting Algorithm based on Temperature Errors Pre-Diction Using Kalman Filter for Building Management Sys-Tem. Latvian Journal of Physics and Technical Sciences, 2021, Vol. 58, No. 5, pp.38-49. ISSN 0868-8257. Available from: doi:10.2478/lpts-2021-0038

Publication language
English (en)