Abstract
For heating and cooling load estimation, weather data are the essentially important input. Mathematically modeled weather data are more attractive than the numeral data known as the reference year for the input, because of containing much stochastic information, achieving data compression and having applicability for the analytical load calculation. In this paper, the mathematical expressions of the component of duirnally periodic temperature, which is the specific feature of weather data, are investigated. We concluded that two expressions are best in the meaning of good applicability for the heat load calculation. One of them is the hourly averaged temperature of a month and another is the duirnally varying temperature made by the amplitude modulation with the solar radiation information. Appropriateness of the two expressions is verified by the more sopfisticated time series models which are based on non-stational time series analysis, which, however, are too complicated to apply for the present purpose.
