Characterization of solar irradiance profiles for photovoltaic system studies through data rescaling in time and amplitude

This paper addresses the representation of the data coming from solar irradiance measurements, to be used in evaluations referring to the operation of photovoltaic systems. Starting from the consideration that for different days of the year the sunrise and sunset timings change and the solar irradiance patterns at clear sky conditions occur with different maximum amplitude, a bi-normalization procedure is applied in order to produce comparable normalized patterns for the various days. The normalized patterns are then subject to clustering in order to obtain a meaningful grouping of similar days. Finally, from the clustering results a day-type succession matrix is constructed, whose entries are interpreted as the conditional probability of finding a given day type providing that the type of the preceding day is known. Data used in the analysis are taken from real sites.