Seasonality Atlas of Solar Radiation in Mexico

Photovoltaic Power Generation Forecasting for Regional Assessment Using Machine Learning

“…Borunda et al [33] used k-means and k-medoids algorithms to perform a cluster analysis of solar radiation in several representative locations in Mexico, obtaining a preliminary seasonality atlas for solar resources.…”

“…In this analysis, time series data are divided into four groups and the number of stations at which the same cluster occurs simultaneously is investigated. It is found that the cluster in which cloudy conditions are maintained for 5 days has peaks of the number of stations that are simultaneously assigned to the cluster in the rainy season and in winter, whereas the cluster with five consecutive clear days has the peaks in spring and summer" [33]/2021 Mexico Annual/Seasonal 2000-2020 "This work performs a cluster analysis to determine the seasonality of the solar radiation of different locations. We use k-means and k-medoids algorithms, and even though both are partitioning algorithms, we end up preferring k-medoids to find the seasonality since the centroids of the clusters belong to data from the dataset and therefore a straightforward interpretation is generated" [34] "K-Means and GMM are both unsupervised clustering techniques but work differently.…”