Photovoltaic panels can consume energy rather than produce it, wherefore a low-cost Data Acquisition System (SAD) was implemented, which is capable of accurately measure and register temperature and irradiance in real time. Subsequently, supported with data mining techniques such as k-means and ID3, a classification and prediction was made, which allows to determine if photovoltaic panels must be connected nor disconnected. The SAD and the data mining methods were developed in Python. The code "interfaz" contains the SAD development, and the codes named "Kmeans" and "ID3" have the data mining techniques used respectively.