A comparison of photovoltaic models for estimating power generation: a case study of Brazilian data

Abstract

This paper presents photovoltaic (PV) generation models used to predict the power output injected into the grid, taking into account the relevant environmental variables, such as irradiance and ambient air temperature. The purpose is to identify the models that have the necessary degree of accuracy and simplicity to be used in studies of technical impacts of connecting PV generation plants to power systems. The paper focuses on semi-empirical models that adopt analytical expressions based on empirical equations and information provided by manufacturers. Using time series of ground measured irradiance and local ambient air temperature, five models are used to generate power output time series in minutely and hourly resolution, which are then compared with actual power measurements using statistical metrics.