A tool for obtaining the LOLP curves for sizing off-grid photovoltaic systems based in neural networks

In this paper a new approach for obtaining LOLP curves is presented. It is based on the neural network called multilayer perceptron (MLP). The method is based on the MLP ability to extract, from a sufficiently general training data set, the existing relationships between variables whose interdependence is unknown a priori. The MLP is trained with LOLP data from locations where they are available, and after the training, the MLP is able to generate LOLP curves for any value.

1. Introduction

There are several methods for sizing off-grid photo voltaic (pv) systems. The more simplistic are called intuitive methods. They are used as a first approach, but they are very inaccurate. Another group of methods are denominated numerical methods. These methods usc system simulations. These ones are more accurate than the previous ones but present the problem that they need to have long time series of solar radiation for the simulations. In a third group there are methods which use equations to describe the pv system size as a function of reliability. These ones arc called numerical methods.