No abstract

No abstract

The optimization of the battery pack in an off-grid Photovoltaic application must consider the minimum sizing that assures the availability of the system under the worst environmental conditions. Thus, it is necessary to predict the evolution of the state of charge of the battery under incomplete daily charging and discharging processes and fluctuating temperatures over day-night cycles. Much of previous development work has been carried out in order to model the short term evolution of battery variables. Many works focus on the on-line parameter estimation of available charge, using standard or advanced estimators, but they are not focused on the development of a model with predictive capabilities. Moreover, normally stable environmental conditions and standard charge-discharge patterns are considered. As the actual cycle-patterns differ from the manufacturer's tests, batteries fail to perform as expected. This paper proposes a novel methodology to model these issues, with predictive capabilities to estimate the remaining charge in a battery after several solar cycles. A new non-linear state space model is proposed as a basis, and the methodology to feed and train the model is introduced. The new methodology is validated using experimental data, providing only 5% of error at higher temperatures than the nominal one.

This paper presents a fault detection system for photovoltaic standalone applications based on Gaussian Process Regression (GPR). The installation is a communication repeater from the Confederación Hidrográfica del Ebro (CHE), public institution which manages the hydrographic system of Aragón, Spain. Therefore, fault-tolerance is a mandatory requirement, complex to fulfill since it depends on the meteorology, the state of the batteries and the power demand. To solve it, we propose an online voltage prediction solution where GPR is applied in a real and large dataset of two years to predict the behavior of the installation up to 48 hour. The dataset captures electrical and thermal measures of the lead-acid batteries which sustain the installation. In particular, the crucial aspect to avoid failures is to determine the voltage at the end of the night, so different GPR methods are studied. Firstly, the photovoltaic standalone installation is described, along with the dataset. Then, there is an overview of GPR, emphasizing in the key aspects to deal with real and large datasets. Besides, three online recursive multistep GPR model alternatives are tailored, justifying the selection of the hyperparameters: Regular GPR, Sparse GPR and Multiple Experts (ME) GPR. An exhaustive assessment is performed, validating the results with those obtained by Long Short-Term Memory (LSTM) and Nonlinear Autoregressive Exogenous Model (NARX) networks. A maximum error of 127 mV and 308 mV at the end of the night with Sparse and ME, respectively, corroborates GPR as a promising tool.