A. Mellit

Artificial Intelligence technique for modelling and forecasting of solar radiation data: a review

Artificial Intelligence (AI) has been used and applied in different sectors, such as engineering, economic, medicine, military, marine, etc. AI has also been applied for modelling, identification, optimisation, prediction, forecasting, and control of complex systems. The main objective of this paper is to present an overview of AI techniques for modelling, prediction and forecasting of solar radiation data. Published literature works presented in this paper show the potential of AI as a design tool for prediction and forecasting of solar radiation data; additionally, they present the advantages of using AI-based prediction solar radiation data in isolated areas where there no instrument for the measurement of this data, especially the parameters related to photovoltaic (PV) systems. Solar radiation plays a very important factor in PV-system performance and sizing.

Least squares support vector machine for short-term prediction of meteorological time series

The prediction of meteorological time series plays very important role in several fields. In this paper, an application of least squares support vector machine (LS-SVM) for short-term prediction of meteorological time series (e.g. solar irradiation, air temperature, relative humidity, wind speed, wind direction and pressure) is presented. In order to check the generalization capability of the LS-SVM approach, a K-fold cross-validation and Kolmogorov–Smirnov test have been carried out. A comparison between LS-SVM and different artificial neural network (ANN) architectures (recurrent neural network, multi-layered perceptron, radial basis function and probabilistic neural network) is presented and discussed. The comparison showed that the LS-SVM produced significantly better results than ANN architectures. It also indicates that LS-SVM provides promising results for short-term prediction of meteorological data.

Modelling of sizing the photovoltaic system parameters using artificial neural network

The objective of this work is to use an artificial neural network (ANN) to predict the sizing parameters of photovoltaic (PV) system with a minimum of input data. A neural network has been trained by using 54 known sizing parameter data corresponding to 54 locations. In this way the network was trained to accept and even handle a number of unusual cases. Known data were subsequently used to investigate the accuracy of prediction. A prediction with maximum deviation of 6% was obtained. This result indicates that the proposed method can successfully be used for the estimation of sizing parameters data for any locations.

Artificial neural network-based modelling and fault detection of partial shaded photovoltaic modules

Abstract In this paper, a fault detection method for photovoltaic module under partially shaded conditions is introduced. It consists to use an artificial neural network in order to estimate the output photovoltaic current and voltage under variable working conditions. The measured data (solar irradiance, cell temperature, photovoltaic current and voltage) at Renewable Energy Laboratory REL, Jijel University (Algeria), have been used. The comparison between the estimated current and voltage with the ones measured gives useful information on the operating state of the considered photovoltaic module. To show the effectiveness of the proposed method, several shading patterns have been investigated. The results showed that the designed method accurately detects the shading effect on the photovoltaic module.

FPGA-based implementation of an intelligent simulator for stand-alone photovoltaic system

Modelling and simulation of stand-alone photovoltaic (SAPV) systems (PV module, battery, regulator, etc.) in real time is crucial for the control, the supervision, the diagnosis and for studying their performances. In this paper, an intelligent simulator for stand-alone PV system was developed. Firstly, a multilayer perceptron (MLP) has been used for modelling and simulating each component of the system, after that the optimal architecture for each component has been implemented and simulated by using the very high-speed description language (VHDL) and the ModelSim. Subsequently, the developed architectures for each component have been implemented under the Xilinx(R) Virtex-II Pro FPGA (XC2V1000) (field programmable gate array). The obtained results showed that good accuracy is found between predicted and experimental data (signal) in a specific location (south of Algeria). The designed intelligent components (PV-MLP generator, MLP-battery and MLP-regulator) of the SAPV system can be used with success for simulating the system in real time (under a specific climatic condition) by predicting the different output signals for each component constituting the system.

Monitoring, Diagnosis, and Power Forecasting for Photovoltaic Fields: A Review

A wide literature review of recent advance on monitoring, diagnosis, and power forecasting for photovoltaic systems is presented in this paper. Research contributions are classified into the following five macroareas: (i) electrical methods, covering monitoring/diagnosis techniques based on the direct measurement of electrical parameters, carried out, respectively, at array level, single string level, and single panel level with special consideration to data transmission methods; (ii) data analysis based on artificial intelligence; (iii) power forecasting, intended as the ability to evaluate the producible power of solar systems, with emphasis on temporal horizons of specific applications; (iv) thermal analysis, mostly with reference to thermal images captured by means of unmanned aerial vehicles; (v) power converter reliability especially focused on residual lifetime estimation. The literature survey has been limited, with some exceptions, to papers published during the last five years to focus mainly on recent developments.

Comparative study between two intelligent MPPT-controllers implemented on FPGA: application for photovoltaic systems

A comparison between two intelligent maximum power point tracking (MPPT) controllers for photovoltaic systems is presented in this article. The presented MPPTs are based on the fuzzy logic controller (FLC) and neuro-fuzzy controller (NFC). Both controllers are designed and implemented on a Xilinx (Virtex-IIV2MB1000) reconfigurable field programmable gate array using hardware description language. Implemented controllers have been simulated and tested under constant and rapid variation of atmospheric conditions. Results show that the NFC performs better than the FLC in the viewpoint efficiency, response time and stability; however, with regard to the simplicity of implementation, the FLC is less complicated than the NFC.

FPGA-based real time simulation and control of grid-connected photovoltaic systems

Abstract Nowadays, FPGA devices are widely used to build control platforms in various fields of real time applications such as wireless telecommunications, image and signal processing, robotics and renewable energy systems. In grid-connected photovoltaic (PV) systems, an effective control strategy is needed for an efficiently use of solar energy as well as for energy supplies optimization. This paper presents an investigation, modelling and FPGA based digital controller design of a two-stage grid connected photovoltaic (GCPV) system. The basic idea of the proposed control structure is to operate the PV system as an active power generator and reactive power regulator. Firstly, a comprehensive study of dynamic behaviour of a two-stage GCPV system using Matlab/Simulink was presented. Subsequently, a digital circuit design of Incremental Conductance (IncCond) algorithm and decoupled active–reactive power control algorithm is carried out using the Hardware Description Language (VHDL). The simulation results confirm the accuracy of the adopted control strategy as well as the feasibility of the overall FPGA based control circuit.

Fault diagnosis in photovoltaic arrays

This paper proposes a simple automatic technique for fault diagnosis in photovoltaic (PV) arrays, based on the analysis of the an omalies observed in the I-V characteristic. Firstly, the I-V characteristic of the PV array is simulated using Matlab/Simscape tool for different faulty conditions; which is experimentally validated by generating different faults on a PV string installed at the Renewable Energy Laboratory of the University of Jijel (Algeria). Subsequently, we compare the I-V characteristic of the PV string under different faults scenarios, in order to identify the anomalies. Finally, six categories are generated: Normal operation, connection fault, connection fault with shadow effect, partial shadow fault, a group of fault which include shadow effect with faults on bypass diode (open circuit bypass diode, inversed bypass diode, shunted bypass diode), and a group of fault which include: bypass diode fault, cell fault, module fault, and shadow effect with shunted by pass diode fault. The results show that the technique can accurately detect and localize faults occurring in the photovoltaic string.

On the impact of photovoltaic module characterization on the prediction of PV plant productivity

A key factor in the design of photovoltaic fields is their yearly productivity and normally the nominal power is the only module parameter taken into account in the productivity predictions. The present paper intends to demonstrate how this approach disregards an important parameter, the nominal fill factor, that may significantly affect module efficiency when working in realistic environmental conditions. A complete calculation method is presented to compute a module productivity in any operating condition, using data sheets information. As an example two commercial module characteristics with different fill factors are compared to illustrate the concept in quantitative terms.