#### DMCA

## Hybrid methodology for hourly global radiation forecasting in Mediterranean area

Citations: | 1 - 1 self |

### Citations

3746 |
Time Series Analysis.
- Hamilton
- 1994
(Show Context)
Citation Context ... therefore different architectures. That is why the impact of the three methods on the quality of prediction is presented in section dedicated to the results. 96.2 The ARMA model The ARMA techniques =-=[1]-=- are reference estimators in the prediction of global radiation field. It is a stochastic process coupling an autoregressive component (AR) to a moving average component (MA). The model is usually the... |

1196 |
Time Series: Theory and Methods.
- Brockwell, Davis
- 1991
(Show Context)
Citation Context ...he original one. The modeling of the series begins with the selection of a suitable mathematical model (or class of models) for the data. Then, it is possible to predict future values of measurements =-=[24]-=-. Before proposing a hybrid ANN/ARMA model, we recall the need to make stationary a time series. We present three approaches to make stationary data available (the comparison of these methodologies wi... |

216 |
Forecasting with artificial neural networks: The state of the art,”
- Zhang, Patuwo, et al.
- 1998
(Show Context)
Citation Context ...chosen to study a particular ANN architecture called Multi Layer Perceptron (MLP) because it has been the most used architecture both in the renewable energy domain and in the time series forecasting =-=[25, 32-34]-=-. A well known, difficult and consuming-time task is to find the best network configuration and its parameters. The more common and important accepted optimization parameters are: number of input neur... |

103 | Artificial Neural Network: a tutorial”
- Jain, Mao, et al.
- 1996
(Show Context)
Citation Context ...y biological neural networks the ANN methodology is able to solve a variety of problems in decision making, optimization, control and obviously prediction and more particularly time series prediction =-=[9, 18-21]-=-. As the univariate ARMA model, ANNs can be considered as one of the referenced predictors among t ime series forecasting methods. The paper is organized as follow: Section 2 describes the data we hav... |

48 | Neural network forecasting for seasonal and trend time series - Zhang, Qi - 2005 |

44 | 25 years of time series forecasting,” - Gooijer, Hyndman - 2006 |

32 | Nearest neighbor method for time series analysis. - Yakowitz - 1987 |

25 | From decision theory to decision aiding methodology
- Tsouki`as
- 2003
(Show Context)
Citation Context ...iance using several MLP connected has to be done. Finally, as noted in sub-section 4.2, it should be interesting to study the possible contribution of Multiple Criteria Decision Aiding (MCDA) methods =-=[36]-=- in order to rank several prediction methods. The use of these methods in the solar energy domain is not new [37-38] and they can be a scientific basis for the definition of decision criteria that hav... |

24 | Testu F. Éléments d’analyse de données. Dunod - Diday, Lemaire, et al. - 1982 |

19 | Simulating climate change scenarios using an improved K-nearest neighbor model. - Sharif, DH - 2005 |

18 |
Kalogirou “Artificial intelligence techniques for photovoltaic applications: A review
- Mellit, Soteris
(Show Context)
Citation Context ...en ANN and different traditional computing technologies in solar energy area. Most of the time, thee technologies used are fuzzy logic, wavelet-based analysis, genetic algorithm, ARMA and ANN methods =-=[10,14-17]-=-. The main reason for this success seems to be the synergy of each computational property that makes them suitable for particular problems and not for others. In our case, we have coupled a linear pro... |

18 | Handbook of Neural Network Signal Processing - Hu - 2000 |

17 | Analyse Discriminante sur Variables Qualitatives - Celeux, Nakache - 1994 |

17 | Stepwise selection of artificial neural networks models for time series prediction - Crone - 2005 |

16 | S.: Artificial intelligence techniques for sizing photovoltaic systems: A review. Renewable and Sustainable Energy Reviews, Volume 13 - Mellit, Kalogirou, et al. - 2009 |

15 | The mathematics of Markov models: what Markov chains can really predict in forest successions - Logofet, Lesnaya - 2000 |

15 | 24-h forecast of solar irradiance using artificial neural network: Application for performance prediction of a grid-connected PV plant at - AM, Pavan |

14 | Univariate and multivariate forecasting of hourly solar radiation with artificial intelligence techniques. - Sfetsos, Coonick - 2001 |

13 | Predicting solar radiation at high resolutions: a comparison of time series forecasts - Reikard - 2009 |

13 |
A broadband simplified version of the Solis clear sky model. Sol Energy 2008;82:758–62
- Ineichen
(Show Context)
Citation Context ...celestial mechanics [2]. The second methodology is based on the clear sky model with the CSI. Several methods allow to determine this model. In our case, we have used the simplified “Solis clear sky” =-=[30]-=- model based on radiative transfer calculations and the Lambert-Beer relation [31]. This expression of the atmospheric transmittance is valid with polychromatic radiations, however when dealing with g... |

11 |
Times Series Forecasting with Neural Networks: A Case Study. Applied Statistics,1998
- Faraday, Chatfield
(Show Context)
Citation Context ... of hour h and hour h-1 63 4 5 Figure 1. The five studied stations marked in the Mediterranean sea: Ajaccio, Bastia , Montpellier , Marseille and Nice. 6 Time series forecasting models A time series =-=[21]-=- is a collection of time ordered observations xt, each one being recorded at a specific time t (period). Time series can appear in a wide set of domains such as Finance, Production or Control, just to... |

11 |
Clear-sky index to separate clear-sky from cloudy-sky situations in climate research
- Marty, Philipona
(Show Context)
Citation Context ...lity of the resource (determinist component). We propose three approaches to make stationary data available: the first is based on the Clearness Index (CI) [28], the second on a Clear Sky Index (CSI) =-=[29]-=-, the third is a variant of the previous one with Periodic Coefficients (PC) [2]. The first methodology based on the Clearness Index (CI(t)) is defined as the ratio between the total ground radiation ... |

11 |
Optimization of an artificial neural network dedicated to the multivariate forecasting of daily global radiation
- Voyant, Muselli, et al.
- 2011
(Show Context)
Citation Context ...zation process. Thus the first step of this process consists in determining the number of endogenous time lags. We have used autocorrelation criteria and a validation test based on the student T-test =-=[35]-=-. Only the lags with an autocorrelation different from zero are considered and put as input of the network. The partial autocorrelations (PACF noted � ), are obtained from theoretical autocorrelation ... |

9 | Design of Experiments on Neural Network's Training for Nonlinear Time Series Forecasting - Balestrassi, Popova, et al. - 2009 |

9 | ANN-based modelling and estimation of daily global solar radiation data: a case study - Benghanem, Mellit, et al. - 2009 |

7 | Neuro-fuzzy dynamic model with Kalman filter to forecast irradiance and temperature for solar energy systems. Renewable Energy - Chaabene, Benammar |

7 |
Cristofari C. Predicting hourly solar irradiations on inclined surfaces based on the horizontal measurements: performances of the association of well-known mathematical models. Energy Conversion and Management 2006;47:1816e29
- Notton, Poggi
(Show Context)
Citation Context ...nomena in an attempt to overcome the seasonality of the resource (determinist component). We propose three approaches to make stationary data available: the first is based on the Clearness Index (CI) =-=[28]-=-, the second on a Clear Sky Index (CSI) [29], the third is a variant of the previous one with Periodic Coefficients (PC) [2]. The first methodology based on the Clearness Index (CI(t)) is defined as t... |

5 | First order Markov chain model for generating synthetic “typical days” series of global irradiation in order to design photovoltaic stand alone systems. Energy Conversion and Management - Muselli, Poggi, et al. |

5 | time-series modeling and forecasting with neural networks - Qi, Zhang, et al. - 2008 |

3 | Forecasting of preprocessed daily solar radiation time series using neural networks
- Paoli, Voyant, et al.
- 2010
(Show Context)
Citation Context ...mization parameters are: number of input neurons, number of hidden layer and neurons, type of activation functions, type of normalization of data and the type of learning algorithm. In recent studies =-=[12, 13, 22, 23]-=-, we have developed an optimization process composed by three independent and chronological subparts: 1. Choice of the endogenous lags (time-delays between network nodes) number: past values of the gl... |

3 |
et al.,Rethinking satellite-based solar irradiance modelling : The SOLIS clear-sky module, Remote Sensing of Environment
- Mueller, Dagestad, et al.
(Show Context)
Citation Context ...th the CSI. Several methods allow to determine this model. In our case, we have used the simplified “Solis clear sky” [30] model based on radiative transfer calculations and the Lambert-Beer relation =-=[31]-=-. This expression of the atmospheric transmittance is valid with polychromatic radiations, however when dealing with global radiation, the Lambert-Beer relation is only an approximation because of the... |

2 |
Solar radiation forecasting using ad-hoc time series preprocessing and neural networks
- Paoli, Voyant, et al.
- 2009
(Show Context)
Citation Context ...mization parameters are: number of input neurons, number of hidden layer and neurons, type of activation functions, type of normalization of data and the type of learning algorithm. In recent studies =-=[12, 13, 22, 23]-=-, we have developed an optimization process composed by three independent and chronological subparts: 1. Choice of the endogenous lags (time-delays between network nodes) number: past values of the gl... |

2 |
Predictability of PV power grid performance on insular sites without weather stations: use of artificial neural networks,” 0905.3569
- Voyant, Muselli, et al.
- 2009
(Show Context)
Citation Context ...mization parameters are: number of input neurons, number of hidden layer and neurons, type of activation functions, type of normalization of data and the type of learning algorithm. In recent studies =-=[12, 13, 22, 23]-=-, we have developed an optimization process composed by three independent and chronological subparts: 1. Choice of the endogenous lags (time-delays between network nodes) number: past values of the gl... |

2 | Methodology for prediction sequences of mean monthly clearness index and daily solar radiation data in remote areas: Application for sizing a stand-alone PV system - Mellit, Kalogirou, et al. |

2 | atlas implementation and multicriteria context of PV plant planning in Corsica Island (France), 24th European photovoltaic solar energy conference and exhibition - Haurant, Muselli, et al. |

2 | Multicriteria decision aiding for selection of photovoltaic plants on farming fields in corsica, 25th European photovoltaic solar energy conference and exhibition - Haurant, Muselli, et al. - 2010 |

1 |
Terra a, Analyse des s ries temporelles : application l Economie et la Gestion, 2 éd
- Bourbonnais
- 2008
(Show Context)
Citation Context ... stationary data available: the first is based on the Clearness Index (CI) [28], the second on a Clear Sky Index (CSI) [29], the third is a variant of the previous one with Periodic Coefficients (PC) =-=[2]-=-. The first methodology based on the Clearness Index (CI(t)) is defined as the ratio between the total ground radiation (called X(t)) and extraterrestrial radiation (H0(t)) : � � CI( t) � X ( t) / H0 ... |

1 |
Optimization of an Artificial Neural Network (ANN) Dedicated to the Daily Global Radiation and
- Voyant, Muselli, et al.
- 2010
(Show Context)
Citation Context |