Short term solar irradiance forecasting via a novel evolutionary multi-model framework and performance assessment for sites with no solar irradiance data

Marzouq, Manal; Fadili, Hakim El; Zenkouar, Khalid; Lakhliai, Zakia; Amouzg, Mohammed

doi:10.1016/j.renene.2020.04.133

Cited by 31 publications

(9 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, based on the linear combination of available past input and output samples, the specific system output is represented. Based on the linear multi-input and multi-output autoregressive exogenous technique of Wu et al [20], the proposed model is generalised as below:…”

Section: Proposed Multi-time Scale Forecastingmentioning

confidence: 99%

“…The Angstrom-Prescott type of strategy has been reported for better irradiance forecasting results for hour ahead prediction that has been implemented with five semi-empirical models [18,19]. The new machine-learning algorithm was proposed to predict solar irradiance to improve prediction by using artificial neural networks [20]. A model based on long-term solar radiation forecasting was reported with hourly time intervals using a feedback backpropagation time-series network to reduce the solar radiance's various influences [21].…”

Section: Introductionmentioning

confidence: 99%

“…For hourly time scale forecasting, the performance of the proposed hybrid CSO-GWO LSTM model has reported 3.99% improved MSE, 5.05% of MAPE and 6.31% of better DA value than the ESO neural network strategy. On comparing with PSO-SVR, the proposed strategy outperformed with20.2% better MSE, 38.32% of MAPE and 38.54% better DA value, respectively. On comparing with the GRU, the proposed strategy outperformed with 32.4% of MSE, 24.1% better MAPE, and 15.9% better DA.…”

mentioning

confidence: 97%

See 2 more Smart Citations

Novel Multi-Time Scale Deep Learning Algorithm for Solar Irradiance Forecasting

Jayalakshmi

Shankar²,

Subramaniam

et al. 2021

Energies

View full text Add to dashboard Cite

Solar irradiance forecasting is an inevitable and most significant process in grid-connected photovoltaic systems. Solar power is highly non-linear, and thus to manage the grid operation efficiently, with irradiance forecasting for various timescales, such as an hour ahead, a day ahead, and a week ahead, strategies are developed and analysed in this article. However, the single time scale model can perform better for that specific time scale but cannot be employed for other time scale forecasting. Moreover, the data consideration for single time scale forecasting is limited. In this work, a multi-time scale model for solar irradiance forecasting is proposed based on the multi-task learning algorithm. An effective resource sharing scheme between each task is presented. The proposed multi-task learning algorithm is implemented with a long short-term memory (LSTM) neural network model and the performance is investigated for various time scale forecasting. The hyperparameter estimation of the proposed LSTM model is made by a hybrid chicken swarm optimizer based on combining the best features of both the chicken swarm optimization algorithm (CSO) and grey wolf optimization (GWO) algorithm. The proposed model is validated, comparing existing methodologies for single timescale forecasting, and the proposed strategy demonstrated highly consistent performance for all time scale forecasting with improved metric results.

show abstract

Section: Proposed Multi-time Scale Forecastingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 97%

See 1 more Smart Citation

Novel Multi-Time Scale Deep Learning Algorithm for Solar Irradiance Forecasting

Jayalakshmi

Shankar²,

Subramaniam

et al. 2021

Energies

View full text Add to dashboard Cite

show abstract

“…The collected raw data refer to a 5.94 kWp grid connected PV plant implemented on the rooftop of National School of Applied Sciences of Safi, Morocco. This location is characterized by hot semi-arid climate [2] . The raw data include: The most important weather variables: global horizontal irradiance, plane of array irradiance, and ambient temperature; The most important electrical variables of each PV technology obtained from inverters: DC power, DC current and voltage, AC power, AC current and voltage, instantaneous and cumulative produced energy, grid injection frequency, operating and feed in time, inverter operating temperature; Module temperature of each PV technology (Tc).…”

Section: Data Descriptionmentioning

confidence: 99%

Data of a PV plant implemented in hot semi-arid climate

et al. 2021

View full text Add to dashboard Cite

This paper presents data collected from a 5.94 kWp grid connected photovoltaic (PV) plant implemented in hot semi-arid climate of Safi region, Morocco. The data include electrical power production and PV module temperature of three PV technologies: mono-crystalline (m-Si), poly-crystalline (p-Si), and amorphous (a-Si); they also include plane of array solar irradiance and ambient temperature. Solar irradiance was measured with calibrated reference cells, inverters provided the produced powers, and the temperatures were obtained by Pt100 probes. The data were measured each 5 min and were remotely accessible through internet. They were preprocessed to eliminate unrepresentative records and were used for the development of simple and accurate models for PV power forecasting [1] . These data are typical for hot semi-arid climate and may be reused for regional forecast of PV power as well as solar energy and PV module temperature predictions.

show abstract

“…Machine learning-based methods, like Artificial Neural Networks (ANNs) [ 22 ], Support Vector Machine (SVM) [ 23 ], and K-Nearest Neighbor (KNN) are widely used and show superior accuracy in short-term predictions. Without the complexity of mathematical and physical relationships, ANNs can learn any nonlinear information and produce accurate short-term predictions [ 24 ]. In time series forecasting, they do have certain drawbacks.…”

Section: Introductionmentioning

confidence: 99%

Attention-Based Models for Multivariate Time Series Forecasting: Multi-step Solar Irradiation Prediction

Sakib,

Mahadi,

Abir

et al. 2024

Heliyon

View full text Add to dashboard Cite

Short term solar irradiance forecasting via a novel evolutionary multi-model framework and performance assessment for sites with no solar irradiance data

Cited by 31 publications

References 49 publications

Novel Multi-Time Scale Deep Learning Algorithm for Solar Irradiance Forecasting

Novel Multi-Time Scale Deep Learning Algorithm for Solar Irradiance Forecasting

Data of a PV plant implemented in hot semi-arid climate

Attention-Based Models for Multivariate Time Series Forecasting: Multi-step Solar Irradiation Prediction

Contact Info

Product

Resources

About