Optimally configured Gated Recurrent Unit using Hyperband for the long-term forecasting of photovoltaic plant

Khan, Ameer Tamoor; Khan, Abdul Rehman; Li, Shuai; Bakhsh, Sunila; Mehmood, Aamir; Jahan, Zaib

doi:10.1016/j.ref.2021.07.002

Cited by 17 publications

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although GRU network has simpler structure comparing to LSTM networks, its internal structure is more complicated than the normal RNN. GRU has one gate less than LSTM, this reduces the matrix multiplication, and consequently, it can save a lot of time without impacting its performance [21,22]. The structure of GRU block is shown in Fig.…”

Section: Gru Networkmentioning

confidence: 99%

A hybrid model of CNN and LSTM autoencoder-based short-term PV power generation forecasting

Ibrahim,

Gharghory,

Kamal

2024

Electr Eng

View full text Add to dashboard Cite

Solar energy is one of the main renewable energies available to fulfill global clean energy targets. The main issue of solar energy like other renewable energies is its randomness and intermittency which affects power grids stability. As a solution for this issue, energy storage units could be used to store surplus energy and reuse it during low solar generation intervals. Also, in order to sustain stable power grid and better grid operation and energy storage management, photovoltaic (PV) power forecasting is inevitable. In this paper, new hybrid model based on deep learning techniques is proposed to predict short-term PV power generation. The proposed model incorporates convolutional neural network (CNN) and long short-term memory (LSTM) autoencoder network. The new model differentiates itself in accomplishing high prediction accuracy by extracting spatial features in time series via CNN layers and temporal features between the time series data through LSTM. The introduced model is tested on dataset of power generation from southern UK solar farm and the weather data corresponding to same location and time intervals; the forecasting performance of the suggested model is evaluated in metrics of root-mean-square error (RMSE) and mean absolute error (MAE). The used model is compared with different models from the literature either of pure type of network such as LSTM and gated recurrent unit (GRU) or hybrid combination of different networks like CNN-LSTM and CNN-GRU. The results show that proposed model provides enhanced results and reduces training time significantly compared to other competitive models, where the performance of the proposed model improved averagely by 5% to 25% in terms of RMSE and MAE performance metrics, and the execution time of training significantly reduced with almost 70% less compared to other models.

show abstract

Section: Gru Networkmentioning

confidence: 99%

A hybrid model of CNN and LSTM autoencoder-based short-term PV power generation forecasting

Ibrahim,

Gharghory,

Kamal

2024

Electr Eng

View full text Add to dashboard Cite

show abstract

“…The gated recurrent unit network is a relatively new variant of RNN which simplifies the architecture of the LSTM network which allows the GRU to be trained faster (Khan et al, 2021). The GRU neuron reduces the four gates of the LSTM into two gates as shown in Fig.…”

Section: Gated Recurrent Unit Networkmentioning

confidence: 99%

“…Narvaez et al trained encoder-decoder GRU and LSTM networks to predict global horizontal irradiance in daily and weekly forecasting horizons, and their results show the LSTM networks outperformed the GRU in both horizons (Narvaez et al, 2021). The GRU network's strengths were also studied in several works involving PV energy forecasting (Abdel-Basset et al, 2021;Khan et al, 2021;Li et al, 2020;Mellit et al, 2021). Al-Ghezi et al (Al-Ghezi et al, 2022) studied the validity of two single and two polynomial linear regression models in measuring daily global horizontal solar radiation (GHSR) using data from different stations -Iraqi Meteorological Authority (IMA) and National Aeronautics and Space Administration (NASA).…”

Section: Introductionmentioning

confidence: 99%

Daily Solar Radiation Forecasting based on a Hybrid NARX-GRU Network in Dumaguete, Philippines

Fuselero

Portus

Doma

2022

Int. J. Renew. Energy Dev.

View full text Add to dashboard Cite

In recent years, solar radiation forecasting has become highly important worldwide as solar energy increases its contribution to electricity grids. However, due to the intermittent nature of solar radiation caused by meteorological parameters, forecasting errors arise, and fluctuations in the power output of photovoltaic (PV) systems become a severe issue. This paper aims to introduce a forecasting hybrid model of daily global solar radiation time series. Meteorological data and solar radiation samples from Dumaguete, Philippines, are used to assess the forecasting accuracy of the proposed nonlinear autoregressive network with exogenous inputs (NARX) – gated recurrent unit (GRU) hybrid model. Four different models were trained using the meteorological and solar radiation data, which are the Optimizable Gaussian Process Regression (GPR), Nonlinear Autoregressive Network (NAR), NARX, and the proposed Hybrid NARX-GRU Network. Results show that the hybrid NARX-GRU model has a root mean square error (RMSE) of ~0.05 and a training time of 33 seconds. The proposed hybrid model has better forecasting performance compared to the three models which obtained RMSE values of 27.741, 39.82, and 28.92, for the GPR, NAR, and NARX, respectively. The simulation results demonstrate that the NARX-GRU model significantly outperforms the regression and single models in terms of statistical metrics and training efficiency. Furthermore, this study shows that the hybridized NARX-GRU model is able to provide an effective estimation for daily global solar radiation, which is important in the operation of PV plants in the country, specifically for unit commitment purposes

show abstract

“…Furthermore, deep learning (DL) models are enhanced based on ANN models with various alternative structures to stimulate the performance as well as be appropriate to different issues. In PV forecast field, apart from ANN‐based models, many DL models are used such as convolutional neural network (CNN) [27], recurrent neural network (RNN) [28], and the improved structures of RNN model, including long short‐term memory (LSTM) [29] and gated recurrent unit (GRU) [30]. Hybrid models are implemented to achieve higher forecast precision by combining the above approaches together or integrating them with optimization algorithms.…”

Section: Introductionmentioning

confidence: 99%

A combination of novel hybrid deep learning model and quantile regression for short‐term deterministic and probabilistic PV maximum power forecasting

Tuyen

Thanh

Huu

et al. 2022

IET Renewable Power Gen

View full text Add to dashboard Cite

Recently, the increasing penetration of renewable energy resources in power system, especially photovoltaic (PV) systems, has caused severe technical issues due to its randomness and dependency on primary sources. Therefore, precise output power forecast is vital for both system operators and PV system owners to improve grid stability and generation quality, respectively. This article proposes a novel hybrid deep learning model named EDSACL comprising four components, namely, convolutional neural network (CNN), long short‐term memory (LSTM), self‐attention mechanism (SAm), and residual learning (RL) strategy for short‐term PV maximum power prediction with forecast horizon of 1, 2, and 4 h. This proposed model initially extracts spatial‐temporal features of input data based on CNN and LSTM before using SAm to achieve hidden features of LSTM layers. RL strategy is employed to maintain the information flow entering the network. Besides, interval prediction is also implemented based on a combination of the EDSACL model and quantile regression with different probabilities. The forecast accuracy of the proposed model is validated based on two real‐world data sets. The predicted results show the superiority of the proposed model over numerous predictive models; in particular, mean absolute percentage error values of the proposed hybrid model are under 6% in all case studies.

show abstract

Optimally configured Gated Recurrent Unit using Hyperband for the long-term forecasting of photovoltaic plant

Cited by 17 publications

References 26 publications

A hybrid model of CNN and LSTM autoencoder-based short-term PV power generation forecasting

A hybrid model of CNN and LSTM autoencoder-based short-term PV power generation forecasting

Daily Solar Radiation Forecasting based on a Hybrid NARX-GRU Network in Dumaguete, Philippines

A combination of novel hybrid deep learning model and quantile regression for short‐term deterministic and probabilistic PV maximum power forecasting

Contact Info

Product

Resources

About