A novel wind power prediction model using graph attention networks and bi-directional deep learning long and short term memory

International Journal of Electrical and Computer Engineering (IJECE)
Vol. 13, No. 6, December 2023, pp. 6847~6854
ISSN: 2088-8708, DOI: 10.11591/ijece.v13i6.pp6847-6854  6847
Journal homepage: http://ijece.iaescore.com
A novel wind power prediction model using graph attention
networks and bi-directional deep learning long and short term
memory
Ibtissame Mansoury, Dounia El Bourakadi, Ali Yahyaouy, Jaouad Boumhidi
LISAC Laboratory, Department of Computer Science, Faculty of Sciences Dhar-Mahraz, Sidi Mohamed Ben Abdellah University,
Fez, Morocco
Article Info ABSTRACT
Article history:
Received May 27, 2023
Revised Jul 9, 2023
Accepted Jul 17, 2023
Today, integrating wind energy forecasting is an important area of research
due to the erratic nature of wind. To achieve this goal, we propose a new
model of wind speed prediction based on graph attention networks (GAT),
we added a new attention mechanism and a learnable adjacency matrix to the
GAT structure to obtain attention scores for each weather variable. The
results of the GAT-based model are merged with the bi-directional deep
learning long and short-term memory (BiLSTM) layer to take advantage of
the geographic and temporal properties of historical weather data. The
experiments and analyzes are carried out using precise meteorological data
collected from wind farms in the Moroccan city of Tetouan. We show that
the proposed model can learn complex input-output correlations of
meteorological data more efficiently than previous wind speed prediction
algorithms. Due to the resulting attention weights, the model also provides
more information about the main weather factors for the evaluated forecast
work.
Keywords:
Bidirectional deep learning
long-short term memory
Graph attention network
Prediction
Renewable energy
Wind power
This is an open access article under the CC BY-SA license.
Corresponding Author:
Ibtissame Mansoury
LISAC Laboratory, Department of Computer Science, Faculty of Science Dhar-Mahraz, Sidi Mohamed
Ben Abdellah University
Fez-3000, Morocco
Email: ibtissame.mansoury@usmba.ac.ma
1. INTRODUCTION
Among the many clean and renewable energy sources, wind energy is significant. Wind energy’s
stochastic and unpredictable character, however, presents considerable difficulties for the power systems’
dependable and secure functioning. Wind energy forecasting that is precise and reliable is essential for the
integration of wind power into the grid. Numerous atmospheric characteristics are employed in numerical
weather prediction (NWP) approaches, which are based on mathematical equations, in the study of weather
forecasting [1]. This technique uses equations drawn from the principles of physics and thermodynamics to
create predictions. While NWP is an effective solution for many forecasting problems, it still has limitations,
such as using an excessive amount of computing power [2]. Additionally, forecasting using NWP may be
susceptible to noise that may be included in the observations of meteorological variables because of the
assumptions about meteorological variables utilized in this technique [3]. Deep learning machine models can
now be trained because of advancements in processing capacity, as an alternative to NWP, deep learning-
based models have therefore been utilized more and more lately in weather forecasting tasks [4], [5]. In
particular, the approaches based on convolutional neural networks have been utilized in the research on

 ISSN: 2088-8708
Int J Elec & Comp Eng, Vol. 13, No. 6, December 2023: 6847-6854
6848
weather prediction models [3], In addition, there are graph convolutional neural networks [6], and restricted
Boltzmann machines [7].
Due to the increase in outstanding results obtained using data-driven approaches, the development
of several deep-learning algorithms for weather forecasting problems has been documented in the literature
[3], [8]. One way to conceptualize meteorological data is as a time series of successive measurements. The
challenges of time series forecasting have often been addressed by recurrent neural networks [9]. For these
tasks, bi-directional deep learning long and short term memory (BiLSTM) models are chosen [10], which
may impair performance throughout the training period. In contrast, convolutional neural networks (CNNs)
are popular and particularly effective for computer vision applications [11]. While vanilla CNNs can handle
the geographic aspects of the data, they cannot effectively exploit the temporal properties that are crucial for
these weather forecasting applications. Salman et al. [3] have developed a 3D-CNN model for forecasting
meteorological components. Lara-Benítez et al. [12] used vanilla CNNs in conjunction with temporal
convolutional networks (TCNs), which is a kind of CNN that utilizes one-dimensional filters to concentrate
on the temporal component of the data, to undertake prediction work in the field of meteorology. With the
integration of TCN, CNNs can analyze spatiotemporal data. For evaluating graph data, graph neural networks
(GNN) [13] have lately grown in prominence. Graph convolutional networks (GCNs ) [14] are a kind of
GNN that combines the fundamentals of GNNs while maintaining a convolutional architecture by employing
graph data rather than assuming that the data is Euclidean. There have been GCN adaptations using a
learnable adjacency matrix in [6], although most GCN implementations focus largely on node classification
tasks that need knowledge of the graph’s nodes and edges. One benefit of this is that normal data can now be
represented using an adjacency matrix, which is typically necessary for GNN implementations. However, to
extract spatial information, it is desirable to learn the adjacency matrix during the training phase. Due to the
transformer design presented in [15], attention methods have lately become quite popular. GNN frameworks
have also integrated the attention mechanism. For example, the graph attention networks (GAT) [16] uses a
self-attention mechanism similar to a transformer model, to determine the attention coefficient between the
vertices of the graph.
Researchers often ignore the impact of spatiotemporal connections between wind turbines in favor
of analyzing and modeling the past data for a wind turbine. Therefore, the development of spatiotemporal
models with a high prediction of wind energy is crucial. In this article, we propose a novel spatiotemporal
wind energy prediction model founded on GAT followed by a BiLSTM algorithm to combine temporal and
geographical information of the data. GAT is a subtype of the GNN created specifically to analyze graphical
data. It is a model for predicting wind power based on a variety of geographical data. It studies the
geographical relationships between different nodes in the graph and how the characteristics of one node can
influence the characteristics of another node. It makes optimal use of the attention mechanism to gather data
from several nodes, while BiLSTM records temporal correlations between meteorological data collected at
distinct periods. Our suggestion to improve the GAT-BiLSTM model makes it possible to calculate the
attention of the nodes which produces GAT form in a more accurate and understandable. In our research, past
weather data from the wind farm of Tetouan City in Morocco is used to estimate wind speed values for many
hours into the future.
This paper is structured as follows: section 2 describes the suggested model. In section 3, you will
find a description of the numerical experiments, a discussion of the results, and a comparison to other
models. Section 4 contains the conclusion.
2. METHOD
2.1. Graph attention networks (GAT)
The spectral-domain GNN and the spatial-domain GNN are two subcategories of the mainstream
GNN. The network topology of the spectral domain GNN is defined by filters, and the spatial domain GNN
analyzes the node correlation to construct the network topology. Currently, the Laplace matrix is the only
method used by the spectral domain GNN to deconstruct the graph and recover its specific information.
Since the neural network now only decomposes and extracts the characteristic information of the graph
using the Laplace matrix, it has disadvantages [17]: i) The noise information in the graph has a simple impact on
the eigenvalue, ii) It is challenging to extend the network since the filter it uses can only be applied to a certain
structure, and iii) Both the time and financial costs of feature computation and analysis are very high.
GAT presented an attention method to solve the difficulty of GCN [18]. As a result, GAT could give
higher weights to important nodes, as neural network parameters and weights are learned simultaneously
[19]. At the moment, researchers create the GAT by collecting numerous attention mechanism networks. The
feature sets and the transformed feature set of the original nodes constitute the network’s input and output for
the graph’s initial attention layer [20]. The fundamental principle of GAT is to find the spatial link of nodes,

Int J Elec & Comp Eng ISSN: 2088-8708 
A novel wind power prediction model using graph attention networks … (Ibtissame Mansoury)
6849
by calculating the weight coefficient of neighboring nodes. The model creates a shared weight matrix W [21]
and uses it to examine and determine the coefficient of attention:
𝑒𝑗
𝑖
= 𝑎(𝑊ℎ𝑖, 𝑊ℎ𝑗) (1)
or ℎ denotes the input characteristics, 𝑊 denotes the weight matrix, and 𝑎 denotes the attention mechanism.
2.2. Bi-directional deep learning long and short-term memory (BiLSTM)
Hochreiter and Schmidhuber [22] proposed the long-short term memory (LSTM) form of the
recurrent neural network (RNN) in 1997, to improve the ability of temporal sequence data to capture long-
term connections. It can get around the drawbacks of the vanishing/exploding gradient that standard RNNs
experience. Recurrent networks may easily learn across many past time steps without gradient difficulties
since the memory of the LSTM helps to maintain the mistake that will be back-propagated via the layers
[23]. Blocks are used to store values with varied durations in the LSTM's structure. Each block has two parts
that allude to memories: the “Cell state” noted ct, which is notable for being able to recall prior information
and displays long-term memory, and the “Hidden state,” which only transmits information for the subsequent
time step and displays short-term memory.
In contrast, BiLSTM, which consists of two LSTM models, is a powerful and enhanced version of
the LSTM processing network. As it comes to the first one, the signal spreads ahead. Whereas in the second,
it spreads oppositely. This implies that when processing the material, it concurrently considers both past and
future context information [24]. Figure 1 illustrates the construction of the BiLSTM network and
demonstrates that each memory block has two LSTM layers. In other words, the output 𝑦𝑡 can be determined
by adding the values of two opposite hidden (ℎ
⃗ 𝑡 and ℎ
⃖⃗𝑡) as you move forward and backward through the
layers. The corresponding formulas are provided as [25]:
ℎ
⃗ 𝑡 = 𝐿𝑆𝑇𝑀(𝑥𝑡, ℎ
⃗ 𝑡−1) (2)
ℎ
⃖⃗𝑡 = 𝐿𝑆𝑇𝑀(𝑥𝑡, ℎ
⃖⃗𝑡+1) (3)
𝑦𝑡 = 𝑊ℎ
⃗
⃗ 𝑦
ℎ
⃗ 𝑡 + 𝑊ℎ
⃖⃗
⃗𝑡
ℎ
⃖⃗𝑡 + 𝑏𝑦 (4)
where, 𝐿𝑆𝑇𝑀(. ) displays the LSTM function, 𝑏𝑦 is the output layer bias, 𝑦𝑡 is the final output of the
BiLSTM network, 𝑊ℎ
⃗
⃗ 𝑦
and 𝑊ℎ
⃖⃗
⃗𝑡
are, respectively, the weights of the front and reverse LSTM.
The amount of information sent across the network increases when LSTM is applied twice in each
of the two directions. As a result, learning twice as much long-term correlation information in the
fundamental architecture of LSTM improves prediction performance [10]. As a result, BiLSTM with dual
formation capability vastly outperforms uni-LSTM in a variety of areas.
Figure 1. The BiLSTM network structure diagram
2.3. Proposed method
Due to the growing global energy demand, renewable energies are becoming more widespread,
especially wind energy. And due to the large instability and fluctuation of wind data, the wind farm is forced
to implement practical procedures to manage the wind rotor and increase energy efficiency. Since wind

 ISSN: 2088-8708
6850
power forecasting is the basic technique to control and distribute the energy of wind farms, it is more
complicated due to the unpredictable and non-linear variation of wind power data. In this article, therefore,
we will study wind energy forecasting in detail. The graphical predictor, which is shown in Figure 2, makes
up the majority of the GAT set that has been provided.
Figure 1. Graph attention network layer
Spatial and temporal wind turbine datasets are pre-processed. The raw wind power data was split
between the training set and the testing set. The spatiotemporal wind power prediction model is integrated
using GAT followed by BiLSTM as the main predictor, thus, all fundamental weight parameters were
defined using the training set. Spatial node feature data from unprocessed datasets are compiled using GAT.
An adjacency matrix is a common way to represent a graph. On the other hand, if the architecture of
the graph is unknown, it is possible to determine the weights of the edges of the graph. It has been made
learnable in previous works [6], and this has been demonstrated to enhance the capacity of the prediction
model. It may also shed light on how the characteristics of one node may be modified by those of other
nodes. The typical GAT model requires knowledge of the adjacency matrix of the graph to work [26].
Comparing this model to the one in [26], we make the adjacency matrix learnable. So, we give the network
the freedom to fully understand the spatial relationships that exist between the nodes in the graph. To
calculate the attention scores for the GAT, this newly acquired information is later utilized. When learning is
underway, the adjacency matrix is subjected to the same processes as in (5) [6].
{
Â = 𝐴 + 𝐼 ,
Â =
Â−Â𝑚𝑖𝑛
Â𝑚𝑎𝑥−Â𝑚𝑖𝑛
. (5)
The identity matrix is therefore added to the adjacency matrix 𝐴 to take into account the potential
self-loop link at each vertex. The updated adjacency matrix 𝐴 is then normalized using the min-max method.
Vertex feature values in the traditional GAT formulation [26] are thought of as being one-dimensional. GAT
computes attention ratings based on the vertices and their one-dimensional properties, which are then used to
create a fresh depiction of each network node [26]. We presume that the collection of node characteristics is
given as ℎ = {ℎ1, ℎ2, . . . , ℎ𝑁}, or 𝑁 is the number of nodes. ℎ𝑖 ∈ ℝ𝑑
where 𝑑 is the total number of
characteristics found in each node. By using a shared weight matrix, initially, a linear conversion of the
attribute values to a specified 𝑑′
number of attributes. A self-attention approach is then used to calculate the
attention scores of each vertex relative to the others, and add to the transformation of the feature values as a
dot product. More exactly, the attention score 𝛼𝑖,𝑗 is calculated in the following manner with the 𝑖−𝑡ℎ
and
𝑗−𝑡ℎ
vertex combination.
𝛼𝑖,𝑗 =
𝑒𝑥𝑝 ( 𝐿𝑒𝑎𝑘𝑦𝑅𝑒𝐿𝑈 (𝑎𝑇[𝑊ℎ𝑖||𝑊ℎ𝑗]))
∑ 𝑒𝑥𝑝 ( 𝐿𝑒𝑎𝑘𝑦𝑅𝑒𝐿𝑈 (𝑎𝑇[𝑊ℎ𝑖||𝑊ℎ𝑘]))
𝑘∈𝒩
(1)

6851
In (6) demonstrates the transformation of the characteristics of the 𝑖−𝑡ℎ
and 𝑗−𝑡ℎ
pair of vertex properties are
transformed into 𝑑’ dimensions and concatenated, i.e [𝑊ℎ𝑖||𝑊ℎ𝑗] ∈ ℝ2𝑑’
. Here, ℎ𝑖 ∈ ℝ𝑑
and ℎ𝑗 ∈ ℝ𝑑
consider the feature vectors of the 𝑖−𝑡ℎ
and 𝑗−𝑡ℎ
vertex respectively. 𝑊 ∈ ℝ𝑑×𝑑’
is the associated weight
matrix that modifies the properties of each vertex linearly. In GAT, a parameterized single-layer feed-
forward neural network models the self-attention process 𝑎 ∈ ℝ2𝑑’
. A scalar is created by the attention
process. The LeakyReLU nonlinearity is then used, with a 0.2 negative slope. The attention score is then
normalized with a softmax. The network adjacency matrix determines the neighborhood of the 𝑗−𝑡ℎ
vertex,
where all other vertices are taken into account while applying the softmax. In contrast, in our case, the
adjacency matrix is learned throughout the learning phase without restricting the connections between the
graph's vertices. As a result, in our example, all of the nodes are linked and the adjacency matrix components
are used to determine the strength of connections between nodes. The attention scores are multiplied by the
linearly changed features of the vertices, and the sum is then calculated over adjacent vertices. Thus, as
opposed to the traditional GAT formulation, the combination of the self-attention coefficients 𝑎 and the
concatenated nodes feature [𝑊ℎ𝑖||𝑊ℎ𝑗] will produce a vector of size 𝑑, or the number of weather variables,
as opposed to a scalar as it does in the traditional GAT formulation [26]. Subsequently, the following is done
to compute the new representation of the 𝑖−𝑡ℎ
vertex:
ĥ𝑖 = 𝜎(∑ 𝑊
𝑗∈𝒩 ℎ𝑗 𝑑𝑖𝑎𝑔(𝛼𝑖𝑗
1
, … , 𝛼𝑖𝑗
𝑑
)) (2)
where 𝛼𝑖𝑗
𝑑
is the difference in attention ratings between the 𝑖−𝑡ℎ
and 𝑗−𝑡ℎ
nodes. The final representation of
the nodes that we obtain after applying K-independent attention techniques and concatenating their outcomes
is as (8).
ĥ𝑖 = ‖𝑘=1
𝐾
𝜎(∑ 𝑊𝑘
𝑗∈𝒩 ℎ𝑗 [𝑑𝑖𝑎𝑔(𝛼𝑖𝑗
1
, … , 𝛼𝑖𝑗
𝑑
)]𝑘
) (3)
The final representation of the 𝑖−𝑡ℎ
node is then determined as (9).
{
𝐷
̂𝑖𝑖 = ∑ Â𝑖𝑗,
𝑗
ℎ
̃𝑖 = ĥ𝑖 (𝐷
̂
−1
2 𝐴
̂𝐷
̂
−1
2 )
(4)
The diagonal matrix D is computed here according to the results of the acquired Â in (5). The
symmetric normalization i.e., 𝐷
̂
−1
2 𝐴
̂𝐷
̂
−1
2 , is then done after this. To determine the new feature representation
ℎ
̃𝑖 the characteristics of the 𝑖−𝑡ℎ
node given by ĥ𝑖 is multiplied by the modified matrix that results. After
traversing a spatial layer composed of a flow GAT layer, the output is modified and routed into a BiLSTM
network. The GAT layers have been described extract spatial information from meteorological data,
Although the BiLSTM is used to enhance the model’s capacity to identify temporal correlations in the data
and how distinct observations at different time steps impact one another.
3. RESULTS AND DISCUSSION
In this section, we will provide the findings and talk about the numerical analysis. We used
information from the wind farm in Tetouan, Morocco, which was gathered on each wind turbine and
primarily included the following elements: wind power, wind turbine status, and weather. The original
meteorological data used to create this article has been pre-processed, mostly adding the contents of wind
speed, wind direction, and ambient temperature. These three categories of meteorological data are essential
for creating a very precise spatiotemporal wind power prediction using the proposed GAT-BiLSTM wind
forecast model.
Our simulation dataset was divided into training and test sets. As a result, we used 70% of the
training phase to introduce the model and 30% of the testing procedure to assess its effectiveness. The
suggested approach is put into practice using Python.
To improve and measure the efficiency of the forecasting model during a winter month, we
conducted comparative research using the three fundamental algorithms: back-propagation (BP) [27], support
vector machine (SVM) [28], and extreme learning machine (ELM) [29], which are the most often used in
wind energy forecasting. We included comparison research for a winter month, to investigate the
effectiveness of the suggested model and its capacity to perform better throughout the crucial season. We
contrasted the results of the BP, SVM, and ELM algorithms, respectively, and the models’ investigated
observations for a month in the winter, and the prediction results are shown in Figures 3 to 6.

 ISSN: 2088-8708
6852
Figure 3. Wind power predicted by BP Figure 4. Wind power predicted by SVM
Figure 5. Wind power predicted by ELM Figure 6. Wind power predicted by GAT-BiLSTM
According to these data, the GAT model provides accurate prediction measures for a month of
winter because its curve closely resembles that which was measured. Table 1 lists all forecasting outcomes
for a month of winter, based on the four models. The table displays the time convergence of each approach in
seconds as well as the MSE score, which can be calculated as (10):
𝑀𝑆𝐸 =
1
𝑛
∑ (𝑦𝑖
𝑛
𝑖=1 − 𝑦
̂𝑖)2
(5)
where the number of instances in the test set is 𝑛, the predicted output is 𝑦
̂𝑖, and the measured output is 𝑦𝑖.
Table 1. Comparison between models for a month in winter
Prediction method MSE Time convergence(s)
GAT-BiLSTM 0.1483 0.6582
ELM 0.2267 0.6416
SVM 0.8007 5.7495
BP 0.9135 6.1578
The results are displayed in Table 1 highlighting the execution of the proposed GAT-BiLSTM
algorithm concerning the MSE error. Due to their ability to significantly distinguish between test values and
projected results of other models, BP and SVM produced the largest MSE. For BP and SVM, the MSEs are
0.9251 and 0.8689 respectively. Also, the ELM error is about 0.3486. All these models have errors greater
than 0.1573 produced by GAT-BiLSTM.
In addition, the biggest time convergence value is reached by BP and SVM, which take a long time
to compute and need several iterations. On the other hand, ELM and the proposed model give the lowest
values, but it still gives a smaller convergence time. Although the suggested GAT-BiLSTM method produced
comparatively better predictions than the persistent models and converged quickly, all of these results
demonstrated the reliability of the approach.

6853
4. CONCLUSION
The most common renewable energy source must be wind energy for the power system to operate
safely, but this is still exceedingly challenging given the volatility of wind speeds and their frequent
interruptions. Because of this, predicting wind speed has become crucial for the efficient use of energy. In
light of this, we have included a power forecasting model for the production of wind energy in this post. The
proposed model attempted to predict the subsequent wind energy generated by the wind turbines using the
GAT and BiLSTM deep leaning, the so-called GAT-BiLSTM to benefit from historical weather data’s spatial
and temporal characteristics. The execution of the recommended GAT-BiLSTM algorithm was emphasized
by the simulation results since it generated comparatively better predictions than the other examined
algorithms. It is an extremely quick, potent, and dynamic learning algorithm. As a consequence, we can
conclude that the GAT-BiLSTM may be used quite flexibly in the field of wind power forecasting. However,
this model can give more accurate results if we optimize the hyperparameters of the whole model. For this
reason, in future works, we think to use an optimization algorithm like a genetic algorithm to improve the
prediction precision. Moreover, we also think to enhance our model by using several wind farms in different
cities.
REFERENCES
[1] P. Bauer, A. Thorpe, and G. Brunet, “The quiet revolution of numerical weather prediction,” Nature, vol. 525, no. 7567,
pp. 47–55, Sep. 2015, doi: 10.1038/nature14956.
[2] X. Chen, M. Wei, and J. Sun, “Workflow-based platform design and implementation for numerical weather prediction models and
meteorological data service,” Atmospheric and Climate Sciences, vol. 07, no. 03, pp. 337–351, 2017, doi:
10.4236/acs.2017.73025.
[3] A. G. Salman, B. Kanigoro, and Y. Heryadi, “Weather forecasting using deep learning techniques,” in 2015 International
Conference on Advanced Computer Science and Information Systems (ICACSIS), Oct. 2015, pp. 281–285, doi:
10.1109/ICACSIS.2015.7415154.
[4] A. Saeed, C. Li, Z. Gan, Y. Xie, and F. Liu, “A simple approach for short-term wind speed interval prediction based on
independently recurrent neural networks and error probability distribution,” Energy, vol. 238, Jan. 2022, doi:
10.1016/j.energy.2021.122012.
[5] Z. Tian, H. Li, and F. Li, “A combination forecasting model of wind speed based on decomposition,” Energy Reports, vol. 7,
pp. 1217–1233, Nov. 2021, doi: 10.1016/j.egyr.2021.02.002.
[6] J. Zhao, Z. Yan, X. Chen, B. Han, S. Wu, and R. Ke, “k-GCN-LSTM: a k-hop graph convolutional network and long–short-term
memory for ship speed prediction,” Physica A: Statistical Mechanics and its Applications, vol. 606, Nov. 2022, doi:
10.1016/j.physa.2022.128107.
[7] R. Hrasko, A. G. C. Pacheco, and R. A. Krohling, “Time series prediction using restricted Boltzmann machines and
backpropagation,” Procedia Computer Science, vol. 55, pp. 990–999, 2015, doi: 10.1016/j.procs.2015.07.104.
[8] D. Qiao, S. Wu, G. Li, J. You, J. Zhang, and B. Shen, “Wind speed forecasting using multi-site collaborative deep learning for
complex terrain application in valleys,” Renewable Energy, vol. 189, pp. 231–244, Apr. 2022, doi: 10.1016/j.renene.2022.02.095.
[9] Z. Duan, H. Liu, Y. Li, and N. Nikitas, “Time-variant post-processing method for long-term numerical wind speed forecasts based
on multi-region recurrent graph network,” Energy, vol. 259, Nov. 2022, doi: 10.1016/j.energy.2022.125021.
[10] S. Siami-Namini, N. Tavakoli, and A. S. Namin, “The performance of LSTM and BiLSTM in forecasting time series,” in 2019
IEEE International Conference on Big Data (Big Data), Dec. 2019, pp. 3285–3292, doi: 10.1109/BigData47090.2019.9005997.
[11] M. Billah, X. Wang, J. Yu, and Y. Jiang, “Real-time goat face recognition using convolutional neural network,” Computers and
Electronics in Agriculture, vol. 194, Mar. 2022, doi: 10.1016/j.compag.2022.106730.
[12] P. Lara-Benítez, M. Carranza-García, J. M. Luna-Romera, and J. C. Riquelme, “Temporal convolutional networks applied to
energy-related time series forecasting,” Applied Sciences, vol. 10, no. 7, Mar. 2020, doi: 10.3390/app10072322.
[13] J. Fox and S. Rajamanickam, “How robust are graph neural networks to structural noise?” arXiv preprint arXiv:1912.10206, Jan.
2020.
[14] C. Zhuang and Q. Ma, “Dual graph convolutional networks for graph-based semi-supervised classification,” in Proceedings of the
2018 World Wide Web Conference on World Wide Web - WWW ’18, 2018, pp. 499–508, doi: 10.1145/3178876.3186116.
[15] Z. Chen et al., “Spatial–temporal short-term traffic flow prediction model based on dynamical-learning graph convolution
mechanism,” Information Sciences, vol. 611, pp. 522–539, Sep. 2022, doi: 10.1016/j.ins.2022.08.080.
[16] X. Wang et al., “Heterogeneous graph attention network,” in The Web Conference 2019 - Proceedings of the World Wide Web
Conference, WWW 2019, Mar. 2019, pp. 2022–2032, doi: 10.1145/3308558.3313562.
[17] H. Peng et al., “Spatial temporal incidence dynamic graph neural networks for traffic flow forecasting,” Information Sciences,
vol. 521, pp. 277–290, Jun. 2020, doi: 10.1016/j.ins.2020.01.043.
[18] Z. Yang and S. Dong, “HAGERec: hierarchical attention graph convolutional network incorporating knowledge graph for
explainable recommendation,” Knowledge-Based Systems, vol. 204, Sep. 2020, doi: 10.1016/j.knosys.2020.106194.
[19] C. Jin et al., “HetGAT: a heterogeneous graph attention network for freeway traffic speed prediction,” Journal of Ambient
Intelligence and Humanized Computing, Jan. 2021, doi: 10.1007/s12652-020-02807-0.
[20] F. Zhou, Q. Yang, T. Zhong, D. Chen, and N. Zhang, “Variational graph neural networks for road traffic prediction in intelligent
transportation systems,” IEEE Transactions on Industrial Informatics, vol. 17, no. 4, pp. 2802–2812, Apr. 2021, doi:
10.1109/TII.2020.3009280.
[21] F. Wang et al., “Graph attention convolutional neural network model for chemical poisoning of honey bees’ prediction,” Science
Bulletin, vol. 65, no. 14, pp. 1184–1191, Jul. 2020, doi: 10.1016/j.scib.2020.04.006.
[22] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, Nov. 1997,
doi: 10.1162/neco.1997.9.8.1735.
[23] D. El Bourakadi, A. Yahyaouy, and J. Boumhidi, “Intelligent energy management for micro-grid based on deep learning LSTM
prediction model and fuzzy decision-making,” Sustainable Computing: Informatics and Systems, vol. 35, Sep. 2022, doi:
10.1016/j.suscom.2022.100709.

 ISSN: 2088-8708
6854
[24] W. Lu, J. Li, J. Wang, and L. Qin, “A CNN-BiLSTM-AM method for stock price prediction,” Neural Computing and
Applications, vol. 33, no. 10, pp. 4741–4753, May 2021, doi: 10.1007/s00521-020-05532-z.
[25] A. Kulshrestha, V. Krishnaswamy, and M. Sharma, “Bayesian BILSTM approach for tourism demand forecasting,” Annals of
Tourism Research, vol. 83, Jul. 2020, doi: 10.1016/j.annals.2020.102925.
[26] S. Brody, U. Alon, and E. Yahav, “How attentive are graph attention networks?” arXiv preprint arXiv:2105.14491, May 2021.
[27] G. Li and J. Shi, “On comparing three artificial neural networks for wind speed forecasting,” Applied Energy, vol. 87, no. 7,
pp. 2313–2320, Jul. 2010, doi: 10.1016/j.apenergy.2009.12.013.
[28] M. Liu, Z. Cao, J. Zhang, L. Wang, C. Huang, and X. Luo, “Short-term wind speed forecasting based on the Jaya-SVM model,”
International Journal of Electrical Power and Energy Systems, vol. 121, Oct. 2020, doi: 10.1016/j.ijepes.2020.106056.
[29] D. El Bourakadi, A. Yahyaouy, and J. Boumhidi, “Multi-agent system based on the extreme learning machine and fuzzy control
for intelligent energy management in microgrid,” Journal of Intelligent Systems, vol. 29, no. 1, pp. 877–893, Dec. 2019, doi:
10.1515/jisys-2018-0125.
BIOGRAPHIES OF AUTHORS
Ibtissame Mansoury obtained a master’s degree in Imaging and Business
Intelligence from Sidi Mohamed Ben Abdellah University, Morocco, in 2017. Now she is a
Ph.D. student at Sidi Mohammed Ben Abdellah University, Morocco, 2017. Her research
interests include artificial intelligence and smart grids. She can be contacted at email:
ibtissame.mansoury@usmba.ac.ma.
Dounia El Bourakadi received her Ph.D. degree in 2020 from the University Sidi
Mohamed Ben Abdellah (USMBA), Fez, Morocco. Her research interests are in artificial
intelligence, optimization, and energy management. She can be contacted at email:
dounia.elbourakadi@usmba.ac.ma.
Ali Yahyaouy is a professor at the Faculty of Sciences Dhar El Mahraz, Sidi
Mohamed Ben Abdellah University, Morocco. He obtained his Ph.D. degree at the University
of Technology of Belfort-Montbéliard, France, and Sidi Mohamed Ben Abdellah University,
Morocco in 2010. His research activities mainly concern machine learning, multi-agent
systems, smart grids, and ITS. He can be contacted at email: ali.yahyaouy@usmba.ac.ma.
Jaouad Boumhidi is a Professor of Computer Science at the Faculty of Sciences,
Fez Morocco. He received his Ph.D. in Computer Science from The University of Sidi Mohamed
ben Abdellah in 2005. His research interests are in machine learning, deep learning, and
intelligent transportation systems. He can be contacted at email: jaouad.boumhidi@usmba.ac.ma.

A novel wind power prediction model using graph attention networks and bi-directional deep learning long and short term memory

More Related Content

Similar to A novel wind power prediction model using graph attention networks and bi-directional deep learning long and short term memory (20)

More from IJECEIAES (20)

Recently uploaded (20)

A novel wind power prediction model using graph attention networks and bi-directional deep learning long and short term memory