=Paper=
{{Paper
|id=Vol-3734/invited8
|storemode=property
|title=Using LSTM Model to predict Short-term wind power
|pdfUrl=https://ceur-ws.org/Vol-3734/paper8.pdf
|volume=Vol-3734
|authors=Zhiqi Wan,Jiali Wang,Jialin Tang,Bo Zhang,Feilin Zhu
|dblpUrl=https://dblp.org/rec/conf/iccic/WanWTZZ24
}}
==Using LSTM Model to predict Short-term wind power==
https://ceur-ws.org/Vol-3734/paper8.pdf
Using LSTM Model to predict Short-term wind power
Zhiqi Wan1, Jiali Wang1, Jialin Tang1, Bo Zhang1 and Feilin Zhu2, β
1 DAYU College, Hohai University, Nanjing, China.
2 College of Hydrology and Water Resources, Hohai University, Nanjing, China
Abstract
Wind energy is an essential renewable energy. For Carbon Neutrality, wind power has received
extensive attention around the world. However, short-term wind power time series are difficult
to predict with complex characteristics such as non-stationary and nonlinear. Therefore, this
paper proposes a short-term prediction method of wind power using the Long Short-term
Memory (LSTM) model. In view of insufficient memory ability, gradient disappearance and
explosion in traditional prediction methods, the strategy of "Data Processing - Autocorrelation
Analysis - Model Prediction" is proposed. Firstly, test the outliers by Z-Score method, and
linearly interpolate the missing values and outliers. Secondly, determine the model input length
through autocorrelation and partial autocorrelation coefficients. Finally, predict each
subsequence with the LSTM model. On the testing set, the root mean square error is 58.55
(MW), mean absolute error is 79.60 (MW), and coefficient of determination is 0.86. In brief,
using LSTM prediction model can obtain a higher accuracy of short-term wind power prediction.
Keywords
Long Short-term Memory, Short-term Wind Power Prediction, Data Processing
1. Introduction
Energy is an essential foundation for human beings. With global climate change,
environmental pollutions and other issues becoming more and more prominent, the
transformation of the energy structure is imminent. Recently, many countries have been
vigorously developing new energy. Wind energy, as an important renewable energy
source, has the advantages of a long history of research, many technological innovations
and broad development prospects, and is crucial for promoting energy transformation.
However, wind energy is characterized by intermittency, volatility, and randomness. The
grid integration of wind power intensifies the pressure of peak regulation. In order to
ensure national energy security, accurate short-term wind power prediction is an
important challenge.
ICCIC 2024: International Conference on Computer and Intelligent Control, June 29β30, 2024, Kuala Lumpur,
Malaysia
β Corresponding author.
2119010405@hhu.edu.cn (Z. Wan); 1093347909@qq.com (J. Wang); tangjialin2003@163.com (J. Tang);
1970649539@qq.com@163.com (B. Zhang); zhufeilin163@163.com (F. Zhu)
0009-0001-6180-0031 (Z. Wan); 0009-00003566-1931 (J. Wang); 0009-0007-7897-1674 (J. Tang); 0009-
0008-0174-5000 (B. Zhang); 0000-0002-9780-9361 (F. Zhu)
Β© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
Many scholars conducted in-depth research on the wind power prediction and
proposed many models and methods. According to the principal classification, they can be
divided into physical methods based on numerical weather predictions and statistical
methods with historical data [1]. The physical prediction method mainly predicts wind
power by screening related physical quantities and establishing space-time physical
equations. This method does not require historical wind power data, but is limited by the
accuracy of measurement data, modeling errors and economic conditions of complex wind
farms. It is mostly applied to new wind farms lacking historical data [2-4]. Statistical
prediction methods mainly include Time-Series Analysis, Kalman filter, Artificial Neural
Network, etc. [5-12], which is based on the learning of historical data and does not need to
consider the complex physical calculation process. It has the characteristics of self-
adaptation, self-adjustment, and self-learning. The model structure is simple and suitable
for wind farms with historical data. Many studies have shown that statistical prediction
methods have higher applicability and accuracy in short-term wind power prediction [13].
As deep learning theories develop, the accuracy of wind power prediction has been
continuously improved [14-16]. Long Short-Term Memory neural network is an
improvement of traditional RNN. It retains important features through the gate function,
and effectively alleviates the problems of insufficient long sequence memory capacity,
gradient disappearance and explosion. For nonlinearly varying wind power time series,
the accuracy of LSTM model is better.
Summarily, according to the idea of "Data Processing - Autocorrelation Analysis - Model
Prediction", this paper proposes a short-term prediction method of wind power based on
LSTM model. Taking the annual hourly data of a power station in Qinghai Province as an
example, the autocorrelation of wind power historical data is proved. The applicability and
accuracy of the model are also analyzed and tested, and the overall idea is shown in Figure
1.
Figure 1: Framework of wind power prediction using LSTM model.
2. Research Methods
2.1 Data Processing
For data-driven short-term prediction models, missing values and outliers in the historical
data can seriously affect the accuracy. Therefore, data processing is required before the
model forecasting.
For missing values, this paper adopts the method of linear interpolation. It is assumed
that the missing values can be represented by the data of its previous and next moments
through a linear relationship, given by the formula:
(1)
In the equation, is the missing value, , are the previous data and next data.
is the moment corresponding to missing value. is the previous period and is
the next period.
For outliers testing, this paper uses the Z-Score method, and the formula as:
(2)
In the equation, is the Z-Score value. is the th data in the time series. is the
mean value of the time series. is the standard deviation of the time series. The threshold
of Z-Score is taken as 3. When , the data value is considered to have a large
difference with other values and is regarded as an outlier. For outliers, the treatment is the
same as for missing values, with the linear interpolation method.
2.2 Correlation Analysis
The core of the predicted model with historical data is to find the patterns hidden in the
time series and then predict the data of next period based on the discovered patterns.
Therefore, before using such prediction models, it is necessary to ensure that the time
series exhibit autocorrelation. To test the autocorrelation of historical wind power data,
autocorrelation coefficient is introduced, like:
(3)
In the equation, is the th autocorrelation coefficient, representing the correlation
between data points separated by time units; is covariance; is variance; is the
lag order; and are the observed values at time moments and . By using
autocorrelation coefficients, a linear relationship between the perdiction point and the
observed point is constructed as follows:
(4)
In the equation, represents the white noise error term. When , it indicates
complete correlation between the prediction point and the observed point, meaning data
at one time point can be entirely predicted by data at another time point; when , it
indicates no correlation between and ; when , it indicates complete
negative correlation between and , implying data at one time point can be entirely
reversely predicted by data at another time point.
Since time series are correlated, autocorrelation coefficients cannot represent the
correlation between and without the influence of other time series. To eliminate
interference from on the correlation between two points, partial
autocorrelation coefficient is introduced as follows:
(5)
Where is the th partial autocorrelation coefficient, representing the correlation
between data points separated by time while considering the situation of the previous
time points; is the th term in the th order partial autocorrelation coefficient;
is the summation index; and is the order of autocorrelation and partial autocorrelation
functions, representing the time interval to be calculated. Based on the definition of partial
autocorrelation, a linear relationship between the and all observed points within the
previous time units is constructed as follows:
(6)
2.3 LSTM Model
Long Short-Term Memory (LSTM) is an improvement of recurrent neural network (RNN).
Compared to traditional neural networks, LSTM neural networks have more effective
memory and forgetting patterns for long time series. By introducing gate mechanisms,
LSTM can better capture long-term dependencies, effectively addressing issues like
gradient disappearance and explosion. Additionally, for nonlinear system time series
prediction, LSTM neural networks also have significant advantages[17].
The key components of LSTM neural networks include forget gate, input gate, cell state,
and output gate, which together form the unique four-layer structure of LSTM, as shown in
Figure 2.
Figure 2: Structure of LSTM .
According to the flow of information within neurons, the four-layer structure is
explained as follows:
1. Forget Gate: Distinguish and forget minor information.
(7)
2. Input Gate: Determine and filter new information stored in the cell state.
(8)
3. Cell State: Update the cell state based on the previous cell state and input gate
information.
(9)
4. Output Gate: Select the information to be input to the next neuron.
(10)
Where π€π€ππ , π€π€ππ , π€π€ππ , π€π€ππ are weight matrices; ππππ , ππππ , ππππ , ππππ are bias vectors; βπ‘π‘ , βπ‘π‘β1
represent the input and output of previous neuron and current neuron; πΆπΆπ‘π‘ , πΆπΆπ‘π‘β1 represent
the cell states of the previous neuron and the current neuron; πΆπΆΜπ‘π‘ is the cell state of the
input gate; is the sigmoid function.
2.4 Performance Evaluation
To quantitatively evaluate the prediction effectiveness, three evaluation metrics are
selected, including Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and
Coefficient of Determination ( ), defined as follows:
(11)
(12)
(13)
Where is the number of samples in the testing set; is the actual wind power; is
the prediction value; and is the mean of the wind power historical sequence.
3. Research Example
This paper collects hourly data of wind power in Qinghai Province throughout the year.
Through the detection of missing values and outliers, there are 5 missing values and 7
outliers. With linear interpolation processing, 8785 normal data can be obtained.
Due to the limitations of data and algorithms, the prediction model often has problems
such as underfitting and overfitting. To improve accuracy, the data set needs to be divided
into a training set and a testing set. The division of the data set and the amount of data are
shown in Table 1.
Table 1
Experimental data sets
Location of the Number of samples
Data type Time span
wind farm (strip)
All data Jan. 1 to Dec. 31 8785
Qinghai Province Training set Jan. 1 to Nov. 26 7902
Testing set Nov. 26 to Dec. 31 883
4. Results and Discussion
4.1 Correlation Analysis
Bring the 8,785 processed data sets into Eq. (1), and set the significance level .
Draw the autocorrelation analysis diagram of historical data, as shown in Figure 3. The
diagram shows that most of the observation points are outside the significance band. It
proves that the historical data of wind power has strong autocorrelation and can be used
as model inputs. With the increase of the time, the autocorrelation coefficient first
decreases and then oscillates around the value of 0, indicating that the closer date to the
prediction point are more likely to affect the prediction results, and wind power has a
certain periodic pattern on the time scale.
Figure 3: Autocorrelation analysis of wind power historical data.
Bring the historical data into Eq (3), set the significance level , and draw the
partial autocorrelation analysis diagram of historical data, as shown in Figure 4. The
diagram shows that with the increase of time, the partial autocorrelation coefficient and
autocorrelation coefficient change in a similar trend, but the partial autocorrelation
coefficient decreases faster and the oscillation peak is smaller. This indicates that if only
consider the partial autocorrelation coefficient between the two points, the prediction
point is more correlated with the observation points in the first 2 time intervals.
Figure 4: Partial autocorrelation analysis of wind power historical data.
Considering the high accuracy and timeliness requirements of short-term prediction
model, and the model inputs have a major influence on the short-term prediction accuracy,
selecting the appropriate length of input data is an important prerequisite. Since the
historical data input to the model is correlated with each other, the autocorrelation
coefficient is mainly considered when selecting the length of input data. However, if the
input data is too long, it will lead to error superposition. Therefore, considering the
autocorrelation and partial autocorrelation coefficients comprehensively, model input
length should not exceed 15 time series.
4.2 Model Training and Prediction
Taking the hourly wind power of Qinghai Province as an example, the predicted step size
is 1, and brought it into LSTM neural network for manual parameter calibration. According
to the above three evaluation indexes, the activation function is set as the "Relu" function,
the time step is 10, the dimension of LSTM layer is 128, epoch=12, batch_size=32. The
model prediction result under this parameter is shown in Figure 4, and the analysis of
indexes is shown in Table 2.
(a) Training Set
(b) Testing Set
Figure 5: Comparison of LSTM model prediction results for training set and testing set
Table 2
Calculation results of evaluation indexes
Training Sets 65.72 55.15 0.79
Testing Sets 58.55 79.60 0.86
From Fig. 5 and Table 2, the LSTM prediction model with historical data has a better
prediction accuracy of wind power. Therefore, when other information is missing and only
historical time series are available, the LSTM prediction model has significant advantages
of high accuracy and convenient calculation.
For the power grid system implementing regulation, the higher the accuracy and the
longer the forecast period, the greater the benefit. For better regulation of the βWind -
Light - Waterβ system, the prediction accuracy is still a work in progress. Based on the
research in this paper, in-depth research, such as input data feature extraction, prediction
model coupling, parameter optimization, can be promoted subsequently.
5. Conclusion
Accurate short-term wind power prediction is a prerequisite for implementing βWind -
Light - Waterβ complementary optimal scheduling. According to the idea of "Data
Processing - Autocorrelation Analysis - Model Prediction", this paper constructs a LSTM
prediction model based on historical data. Taking the annual hourly data of a power
station in Qinghai Province as an example, this paper analyzes and tests the applicability
and accuracy, and the following conclusions are obtained:
(1) The historical data of wind power has autocorrelation and periodicity, making it
suitable for use as input data in prediction models..
(2) If the interval between the predicted point and the observed point is smaller, the
autocorrelation coefficient is larger and the autocorrelation is stronger. Therefore, several
observation points close to the front of the prediction point should be selected for
prediction. For each prediction point, the number of input observation points should not
exceed 15.
(3) When the available data is limited, the LSTM prediction model can obtain the single-
value prediction of wind power with high precision in a short time.
Acknowledgment
This research was funded by the College Student Innovation and Entrepreneurship
Training Program of Hohai University: Ultra-short-term probabilistic prediction method of
wind and photovoltaic power in watersheds under the background of "dual-carbon"
strategy Nexus (No. 202310294022Z).
References
[1] J. Yan, Y. Liu, S. Han, et al., βReviews on uncertainty analysis of wind power
forecasting,β Renewable and Sustainable Energy Reviews, pp. 521322-1330, 2015.
[2] G. Giebel, J. Badger, I. MartiP, βShort-term forecasting using advanced physical
modeling-the results of the anemos project,β European Wind Energy Conference &
Exhibition, Athens, pp. 236-264, 2006.
[3] L. L. Li, X. Zhao, M. L. Tseng, et al., βShot-term wind power forecasting based on
support vector machine with improved dragonfly algorithm,β Journal of Cleaner
Production, vol. 242 pp. 38-47, 2020.
[4] L. Ye, Y. Zhao, C. Zeng, et al., βShort-term wind power prediction based on spatial
model,β Renew Energy, vol. 101 pp. 1067-74, 2017.
[5] E. Erden, J. Shi, βARMA based approaches for forecasting the tuple of wind speed and
direction,β Applied Energy, vol. 88, no. 4, pp. 1405-1414, 2011.
[6] D. A. Bechrakis, P. D. Sparis, βWind speed prediction using artificial neural networks,β
Wind Engineering, vol. 22, no. 6, pp. 287-295, 1998.
[7] E. A. Bossanyi, βShort-term wind prediction using Kalman filters,β Wind Engineering,
vol. 9, no. 1, pp. 1-8, 1985.
[8] E. T. Renani, M. F. M. Elias, N. A. Rahim, βUsing data-driven approach for wind power
prediction: A comparative study,β Energy Conversion & Management, vol. 118, pp.
193-203, 2016.
[9] A. E. Saleh, M. S. Moustafa, K. M. Abo-Al-Ez, et al., βA hybrid neuro-fuzzy power
prediction system for wind energy generation,β International Journal of Electrical
Power & Energy Systems, vol. 74, pp. 384-395, 2016.
[10] S. Wang, M. Li, L. Zhao, et al., βShort-term wind power prediction based on improved
small-world neural network,β Neural Computing & Applications, vol. 31, pp. 3173-
3185, 2019.
[11] Y. Hao, L. Dong, X. Liao, et al., βA novel clustering algorithm based on mathematical
morphology for wind power generation prediction,β Renewable Energy, vol. 136, pp.
572-585, 2019.
[12] C. D. Zuluaga, M. A. Γlvarez, E. Giraldo, βShort-term wind speed prediction based on
robust Kalman filtering: An experimental comparison,β Applied Energy, vol. 156, 2015.
[13] J. Duan, P. Wang, W. Ma, et al., βA novel hybrid model based on nonlinear weighted
combination for short-term wind power forecasting,β International Journal of
Electrical Power and Energy Systems, vol. 134, 2022.
[14] M. Gong, C. Yan, W. Xu, et al., βShort-term wind power forecasting model based on
temporal convolutional network and Informer,β Energy, vol. 283, 2023.
[15] S. Wang, J. Shi, W. Yang, et al., βHigh and low frequency wind power prediction based
on Transformer and BiGRU-Attention,β Energy, vol. 288, pp. 129753, 2024.
[16] L. Xiang, X. Fu, Q. Yao, et al., βA novel model for ultra-short term wind power
prediction based on Vision Transformer,β Energy, vol. 294, pp. 294130854, 2024.
[17] S. Hochreiter, J. Schmidhuber, βLong short-term memory,β Neural computation, vol. 9,
no. 8, pp. 1735-1780, 1997. DOI:10.1162/ neco.1997. 9. 8. 1735.