The operation complexity of the distribution system increases as a large number of distributed generators (DG) and electric vehicles were introduced, resulting in higher demands for fast online reactive power optimization. In a power system, the characteristic selection criteria for power quality disturbance classification are not universal. The classification effect and efficiency needs to be improved, as does the generalization potential. In order to categorize the quality in the power signal disturbance, this paper proposes a multilayer severe learning computer autoencoder to optimize the input weights and extract the characteristics of electric power quality disturbances. Then, a multilabel classification algorithm based on rating is proposed to understand the relationship between the labels and identify the various power quality disturbances. The two algorithms are combined to construct a multilabel classification model based on a multilevel extreme learning machine, and the optimal network structure of the multilevel extreme learning machine as well as the optimal multilabel classification threshold are developed. The proposed method can be used to classify the single and compound power quality disturbances with improved classification effect, reliability, robustness, and antinoise performance, according to the experimental results. The hamming loss obtained by the proposed algorithm is about 0.17 whereas MLRBF, SVM and MLKNN schemes have 0.28, 0.23 and 0.22 respectively at a noise intensity of 20 dB. The average precision obtained by the proposed algorithm 0.85 whereas the MLRBF, SVM and MLKNN schemes indicates 0.7, 0.77 and 0.78 respectively.
With the continuous development of the power system and the diversification of power access forms, the power quality is getting worse and worse. At the same time, various electrical equipment has extremely high requirements for power quality standards. Therefore, identifying and classifying the power quality disturbance signals accurately and quickly is a prerequisite for ensuring stable, safe, and efficient operation of the power grid [
Machine learning has advanced rapidly in the academia and industry in recent years. Its ability to perform complex recognition tasks has been demonstrated by a large increase in recognition rate on many typical recognition tasks. A significant number of academics have flocked to research its hypotheses and applications. Deep learning is being used in a variety of fields to solve some of the problems in this area. Machine learning research has started to appear in the area of regulation [
Precise power consumption forecasting not only plays a decisionmaking role in the realtime power dispatching, but also an effective way to ensure the grid stability, solve power deviations, and save the energy. At present, the state is advancing to achieve the power deviation control from demand as the entry point, and the focus of power sales companies is on power consumption forecasting. The realtime and accuracy of the nexttime user power consumption forecast is not only one of the realization goals of demandside management, but also has positive significance for the safe and stable operation of the power system and the improvement of economic benefits. Due to the characteristics of randomness and uncertainty in shortterm power consumption of users, the choice of forecasting method and its ability to overcome the randomness directly affect the implementation quality of realtime and accurate power dispatch [
Because of the rapid growth of largescale wind farms, wind energy is playing an increasingly important role in domestic and international power markets as a sustainable and costeffective renewable energy source. Wind's highly unpredictable capacity, on the other hand, can trigger nonlinear characteristics in the wind power, which can have a number of negative consequences for the wind power system's reliability [
With the use of a large number of nonlinear and unbalanced loads in the power system, the impact of power quality on the safe and stable operation of the power system has become more obvious. The realtime and effective classification of power quality disturbances has become the basis for further improvement of power quality. In recent years, the classification of power quality disturbances has been well studied [
In the field of quality disturbance classification, commonly used feature extraction algorithms include Fourier transform [
The above algorithms require feature selection of the signal, which is prone to feature redundancy, and some other features will be lost, resulting in a decrease in recognition rate and reduced antinoise performance. In order to avoid the process of selecting power quality disturbance features, literature [
Power quality disturbance classification in distribution networks is one of the important aspect which is in the focus of both academia and industry. Various methods have been proposed but their classification efficiency is lower. In order to improve the classification efficiency and improve the power quality, this paper focuses on the improvement of classification accuracy and efficiency as the research goal, and proposed a multilayer extreme learning machine selfencoding network structure, and rankingbased power quality disturbance classification model, and compares it with other classification methods on power quality disturbance classification. The comparison verifies the effectiveness of the proposed method.
The remaining of the paper is organized as follows. In Section 2, the system model is described. In Section 3, the numerical results are provided in detail. Section 4 provides the application scenario while Section 5 concludes the paper.
Multilayer extreme learning machine based on selfencoding, which can effectively represent the complex functions, has high prediction accuracy and generalization ability. The rankingbased multilabel classification algorithm, taking into account the correlation between the labels, is suitable for the classification of various power quality disturbances, and has a strong antinoise ability. This paper combines the multilayer extreme learning machine with the multilabel classification algorithm, and proposes a multilabel classification model based on the multilayer extreme learning machine.
The extreme learning machine autoencoder (ELMAE) optimizes the input weight, improves the classification accuracy and generalization ability, overcomes the problem of neuron invalidity caused by the random weight and hidden layer threshold of the extreme learning machine, and improves the classification efficiency [
The network structure of the extreme learning machine autoencoder is shown in
Suppose there is an arbitrary sample space
Introduce the orthogonal random weight matrix
By reconstructing the matrix
The
The multilayer extreme learning machine autoencoder (MLELMAE) performs stacking operations on the basis of ELMAE [
The first step is to use the set of input data samples to calculate the output matrix
In the second step,
The third step, and so on, takes the output matrix
In multilabel classification algorithm, a single data sample belongs to the multiple category labels, and fully considering the correlation between the labels [
Assuming that the sample space is
Based on the rankbased multilabel classification algorithm [
For the sample
It can be observed that the predicted label set of the sample
Evaluation indicators to measure the effect of multilabel classification algorithms [
Hamming loss
A one type of error
The ranking loss
The coverage rate
In this paper, the selfencoding structure of the multilayer extreme learning machine and the rankingbased multilabel classification algorithm are combined to construct the basic structure of the classifier model as shown in
According to the model in
In the process of classifying the power quality disturbances, the MLELMAE algorithm determines the optimal network structure through experiments based on the set of input data samples, and generates the output matrix and weight matrix of each hidden layer. The rankingbased multilabel classification algorithm determines the optimal classification threshold through experiments [
This paper uses the MATLAB simulator to compute the simulation experiment, using average accuracy, hamming loss, firstclass error, ranking loss, and coverage as the evaluation indicators of the classification effect. In addition, the experiment also increases the training time and test time of the classifier as a measure of classification efficiency evaluation index. In order to reduce the error, each experimental data is the arithmetic mean of the algorithm program running 20 times under the same experimental conditions.
The common single power quality disturbance signal model [
Type  Mathematical model  Parameter 
Standard signal  
Voltage dip  
Voltage swell  
Voltage interruption  
Harmonic  
Transient oscillation  


Transient pulse  1 ms 

Voltage fluctuation  0.05 

0.1 
The signal in the experiment is sampled with 3200 Hz as the sampling frequency, 30 cycles are sampled in total, and the total number of sampling points is 1921. All signals are randomly generated 200 samples according to the mathematical model, 100 are used as training data, and the other 100 are used as test data. The training data set and the test data set each contain 4800 samples.
In this paper, the optimal network structure of MLELMAE is evaluated through experiments, and the multilabel mapping function
In this paper, the number of hidden layers of MLELMAE is limited to 9 layers, the number of nodes in each layer is set to 50∼2000, and the number of nodes is changed by 50. The Hamming loss is used as the basis for the evaluation of the classification effect, and the training time and the test time are referred to at the same time, and the network structure parameters of the multilayer extreme learning machine with good classification effect and high efficiency are selected.
In order to obtain the optimal number of nodes in each layer when the total number of layers is different, the following experimental steps are designed:
Set the number of hidden layers of MLELMAE to 1.
Set the number of hidden layer nodes to change according to the law of 50 +
Set the number of layers of MLELMAE to 2, where the number of nodes in the first layer has been fixed by steps 1 and 2, repeat step 2, get the optimal
Set the number of layers of MLELMAE to 3∼9 in turn, and get the optimal number of nodes for each layer.
No. of node  

1  350  –  –  –  –  –  –  –  – 
2  350  1150  –  –  –  –  –  –  – 
3  350  1150  350  –  –  –  –  –  – 
4  350  1150  350  750  –  –  –  –  – 
5  350  1150  350  750  750  –  –  –  – 
6  350  1150  350  750  750  750  –  –  – 
7  350  1150  350  750  750  750  750  –  – 
8  350  1150  350  750  750  750  750  750  – 
9  350  1150  350  750  750  750  750  750  750 
Based on the above experimental results, we can observe that, there is not a single quantitative relationship between the number of hidden layers and the number of nodes in MLELMAE and the classification results. There is an optimal number of hidden layers and nodes, which saves training and testing time while ensuring the classification effect and improved efficiency. In this paper, when the number of hidden layers of MLELMAE is 4, the power quality disturbance classification effect is the best and the efficiency is higher. Therefore, this paper sets the multilayer extreme learning machine as “input layerhidden layer (350115035075)” structure of the output layer.
The experiment adopts the optimal 4layer MLELMAE network structure, and the classification threshold of the multilabel classifier is set in the range of 0.05 to 0.95, with 0.05 as the step interval. The Hamming loss, ranking loss, firstclass error, coverage rate, and average accuracy are used as the evaluation basis for the classification effect.
Based on the above experimental results, it can be obtained that the classification threshold has a more obvious impact on the Hamming loss, the coverage rate and average accuracy are affected by the change of the classification threshold to a certain extent, and the first type of errors and ranking errors are not significantly affected by the classification threshold. In this paper, the classification threshold of the multilabel classifier is set to 0.6.
The multilabel classification model of the multilayer extreme learning machine is deployed to predict the disturbance category contained in the power quality disturbance signal [
In the first step, the power quality disturbance signal is subjected to discrete wavelet transform (DWT) to obtain the decomposition coefficients of each layer, and the decomposition coefficients are divided into training data sets and test data sets.
The second step is to randomly generate the weights of each layer of the orthogonalized MLELMAE network.
The third step is to train the multilayer extreme learning machine network with the training data set as input, and adjust the weights of each layer of the network.
The fourth step is to obtain the classification function of the multilabel classification algorithm according to the mapping relationship of MLELMAE.
Finally, the test data set is used as input, and the multilabel classification model of the trained multilayer extreme learning machine is used to predict the disturbance category.
The experimental research objects are a single disturbance signal and a composite disturbance signal, and noise disturbances with a signaltonoise ratio (SNR) of 50, 40, 30 and 20 dB are superimposed on each power quality disturbance signal. The experimental data is the average of the experimental results of various disturbance signals under the same noise intensity. In order to further verify the performance of the power quality disturbance classification method proposed in this paper, the SVM, MLKNN and MLRBF schemes are used to complete the disturbance signal classification, and the results are compared with the method proposed in this paper.
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance  0.118  0.117  0.138  0.161 
Compound disturbance of the first type  0.114  0.121  0.137  0.169 
Compound disturbance of the second type  0.125  0.127  0.145  0.195 
Compound disturbance of the third type  0.128  0.132  0.148  0.215 
Average value  0.121  0.124  0.142  0.185 
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance  0.076  0.069  0.085  0.126 
Compound disturbance of the first type  0.077  0.077  0.094  0.135 
Compound disturbance of the second type  0.087  0.080  0.102  0.160 
Compound disturbance of the third type  0.092  0.087  0.108  0.178 
Average value  0.083  0.078  0.097  0.15 
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance  0.058  0.062  0.070  0.120 
Compound disturbance of the first type  0.059  0.063  0.069  0.104 
Compound disturbance of the second type  0.062  0.066  0.074  0.134 
Compound disturbance of the third type  0.064  0.069  0.075  0.150 
Average value  0.061  0.065  0.072  0.127 
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance  1.899  1.823  1.851  2.278 
Compound disturbance of the first type  1.946  1.852  2.059  2.390 
Compound disturbance of the second type  2.045  2.100  2.150  2.762 
Compound disturbance of the third type  2.113  2.201  2.326  3.039 
Average value  2.001  1.994  2.097  2.617 
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance  0.959  0.963  0.933  0.925 
Compound disturbance of the first type  0.936  0.932  0.915  0.882 
Compound disturbance of the second type  0.858  0.887  0.863  0.778 
Compound disturbance of the third type  0.884  0.886  0.868  0.766 
Average value  0.909  0.912  0.895  0.838 
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance (s)  4.816  4.484  4.838  4.606 
Compound disturbance of the first type (s)  4.673  4.594  4.694  4.720 
Compound disturbance of the second type (s)  5.098  4.828  5.121  4.960 
Compound disturbance of the third type (s)  5.226  4.989  5.250  5.125 
Average value (s)  4.953  4.724  4.976  4.853 
Disturbance signal  Interference noise (dB)  

50  40  30  20  
Single disturbance (s)  1.077  1.119  1.085  1.122 
Compound disturbance of the first type (s)  1.104  1.086  1.112  1.089 
Compound disturbance of the second type (s)  1.160  1.185  1.168  1.188 
Compound disturbance of the third type (s)  1.199  1.214  1.207  1.218 
Average value (s)  1.135  1.151  1.143  1.154 
Based on the above experimental results, the classifier model proposed in this paper has lower coverage rate, higher average accuracy, and Hamming loss, firstclass error, ranking loss, and higher coverage than other classification methods. The proposed algorithm has a good classification effect and the training time and test time for classification are relatively low, showing higher classification efficiency.
First, the proposed algorithm includes the multiple hidden layers with strong nonlinear mapping capabilities and generalization capabilities, which can effectively characterize the complex interference signals and improve the average accuracy of classification.
Second, the proposed algorithm can extract a large amount of feature information, avoiding the contingency of random assignment of the algorithm. Under the interference of different intensities of noise, the variation of various performance indicators is small, and it has good generalization ability and robustness.
Third, the rankingbased multilabel classification algorithm fully considers the correlation between each label, and is suitable for the classification of single disturbance and compound disturbance, and has good antinoise ability and classification accuracy.
Fourth, the proposed algorithm does not require iteration, reducing training time and testing time, and has obvious advantages when dealing with a large amount of power quality interference.
The power load of the grid in sTarbela is responsible for the power supply of 13 cities and regions and part of the power exchange between the three provinces of Punjab, Sindh and Balochistan. It has 11000 kV substation, 4500 kV substations, 28220 kV substations, and 110 kV substation 103. There are 18235 kV substations with a power supply capacity of 5.5 million kW. There are 621 35220 kV lines with a line length of 8443.83 km.
This paper takes the electric energy disturbance as an example to study the application effect of the classification algorithm in the actual grid. According to the classification model and parameter range shown in the article, 100 training samples and 100 test samples are randomly selected for each type of power quality disturbance signal.
Disturbance signal  Classification result  

Single disturbance  Compound disturbance of the first type  Compound disturbance of the second type  Compound disturbance of the third type  
Single disturbance  95  5  0  0 
Compound disturbance of the first type  5  92  2  1 
Compound disturbance of the second type  0  11  85  4 
Compound disturbance of the third type  0  6  9  85 
In this paper, combining the multilayer extreme learning machine based on selfencoding and the multilabel classification algorithm based on ranking, a new power quality disturbance classification method is proposed, and the structure model and classification process of the classifier are explained. Experiments show that the proposed algorithm performs well in terms of Hamming loss, average accuracy, and coverage of the classification results, and has good noise resistance and robustness. It also significantly reduces the training time and test time, and the advantage is obvious when there are more interference signals. The proposed results provide effective means to determine the classification of disturbance in power signals quality. It can be generally deployed in relevant distribution networks that has to be evaluated in the context of disturbance analysis. Moreover, the proposed solution is optimal and has lower hamming loss, higher accuracy and improved ability to classify the disturbance. It has good antinoise ability and high classification accuracy [