Research on diagnosis method of series arc fault of three-phase load based on SSA-ELM

Bin Li,Shihao Jia

doi:10.1038/s41598-021-04605-w

Abstract

Arc fault in the three-phase load circuit may cause fire, resulting in production interruption and even worse, it will cause casualties. In order to effectively detect the arc fault in the three-phase circuit, series arc fault experiments of three-phase motor load and frequency converter were carried out under different current conditions. Firstly, variational mode decomposition (VMD) was performed for each cycle of A-phase current, and then the VMD energy entropy and sample entropy were calculated. Secondly, the noise-dominated component was removed according to the permutation entropy, then the average value after first-order difference of the half-cycle reconstructed signal was obtained. An arc fault diagnosis model of extreme learning machine (ELM) optimized by sparrow search algorithm (SSA) was established. The feature vectors were divided into training group and test group to train the model and test its fault diagnosis accuracy. Compared with GA-ELM, PSO-ELM, support vector machine (SVM) and SSA-SVM, the experimental results show that the proposed method can identify the series arc fault accurately and more quickly.

Highlights

Arc fault in the three-phase load circuit may cause fire, resulting in production interruption and even worse, it will cause casualties
Qu et al input the characteristics in time domain and frequency domain of the current into the learning vector quantization neural network (LVQ-NN) to determine the load type, and detected the arc fault through particle swarm optimization optimized support vector machine (PSO-SVM)[12]
sparrow search algorithm (SSA) is used to optimize the number of neurons in the hidden layer of extreme learning machine (ELM) to improve its ability of arc fault diagnosis, the specific steps are as follows: (1) Input the collected and classified feature vectors into ELM

Summary

Number of current sampling points

Taking the frequency converter running at 20A as an example, Figs. 4 and 5 show VMD results under normal and fault state respectively. The frequency of each IMF component shows randomness, and the VMD energy entropy increases significantly[20,21]. The current time series will be complicated, resulting in an increase in sample entropy. When the frequency converter is running, the sample entropy curves of normal and fault states sometimes cross. It is found in the experiment that the curves of VMD energy entropy and sample entropy under normal or fault conditions will occasionally cross, which makes machine learning difficult, so Figure 8. In this paper, VMD energy entropy, sample entropy of each cycle and the average value after first-order difference of half cycle were used to form the three-dimensional feature vector E = [HEE, HSE, X]. When the sparrows find danger, they will update their position according to the following equation: Xbtest + Xitj + K β

Xbt est fi

Iteration times

Conclusion

Findings

Additional information