Research Article

Utilization of Stockwell Transform and Random Forest Algorithm for Efficient Detection and Classification of Power Quality Disturbances

Table 12

Comparing the effectiveness of the proposed method in view of other recently published articles.

Type of classificationNumber of classesNumber of featuresSampling frequencyOverall classification accuracy (%) with noise

SSD + HD [26]7520 kHz95.4
FFT + ELM [27]1276.4 kHz95.38
HHT + WBELM [25]1643.2 kHz95.6
EWT + SVM [21]1666.4 kHz95.56
DWT + ML [28]61025 kHz95.83
RF + DWT [29]2193.2 kHz96.48
WT + PCA + SVM [30]5596.2
Proposed ST + RF1793.2 kHz99.01

Note. SSD: sparse signal decomposition; HD: hybrid dictionary; HHT: Hilbert–Huang transform; WBELM: weighted bidirectional-based extreme learning machine; ELM: extreme learning machine; PCA: principal component analysis; SVM: support vector machine; FFT: fast Fourier transform; ST: S-transform; RF: random forest.