Research Article
Feature Extraction and Classification of Power Quality Disturbances Using Optimized Tunable-Q Wavelet Transform and Incremental Support Vector Machine
Table 5
Performance comparison of OTQWT-ISVM for different values of Q and r.
| Classes | OTQWT-ISVM (with decomposition parameters, Q = 2.5 and r = 3) | OTQWT-ISVM (with decomposition parameters, Q = 3.5 and r = 3) | Training time (s) | Testing time (s) | Training RMSE | Testing RMSE | Accuracy (%) | Training time (s) | Testing time (s) | Training RMSE | Testing RMSE | Accuracy (%) | Noise free-case | 20 dB | Noise free-case | 20 dB |
| PCL1 | 0.0179 | 0.0130 | 0.0039 | 0.0122 | 100 | 100 | 0.0296 | 0.0197 | 0.0009 | 0.0057 | 100 | 100 | PCL2 | 0.0158 | 0.0105 | 0.0061 | 0.0227 | 100 | 100 | 0.0275 | 0.0172 | 0.0031 | 0.0161 | 100 | 100 | PCL3 | 0.0163 | 0.0145 | 0.0038 | 0.0429 | 100 | 98.57 | 0.0280 | 0.0212 | 0.0008 | 0.0361 | 100 | 98.57 | PCL4 | 0.0160 | 0.0130 | 0.0041 | 0.1325 | 98.57 | 98.57 | 0.0277 | 0.0197 | 0.0011 | 0.1151 | 100 | 100 | PCL5 | 0.0164 | 0.0108 | 0.0118 | 0.1812 | 100 | 100 | 0.0281 | 0.0175 | 0.0088 | 0.1707 | 98.57 | 100 | PCL6 | 0.0177 | 0.0120 | 0.0072 | 0.1387 | 98.57 | 98.57 | 0.0294 | 0.0187 | 0.0042 | 0.1260 | 100 | 98.57 | PCL7 | 0.0169 | 0.0133 | 0.0105 | 0.1843 | 100 | 100 | 0.0286 | 0.0199 | 0.0075 | 0.1736 | 100 | 100 | PCL8 | 0.0166 | 0.0117 | 0.0162 | 0.3052 | 98.57 | 98.57 | 0.0283 | 0.0184 | 0.0132 | 0.2869 | 100 | 98.57 | PCL9 | 0.0164 | 0.0139 | 0.0185 | 0.3015 | 100 | 100 | 0.0281 | 0.0206 | 0.0155 | 0.2887 | 98.57 | 100 | PCL10 | 0.0183 | 0.0145 | 0.0131 | 0.0635 | 100 | 100 | 0.0299 | 0.0212 | 0.0101 | 0.0591 | 100 | 100 | PCL11 | 0.0179 | 0.0120 | 0.0039 | 0.0118 | 98.57 | 98.57 | 0.0296 | 0.0187 | 0.0009 | 0.0056 | 98.57 | 98.57 | PCL12 | 0.0185 | 0.0145 | 0.0079 | 0.1676 | 98.57 | 98.57 | 0.0302 | 0.0212 | 0.0049 | 0.1625 | 100 | 100 | Mean | 0.0170 | 0.0128 | 0.0089 | 0.1303 | 99.40 | 99.28 | 0.0287 | 0.0195 | 0.0059 | 0.1205 | 99.64 | 99.52 |
|
|