Page 531 - Decision Making Applications in Modern Power Systems

490 Decision Making Applications in Modern Power Systems

TABLE 18.7 Test results and confusion matrix of wavelet packet energy
and bagged decision tree-3 module for detection of a fault condition
during the power swing.
Case Data segmented Predicted Actual class Accuracy
study (%) class (%)
0 1
Case-1 90 10 0 3948 52 99.37
1 10 5840
Case-2 80 20 0 3947 53 99.41
1 5 5845
Case-3 70 30 0 3972 28 99.70
1 1 5849
Case-4 60 40 0 3974 26 98.22
1 2 5848
Case-5 50 50 0 3972 28 99.68

FIGURE 18.9 Performance of WPE_BDT-3 module in case of shunt fault (phase-A1 to
ground) during a stable power swing: (A) Three phase instantaneous currents of circuit-I (B)
Three phase instantaneous currents of circuit-II (C) Three phase instantaneous voltages at
sending-end bus (D) Three stepped MHO relay characteristics (E) Fault detection using
WPE_BDT-3 module. WPE_BDT, Wavelet packet energy and bagged decision tree.

and identify the faulted phase(s). To examine the efficacy of proposed
decision-making algorithm, WPE_BDT-4 module for fault classification has
been tested here with new fault cases involving normal shunt faults and
faults during PS by varying different fault parameters as given in Table 18.5.