Table 2 HVDC parameters for training datasets.
From: Bayesian-optimized LSTM-DWT approach for reliable fault detection in MMC-based HVDC systems
Variable | Values |
|---|---|
Location (km) for internal faults | 1, 10, 20, 30, 40, 50, 60, 70, 80, 90, 110, 120, 130, 140, 150, 160, 170, 180, 190, 198, Total number of fault location values: L = 20 |
Fault-resistance (Ω) for internal faults | 0.01, 0.1, 0.2, 0.5, 1, 5, 15, 25, 55, 75, 100, 125, 175, 200, 250, 275, 300, 325, 350, 375, 400. Total number of fault resistance values: R = 21 |
External-fault types (AC faults) Each character set except G (stands for ground) represent phase AC faults occurred at AC wind farm 1 (Converter 1) | Three-Phase to Ground (ABCG) with fault resistance range (0.01, 0.1, 1, 5, 10, 15, 25, 50, 75,100, 125, 175), Two-Phase to Ground (ABG, ACG, BCG), Single-Phase to Ground (AG, BG, CG), Phase to Phase (AB, AC, BC) with fault resistances: 0.1, 5,10, 15, 20, 25, 30. Total AC faults = 12 + 21 + 21 + 21 = 75 |
External-fault types (DC bus faults) | Fault resistances: 0.01, 0.1, 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50,…, 375, 380, 385, 390, 395, 400,…, 595,600, 605, 610. Total DC bus faults are 125 |
Noise-(dB) | 20, 25, 30,45 |
Healthy State (100) + External fault types (125 + 75) = 100 + 200 = 300 Total Internal faults (Fint) = (kI × 3) = 420 × 3 = 1260 | kI = L × R = 20 × 21 = 420/each Fint fault type, 3 Fint faults types (i.e. N-PTG = 420, P-PTG = 420, PTP = 420), Training samples = [(Fint = kI × 3) + 300] = N = 1560, Training samples = 1560 for training only |
