Research Article

Artificial Neural Network-Based Fault Distance Locator for Double-Circuit Transmission Lines

Table 6

Comparison of neural network-based fault location schemes.

Schemes suggested
by authors
Fault locator inputsLine configurationFault resistance
range ( )
Fault inception
angle (°)
Other factors
considered
Response time
and accuracy

Mahanty and Gupta [13]Samples of 3-phase V and ISingle-circuit line for LG and LL faults only0–2000–90°Other types of faults and wide variation in inception angle not considered.Response time not indicated and error is 6%.
Mazon et al. [9]Samples of 50 Hz components of 3-phase voltages and currents of each circuitsDouble-circuit line for LG faults only0–20Other types of faults and variation in inception angle not considered.Response time not indicated and error is 0.19%.
Bhalja and Maheshwari [23]Δ , δ , and resistanceDouble-circuit line for LG faults only0–200Mutual coupling, remote source infeed. Not indicated.
Singular distance locator (by Jain et al.) [25]Samples of 50 Hz components of 3-phase voltages and currents of each circuitsDouble-circuit line for all 10 types of faults in both the circuits (total 20 types of faults)0–1000–360°Mutual coupling, remote source infeed, and all 10 types of faults in each circuit.1-cycle time from inception of faults and % error is from −7% to +1.97%.
Proposed scheme (modular distance locator)Samples of 50 Hz components of 3-phase voltages and currents of each circuitsDouble-circuit line for all 10 types of faults in both the circuits (total 20 types of faults)0–1000–360°Mutual coupling, remote source infeed, all 10 types of faults in each circuit, source strength variation, CT saturation, and single-circuit operation.1-cycle time from inception of faults and % error is from −1.362% to +1.201%.