Accurate Fault Classifier and Locator for EHV Transmission Lines Based on Artificial Neural Networks

Table 4

Results of the fault type classifier using single and modular ANN (FC₁ and FC₂).

Fault type	Fault location (Km)	Fault inception angle (°)	Fault resistance (Ω)	Network	Desired output				Actual output
					A	B	C	D	A	B	C	D
𝐴-𝐺	12	25	0	Single ANN					1.1112	0.1281	0.1187	0.9287
				Modular ANN	1	0	0	1	1.0013	−0.0071	0.0004	1.0005
	94	225	199	Single ANN					0.9051	0.1266	0.1108	1.1290
				Modular ANN					1.0071	−0.0003	0.0037	1.0085
𝐶-𝐺	12	25	0	Single ANN					−0.0784	0.1420	1.1128	1.2940
				Modular ANN	0	0	1	1	0.0021	−0.0041	1.0093	0.9941
	94	225	199	Single ANN					0.1328	−0.1007	1.2871	1.1008
				Modular ANN					−0.0074	−0.0032	1.0078	0.9814
𝐴-𝐵-𝐺	12	25	0	Single ANN					1.1850	0.9072	0.0088	0.9653
				Modular ANN	1	1	0	1	1.0085	1.0066	−0.0042	1.1473
	94	225	199	Single ANN					1.0951	0.9552	0.1178	1.1008
				Modular ANN					1.0082	1.0077	−0.0004	0.9993
𝐵-𝐶-𝐺	12	25	0	Single ANN					0.1136	1.0951	0.9867	1.1000
				Modular ANN	0	1	1	1	−0.0031	0.9972	1.0007	1.0081
	94	225	199	Single ANN					0.1807	1.1009	1.1150	1.087
				Modular ANN					−0.0001	1.0023	1.0744	1.0003
𝐴-𝐵-𝐶	12	25	—	Single ANN					0.9962	1.1007	1.1105	−0.1042
				Modular ANN	1	1	1	0	1.0087	1.0025	0.9992	0.0035
	94	225	—	Single ANN					0.9198	0.9289	1.1008	−0.1008
				Modular ANN					1.0008	1.0036	1.0089	0.0429