Table 6.

Most common FDD algorithms for PV plants.

Reference	Type of Algorithm	Function	Type of Fault	PV-System used	Accuracy
[227]	BPNN-GA	Detection	local material ageing, shadowing, open/short circuit	PV Array	–
[235]	Fuzzy logic	Detection	Arc fault	PV Array	96 %
[236]	LAPART algorithm	Detection	Power Reduction	PV Module	86 %
[237]	fuzzy inference system	Detection & identification	DC side short-circuits	PV Array	94 %
[238]	neuro-fuzzy classifier	Detection & Classification	series losses, faulty by-pass diode, blocking diode	PV Array	90–98 %
[231]	ANN	Identification	By-pass diode short circuit, Connection fault	PV System	–
[232]	RBF-ELM	Classification	short circuit, ageing	PV Array	88.5 %
[239]	M-SVM	Identification & classification	Degradation, line-to-line fault	PV Module	98 %
[240]	PNN	Detection & Classification	short circuit, disconnected string	PV Array	100 %
[241]	fuzzy C-mean (FCM)	Detection & Classification	short, open, partial occlusion, and other defects	PV Array	96 %
[242]	decision tree	Detection & Classification	string, short-circuit, or line-line fault	PV Array	99 %
[243]	Improved GA	Identification & localization	short-circuit,	PV String	95 %
[244]	random forest (RF)	Detection & Classification	line faults, deterioration, open circuit, fractional shading	PV Array	99.13 %