Composite Fault Diagnosis of Rolling Bearing Based on Optimized Wavelet Packet AR Spectrum Energy Entropy Combined with Adaptive No Velocity Term PSO-SOM-BPNN

Table 8

Parameters of ANSVTPSO-SOM-BPNN.

ANVTPSO parameters	Value	SOM parameters	Value	BPNN parameters	Value
Number of population particles	20	Input node	9	Input node	10
Spatial dimension	199	Network competition layer		Hidden layer node	13
Position	[-1, 1]	Topological function	Hextop	Output node	4
Learning factor	Undetermined	Distance function	Linkdist	Maximum training times	1000
Learning factor	Undetermined	Classification stage learning rate	0.9	Training objectives	10^-8
Inertia weight	0.9	Classification stage learning step	1000	Learning rate	0.001
Inertia weight	0.4	Learning rate in tuning phase	0.02	Hidden layer transfer function	Tansig
Maximum number of iterations	20	Tuning phase domain distance	1	Output layer transfer function	Purelin