Page 111 - IJOCTA-15-4

Rolling bearing fault diagnosis method based on GJO–VMD, multiscale fuzzy entropy, and GSABO–BP...

Figure 1. A flowchart illustrating the optimization of key parameters of variational mode decomposition
(VMD) using the golden jackal optimization

component, where E[] represents the an- (v) Compute the overall assessment fac-
ticipated value of the variable. tor (δ i ) to reorder the weights of the
(iii) Use the subsequent formula to compute the time–frequency domain of each subcarrier
time–frequency domain weighting value λ i : in descending sequence and form a vector
({δ} = {δ 1 , δ 2 , . . . , δ K }).
(vi) Measure how far off the complete evalua-
λ i = β · e i + γ · kur i (5) tion criteria are between adjacent compo-
nents.
In this equation, β and γ serve as weight-
ing coefficients, subject to the constraint
that their sum equals. Given that the sig- d i = δ i − δ i+1 (7)
nal’s representation in the frequency do- (vii) The largest dissimilarity index (m) should
main is more resistant to noise and various be defined. The first m entries of the re-
disturbances, it was set that β = 0.6 and ordered set ({δ}) under the comprehen-
γ = 0.4.
sive judgment criterion represent the sensi-
(iv) Calculate the comprehensive evaluation
tive mode fragments conveying the primary
factor (δ i ):
data information of the signal.
The invention relates to an efficient algo-
λ i − min(λ) rithm, which is specially used for distinguishing
δ i = (6)
max(λ) − min(λ) and identifying IMF components. The algorithm
653