Algorithm 2 Signal Feature Extraction.

Input: Measured signal x

Signal period T

Feature frequency set fn = [0]

Cycle judgment index m = 1

Output: Feature frequency fc

Construct the target vector t through T;

While m ≠ 0

y = MOMEDA(x)  Equation (3)–(7)

s = extension(y)    extend waveform [24]

Amp = FFT(s)     Obtain the envelope spectrum

fc = max(Amp)    Obtain the feature frequency fc

fd = fn − fc

m = min(abs(fd))  Feature frequency repetition recognition

fn = [fn fc]

x = s

end