CN104964822A - No-rotation-speed-signal time domain synchronization averaging method for main reducer gearbox - Google Patents

Abstract

The invention discloses a no-rotation-speed-signal time domain synchronization averaging method for a main reducer gearbox. The method comprises the following steps: 1) carrying out Fourier transform on vibration signals of the main reducer gearbox to obtain a Fourier transform spectrogram; 2) carrying out detecting to obtain peak values of front three order gear engagement frequency harmonics according to the given approximate rotation speed and gear tooth numbers; 3) detecting local signal to noise ratio SNRk and cut-off order number Korder of the front five order engagement frequency harmonics, and selecting the engagement frequency harmonic, of which the cut-off order number is the highest, to carry out next-step processing; 4) carrying out narrow-band demodulation on the selected engagement frequency harmonic and calculating the angle position of a rotation shaft; and 5) according to the angle position of the rotation shaft, calculating rotation speed moment, segmenting the original vibration data according to the rotation period of the rotation shaft, and carrying out re-sampling on the data of each rotation period. The method can carry out time domain synchronization averaging automatically only with the given approximate rotation speed value and gear tooth numbers and without timing signals.

Description

A kind of for Driven Gear of Final Reduction Gear case without tach signal Synchronous time average method

Technical field

The invention belongs to machine performance monitoring and fault diagnosis technology field, relate to a kind of Synchronous time average method without tach signal, especially a kind of for Helicopter Main Reducer gear case without tach signal Synchronous time average method.

Background technology

Driven Gear of Final Reduction Gear case is as the critical component of Helicopter Transmission System, complex structure, under being in the rugged surroundings of high speed, alternating overload for a long time, again due to its irredundant design, easy generation fault and damage, its health status is directly connected to the flight safety of helicopter.Therefore, occur just to carry out condition monitoring and fault diagnosis to gear case to improve helicopter safety in operation and Accident prevention.

Extract the prerequisite that effective fault signature is Fault Diagnosis of Gear Case, it directly constrains efficiency and the accuracy of Fault Diagnosis of Gear Case.At present, based on time target Synchronous time average be the most effective gear case vibration signal characteristics extraction process method, the method effectively can eliminate random disturbance and the noise of Complicated Periodic signal, extract from the vibration signal of complication system and turn frequently relevant periodic signal, be widely applied, be particularly suitable for gearbox fault analysis.But this method not only needs pickup vibration analysis signal, also needs the timing signal of being picked up revolving shaft by speed probe or optical encoder, locks the starting point of each signal segment simultaneously.Helicopter Main Reducer gear case cannot install timing signal sensor (speed probe or optical encoder) due to reasons in structure, limits the application of Synchronous time average method in Helicopter Main Reducer gear condition monitoring and fault diagnosis.Therefore, the research tool carried out without tach signal Synchronous time average method is of great significance.

Summary of the invention

The object of the invention is to the shortcoming overcoming existing Synchronous time average method, there is provided a kind of for Helicopter Main Reducer gear case without tach signal Synchronous time average method, the method only needs known approximate tachometer value and the gear number of teeth, without the need to timing signal, automatically Synchronous time average can be carried out to original vibration signal, have and can effectively eliminate random disturbance and noise, execution speed is fast, without the need to the feature of human intervention.

In order to achieve the above object, solution of the present invention is:

For Driven Gear of Final Reduction Gear case without a tach signal Synchronous time average method, comprise the steps:

Step one, Fourier transform is carried out to Driven Gear of Final Reduction Gear case vibration signal, obtain Fourier transform spectrogram;

Step 2, according to known approximate rotating speed and the gear number of teeth, detect and obtain the peak value of first three rank gear mesh frequency harmonic wave, wherein detect bin width be gear mesh frequency harmonic wave ± 5%;

Step 3, calculate the local SNR SNR of the first five rank meshing frequency harmonic wave _kwith cut-off exponent number K _order, the meshing frequency harmonic wave that selective cut-off exponent number is the highest hands over next step to process;

The meshing frequency harmonic wave that step 4, narrowband demodulation are chosen, calculates the angle position of rotating shaft;

Step 5, judge whether the fluctuation of speed exceeds maximum permissible value, if do not exceeded, then perform next step, if exceeded, then select the engagement harmonic frequency of lower single order, re-execute step 4;

Step 6, calculate the rotating speed moment according to the angle position of rotating shaft, by rotating shaft swing circle by original vibration data segmentation, resampling is carried out to the data of each swing circle.

According to above-mentioned feature, the formula of described Fourier transform is:

$FFT\left(f\right)={\∫}_{-\mathrm{\∞}}^{+\mathrm{\∞}}x\left(t\right)e^{-j2\mathrm{\π}ft}dt.$

According to above-mentioned feature, described local SNR SNR _kformula be:

${\mathrm{SNR}}_k=10\log \left(\frac{P_{kTot}-P_{kb}}{P_{kb}}\right)$

Wherein, P _kTotfor meshing frequency harmonic wave place, kth rank bandwidth is total power signal in 2B, P _kbfor noise power in the meshing frequency harmonic wave place bandwidth 2B of kth rank, band bandwidth B:

$B=\min (f_1,f_2)/2=\frac{f_m}{2max(N_1,N_2)}$

Wherein, f ₁, f ₂for meshing gear two rotary shaft rotating speed, N ₁, N ₂for engaging the number of teeth of two gear pairs.

According to above-mentioned feature, described cut-off exponent number K _orderformula:

$K_{order}=\frac{O_c^{\left(k\right)}}{0.83\sqrt{2}{N}_{sh}{10}^{{\mathrm{SNR}}_k/20}}$

In formula, for 3dB, N _sh=N ₁or N ₂, determine according to chosen axis; N ₁, N ₂for engaging the number of teeth of two gear pairs.

According to above-mentioned feature, the resampling formula in described step 6 is:

$X_r=\underset{n=1}{\overset{M}{\Σ}}{x}_t\frac{sin\[\mathrm{\π}(r\mathrm{\δ}-n\mathrm{\Δ})/\mathrm{\Δ}\]}{\mathrm{\π}(r\mathrm{\δ}-\mathrm{\Δ})/\mathrm{\Δ}},r=0,1,2,...,N$

δ＝△(M/N)

Wherein, N is counting after resampling, M be resampling before the counting of raw data section, X _rfor data after resampling, x _tfor according to rotating shaft swing circle to the data after original vibration data segmentation.

After adopting such scheme, the present invention adopts the local SNR at assessment meshing frequency harmonic wave place and cut-off exponent number automatically to choose certain rank meshing frequency harmonic wave of satisfying condition centre frequency as narrow-band filtering, without the need to any timing signal, and effectively can improve the signal to noise ratio (S/N ratio) of original signal, for Helicopter Main Reducer monitoring state of gear case and fault diagnosis provide foundation.

Accompanying drawing explanation

Fig. 1 process flow diagram of the present invention;

Fig. 2 gear mesh frequency harmonic wave narrow-band filtering spectrogram

Embodiment

Below in conjunction with accompanying drawing, technical scheme of the present invention is described in detail.

In order to effectively extract useful vibration performance from Helicopter Main Reducer gear case vibration signal, the present invention chooses the particular harmonic of gear mesh frequency automatically, and narrowband demodulation is carried out to it, thus estimate the angle position (can markers be changed into) of rotating shaft, according to rotating shaft swing circle markers, segmentation and resampling are carried out to original vibration signal again, effectively vibration noise is reduced eventually through time domain average, for the fault signature of Helicopter Main Reducer gear case provides method, for succeeding state monitoring and fault diagnosis are laid a good foundation.

As shown in Figure 1, the invention provides a kind of for Helicopter Main Reducer gear case without tach signal Synchronous time average method, comprise the steps:

1) Fourier transform is carried out to Driven Gear of Final Reduction Gear case vibration signal, obtain Fourier transform spectrogram.The formula of Fourier transform is:

$FFT\left(f\right)={\∫}_{-\mathrm{\∞}}^{+\mathrm{\∞}}x\left(t\right)e^{-j2\mathrm{\π}ft}dt.$

Wherein, x (t) is Driven Gear of Final Reduction Gear case vibration signal.

2) according to known approximate rotating speed and the gear number of teeth, detect the peak value obtaining first three rank gear mesh frequency harmonic wave, this peak value can be used for estimating engagement harmonic wave occurrence, and the estimation of next step local SNR.Wherein detect bin width be gear mesh frequency harmonic wave ± 5%.

3) the local SNR SNR of the first five rank meshing frequency harmonic wave is estimated _kwith cut-off exponent number K _order, the meshing frequency harmonic wave that selective cut-off exponent number is the highest.

The local SNR SNR of gear kth rank meshing frequency harmonic wave _k:

${\mathrm{SNR}}_k=10log\left(\frac{P_{kTot}-P_{kb}}{P_{kb}}\right)$

Wherein, P _kTotfor meshing frequency harmonic wave place, kth rank bandwidth is total power signal in 2B, P _kbfor noise power in the meshing frequency harmonic wave place bandwidth 2B of kth rank, band bandwidth B:

$B=min(f_1,f_2)/2=\frac{f_m}{2max(N_1,N_2)}$

Wherein, f ₁, f ₂for meshing gear two rotary shaft rotating speed, N ₁, N ₂for engaging the number of teeth of two gear pairs.

In above-mentioned steps (2), cut-off exponent number K _order:

$K_{order}=\frac{O_c^{\left(k\right)}}{0.83\sqrt{2}{N}_{sh}{10}^{{\mathrm{SNR}}_k/20}}$

In formula, for 3dB, N _sh=N ₁or N ₂(determining according to chosen axis).

4) the meshing frequency harmonic wave chosen of narrowband demodulation, estimates the angle position of rotating shaft.

By the bandpass filter centered by the meshing frequency harmonic wave of kth rank, obtain signal adopt Hilbert transfer pair y (t) to carry out demodulation and obtain the fluctuation of speed.

5) judge whether the fluctuation of speed exceeds maximum permissible value, if do not exceeded, then perform next step, if exceeded, then select the engagement harmonic frequency of lower single order, re-execute step 4.

6) estimate the rotating speed moment, by rotating shaft swing circle by original vibration data segmentation, resampling is carried out to the data of each swing circle, then carries out superposed average.

Resampling formula is:

$X_r=\underset{n=1}{\overset{M}{\Σ}}{x}_t\frac{sin\[\mathrm{\π}(r\mathrm{\δ}-n\mathrm{\Δ})/\mathrm{\Δ}\]}{\mathrm{\π}(r\mathrm{\δ}-\mathrm{\Δ})/\mathrm{\Δ}},r=0,1,2,...,N$

δ＝△(M/N)

Wherein, N is counting after resampling, M be resampling before the counting of raw data section, △ is the sampling interval of original signal, x _tfor according to swing circle to the data after original vibration data segmentation.

Be understandable that, for those of ordinary skills, can be equal to according to technical scheme of the present invention and inventive concept thereof and replace or change, and all these change or replace the protection domain that all should belong to the claim appended by the present invention.

Claims (5)

1. for Driven Gear of Final Reduction Gear case without a tach signal Synchronous time average method, it is characterized in that comprising the steps:

Step one, Fourier transform is carried out to Driven Gear of Final Reduction Gear case vibration signal, obtain Fourier transform spectrogram;

Step 2, according to known approximate rotating speed and the gear number of teeth, detect and obtain the peak value of first three rank gear mesh frequency harmonic wave, wherein detect bin width be gear mesh frequency harmonic wave ± 5%;

Step 3, calculate the local SNR SNR of the first five rank meshing frequency harmonic wave _kwith cut-off exponent number K _order, the meshing frequency harmonic wave that selective cut-off exponent number is the highest hands over next step to process;

The meshing frequency harmonic wave that step 4, narrowband demodulation are chosen, calculates the angle position of rotating shaft;

Step 5, judge whether the fluctuation of speed exceeds maximum permissible value, if do not exceeded, then perform next step, if exceeded, then select the engagement harmonic frequency of lower single order, re-execute step 4;

Step 6, calculate the rotating speed moment according to the angle position of rotating shaft, by rotating shaft swing circle by original vibration data segmentation, resampling is carried out to the data of each swing circle.

2. according to claim 1 without tach signal Synchronous time average method, it is characterized in that the formula of described Fourier transform is:

$FFT\left(f\right)={\∫}_{-\mathrm{\∞}}^{+\mathrm{\∞}}x\left(t\right)e^{-j2\mathrm{\π}ft}dt.$

3. according to claim 1 without tach signal Synchronous time average method, it is characterized in that described local SNR SNR _kformula be:

${\mathrm{SNR}}_k=10log\left(\frac{p_{kTot}-p_{kb}}{P_{kb}}\right)$

Wherein, P _kTotfor meshing frequency harmonic wave place, kth rank bandwidth is total power signal in 2B, P _kbfor noise power in the meshing frequency harmonic wave place bandwidth 2B of kth rank, band bandwidth B:

$B=min(f_1,f_2)/2=\frac{f_m}{2max(N_1,N_2)}$

Wherein, f ₁, f ₂for meshing gear two rotary shaft rotating speed, N ₁, N ₂for engaging the number of teeth of two gear pairs.

4. according to claim 1 without tach signal Synchronous time average method, it is characterized in that described cut-off exponent number K _orderformula:

$K_{order}=\frac{O_c^{\left(k\right)}}{0.83\sqrt{2}{N}_{sh}{10}^{{\mathrm{SNR}}_k/20}}$

In formula, for 3dB, N _sh=N ₁or N ₂, determine according to chosen axis; N ₁, N ₂for engaging the number of teeth of two gear pairs.

5. according to claim 1 without tach signal Synchronous time average method, it is characterized in that the resampling formula in described step 6 is:

$X_r=\underset{n=1}{\overset{M}{\Σ}}{x}_t\frac{sin\[\mathrm{\π}(r\mathrm{\δ}-n\mathrm{\Δ})/\mathrm{\Δ}\]}{\mathrm{\π}(r\mathrm{\δ}-\mathrm{\Δ})/\mathrm{\Δ}}$ r＝0,1,2,...,N

δ＝△(M/N)

Wherein, N is counting after resampling, M be resampling before the counting of raw data section, X _rfor data after resampling, x _tfor according to rotating shaft swing circle to the data after original vibration data segmentation.