CN112525507B - Adaptive acquisition method for vibration information in process of starting and stopping rotor system - Google Patents

Adaptive acquisition method for vibration information in process of starting and stopping rotor system Download PDF

Info

Publication number
CN112525507B
CN112525507B CN202011314957.5A CN202011314957A CN112525507B CN 112525507 B CN112525507 B CN 112525507B CN 202011314957 A CN202011314957 A CN 202011314957A CN 112525507 B CN112525507 B CN 112525507B
Authority
CN
China
Prior art keywords
vibration signal
frequency
rotor system
original
demodulated
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202011314957.5A
Other languages
Chinese (zh)
Other versions
CN112525507A (en
Inventor
张西宁
闵佳信
李霖
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian Jiaotong University filed Critical Xian Jiaotong University
Priority to CN202011314957.5A priority Critical patent/CN112525507B/en
Publication of CN112525507A publication Critical patent/CN112525507A/en
Application granted granted Critical
Publication of CN112525507B publication Critical patent/CN112525507B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
    • G01H11/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by detecting changes in electric or magnetic properties
    • G01H11/02—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by detecting changes in electric or magnetic properties by magnetic means, e.g. reluctance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
    • G06F17/148—Wavelet transforms

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)

Abstract

A self-adaptive obtaining method of vibration information in a rotor system start-stop process includes the steps that firstly, original vibration signals in the rotor system start-stop process are obtained, then estimated instantaneous frequency is obtained on the basis of nonlinear short-time Fourier transform, then according to the estimated instantaneous frequency, the original vibration signals are subjected to generalized demodulation to obtain demodulated vibration signals, and then fundamental frequency components are extracted through self-adaptive decomposition of the demodulated vibration signals through Empirical Wavelet Transform (EWT); finally, inverse operation of generalized demodulation is carried out on the fundamental frequency component, and a fundamental frequency vibration signal is reconstructed; the invention can adaptively decouple and separate and extract the coupling signal, has good decomposition effect on the local coupling component, is almost equivalent to the current VKF algorithm in precision, and avoids the problem of difficult parameter design of the VKF algorithm.

Description

Adaptive acquisition method for vibration information in process of starting and stopping rotor system
Technical Field
The invention belongs to the field of rotor system fault diagnosis, and particularly relates to a vibration information self-adaptive acquisition method in a process of starting and stopping a rotor system.
Background
As an important component of national economy and national defense technology industry, rotary machines have a very important position in various industries such as aerospace, chemical engineering, energy and the like. For rotating machinery, the fact that a rotor system can run stably is the key for guaranteeing working efficiency and stability, the design is not proper, and the mass imbalance of the rotor system caused by errors such as manufacturing and assembling is the main reason for noise and vibration. However, in practice, due to the misalignment of the rotor system, oil film oscillation, cracks and other faults, a large amount of nonlinear information always exists in the rotor system, a large amount of harmonic signals exist in the signals, the signals are difficult to obtain in many working conditions, and the rotor system is also influenced by starting and stopping for many times in order to obtain the vibration information. Therefore, how to accurately extract the interesting vibration component in the process of starting and stopping the vehicle as little as possible is a problem to be solved urgently in the field.
At present, the research on the extraction of signal components in the process of changing the rotating speed is widely applied to the field of fault diagnosis. Such as order analysis technology, Empirical Mode Decomposition (EMD), and Variational Mode Decomposition (VMD), but they all have limitations, such as the order analysis technology relying on a key phase signal and low calculation accuracy for non-stationary transient signals; the EMD method has end point effect and mode confusion, and lacks of deeper research and application in the field of fault diagnosis under the working condition of variable rotating speed; the VMD method excessively penalizes the domain boundary and the internal jump, has a problem of boundary effect, and requires presetting parameters, which also brings inconvenience to its use. The second generation Vold-Kalman filtering (VKF) algorithm is the most common order tracking method capable of realizing coupling component decoupling at present, but the VKF algorithm has the problem of difficult parameter setting, requires a great deal of experience to obtain parameters adapted to the non-stationary characteristics of signals, and has too large limitation in application.
Empirical Wavelet Transform (EWT) combines the advantages of EMD and wavelet transform, has a complete theoretical basis and fast computation speed, and can avoid end-point effects in EMD. However, the conventional EWT is based on fourier spectrum division, and the conventional EWT appears to be unwieldy for the case where there is a frequency domain crossing in each modality; the rotor start-up vibration signal is a dynamic response under wide frequency excitation, with components overlapping each other in the fourier spectrum. Therefore, the traditional EWT obviously cannot adaptively decompose the variable-rotation-speed vibration signal in the starting and stopping process of the rotor, and how to realize the decoupling and separation of each component of the vibration signal is very important for the complex vibration signal of the non-steady-state rotor system.
Disclosure of Invention
In order to overcome the defects of the existing method, the invention aims to provide a self-adaptive obtaining method of vibration information in the process of starting and stopping a rotor system, which can self-adaptively decouple and separate and extract a coupling signal, has a good decomposition effect on a local coupling component, is almost equivalent to the current VKF algorithm in precision, and avoids the problem of difficult parameter design of the VKF algorithm.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a self-adaptive obtaining method for vibration information in a rotor system start and stop vehicle process comprises the following steps:
1) acquiring an original vibration signal x (t) in the process of starting and stopping a rotor system:
2) carrying out nonlinear short-time Fourier transform (NLSTFT) on an original vibration signal x (t), wherein the expression is as follows:
Figure BDA0002791063820000021
wherein t represents time, w (u-t) represents the selected window function, h represents the window length, S (t, h) is the energy distribution of the signal in the time-frequency domain, and c (t) is the first derivative of the instantaneous frequency;
adopting a recursive calculation method, enabling c (t) to be 0 during the first calculation, then obtaining a peak value corresponding to the time S (t, w) by using a frequency domain peak value search algorithm, obtaining a first estimated instantaneous frequency through fitting, then taking a first derivative of the first estimated instantaneous frequency as a value of c (t) during the second iteration, setting an iteration termination condition that the iteration is stopped when the variation of the estimated instantaneous frequency is less than 5% before and after the two iterations, and obtaining an estimated instantaneous frequency I (t);
3) according to the estimated instantaneous frequency I (t), carrying out generalized demodulation on the original vibration signal x (t) to obtain a demodulated vibration signal x' (t), wherein the expression is as follows:
x′(t)＝x(t)·exp(-j2π(I(t)) (2)
4) respectively combining the demodulated vibration signal x' (t) with empirical wavelet function
Figure BDA0002791063820000034
And empirical scale function
Figure BDA0002791063820000035
Performing inner product operation to obtain an approximation coefficient Wx(0, t) and detail coefficient Wx(n, t), the demodulated vibration signal x '(t) is decomposed into different natural modes, and then the demodulated vibration signal x' (t) is decomposed into:
Figure BDA0002791063820000031
therein Ψn(t)、φ1(t) are empirical wavelet functions, respectively
Figure BDA0002791063820000032
And empirical scale function
Figure BDA0002791063820000033
Fourier transform of (W)x(0，t)×φ1(t) represents the low frequency component in the demodulated vibration signal x' (t), i.e. the fundamental frequency component in the original vibration signal x (t);
5) and performing inverse operation of generalized demodulation on the fundamental frequency component to reconstruct a fundamental frequency vibration signal.
The original vibration signal x (t) is an experimental vibration signal or a simulated vibration signal.
The invention has the beneficial effects that:
according to the invention, the time-frequency analysis is carried out by adopting an NLSTFT method, and the rotating speed information of the rotor system can be accurately obtained under the condition without a tachometer through multiple iterations;
the invention carries out generalized demodulation on the original vibration signal x (t) to obtain the demodulated vibration signal x' (t), so that the original vibration signal x (t) containing various interferences is rearranged in a frequency spectrum, and fundamental frequency components and harmonic components causing the interferences are separated from each other, thereby solving the problem that the EWT can not process frequency domain cross signals, and simultaneously, because the components of the signals are separated from each other, the EWT can be helped to determine decomposition mode numbers and boundary lines, and the algorithm is simplified;
the invention carries out self-adaptive decomposition and extraction on the demodulated vibration signal x' (t) through the EWT, thereby avoiding the problem of difficult parameter design of the traditional method, and simultaneously has the advantages of self-adaptation and high precision due to the compactness of the EWT, higher precision and obvious signal characteristic extraction.
Drawings
FIG. 1 is a flow chart of the present invention.
Fig. 2 is a model schematic diagram of a rotor system in embodiment 1.
FIG. 3 (a) shows x in the original vibration signal x (t) in example 11、x2、x3A time domain waveform map of the component; (b) is a time domain waveform diagram of an original vibration signal x (t); (c) is the frequency domain waveform of the original vibration signal x (t).
Fig. 4 is a time-frequency distribution diagram obtained by NLSTFT of the original vibration signal x (t) in example 1.
Fig. 5 is a frequency domain waveform comparison of the demodulated vibration signal x' (t) and the original vibration signal x (t) in example 1.
FIG. 6 (a) shows the original vibration signal x (t) obtained in example 11′、 x2′、x3' component time domain waveform diagrams; (b) and original x1、x2、x3Reconstruction error of the relative ratio.
FIG. 7 shows the rotor test stand RK4 in accordance with example 2.
Fig. 8 is an original vibration signal x (t) of the rub-impact rotor system collected in embodiment 2, where (a) is a time-domain waveform of the original vibration signal x (t) and a partially enlarged view of a marked position in a rectangular frame; (b) is the amplitude frequency response of the original vibration signal x (t); (c) is the original vibration signal x (t) phase frequency response.
Fig. 9 (a) is a time-frequency distribution obtained by NLSTFT processing the original vibration signal x (t) in embodiment 2; (b) is a comparison of the original instantaneous frequency with the estimated instantaneous frequency i (t).
Fig. 10 is a schematic diagram of the frequency domain waveform of the demodulated vibration signal x' (t) and the EWT adaptive decomposition in embodiment 2.
Fig. 11 (a) is a partially enlarged view of the time domain waveform of the fundamental frequency vibration signal reconstructed in embodiment 2 and the marked positions in the rectangular frame; (b) the amplitude-frequency response of the fundamental frequency vibration signal is obtained; (c) the phase-frequency response of the fundamental frequency vibration signal.
Fig. 12 (a) is a time-domain waveform of a fundamental frequency vibration signal reconstructed by the VKF method in example 2; (b) the local amplification comparison of the fundamental frequency vibration signals reconstructed by the method and the VKF method is carried out at the marked rectangular frame.
FIG. 13 (a) shows an original vibration signal x2(t) a time domain waveform; (b) processing of the original vibration signal x for the inventive method and the VKF method with different parameters2(t) local magnification contrast at the marked rectangular box.
Detailed Description
The invention is described in detail below with reference to the figures and examples.
Example 1: as shown in fig. 1, a method for adaptively acquiring vibration information during a vehicle start and stop process of a rotor system includes the following steps:
1) acquiring an original vibration signal x (t) in the starting and stopping process of a rotor system;
in embodiment 1, a simulated vibration signal of a rotor system model is used as an original vibration signal X (t), the rotor system model is as shown in fig. 2, the mass of a rotating shaft is not counted, a disk with the mass m is fixed in the middle, the rotating shaft is supported by bearings at the two ends of A, B, C is the mass center of the disk, D is the rotation center of the disk, the eccentricity CD is equal to e, the rigidity of the rotor in the X and Y directions is assumed to be k, the damping in the X and Y directions is assumed to be C, the rotating speed is ω t, and the initial angular phase is ω t
Figure BDA0002791063820000051
The vibration equation during start-up is approximated as:
Figure BDA0002791063820000061
wherein m represents the mass of the rotating disc/kg; c represents damping/Ns.multidot.mm-1; k represents stiffness/N.mm-1; m represents an eccentric mass/kg; e represents the eccentricity/mm; omega represents the instantaneous rotating speed/rad & s < -1 >; θ represents instantaneous phase/rad;
the vibration characteristics of the isotropic symmetry degree rotor in different directions are similar, and the obtained vibration response expression is as follows:
Figure BDA0002791063820000062
where λ ═ ω/ωn，ξ＝c/2mω，ωnThe nth order critical rotating speed of the rotor system;
in practice, however, due to various faults of the rotor system, a great deal of harmonic components always exist in the vibration response of the rotor system, and a linear superposition signal is adopted to simulate the original vibration signal x (t) in the starting process of the rotor
x(t)＝x1+x2+x3 (6)
Wherein x1Representing the fundamental frequency component in the original vibration signal x (t); x is the number of2Represents a double frequency component; x is the number of3Represents a triple frequency component;
the rotor system model parameters are set as follows: m is 50 kg; e-5 × 10-5m；k＝1×107 N/m；c＝1500N·s·m-1，
Figure BDA0002791063820000063
The amplitudes of the second frequency multiplication and the third frequency multiplication are respectively 0.6 time and 0.4 time of the fundamental frequency component; the obtained characteristic of the original vibration signal x (t) is shown in fig. 3, and by observing the time domain waveform and the frequency domain waveform of the original vibration signal x (t), the original vibration signal can be found to be represented by aliasing of each component in the time domain, and the frequency spectrum cross appears in the frequency domain, so that the vibration information can not be effectively obtained;
2) obtaining an estimated instantaneous frequency i (t) based on a nonlinear short-time fourier transform NLSTFT:
carrying out nonlinear short-time Fourier transform on an original vibration signal x (t), wherein the expression is as follows:
Figure BDA0002791063820000071
on the basis, a recursive calculation method is adopted, so that c (t) is 0 during the first calculation, then a frequency domain peak value search algorithm is utilized to obtain a peak value corresponding to the time S (t, w), a first estimated instantaneous frequency is obtained through fitting, then a first derivative of the first estimated instantaneous frequency is used as a value of c (t) during the second iteration, and the iteration is stopped when the change amount of the estimated instantaneous frequency before and after two iterations is less than 5% under the set iteration termination condition, so that the final estimated instantaneous frequency I (t) can be obtained; fig. 4 shows the result of NLSTFT on the original vibration signal x' (t), which clearly results in the estimated instantaneous frequency i (t) of the original vibration signal x (t);
3) preprocessing an original vibration signal x (t) based on generalized demodulation:
according to the estimated instantaneous frequency I (t), carrying out generalized demodulation on the original vibration signal x (t) to obtain a demodulated vibration signal x' (t), wherein the expression is as follows:
x′(t)＝x(t)·exp(-j2π(I(t)) (2)
after the generalized demodulation, the energy equivalent to each component in the original vibration signal x (t) is concentrated at the difference between the self frequency and the estimated instantaneous frequency i (t), wherein the fundamental frequency component is compressed to be near 0Hz, and each order harmonic component is rearranged on the frequency spectrum, so that the separation of the harmonic component and the fundamental frequency vibration component is realized; FIG. 5 is a frequency domain waveform comparison of a demodulated vibration signal x' (t) with an original vibration signal x (t);
4) adaptive decomposition extraction of the demodulated vibration signal x' (t) by the empirical wavelet transform EWT:
respectively combining the demodulated vibration signal x' (t) with empirical wavelet function
Figure BDA0002791063820000072
And empirical scale function
Figure BDA0002791063820000073
Performing inner product operation to obtain an approximation coefficient Wx(0, t) and detail coefficient Wx(n, t), the demodulated vibration signal x '(t) is decomposed into different natural modes, and then the demodulated vibration signal x' (t) is decomposed into:
Figure BDA0002791063820000081
therein Ψn(t)、φ1(t) are empirical wavelet functions, respectively
Figure BDA0002791063820000082
And empirical scale function
Figure BDA0002791063820000083
Fourier transform of (W)x(0，t)×φ1(t) represents the low frequency component in the demodulated vibration signal x' (t), i.e. the fundamental frequency component in the original vibration signal x (t); as shown in FIG. 5, the components in which the frequency is close to 0Hz are the generationsThe required fundamental frequency components are shown, and because an EWT method is adopted in the step, amplitude modulation and frequency modulation components with a tightly-supported Fourier spectrum can be extracted, and the accuracy of separation and extraction is ensured;
5) and (3) reconstructing a fundamental frequency vibration signal:
after the fundamental frequency component is obtained, inverse operation of generalized demodulation is carried out on the fundamental frequency component, and a fundamental frequency vibration signal of a rotary subsystem is reconstructed; FIG. 6 shows the x in the original vibration signal x (t)1、x2、x3Three reconstructed signals x 'obtained after being respectively processed'1、x′2、x3' and with original x1、x2、x3Reconstruction error of the relative ratio.
Example 2: as shown in fig. 1, a method for adaptively acquiring vibration information during a vehicle start and stop process of a rotor system includes the following steps:
1) acquiring an original vibration signal x (t) in the process of starting and stopping a rotor system:
in example 2, a rotor experiment table as shown in fig. 7 is used, and an experimental vibration signal is used as an original vibration signal x (t), the rotor experiment table is driven by a direct current motor, and the speed regulation of a rotor system is realized by a motor speed regulator; two wheel discs A, B are mounted on the rotating shaft, unbalanced excitation of a rotor system can be realized by adding unbalanced mass on the wheel disc A, B, and the symmetry of the rotor system can be changed by changing the position of the wheel disc A, B, so that a complex vibration experiment is realized; the whole rotor test bed is provided with 5 eddy current sensors, wherein two eddy current sensors (3, 4, 5 and 6 in the figure) are vertically arranged at the outer sides of the two wheel discs A, B close to bearings at two ends respectively to obtain original vibration signals of different positions and directions of a rotor system; installing a key phase sensor ( position 1 or 2 in the figure) at the output shaft end of the direct current motor to obtain the rotating speed information of the rotor system and recording the rotating speed information as the original instantaneous frequency;
fig. 8 (a) is the original vibration signal x (t) collected under the condition of rotor rub-impact fault, it can be clearly seen that there is a lot of interference irrelevant to the fundamental frequency component in the original vibration signal x (t), and the time domain waveform and the mark position (v) in the rectangular frame are observed1、v2、v3) The partial enlarged view shows that the harmonic components are overlapped with the required fundamental frequency components, and the amplitude-frequency and phase-frequency response curves shown in (b) and (c) of fig. 8 show that the vibration information is submerged in various interferences and cannot be extracted, and the vibration state of the rotor cannot be accurately known;
2) obtaining an estimated instantaneous frequency i (t) based on a nonlinear short-time fourier transform NLSTFT:
carrying out nonlinear short-time Fourier transform on an original vibration signal x (t), wherein the expression is as follows:
Figure BDA0002791063820000091
on the basis, a recursive calculation method is adopted, so that c (t) in the first calculation is 0, then a frequency domain peak value search algorithm is utilized to obtain a peak value corresponding to the time S (t, w), a first estimated instantaneous frequency is obtained through fitting, then a first derivative of the first estimated instantaneous frequency is used as a value of c (t) in the second iteration, the termination condition of the iteration is set, the iteration is stopped when the change amount of the estimated instantaneous frequency is less than 5% before and after two iterations, the estimated instantaneous frequency I (t) can be obtained, and the time-frequency distribution of an original vibration signal x (t) is obtained as shown in (a) in fig. 9, wherein (b) in fig. 9 is the comparison of the original instantaneous frequency and the estimated instantaneous frequency I (t), it can be seen that the relative error can be kept within 5% at a low frequency, and the accuracy can reach more than 99% as the frequency conversion increases;
3) preprocessing an original vibration signal x (t) based on generalized demodulation:
according to the obtained estimated instantaneous frequency I (t), carrying out generalized demodulation on the original vibration signal x (t) to obtain a demodulated vibration signal x' (t), wherein the expression is as follows:
x′(t)＝x(t)·exp(-j2π(I(t)) (2)
after the generalized demodulation, the energy equivalent to each component in the original vibration signal is concentrated at the difference between the self frequency and the estimated instantaneous frequency I (t), wherein the fundamental frequency component is compressed to be near 0Hz, each order harmonic component is rearranged on the frequency spectrum, and the like, so that the separation of the harmonic component and the fundamental frequency vibration component is realized; FIG. 10 is a spectral diagram of a demodulated vibration signal x' (t), where it can be seen that fundamental frequency components and harmonic components do not spectrally cross-alias;
4) the empirical wavelet transform EWT performs adaptive decomposition extraction on the demodulated vibration signal x' (t):
respectively combining the demodulated vibration signal x' (t) with empirical wavelet function
Figure BDA0002791063820000104
And empirical scale function
Figure BDA0002791063820000105
Performing inner product operation to obtain an approximation coefficient Wx(0, t) and detail coefficient Wx(n, t), i.e. the demodulated vibration signal x '(t) is decomposed into different natural modes, the demodulated vibration signal x' (t) will be decomposed into:
Figure BDA0002791063820000101
therein Ψn(t)、φ1(t) are empirical wavelet functions, respectively
Figure BDA0002791063820000102
And empirical scale function
Figure BDA0002791063820000103
Fourier transform of (W)x(0，t)×φ1(t) represents the low frequency component in the demodulated vibration signal x' (t), i.e. the fundamental frequency component in the original vibration signal x (t); FIG. 10 is a schematic diagram of an EWT adaptively decomposing a demodulated vibration signal x' (t), which is capable of adaptively separating and extracting a desired fundamental frequency component from a harmonic component of interference, and obtaining a fundamental frequency component near 0 Hz;
5) reconstruction of vibration information component:
the fundamental frequency component is subjected to inverse operation of generalized demodulation to reconstruct the fundamental frequency vibration signal, as shown in (a) of FIG. 11, and the time domain waveform and moment are observedMarking position in shape frame (v'1、v2′、v3') the partial enlarged view, it can be seen that the interference in the original vibration signal x (t) is effectively removed, the harmonic component excited by the rub-impact fault is obviously weakened, the subharmonic resonance peak appearing at the critical rotation speed 1/2 is also effectively weakened, and meanwhile, the signal waveform also presents the characteristics of regularity and flatness; as can be seen from the amplitude-frequency-phase-frequency response curves shown in (b) and (c) of fig. 11, the amplitude-frequency-phase-frequency response curve of the obtained fundamental frequency vibration signal can clearly indicate the amplitude and phase information of the vibration of the rotor system.
In order to better evaluate the effectiveness and the innovativeness of the method, the original vibration signals x (t) are respectively processed by the method and a traditional second generation VKF algorithm, and as shown in FIG. 12, VKF filtering also has good performance in obtaining fundamental frequency vibration signals and certain decoupling capacity. Comparing the local enlarged images of the fundamental frequency vibration signals obtained by the two methods, the comparison effect is shown in (b) in fig. 12, and it can be seen that the method of the present invention is equivalent to the traditional VKF algorithm in terms of calculation accuracy; further setting different faults for experimental verification to obtain an original vibration signal x in the starting process of the rotor system with the non-centering fault2(t), as shown in fig. 13, it can be seen that the influence of parameter setting on the decoupling capability of VKF is very large, and when the filter bandwidth is changed from 5Hz (labeled m) to 8Hz (labeled n), the decoupling capability of VKF is greatly reduced, which brings great difficulty to practical use; the method has the advantages of solving the problem of parameter design, along with good decoupling, separation and extraction capabilities and stronger applicability.
The above examples are only for illustrating the technical idea and features of the present invention, and are not to be construed as limiting the scope of the present invention. It will be appreciated by those skilled in the art that various modifications and changes may be made without departing from the spirit of the invention.

Claims (2)

1. A self-adaptive obtaining method for vibration information in a rotor system start-stop vehicle process is characterized by comprising the following steps:
1) acquiring an original vibration signal x (t) in the starting and stopping process of a rotor system;
2) carrying out nonlinear short-time Fourier transform on an original vibration signal x (t), wherein the expression is as follows:
Figure FDA0002791063810000011
wherein t represents time, w (u-t) represents the selected window function, h represents the window length, S (t, h) is the energy distribution of the signal in the time-frequency domain, and c (t) is the first derivative of the instantaneous frequency;
adopting a recursive calculation method, enabling c (t) to be 0 during the first calculation, then obtaining a peak value corresponding to the time S (t, w) by using a frequency domain peak value search algorithm, obtaining a first estimated instantaneous frequency through fitting, then taking a first derivative of the first estimated instantaneous frequency as a value of c (t) during the second iteration, setting an iteration termination condition that the iteration is stopped when the variation of the estimated instantaneous frequency is less than 5% before and after the two iterations, and obtaining an estimated instantaneous frequency I (t);
3) according to the estimated instantaneous frequency I (t), carrying out generalized demodulation on the original vibration signal x (t) to obtain a demodulated vibration signal x' (t), wherein the expression is as follows:
x′(t)＝x(t)·exp(-j2π(I(t)) (2)
4) respectively combining the demodulated vibration signal x' (t) with empirical wavelet function
Figure FDA0002791063810000012
And empirical scale function
Figure FDA0002791063810000013
Performing inner product operation to obtain an approximation coefficient Wx(0, t) and detail coefficient Wx(n, t), the demodulated vibration signal x '(t) is decomposed into different natural modes, and then the demodulated vibration signal x' (t) is decomposed into:
Figure FDA0002791063810000014
therein Ψn(t)、φ1(t) are empirical wavelet functions, respectively
Figure FDA0002791063810000021
And empirical scale function
Figure FDA0002791063810000022
Fourier transform of (W)x(0，t)×φ1(t) represents the low frequency component, i.e., fundamental frequency component, in the detuned vibration signal x' (t);
5) and performing inverse operation of generalized demodulation on the fundamental frequency component to reconstruct a fundamental frequency vibration signal.
2. The self-adaptive obtaining method of the vibration information in the process of starting and stopping the vehicle by the rotor system according to claim 1, characterized in that: the original vibration signal x (t) is an experimental vibration signal or a simulated vibration signal.
CN202011314957.5A 2020-11-21 2020-11-21 Adaptive acquisition method for vibration information in process of starting and stopping rotor system Active CN112525507B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011314957.5A CN112525507B (en) 2020-11-21 2020-11-21 Adaptive acquisition method for vibration information in process of starting and stopping rotor system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011314957.5A CN112525507B (en) 2020-11-21 2020-11-21 Adaptive acquisition method for vibration information in process of starting and stopping rotor system

Publications (2)

Publication Number Publication Date
CN112525507A CN112525507A (en) 2021-03-19
CN112525507B true CN112525507B (en) 2022-01-07

Family

ID=74982144

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011314957.5A Active CN112525507B (en) 2020-11-21 2020-11-21 Adaptive acquisition method for vibration information in process of starting and stopping rotor system

Country Status (1)

Country Link
CN (1) CN112525507B (en)

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102565704B (en) * 2012-01-06 2013-12-04 武汉科技大学 Method for identifying frequency-converting speed-adjusting dynamic running mode of electric motor based on time-frequency analysis
CN102735442A (en) * 2012-07-17 2012-10-17 华东理工大学 Method for online monitoring and fault diagnosis of rotor
CN105067105B (en) * 2015-05-04 2018-04-17 西安交通大学 Utilize the kinetic parameter identification method of rotating machinery start and stop car transient signal feature
CN106769009A (en) * 2016-12-12 2017-05-31 新疆大学 Start and stop car fault signature based on Parametric Time-frequency Analysis is extracted and diagnostic method
CN108917918A (en) * 2018-06-20 2018-11-30 沈阳大学 One kind being directed to bearing vibration signal instantaneous frequency analysis and diagnosis method
CN110503025B (en) * 2019-08-19 2023-04-18 重庆大学 Analog circuit early fault diagnosis method based on semi-supervised cooperative training
CN110671613B (en) * 2019-10-15 2021-03-16 重庆邮电大学 Fluid pipeline leakage signal time delay estimation method based on improved empirical wavelet transform
CN111307426A (en) * 2019-11-20 2020-06-19 李嘉诚 Rotating machinery fault feature extraction method based on FrFT-EWT principle

Also Published As

Publication number Publication date
CN112525507A (en) 2021-03-19

Similar Documents

Publication Publication Date Title
Huang et al. Time-frequency squeezing and generalized demodulation combined for variable speed bearing fault diagnosis
Wang et al. Multiscale filtering reconstruction for wind turbine gearbox fault diagnosis under varying-speed and noisy conditions
Wang et al. Rolling element bearing fault diagnosis via fault characteristic order (FCO) analysis
CN110617964A (en) Synchronous compression transformation order ratio analysis method for fault diagnosis of rolling bearing
Feng et al. Time–frequency analysis of time-varying modulated signals based on improved energy separation by iterative generalized demodulation
Wang et al. Multi-scale enveloping order spectrogram for rotating machine health diagnosis
CN110763462B (en) Time-varying vibration signal fault diagnosis method based on synchronous compression operator
Zhang et al. Enhancement of time-frequency post-processing readability for nonstationary signal analysis of rotating machinery: Principle and validation
Feng et al. A novel order spectrum-based Vold-Kalman filter bandwidth selection scheme for fault diagnosis of gearbox in offshore wind turbines
Ming et al. Fault feature extraction and enhancement of rolling element bearing in varying speed condition
CN113009334B (en) Motor fault detection method and system based on wavelet packet energy analysis
Zhang et al. Enhancement of adaptive mode decomposition via angular resampling for nonstationary signal analysis of rotating machinery: Principle and applications
Patel et al. Induction motor bearing fault identification using vibration measurement
Zhang et al. Improved local cepstrum and its applications for gearbox and rolling bearing fault detection
Cheng et al. Fault diagnosis of wind turbine gearbox using DFIG stator current analysis
Ren et al. ISVD-based in-band noise reduction approach combined with envelope order analysis for rolling bearing vibration monitoring under varying speed conditions
CN111307426A (en) Rotating machinery fault feature extraction method based on FrFT-EWT principle
Song et al. The MFBD: A novel weak features extraction method for rotating machinery
CN112525507B (en) Adaptive acquisition method for vibration information in process of starting and stopping rotor system
Chen et al. Application of reassigned wavelet scalogram in wind turbine planetary gearbox fault diagnosis under nonstationary conditions
Qian Gabor expansion for order tracking
Zhang et al. Wind turbine planetary gearbox fault diagnosis via proportion-extracting synchrosqueezing chirplet transform
Wen et al. Planetary gearbox fault diagnosis using envelope manifold demodulation
Song et al. Multispectral Balanced Automatic Fault Diagnosis for Rolling Bearings under Variable Speed Conditions
CN112781723B (en) Harmonic component detection method based on frequency spectrum variance

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant