A non-steady signal's non-rotational speed self-adaptive data processing method and system

Abstract

The present invention belongs to the technical field of digital signal processing, and particularly relates to a non-steady signal non-rotational speed adaptive data processing method and system, belonging to a data-driven adaptive processing strategy. First, calculate the time-frequency distribution spectrogram of the signal through short-time Fourier transform, design a time-varying frequency window, and determine the number of points within the window according to the frequency domain width and sampling frequency of the signal. Then, calculate the maximum energy within each frequency window, dynamically adjust the window center frequency, extract the frequency ridge line within the optimal time-varying frequency window, and obtain the instantaneous frequency distribution of the original signal. Finally, construct a secondary sampling time series according to the instantaneous frequency distribution, and resample the original signal using cubic spline interpolation to obtain a new steady-state signal. The present invention solves the signal processing problem under non-rotational speed conditions, significantly improves the accuracy and robustness of non-steady signal processing, and is applicable to scenarios where it is difficult to accurately synchronize rotational speed data with vibration or sound data in a wireless monitoring system.

Description

Technical Field

[0001] The present invention belongs to the technical field of digital signal processing, and particularly relates to a non-steady signal non-rotational speed adaptive data processing method and system. Background Art

[0002] At present, the fault diagnosis of rotating equipment is mainly based on the analysis and processing of steady signals, such as spectrum analysis, multiple frequency analysis, sideband analysis, band energy, etc. Such analysis methods cannot fully utilize the information contained in non-steady signals generated during the processes of equipment acceleration, deceleration, and speed change. In addition, when diagnosing mechanical faults, we often encounter a large number of non-steady signals, usually because load fluctuations, power supply fluctuations, speed fluctuations, etc. occur during the operation of rotating equipment, resulting in the non-steadiness of the collected sound or vibration signals themselves. If Fourier analysis is directly performed on non-steady signals to obtain spectra, serious frequency ambiguity phenomena will occur, resulting in the complete masking of fault characteristics and the inability to perform effective diagnosis.

[0003] In response to such problems, many experts and scholars have proposed various solutions and technical paths. The currently more mature method is to install a rotational speed sensor on the equipment, and perform order sampling on the original sound or vibration signal by encrypting the collected rotational speed information (H. Hotmanspoth order analysis system CN107192550B). On the hardware side, through the fusion calculation of speed data and vibration or sound signals, direct equal-angle order sampling of periodic motion equipment is realized. Another type of method is also to collect the rotational speed information of the equipment through a rotational speed sensor, draw an angular change curve based on the rotational speed, and perform equal-angle division on the angular change curve to determine the time series of equal-angle resampling to achieve interpolation resampling from software (Zhao Hongshan, a method for fault diagnosis of wind turbine bearings under variable rotational speed CN105784366A).

[0004] Another strategy is to obtain the keyphasor or rotational speed information of the device by deeply analyzing the original signal without installing a rotational speed sensor. Jiang Zhansi et al. proposed a bearing fault diagnosis method based on the comprehensive analysis of order and envelope spectrum (Jiang Zhansi, Rotating Machinery Rolling Bearing Fault Diagnosis Method Based on Order Spectrum and Envelope Spectrum, CN109520738B), but the method does not clearly point out the calculation steps of its order spectrum (COT calculation order tracking algorithm), and problems such as accuracy and divergence existing in the process of performing order analysis. There is also a method to approximately determine the phase information of the signal by extracting the peak sequence in the time-domain signal and further realize the order analysis of non-steady signals (Zeng Zhijie, Wind Power Gear Vibration Signal Order Analysis Method Using Time Series Peak Search, CN109063387B). This method obtains the peak time sequence of the narrowband signal through band-pass filtering. When the original signal has complex frequency components and there are multiple frequency components in a single narrowband, this method will not be able to achieve effective order analysis. Due to the uncertainty of time-domain features, Chen Jian adopted a method of extracting the time-frequency ridge line during fault feature interpretation to obtain the time-frequency distribution of the original signal and further complete the order resampling analysis (Chen Jian, A Method and Device for Extracting Time-Frequency Ridge Lines for Fault Feature Interpretation, CN 111797789A). Although this method can ensure the convergence of the algorithm, extracting multiple ridge lines in the global spectrogram of the time-frequency distribution and simultaneously using these ridge lines for order analysis cannot guarantee obtaining the optimal solution for steady-state signal recovery, and extracting all the ridge lines may result in the situation of some invalid ridge lines, bringing misdiagnosis to the final fault analysis.

[0005] Among the existing several methods, they can be roughly divided into two technical paths: one is to obtain the instantaneous rotational speed information of the device by adding a rotational speed sensor through an external acquisition hardware method and perform order analysis; the other method is to obtain the phase or instantaneous frequency of the original signal from the time-domain waveform or time-frequency distribution of the signal through a data-driven method and perform secondary sampling in software to obtain an approximately steady-state signal. However, the existing methods all have one or more of the following defects and deficiencies:

[0006] 1. When installing a rotational speed sensor on the device, different from the vibration sound sensor, most rotational speed sensors adopt the principle of distance change, so a relatively stationary fixed point is required during installation, and the probe needs to directly point to the rotating part, which brings great difficulties to on-site construction and installation; currently, the commonly used optoelectronic rotational speed sensors have certain requirements for the environment, and with the accumulation of dust in harsh environments, it may lead to acquisition failure, while another Hall rotational speed sensor needs to bond or weld a metal keyphasor on the rotating part, which is often difficult to achieve on-site.

[0007] 2. On the other hand, in the existing equipment detection system, wireless monitoring has gradually replaced the original wired monitoring system and is widely used due to its own advantages. The wireless monitoring system generally consists of wireless sensors and a wireless acquisition gateway. The sensors usually use batteries for power supply, and absolute time is used to synchronize between each measurement point. This synchronization method has low accuracy and is currently difficult to achieve in terms of hardware for order analysis that requires high-precision synchronous acquisition of rotational speed and vibration signals.

[0008] 3. Currently, in terms of non-rotational speed order analysis, there are two major categories of mainstream technical solutions. The first is through precise analysis of time-domain signals to extract the phase information of the original signal through waveforms. However, this type of method requires the original signal to have a high signal-to-noise ratio or obvious time-domain periodic characteristics, such as periodic impulses, periodic pulse signals, etc. However, vibration or sound signals in engineering applications basically contain a lot of background noise and the time-domain characteristics are not obvious. In this case, this type of method is often difficult to achieve.

[0009] 4. Another type of method is based on the time-frequency analysis of the original signal, such as short-time Fourier transform, wavelet transform, Gabor transform, etc. Before performing time-frequency analysis, this type of method generally needs to filter the original signal. Whether it is choosing the envelope signal in the low-frequency band or the resonance band for time-frequency analysis, it is necessary to involve the design of filter parameters, which will bring certain uncertainty to the result of frequency extraction. In addition, most of the current time-frequency ridge extraction processes require manual intervention, and the ridges extracted by simply optimizing the peaks may come from different frequency components in the original signal, resulting in the failure of order analysis. Summary of the Invention

[0010] The object of the present invention is to provide a non-steady signal non-rotational speed adaptive data processing method to solve the problem in the prior art of how to adaptively process non-steady vibration or sound signals, extract instantaneous frequency characteristics and restore spectral line characteristics in the spectrum through a completely data-driven method in the case of non-rotational speed data, so as to overcome the frequency ambiguity problem caused by the lack of rotational speed information in traditional methods.

[0011] The present invention realizes the above object through the following technical solutions:

[0012] In the first aspect, the present invention proposes a non-steady signal non-rotational speed adaptive data processing method, and the method includes:

[0013] Receiving the original non-steady signal of the device to be measured and calculating its time-frequency distribution spectrum;

[0014] Designing a time-varying frequency window in the frequency dimension of the time-frequency distribution spectrum, and the number of points within the window of the time-varying frequency window is determined according to the frequency domain width and sampling frequency of the original non-steady signal;

[0015] Calculate the maximum energy within the current frequency window and adjust the center frequency of the next frequency window according to the maximum energy;

[0016] Obtain the sum of the maximum frequency energies of all time-varying frequency windows and determine the optimal time-varying frequency window according to the sum of the maximum frequency energies;

[0017] Extract the frequency ridge line within the optimal time-varying frequency window and obtain the instantaneous frequency distribution of the original non-stationary signal;

[0018] Construct a decimated time series according to the instantaneous frequency distribution and resample the original non-stationary signal to obtain a new stationary signal of the device under test.

[0019] Further, the original non-stationary signal is any one of a vibration signal, a sound signal, a current signal, and a voltage signal.

[0020] Further, the time-frequency distribution spectrum is obtained by short-time Fourier transform. The calculation parameters of the short-time Fourier transform include a window function, a window length, an overlap rate, and a sampling frequency. The window function is a rectangular window.

[0021] Further, designing the time-varying frequency window in the frequency dimension of the time-frequency distribution spectrum includes:

[0022] Construct a time-varying frequency-domain window function with a fixed window length and a center frequency varying with time , The window length and the center frequency of which satisfy the following conditions:

[0023] ;

[0024] In the above formula, is the index after discretization in the time domain and the frequency domain, represents the frequency-domain window function with the starting center frequency of at time , is the fixed window length, is the center frequency of the window at time , is the window length of the window function, is the frequency in the time-frequency distribution.

[0025] Further, the method for determining the optimal time-varying frequency window includes:

[0026] Calculate the maximum energy within the time-varying frequency window corresponding to each starting center frequency;

[0027] Sum up the maximum energies of all time-varying frequency windows to obtain the energy sum corresponding to each starting center frequency , as shown in the following formula:

[0028] ;

[0029] Select the time-varying frequency window corresponding to the energy and the maximum starting center frequency as the optimal time-varying frequency window, as shown in the following formula:

[0030] ;

[0031] In the above formula, is the sampling frequency of the original signal.

[0032] Furthermore, constructing the secondary sampling time series according to the instantaneous frequency distribution includes:

[0033] Determine the characteristic expression of the optimal instantaneous frequency of the original signal, as shown in the following formula:

[0034] ;

[0035] According to the instantaneous frequency calculate the sampling interval at each moment, as shown in the following formula:

[0036] ;

[0037] In the above formula, and respectively represent the secondary sampling instantaneous sampling frequency and sampling interval at the moment, is the number of sampling points, is the fixed sampling interval, is the average sampling frequency, is the sampling frequency of the original signal;

[0038] Use the sampling interval to reverse cubic spline to establish the secondary sampling time series , as shown in the following formula:

[0039] ;

[0040] In the above formula, to represent the regenerated uniform time series, with a total of time points, represents the corresponding moment, is the fixed sampling interval.

[0041] Furthermore, resampling the original non-steady signal to obtain a new steady signal of the device under test includes:

[0042] Obtain the time series The signal amplitude corresponding to each moment in

[0043] Using cubic spline interpolation, combined with the signal amplitude, for the original unsteady signal Interpolate to obtain a new steady signal of the device under test , as follows:

[0044] .

[0045] In a second aspect, the present invention proposes a non-steady signal non-rotational speed adaptive data processing system, which is applied to execute the non-rotational speed adaptive data processing method described in any one of the above, and the system includes:

[0046] A signal acquisition module for receiving the original unsteady signal of the device under test;

[0047] A time-frequency distribution calculation module for calculating the time-frequency distribution spectrum of the original unsteady signal by using short-time Fourier transform;

[0048] A time-varying frequency window design module for designing a time-varying frequency window in the time-frequency distribution spectrum in the frequency dimension, and the number of points within the window of the time-varying frequency window is determined according to the frequency domain width and sampling frequency of the original unsteady signal;

[0049] A maximum energy calculation module for calculating the maximum energy within the current frequency window and adjusting the center frequency of the next frequency window according to the maximum energy;

[0050] An optimal time-varying frequency window determination module for obtaining the sum of the maximum frequency energies of all time-varying frequency windows and determining the optimal time-varying frequency window according to the sum of the maximum frequency energies;

[0051] An instantaneous frequency extraction module for extracting the frequency ridge line within the optimal time-varying frequency window and obtaining the instantaneous frequency distribution of the original unsteady signal;

[0052] A resampling module for constructing a secondary sampling time series according to the instantaneous frequency distribution and resampling the original unsteady signal to obtain a new steady signal of the device under test.

[0053] Furthermore, a design rule is preset in the time-varying frequency window design module, and the design rule is specifically that the center frequency of the time-varying frequency window changes with time and the window length of the time-varying frequency window is fixed.

[0054] The beneficial effects of the present invention are as follows:

[0055] The present invention proposes a fully data-driven adaptive processing strategy for unsteady vibration or sound signals in the absence of rotational speed data. By calculating the time-frequency distribution spectrogram of the signal, extracting the optimal frequency ridge line in the time-frequency diagram, obtaining the instantaneous frequency characteristics of the original signal, and then performing secondary sampling to obtain a new steady-state signal. This method effectively solves the problem of frequency ambiguity caused by the lack of rotational speed information in traditional methods, significantly improves the accuracy and reliability of unsteady signal processing, and is particularly suitable for scenarios where it is difficult to accurately synchronize rotational speed data with vibration or sound data in wireless monitoring systems.

[0056] The present invention adopts a method of extracting the frequency ridge line in the time-frequency domain, which has stronger robustness and universality compared with the method of extracting the phase only through the time-domain signal. By constructing a time-varying frequency window and based on the maximum time-frequency spectrum energy and optimized window function, the accuracy and convergence of the frequency ridge line extraction are ensured. This method not only avoids the ridge line extraction error caused by improper selection of the frequency window in traditional methods, but also significantly improves the robustness and accuracy of unsteady signal processing, and is applicable to various unsteady signals reflecting the equipment state, such as vibration, sound, current, and voltage signals. Description of the Drawings

[0057] Figure 1 It is a schematic flow chart of a method for processing non-steady signals without rotational speed adaptation data provided by an embodiment of the present application;

[0058] Figure 2 It is another schematic flow chart of a method for processing non-steady signals without rotational speed adaptation data provided by an embodiment of the present application;

[0059] Figure 3 It is the original waveform and its spectrum of the signal in an example of the specific implementation manner of the present application;

[0060] Figure 4 It is the time-frequency distribution spectrum of the original signal in an example of the specific implementation manner of the present application;

[0061] Figure 5 It is the distribution diagram of the center frequency of the time-varying window corresponding to a certain starting center frequency and the maximum energy in the frequency domain window in an example of the specific implementation manner of the present application;

[0062] Figure 6 It is the optimal time-varying window function group and the local time-frequency distribution diagram in the window in an example of the specific implementation manner of the present application;

[0063] Figure 7 It is the instantaneous frequency ridge line and the constructed new sampling time series in an example of the specific implementation manner of the present application;

[0064] Figure 8 It is the waveform and spectrum of the approximately steady-state signal after processing in an example of the specific implementation manner of the present application;

[0065] Figure 9 It is the time-frequency distribution spectrum of the approximately steady-state signal after processing in the specific embodiment of this application. Specific embodiment

[0066] The following further describes this application in detail with reference to the accompanying drawings. It is necessary to point out here that the following specific embodiments are only used to further illustrate this application and cannot be understood as limiting the protection scope of this application. Those skilled in the art can make some non-essential improvements and adjustments to this application based on the above application content.

[0067] Embodiment 1

[0068] As Figure 1-2 shown, this embodiment proposes a non-steady signal non-rotational speed adaptive data processing method. "Non-rotational speed" means that during the signal processing, it is not necessary to rely on a rotational speed sensor or other external devices to obtain the rotational speed information of the rotating machinery. The method includes the following steps: receiving the original non-steady signal of the device to be measured, and calculating its time-frequency distribution spectrum by using STFT (Short-Time Fourier Transform); designing a time-varying frequency window in the time-frequency distribution spectrum, and the number of points within the window of the time-varying frequency window is determined according to the frequency domain width and sampling frequency of the original non-steady signal; calculating the maximum energy within the current frequency window, and adjusting the center frequency of the next frequency window according to the maximum energy; obtaining the sum of the maximum frequency energies of all time-varying frequency windows, and determining the optimal time-varying frequency window according to the sum of the maximum frequency energies; extracting the frequency ridge line within the optimal time-varying frequency window, obtaining the instantaneous frequency distribution of the original non-steady signal; constructing a secondary sampling time series according to the instantaneous frequency distribution, and resampling the original non-steady signal to obtain a new steady signal of the device to be measured.

[0069] In a specific embodiment, the original non-steady signal is any one of vibration signal, sound signal, current signal, and voltage signal.

[0070] In a specific embodiment, the time-frequency distribution spectrum is obtained by short-time Fourier transform. The calculation parameters of the short-time Fourier transform include window function, window length, overlap rate, and sampling frequency, and the window function is a rectangular window.

[0071] Specifically, when calculating the time-frequency distribution of the original signal, other time-frequency analysis methods such as Hilbert-Huang transform, wavelet transform, and Gabor transform can also be used.

[0072] More specifically, taking to represent the original input signal, according to the following short-time Fourier transform formula, calculate the time-frequency distribution result of the original signal:

[0073] ;

[0074] where Unwindowed function, When using this formula to calculate the time-frequency distribution of discrete signals, it is necessary to determine parameters including window function, window length, overlap rate, sampling frequency, etc.

[0075] In the above scheme, a time-varying frequency window is designed in the frequency dimension in the time-frequency distribution spectrum, including: constructing a time-varying frequency domain window function with a fixed window length and a center frequency that varies with time , The window length and center frequency satisfy the following conditions:

[0076] ;

[0077] In the above formula, is the discretized index of time domain and frequency domain, Indicates the starting center frequency is At the moment The frequency domain window function when For fixed window length, for time The center frequency of the window, is the window length of the window function, is the frequency in the time-frequency distribution.

[0078] The method for determining the optimal time-varying frequency window includes: for each starting center frequency A corresponding set of time-varying window functions is calculated, and the maximum energy in the time-varying frequency window corresponding to each starting center frequency is calculated under each window function; the maximum energy of all time-varying frequency windows is summed to obtain the energy sum corresponding to each starting center frequency. , as follows:

[0079] ;

[0080] Finally calculate all The maximum value in , obtain the optimal time-varying window function group , select energy and The time-varying frequency window corresponding to the maximum starting center frequency is used as the optimal time-varying frequency window, as shown in the following formula:

[0081] ;

[0082] In the above formula, is the sampling frequency of the original signal.

[0083] It should be emphasized that, in the implementation process of the present invention, when constructing the time-varying frequency window, the maximum time-frequency spectrum energy in each group of time-varying windows is used as the optimization index of the window function. Specifically, for each starting center frequency A corresponding set of time-varying window functions, calculate the maximum energy in the frequency domain under each window function, and sum the maximum energies within all time-varying windows to obtain the energy sum corresponding to each starting center frequency. . By comparing all values, select the time-varying frequency window corresponding to the starting center frequency with the maximum energy sum as the optimal time-varying frequency window.

[0084] This optimization index ensures that the time-varying frequency window can adaptively focus on the most significant energy change region in the signal, thereby improving the accuracy and reliability of frequency ridge extraction. This method not only avoids the ridge extraction error caused by improper selection of the frequency window in the traditional method, but also significantly improves the robustness and accuracy of non-stationary signal processing.

[0085] In a specific embodiment, extract the frequency ridge within the above-mentioned optimal time-varying window to obtain the optimal instantaneous frequency feature expression of the original signal, and construct a sub-sampled time series according to the instantaneous frequency distribution, including: determining the optimal instantaneous frequency of the original signal, and the feature expression is as follows:

[0086] ;

[0087] In the implementation process of the present invention, the time-frequency ridge and instantaneous frequency distribution of the original signal are extracted through the optimal time-varying window function group, ensuring the effectiveness and convergence of the extraction results. Specifically, first determine the optimal time-varying frequency window according to the maximum energy sum in the time-frequency distribution spectrogram, and then extract the frequency ridge within this optimal window to obtain the instantaneous frequency distribution of the original signal. Since the optimal time-varying window function group can adaptively focus on the most significant energy change region in the signal, it avoids the ridge extraction error caused by improper selection of the frequency window in the traditional method. Through this method, the instantaneous frequency characteristics of the original signal can be effectively extracted, ensuring the accuracy of subsequent resampling and signal processing. This method not only improves the robustness of non-stationary signal processing, but also significantly improves the convergence of frequency ridge extraction.

[0088] In non-stationary signals, since the signal frequency changes with time, if the signal is sampled at a fixed sampling interval , then the number of sampling points within each period will also change with time. In order to restore the instantaneous frequency to the steady-state signal frequency, it is necessary to adopt a non-fixed time interval to ensure that there is a fixed number of sampling points within each period. Then, the time-varying sampling frequency and the signal instantaneous frequency should satisfy the following relationship:

[0089] ;

[0090] In the above formula, and respectively represent the secondary sampling instantaneous sampling frequency and sampling interval at time is the number of sampling points, is a fixed sampling interval, is the average sampling frequency, is the sampling frequency of the original signal;

[0091] Establish a secondary sampling time series using the inverse cubic spline of the sampling interval , as follows:

[0092] ;

[0093] In the above formula, to represent the regenerated uniform time series, with a total of time points, represents the corresponding time, is a fixed sampling interval.

[0094] In a specific embodiment, resample the original non-steady signal to obtain a new steady signal of the device under test, including: obtaining the signal amplitude corresponding to each time in the time series ; using cubic spline interpolation and combining the signal amplitude to interpolate the original non-steady signal to obtain a new steady signal of the device under test, as follows:

[0095] .

[0096] Based on the same inventive concept, this embodiment also proposes a non-steady signal non-rotational speed adaptive data processing system, which is applied to execute the non-rotational speed adaptive data processing method as described above. The system includes:

[0097] A signal acquisition module for receiving the original non-steady signal of the device under test;

[0098] A time-frequency distribution calculation module for calculating the time-frequency distribution spectrum of the original non-steady signal using the short-time Fourier transform;

[0099] A time-varying frequency window design module for designing a time-varying frequency window in the time-frequency distribution spectrum in the frequency dimension. The number of points within the window of the time-varying frequency window is determined according to the frequency domain width and sampling frequency of the original non-steady signal;

[0100] A maximum energy calculation module for calculating the maximum energy within the current frequency window and adjusting the center frequency of the next frequency window according to the maximum energy;

[0101] The optimal time-varying frequency window determination module is used to obtain the sum of the maximum frequency energies of all time-varying frequency windows and determine the optimal time-varying frequency window according to the sum of the maximum frequency energies;

[0102] The instantaneous frequency extraction module is used to extract the frequency ridge line within the optimal time-varying frequency window and obtain the instantaneous frequency distribution of the original non-steady-state signal;

[0103] The resampling module is used to construct a secondary sampling time series according to the instantaneous frequency distribution, resample the original non-steady-state signal, and obtain a new steady-state signal of the device under test.

[0104] In a specific embodiment, a design rule is preset in the time-varying frequency window design module. The design rule is specifically that the center frequency of the time-varying frequency window changes with time, and the window length of the time-varying frequency window is fixed.

[0105] It should be noted here that each module in the above non-rotational speed adaptive data processing system corresponds to each step in implementing the above non-rotational speed adaptive data processing method. The examples and application scenarios implemented by multiple modules and the corresponding steps are the same, but are not limited to the content disclosed in the above Embodiment 1.

[0106] To more clearly illustrate the present invention and its advantages, the following will further explain the method provided by the present invention in combination with specific examples and relevant partial figures.

[0107] 1. Install an acceleration sensor on the surface of the device to be monitored and sample at a certain sampling frequency to obtain a vibration signal (In the actual application of the present invention, it is not limited to vibration signals, and can also be non-steady-state process signals such as sound, current, voltage, etc. that can reflect the frequency conversion operation state of the device). The main acquisition parameters used in this example are: sampling frequency sampling points .

[0108] The following Figure 3 Taking the original signal shown as an example, the specific implementation steps of the present invention are further elaborated. The upper figure is the waveform of the original signal, and the lower figure is the corresponding spectrum. It can be clearly seen from the spectrum that since the device is in a variable rotational speed state during operation, the original frequency of this signal is broadened, making it difficult to accurately distinguish the original frequency components.

[0109] Figure 4 is the time-frequency distribution spectrogram of the original non-steady-state signal. The bright colors in the figure represent the positions with high spectral energy, and the dark colors indicate low spectral energy. It can be seen that there are several obvious frequency components between 1000 Hz and 1400 Hz, but these frequencies change with time, indicating that the original signal has frequency fluctuations during this time period and is a non-steady-state signal. This phenomenon leads toFigure 2 Frequency broadening appears in the spectrum. The time-frequency analysis parameters used in this example are: rectangular window, window length 256, overlap rate 255, analysis frequency 2000 Hz.

[0110] 2. Time-frequency distribution of the original signal , take a fixed frequency domain window width , starting center frequency , randomly select the starting center frequency . The frequency center sequence of the time-varying window function corresponding to it is as shown in Figure 5 . It can be seen that the center frequency of the frequency window at different times has a certain fluctuation. Further calculate the maximum energy in each window in the frequency domain, and sum all the maximum energies to obtain as shown in Figure 5 below. It can be seen that a maximum value will appear at a certain center frequency.

[0111] 3. According to the result of the previous step , calculate its maximum value to obtain the optimal time-varying window function group as shown in Figure 6 above. The local time-frequency distribution within the frequency window is as shown in Figure 6 below.

[0112] Extract the time-frequency distribution ridge line within the window function group, and finally obtain the approximate instantaneous frequency distribution of the original signal ( Figure 7 above).

[0113] 4. After obtaining the instantaneous frequency , use the instantaneous frequency integral to calculate how many signal cycles there are in total within the actual sampling time interval , and then use the inverse cubic spline to establish a new sampling time sequence .

[0114] The time difference between the new sampling time sequence and the original sampling time sequence is as shown in Figure 7 below. It can be seen that when the frequency is higher than , the secondary sampling interval becomes shorter and the sampling frequency increases. When the frequency is lower than , the secondary sampling interval becomes longer and the sampling frequency decreases.

[0115] 5. Finally, use the resampled time sequence combined with the cubic spline interpolation of the original signal to obtain a new approximate steady-state signal as shown in Figure 8 . The upper figure is the time-domain waveform, and the lower figure is the spectrum. It can be seen from the spectrum that the broadening phenomenon of the original several main frequency components has been well eliminated, and several frequency components are clearly visible. At Figure 9It can also be seen from the time-frequency spectrum that the main frequency components no longer change with time but basically become a straight line, indicating that the original non-steady signal has become an approximate steady signal after processing, which is convenient for subsequent further diagnosis and analysis.

[0116] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0117] In addition, in each embodiment of the present application, each functional module can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0118] The above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A non-steady signal non-rotational speed self-adaptive data processing method, characterized in that The method includes: Receiving the original non-steady-state signal of the device under test and calculating its time-frequency distribution spectrum; Designing a time-varying frequency window in the time-frequency distribution spectrum in the frequency dimension, where the number of points within the window of the time-varying frequency window is determined according to the frequency domain width and sampling frequency of the original non-steady-state signal; Calculating the maximum energy within the current frequency window and adjusting the center frequency of the next frequency window according to the maximum energy; Obtaining the sum of the maximum frequency energies of all time-varying frequency windows and determining the optimal time-varying frequency window according to the sum of the maximum frequency energies; Extracting the frequency ridge within the optimal time-varying frequency window and obtaining the instantaneous frequency distribution of the original non-steady-state signal; Constructing a secondary sampling time series according to the instantaneous frequency distribution and resampling the original non-steady-state signal to obtain a new steady-state signal of the device under test; The designing of the time-varying frequency window in the time-frequency distribution spectrum in the frequency dimension includes: A time-frequency domain window function with a fixed window length and a center frequency varying with time , The window length and center frequency of which satisfy the following conditions: ； In the above formula, is the index after discretization in the time domain and frequency domain, represents that the starting center frequency is at the moment the frequency domain window function, is the fixed window length, is the moment the center frequency of the window, is the window length of the window function, is the frequency in the time-frequency distribution.

2. The non-steady signal non-rotational speed adaptive data processing method according to claim 1, characterized in that, The original non-steady-state signal is any one of a vibration signal, a sound signal, an electric current signal, and a voltage signal.

3. The non-steady signal non-rotational speed adaptive data processing method according to claim 1, characterized in that, The time-frequency distribution spectrum is obtained through short-time Fourier transform, and the calculation parameters of the short-time Fourier transform include a window function, a window length, an overlap rate, and a sampling frequency, and the window function is a rectangular window.

4. A non-steady signal non-rotational speed adaptive data processing method according to claim 1, characterized in that The method for determining the optimal time-varying frequency window includes: Calculating the maximum energy within the time-varying frequency window corresponding to each starting center frequency; Sum the maximum energies of all time-varying frequency windows to obtain the energy sum corresponding to each starting center frequency , as shown in the following equation: ； In the above formula, is the sampling frequency of the original signal.

5. A non-steady signal non-rotational speed adaptive data processing method according to claim 4, characterized in that, The constructing of the secondary sampling time series according to the instantaneous frequency distribution includes: Determine the optimal instantaneous frequency of the original signal The characteristic expression is as follows: ； According to the instantaneous frequency Calculate the sampling interval at each moment as follows: ； In the above formula, and respectively represent the instantaneous sampling frequency and sampling interval of the secondary sampling at time is the number of sampling points, is the fixed sampling interval, is the average sampling frequency, is the sampling frequency of the original signal; Establish a secondary sampling time series using the inverse cubic spline of the sampling interval , as follows: ； In the above formula, to represent the regenerated uniform time series, with a total of time points, represents the corresponding moment, represents the corresponding moment, is the fixed sampling interval.

6. A non-steady signal non-rotational speed adaptive data processing method according to claim 5, characterized in that, The resampling of the original non-steady-state signal to obtain a new steady-state signal of the device under test includes: Obtain the time series and the signal amplitude corresponding to each moment; Using cubic spline interpolation and combining with the signal amplitude, interpolate the original unsteady signal to obtain a new steady signal of the device under test as follows: 。 7. A non-steady signal non-rotational speed adaptive data processing system, which is applied to execute the non-rotational speed adaptive data processing method described in any one of claims 1-6, and is characterized in that, The system includes: A signal acquisition module for receiving the original non-steady-state signal of the device under test; A time-frequency distribution calculation module for calculating the time-frequency distribution spectrum of the original non-steady-state signal by using short-time Fourier transform; A time-varying frequency window design module for designing a time-varying frequency window in the time-frequency distribution spectrum in the frequency dimension, where the number of points within the window of the time-varying frequency window is determined according to the frequency domain width and sampling frequency of the original non-steady-state signal; A maximum energy calculation module for calculating the maximum energy within the current frequency window and adjusting the center frequency of the next frequency window according to the maximum energy; An optimal time-varying frequency window determination module for obtaining the sum of the maximum frequency energies of all time-varying frequency windows and determining the optimal time-varying frequency window according to the sum of the maximum frequency energies; An instantaneous frequency extraction module for extracting the frequency ridge within the optimal time-varying frequency window and obtaining the instantaneous frequency distribution of the original non-steady-state signal; A resampling module for constructing a secondary sampling time series according to the instantaneous frequency distribution and resampling the original non-steady-state signal to obtain a new steady-state signal of the device under test.

8. An astatic signal non-rotating speed adaptive data processing system according to claim 7, characterized in that, A design rule is preset in the time-varying frequency window design module, and the design rule is specifically that the center frequency of the time-varying frequency window changes with time and the window length of the time-varying frequency window is fixed.

Images