A method, device, apparatus and medium for extracting an impact vibration signal

Abstract

This application discloses a method, device, apparatus, and medium for extracting impact vibration signals, relating to the field of vibration detection technology. The method involves obtaining the phase-locked center frequency (PLC) of the high-frequency amplification region within the frequency response curve of a vibration sensor, utilizing the sensor's frequency response amplification region to naturally enhance impact ringing, thereby improving the detectability of weak impacts within the selected high-frequency band. An in-phase reference signal and a quadrature reference signal corresponding to the PLC CSC are generated. The vibration signal output from the vibration sensor is simultaneously demodulated with the two reference signals using a switching method to obtain in-phase and quadrature components, effectively forming a narrowband synchronous detection around the PLC CSC, improving frequency selectivity and reducing dependence on ultra-wide flat-band sensors. Finally, a phase-robust impact envelope is determined based on the in-phase and quadrature components to output the impact vibration signal, making the envelope result less susceptible to changes in the impact arrival phase and improving detection stability.

Description

Technical Field

[0001] This application relates to the field of vibration detection technology, and in particular to a method, equipment, apparatus and medium for extracting impact vibration signals. Background Technology

[0002] In high-speed rotating equipment such as aero engines, there are strong low-frequency large-amplitude vibration signals during normal operation, such as alternating vibration, rotational frequency and its harmonics, aerodynamic excitation and installation structure vibration. However, the impact signals generated by early wear of components, minor chipping, cracks or local defects have small amplitude and short duration, and are often overwhelmed by the above-mentioned low-frequency large-amplitude vibration signals.

[0003] In addition, traditional vibration monitoring usually collects signals within the flat frequency band of the vibration sensor to ensure the calibrability of the signal amplitude and frequency response; however, the flat frequency band of the sensor is usually difficult to increase indefinitely. If a wider frequency band sensor and a higher sampling rate are used to capture the high-frequency components excited by the impact, it will result in higher hardware costs, larger data volume, and higher real-time processing pressure.

[0004] Therefore, how to extract the weak impact vibration signal that has been submerged without relying on sensors with ultra-wide flat frequency bands and high sampling rates is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a method, device, apparatus, and medium for extracting impact vibration signals, in order to solve the problems of conventional solutions relying on sensors with ultra-wide flat frequency bands and high sampling rates, and the impact vibration signals being easily submerged.

[0006] To address the aforementioned technical problems, this application provides a method for extracting impact vibration signals, comprising: Obtain the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor; Generate in-phase and quadrature reference signals corresponding to the phase-locked center frequency; The vibration signal output by the vibration sensor is simultaneously demodulated with the in-phase reference signal and the quadrature reference signal in a switching manner to obtain the in-phase component and the quadrature component. The impact envelope is determined based on the in-phase component and the quadrature component to output the impact vibration signal.

[0007] On the one hand, the high-frequency amplification region is the high-frequency band in the frequency response curve where the output sensitivity is higher than the average sensitivity of the flat frequency band, the phase-locked loop center frequency is located within the high-frequency amplification region, and at least one of the following conditions is met during the determination of the phase-locked loop center frequency: The sensitivity corresponding to the phase-locked center frequency is greater than the average sensitivity corresponding to the flat frequency band within the frequency response curve. The frequency difference between the low-frequency converter and the corresponding preset harmonic is greater than the preset frequency difference; It is confirmed that the vibration sensor and its corresponding front-end circuit do not saturate; High-frequency ringing caused by covering impact vibration signals.

[0008] On the other hand, the process of determining the phase-locked loop center frequency includes: Obtain the flat frequency band and corresponding average sensitivity of the vibration sensor; The first boundary of the high-frequency amplification region is defined in advance; The high-frequency amplification region is determined based on the first boundary, the average sensitivity, and the resonant frequency of the vibration sensor; The candidate frequency evaluation function is determined based on the safety penalty item, impact ringing energy, rotor interference energy, and noise energy. Within the high-frequency amplification region, the phase-locked center frequency is determined based on the candidate frequency evaluation function.

[0009] On the other hand, before performing switching synchronous demodulation with the in-phase reference signal and the quadrature reference signal, the method further includes: The vibration signal output by the vibration sensor is subjected to anti-low-frequency submersion front-end processing to obtain the final vibration signal; The final vibration signal is used as the input signal and then subjected to a switch-type synchronous demodulation step with the in-phase reference signal and the quadrature reference signal, respectively.

[0010] On the other hand, the anti-low-frequency flooding front-end processing includes at least one or more of the following: high-pass filtering, band-pass filtering, notch filtering, limiting, automatic gain control, differential input, isolation amplification, and anti-aliasing filtering.

[0011] On the other hand, the specific process of the anti-low-frequency flooding front-end processing includes: The vibration signal is subjected to high-pass filtering to obtain the filtered first vibration signal; The first vibration signal is subjected to a bandpass filter to obtain the filtered second vibration signal; The second vibration signal is subjected to a second bandpass filter to obtain the filtered third vibration signal; The third vibration signal is processed by automatic gain control to obtain the final vibration signal.

[0012] On the other hand, the phase difference between the in-phase reference signal and the quadrature reference signal is 90°, and the in-phase reference signal and the quadrature reference signal are square wave switching signals that correspond to the phase-locked loop center frequency and switch between +1 and -1, or sine reference signal and cosine reference signal that correspond to the phase-locked loop center frequency.

[0013] On the other hand, the switch-type synchronous demodulation is achieved by in-phase switch phase-locked channels and quadrature switch phase-locked channels, each of which includes an analog switch, a resistor network, and a low-pass filter; The control terminal of the analog switch is connected to the corresponding reference signal, the input terminal of the analog switch is connected to the vibration signal, and the output terminal of the analog switch is connected to the low-pass filter through the resistor network, so that the low-pass filter outputs the corresponding in-phase component or quadrature component.

[0014] On the other hand, the vibration signal output by the vibration sensor is subjected to switching synchronous demodulation with the in-phase reference signal and the quadrature reference signal, respectively, to obtain the in-phase component and the quadrature component, including: The vibration signal is multiplied or chopper-processed with the in-phase reference signal and the quadrature reference signal respectively to obtain the processed first signal and second signal. The first signal and the second signal are filtered to obtain the corresponding in-phase component and quadrature component.

[0015] On the other hand, determining the impact envelope based on the in-phase component and the quadrature component to output the impact vibration signal includes: The impact envelope is calculated based on the in-phase component and the quadrature component. The impact envelope is subjected to fault analysis processing to output the impact vibration signal.

[0016] On the other hand, calculating the impact envelope based on the in-phase component and the quadrature component includes: The in-phase component and the quadrature component are squared respectively to obtain the first squared component and the second squared component. The first squared component and the second squared component are summed to obtain the third squared component; The impact envelope is obtained by taking the square root of the third square component.

[0017] On the other hand, the fault analysis and processing method includes at least one or more of the following processing methods: smoothing processing, threshold detection processing, impact counting processing, impact interval statistical processing, and fault classification processing.

[0018] On the other hand, fault analysis processing is performed on the impact envelope to output the impact vibration signal, including: Envelope spectrum analysis was performed on the impact envelope to obtain the envelope spectrum of the impact envelope; The fault type is determined based on the spectral peak characteristic parameters corresponding to the fault characteristic frequency in the envelope spectrum, so as to output the impact vibration signal.

[0019] On the other hand, the frequency response curve corresponding to the high-frequency amplification region comes from the inherent frequency response of the vibration sensor itself, or from the combined frequency response formed by the mounting base of the vibration sensor, local resonance of the mechanical structure, and additional sensitive structures.

[0020] On the other hand, the bandwidth of the bandpass filter in the anti-low-frequency flooding front-end processing method covers the preset energy value of the impact ringing, and the bandwidth is selected when it covers the preset energy value of the impact ringing.

[0021] On the other hand, the front-end gain of the automatic gain control processing of the anti-low-frequency flooding front-end processing method is obtained by matching the range of low-frequency vibration amplitude and impact amplitude.

[0022] On the other hand, it also includes: Acquire the rotational speed signal of the rotating equipment; If the rate of change of the rotational speed signal is greater than a preset rate of change, the impact envelope is resampled according to the rotational speed signal; or, the low-pass bandwidth corresponding to the in-phase component and the quadrature component, as well as the diagnostic threshold used when outputting the impact vibration signal, are adjusted according to the operating conditions of the rotating equipment.

[0023] On the other hand, the cutoff frequency of the low-pass filter is greater than the impact repetition frequency and fault characteristic frequency corresponding to the impact vibration signal, and less than the phase-locked center frequency.

[0024] To address the aforementioned technical problems, this application also provides an impact vibration signal extraction device, comprising: The acquisition module is used to acquire the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor; The generation module is used to generate in-phase reference signals and quadrature reference signals corresponding to the phase-locked center frequency; The demodulation module is used to perform on / off synchronous demodulation of the vibration signal output by the vibration sensor with the in-phase reference signal and the quadrature reference signal, respectively, to obtain the in-phase component and the quadrature component; The output module is used to determine the impact envelope based on the in-phase component and the quadrature component, so as to output the impact vibration signal.

[0025] To address the aforementioned technical problems, this application also provides an impact vibration signal extraction device, comprising: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the method for extracting impact vibration signals as described above.

[0026] To address the aforementioned technical problems, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the impact vibration signal extraction method as described above.

[0027] This application provides a method for extracting impact vibration signals. First, the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor is obtained. The phase-locked operating frequency is determined within the high-frequency amplification region (rather than the flat frequency band) of the sensor's frequency response curve. The high sensitivity of the sensor to this frequency band enhances the high-frequency ringing excited by the impact. This eliminates the need to use a wider flat frequency band sensor and a higher sampling rate to capture the high-frequency impact component, thereby reducing hardware costs, data volume, and real-time processing pressure, and overcoming the problems of relying on ultra-wide flat frequency band sensors and high sampling rates. Simultaneously, frequency selection is performed before demodulation, pre-locking the target frequency band for subsequent narrowband synchronous detection. Second, in-phase and quadrature reference signals corresponding to the phase-locked center frequency are generated, providing two (in-phase and quadrature) reference signals corresponding to the operating frequency. This lays the foundation for phase-robust dual-channel synchronous detection, enabling the simultaneous acquisition of in-phase and quadrature components, avoiding missed detections or inconsistent amplitude estimations in single-phase detection when the impact ringing reaches a random phase. Next, the vibration signal output from the vibration sensor is simultaneously demodulated with both in-phase and quadrature reference signals using a switching method to obtain in-phase and quadrature components. The impact ringing near this frequency is shifted to a low-frequency baseband, while suppressing low-frequency large-amplitude vibrations, frequency harmonics, and broadband noise deviating from this frequency. This allows for narrow-band, highly selective extraction of the submerged impact component within a strong low-frequency background. Finally, the impact envelope is determined based on the in-phase and quadrature components to output the impact vibration signal. The resulting impact envelope, synthesized from the in-phase and quadrature components, is independent of the specific phase at impact arrival, improving phase robustness and detection stability. Based on this, an impact vibration signal suitable for early fault identification is output, containing interpretable information such as the impact occurrence time and repetition period, facilitating subsequent threshold detection and envelope spectrum analysis.

[0028] In addition, this application also provides an impact vibration signal extraction device, apparatus, and medium, which have the same beneficial effects as the impact vibration signal extraction method described above. Attached Figure Description

[0029] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 A flowchart illustrating a method for extracting impact vibration signals provided in an embodiment of this application; Figure 2 A schematic diagram of sensor frequency response curve and phase-locked loop center frequency selection provided for embodiments of this application; Figure 3 A schematic diagram of a vibration sensor and a simulated low-frequency flooding front-end circuit is provided for embodiments of this application; Figure 4 A schematic diagram of a switch-type dual-phase phase-locked demodulation circuit provided in this application embodiment; Figure 5 A structural diagram of an impact vibration signal extraction device provided in an embodiment of this application; Figure 6 A timing diagram for a two-phase switch phase-locked loop provided in an embodiment of this application; Figure 7 A schematic diagram of another impact vibration signal extraction process provided in an embodiment of this application; Figure 8 A structural diagram of an impact vibration signal extraction device provided in an embodiment of this application; Figure 9 A structural diagram of an impact vibration signal extraction device provided in an embodiment of this application; Figure 10 A schematic diagram showing the frequency response curve of a sensor and the markings of f1, f2, f3, f4, resonant frequency, and phase-locked loop operating frequency provided for embodiments of this application; Figure 11 This application provides a schematic diagram of a local magnification of the resonant region of the frequency response curve and the selection of the phase-locked loop operating frequency. Figure 12 A schematic diagram illustrating the time-domain superposition relationship between low-frequency large-amplitude vibration and weak impact response, provided for an embodiment of this application; Figure 13 A schematic diagram illustrating the magnitude difference between low-frequency large-amplitude vibration and weak impact response under a dual-coordinate axis, provided for an embodiment of this application; Figure 14 This is a schematic diagram illustrating the frequency domain behavior of a low-frequency large-amplitude vibration after incorporating a weak impact, as provided in an embodiment of this application. Figure 15A schematic diagram illustrating how a weak impact in the original sensor output signal is masked by a low-frequency large-amplitude vibration, as provided in an embodiment of this application. Figure 16 A schematic diagram showing a spectrum comparison before low-frequency flooding front-end processing, provided for an embodiment of this application; Figure 17 A schematic diagram showing a comparison of the spectrum after anti-low-frequency flooding front-end processing, provided in an embodiment of this application; Figure 18 This application provides an embodiment of a switch-type two-phase phase-locked loop demodulated in-phase circuit. I Channels and Orthogonals Q Channel baseband output schematic diagram; Figure 19 A schematic diagram of the time-domain result of the impact envelope obtained by synthesizing in-phase / positive traffic channels, provided in an embodiment of this application; Figure 20 A frequency domain result of extracting the envelope of a switch-type dual-phase phase-locked loop and a magnified display diagram of the 270Hz fault frequency are provided for embodiments of this application; Figure 21 A schematic diagram comparing the envelope spectrum of a switch-type dual-phase lock-in circuit with that of a conventional Hilbert envelope spectrum, provided for an embodiment of this application; Figure 22 A time-domain comparison diagram of a phase-locked envelope and a normal Hilbert envelope provided for embodiments of this application; Figure 23 A comparison diagram of frequency domain details in the 0-2000Hz range between phase-locked envelope and ordinary Hilbert envelope, provided for embodiments of this application; Figure 24 A time-domain waveform diagram of a weak impact response provided in an embodiment of this application; Figure 25 This is a magnified view of a weak impact response spectrum and a local area near the impact ringing frequency, provided for an embodiment of this application. Detailed Implementation

[0031] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0032] The core of this application is to provide a method, device, apparatus, and medium for extracting impact vibration signals, in order to solve the problems of conventional solutions that rely on sensors with ultra-wide flat frequency bands and high sampling rates, and where impact vibration signals are easily submerged.

[0033] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0034] To facilitate understanding, the following is a brief explanation of the terminology and background knowledge involved in this application. The frequency response curve of a vibration sensor represents the relationship between the output sensitivity of the vibration sensor and vibration inputs of different frequencies. Conventional measurements generally prioritize using a relatively flat operating frequency band to ensure accurate amplitude measurement. However, in the region close to the sensor's natural frequency or the resonant frequency of the mounting structure, sensitivity may increase; this is referred to as the frequency response amplification region. Low-frequency, large-amplitude vibration signals refer to vibration signals with large amplitudes, low frequencies, or concentrated near low-order harmonics, generated by factors such as rotor imbalance, aerodynamic excitation, foundation vibration, and periodic alternating loads. These signals may drown out weak impact signals generated by early faults. Weak impact vibration signals refer to short-term impacts caused by local anomalies such as bearing rolling element defects, minor blade chipping, localized wear, crack propagation, and tooth surface pitting. The impact itself has a small amplitude but usually excites high-frequency damped ringing. In rotating equipment, the impacts of early faults are often very weak and easily masked by low-frequency, large-amplitude vibrations when observed directly in the time domain or ordinary frequency spectrum. However, when an impact occurs, it excites the sensor, mounting base, or the structure under test to generate high-frequency ringing. If the ringing envelope can be amplified and extracted within this high-frequency sensitive band, weak impacts can be identified without increasing the cost of full-bandwidth sampling and processing. The high-frequency setting is between 30kHz and 60kHz, and the low-frequency setting is between 0 and 10kHz.

[0035] Figure 1 A flowchart illustrating a method for extracting impact vibration signals provided in this application embodiment is shown below. Figure 1 As shown, the method includes: S11: Obtain the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor; S12: Generate in-phase and quadrature reference signals corresponding to the phase-locked center frequency; S13: The vibration signal output by the vibration sensor is synchronously demodulated with the in-phase reference signal and the quadrature reference signal in a switching manner to obtain the in-phase component and the quadrature component. S14: Determine the impact envelope based on the in-phase and quadrature components to output the impact vibration signal.

[0036] Specifically, in step S11, the vibration sensor is a device installed on the rotating equipment under test, used to collect raw vibration signals containing low-frequency large-amplitude vibrations and weak impact vibrations. The frequency response curve is a curve representing the relationship between the output sensitivity of the vibration sensor and vibration inputs of different frequencies. It can be obtained during factory calibration, offline calibration after installation, or self-calibration during the initial operation of the equipment. The high-frequency amplification region refers to the area where the output sensitivity of the vibration sensor or sensor-mounting structure assembly is significantly higher than the average sensitivity of the flat frequency band within a certain high-frequency range. Although this region may not be suitable for absolute amplitude calibration measurements, it can be used to enhance the high-frequency ringing components excited by weak impacts. The phase-locked loop center frequency refers to the frequency selected within the high-frequency amplification region as the center of the target frequency band for extracting impact ringing, denoted as . Select To ensure its sensitivity is higher than the average sensitivity of the flat frequency band, far from the main low-frequency transitions and their strong harmonics, without causing saturation of the preamplifier or acquisition circuit, and capable of covering the high-frequency ringing excited by the impulse signal, its effective bandwidth is denoted as . B .

[0037] In step S12, an in-phase reference signal and a quadrature reference signal corresponding to the phase-locked loop center frequency are generated, both of which have frequencies corresponding to the phase-locked loop center frequency. Correspondingly, the phase difference is 90°. In the switching implementation, the reference signal can be a square wave switching signal with values ​​of +1 and -1; in the digital implementation, a sign multiplication or lookup table sine / cosine reference signal can also be used. By using two reference signals with a 90° phase difference to synchronously detect the input signal, in-phase and quadrature components can be obtained. The amplitude envelope is then calculated from the two components, avoiding missed detections in single-phase lockout due to the unstable phase of the impact signal.

[0038] Step S13, the switch-type synchronous demodulation, is a synchronous detection process. Its core is not ordinary amplification, but rather multiplying the input signal with a reference switch signal or performing switch chopping, equivalent to frequency selection of the input signal. Subsequently, low-pass filtering preserves narrowband information near the reference frequency while suppressing other frequency components. This application employs dual-phase lock-in demodulation, which splits the pre-processed vibration signal into two independent switching channels: the in-phase channel is controlled by a reference square wave with a phase of 0°, and the quadrature channel is controlled by a reference square wave with a phase of 90°. This dual-channel parallel processing simultaneously extracts the amplitude and phase information of the signal, obtaining the in-phase detection signal and the quadrature detection signal, respectively. These are then low-pass filtered or integrated to obtain the in-phase component. and orthogonal components This process makes The nearby high-frequency impact ringing components are shifted to the low-frequency baseband, while... Unrelated low-frequency large-amplitude vibrations, noise, and broadband interference are suppressed in subsequent low-pass filtering.

[0039] In step S14, the impact envelope refers to the outer envelope of the high-frequency impact ringing amplitude as it changes over time. This envelope typically contains information such as the impact occurrence time, the impact repetition period, and the fault characteristic frequency. Based on the in-phase component... and orthogonal components The impact envelope is calculated, which is related to the high-frequency impact ringing energy and does not depend on the specific phase when the impact arrives. Then, the impact envelope is smoothed, thresholded, impact intervals are statistically analyzed, envelope spectrum is analyzed, or faults are classified. This allows the output of the existence of weak impacts, the time of impact occurrence, the impact repetition frequency, or the equipment fault status, thus realizing the extraction and output of impact vibration signals.

[0040] Furthermore, assuming the original vibration signal output by the sensor includes low-frequency large-amplitude vibration, weak impact vibration, and noise, it can be expressed as: ,in Indicates low-frequency, large-amplitude vibration. Indicates weak impact vibration. This represents noise. When a weak impulse occurs, it can excite damped ringing near the high-frequency amplification region, thus affecting the phase-locked loop center frequency. There are extractable narrowband high-frequency components nearby. The two-phase lock-in channels are calculated separately. Where LPF represents low-pass filtering or integration. For in-phase reference signal, For orthogonal reference signals, The vibration signal entering the phase-locked channel is used; finally, the impact envelope is calculated. Due to the envelope... Simultaneously utilizing both in-phase and positive traffic channels can reduce detection errors caused by the uncertainty in the arrival phase of the impact signal; low-frequency large-amplitude vibrations, because they are not in... The signal is located nearby and is suppressed during phase-locked low-pass processing, making it difficult to form a stable baseband output.

[0041] The method for extracting impact vibration signals in this embodiment can be applied to the detection of weak impact faults in rotating machinery such as aero-engines, rail vehicle running gears, bearings, gearboxes, wind turbines, and gas turbines. It can also be applied to other scenarios where weak impacts need to be detected under strong interference backgrounds. No limitation is made here, and it can be set according to the actual situation.

[0042] This application provides a method for extracting impact vibration signals. First, the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor is obtained. The phase-locked operating frequency is determined to be within the high-frequency amplification region (rather than the flat frequency band) where the sensor's frequency response curve is elevated. The high sensitivity of the sensor to this frequency band enhances the high-frequency ringing excited by the impact. This eliminates the need to use a wider flat frequency band sensor and a higher sampling rate to capture the high-frequency impact component, thereby reducing hardware costs, data volume, and real-time processing pressure, and overcoming the problem of relying on ultra-wide flat frequency band sensors and high sampling rates. Simultaneously, frequency selection is performed before demodulation, pre-locking the target frequency band for subsequent narrowband synchronous detection. Second, an in-phase reference signal and a quadrature reference signal corresponding to the phase-locked center frequency are generated, providing two (in-phase and quadrature) reference signals corresponding to the operating frequency. This lays the foundation for phase-robust dual-channel synchronous detection, enabling the simultaneous acquisition of in-phase and quadrature components, avoiding missed detections or inconsistent amplitude estimations in single-phase detection when the impact ringing reaches a random phase. Next, the vibration signal output from the vibration sensor is simultaneously demodulated with both in-phase and quadrature reference signals using a switching method to obtain in-phase and quadrature components. The impact ringing near this frequency is shifted to a low-frequency baseband, while suppressing low-frequency large-amplitude vibrations, frequency harmonics, and broadband noise deviating from this frequency. This allows for narrow-band, highly selective extraction of the submerged impact component within a strong low-frequency background. Finally, the impact envelope is determined based on the in-phase and quadrature components to output the impact vibration signal. The resulting impact envelope, synthesized from the in-phase and quadrature components, is independent of the specific phase at impact arrival, improving phase robustness and detection stability. Based on this, an impact vibration signal suitable for early fault identification is output, containing interpretable information such as the impact occurrence time and repetition period, facilitating subsequent threshold detection and envelope spectrum analysis.

[0043] In some embodiments, the high-frequency amplification region is a high-frequency band in the frequency response curve where the output sensitivity is higher than the average sensitivity of the flat frequency band, the phase-locked loop center frequency is located within the high-frequency amplification region, and at least one of the following conditions is met during the determination of the phase-locked loop center frequency: The sensitivity corresponding to the center frequency of the phase-locked loop is greater than the average sensitivity corresponding to the flat frequency band within the frequency response curve. The frequency difference between the low-frequency converter and the corresponding preset harmonic is greater than the preset frequency difference; Ensure that the vibration sensor and its corresponding front-end circuit do not saturate; High-frequency ringing caused by covering impact vibration signals.

[0044] Specifically, the selected phase-locked center frequency Its sensitivity is higher than the average sensitivity of the flat frequency band, so as to take advantage of the frequency response amplification region to naturally enhance the impact ringing. Maintaining a frequency difference greater than a preset value between the vibration sensor and its low-frequency switching frequency and low-order harmonics is to avoid the main switching frequency, low-order harmonics, and other known strong interference frequencies, thereby reducing rotor interference entering the phase-locked loop. Ensuring that the vibration sensor and its corresponding front-end circuit do not saturate is to avoid introducing excessive nonlinearity or saturation risks by selecting a frequency point too close to the resonance peak. Covering the high-frequency ringing excited by the impact signal ensures that the target impact energy falls within the target frequency band extracted by the phase-locked loop.

[0045] Figure 2 A schematic diagram of sensor frequency response curve and phase-locked loop center frequency selection provided for embodiments of this application is shown below. Figure 2 As shown in the figure, the horizontal axis represents the resonant frequency, and the vertical axis represents the relative sensitivity or gain. Along the frequency direction, the low-frequency large-amplitude vibration region, the flat frequency band, the frequency response amplification region, and the resonant frequency are marked sequentially. Conventional vibration measurements typically emphasize the use of the flat frequency band, while this application utilizes the high-frequency response amplification region to enhance the damped ringing excited by weak impacts, i.e., selecting the phase-locked center frequency within the frequency response amplification region. This allows the selected frequency point to be located near the resonant frequency while avoiding excessive nonlinearity or saturation risks, thus making it more sensitive to impact ringing. It should be noted that... Figure 2 This application is used to explain that the phase-locked loop center frequency is determined based on the frequency response amplification region of the sensor or sensor-mounting structure combination, rather than being arbitrarily set according to a fixed empirical frequency, so that the selected phase-locked loop operating frequency has a physical basis.

[0046] The constraints on the high-frequency amplification region and the phase-locked loop center frequency provided in this embodiment ensure that the phase-locked loop center frequency is not arbitrarily set, but is determined based on the frequency response amplification region of the sensor or sensor-mounting structure combination. This enhances weak impact ringing while avoiding strong interference, preventing saturation, and covering the target ringing energy. As a result, it improves the detectability of weak impacts while reducing rotor interference entering the phase-locked loop channel, thereby improving the selectivity and extraction stability of the target frequency band.

[0047] In some embodiments, the process of determining the phase-locked loop center frequency includes: Obtain the flat frequency band and corresponding average sensitivity of the vibration sensor; Predefine the first boundary of the high-frequency amplification region; The high-frequency amplification region is determined based on the first boundary, average sensitivity, and resonant frequency of the vibration sensor; The candidate frequency evaluation function is determined based on the safety penalty item, impact ringing energy, rotor interference energy, and noise energy. Within the high-frequency amplification region, the phase-locked loop center frequency is determined based on the candidate frequency evaluation function.

[0048] Specifically, the selection of the frequency response amplification region is the first step that distinguishes the proposed method from traditional envelope demodulation. Traditional envelope demodulation typically selects the resonant band after digitization, while the proposed method determines the phase-locked loop operating frequency before analog demodulation.

[0049] Let the flat response band of the sensor be ( The average sensitivity within this frequency band is defined as: ; The frequency response amplification region used for phase-locked loop detection is defined as: ; in, To amplify the threshold, This is the lower boundary of the high-frequency amplification region. The resonant frequency of the sensor or mounting structure.

[0050] To avoid nonlinear amplification or front-end saturation caused by the operating frequency being too close to the resonant peak, a safety penalty term is introduced: ; in, It is close to the upper boundary of the resonant region. The penalty coefficient is... .

[0051] The candidate frequency evaluation function is defined as follows: ; in, For frequency Nearby impact ringing energy, For rotor interference energy, For noise energy, It is a small positive number.

[0052] The phase-locked loop operating frequency is selected as follows: .

[0053] In other words, the flat frequency band and average sensitivity of the sensor are first obtained as the benchmark for determining the frequency response amplification region. A first boundary of the high-frequency amplification region is predefined, for example, the frequency point where the sensitivity increases relative to the average sensitivity of the flat frequency band by a preset proportion is used as the lower boundary. Then, combining the first boundary, average sensitivity, and sensor resonant frequency, the frequency band located above the flat frequency band, near the resonant frequency, and with increased sensitivity is determined as the high-frequency amplification region. Within this high-frequency amplification region, a candidate frequency evaluation function is constructed, consisting of a safety penalty term, impact ringing energy, rotor interference energy, and noise energy. A higher impact ringing energy, lower rotor interference energy and noise energy, and greater distance from saturation risk (lower safety penalty term) result in a better evaluation. Finally, the frequency point within the high-frequency amplification region that optimizes the evaluation function is selected as the phase-locked loop center frequency.

[0054] The phase-locked loop center frequency determination process provided in this embodiment combines the determination of the frequency response amplification region boundary with the candidate frequency, so that the selection of the phase-locked loop center frequency takes into account the enhancement of impact ringing, rotor interference and noise suppression, and front-end safety, and realizes the quantification of the phase-locked loop center frequency. This allows for the selection of a working frequency point within the frequency response amplification region that is more sensitive to impact ringing and more robust to interference, further improving the extraction effect and detection stability of weak impacts.

[0055] In some embodiments, prior to switching synchronous demodulation with the in-phase reference signal and the quadrature reference signal, the method further includes: The vibration signal output by the vibration sensor is subjected to anti-low-frequency submersion front-end processing to obtain the final vibration signal; The final vibration signal is used as the input signal and then subjected to a switch-type synchronous demodulation step with the in-phase reference signal and the quadrature reference signal, respectively.

[0056] Specifically, the anti-low-frequency submersion front-end refers to the analog or digital preprocessing link arranged before phase-locked demodulation. It is used to suppress low-frequency large-amplitude vibrations before phase-locked demodulation and analog-to-digital conversion, avoiding full dynamic range or front-end saturation. After the sensor output signal is processed by the anti-low-frequency submersion front-end, the final vibration signal is obtained. This final vibration signal is then used as the input signal to enter the switch-type synchronous demodulation step, improving the visibility of weak impacts from the source.

[0057] The low-frequency flooding front-end processing provided in this embodiment reduces the occupation of the back-end dynamic range by low-frequency large-amplitude vibrations before phase-locked demodulation, reduces amplitude compression, saturation and insufficient quantization resolution of sensors, preamplifiers or analog-to-digital converters, provides more suitable input for the phase-locked channel, thereby avoiding weak impacts being compressed or masked before entering the algorithm, and improving the effectiveness of subsequent phase-locked extraction.

[0058] In some embodiments, the anti-low-frequency flooding front-end processing includes at least one or more of the following: high-pass filtering, band-pass filtering, notch filtering, limiting, automatic gain control, differential input, isolation amplification, and anti-aliasing filtering.

[0059] Specifically, high-pass filtering can be used to remove low-frequency high-amplitude components, and a high-pass filter can be used to remove low-frequency high-amplitude components. The system uses a bandpass filter to limit the frequency band entering the phase-locked loop (PLL) module, a notch filter to suppress known frequency harmonics, and amplitude limiting and automatic gain control to prevent front-end saturation. Depending on the application scenario, one or more of the following can be selected and combined: differential input, isolation amplification, or anti-aliasing filtering. These processing methods can be flexibly selected according to the actual application scenario, and can be used individually or in combination.

[0060] The diverse combinations of anti-low-frequency flooding front-end processing methods provided in this embodiment can flexibly configure the front-end link for different devices, different installation environments and different interference characteristics. While preserving high-frequency impact ringing components as much as possible, it can fully suppress low-frequency large-amplitude vibrations and known interferences, reduce the risk of front-end saturation, and improve the adaptability of the solution to complex working conditions and the utilization rate of dynamic range.

[0061] In some embodiments, the specific process of the anti-low-frequency flooding front-end processing includes: The vibration signal is subjected to high-pass filtering to obtain the first filtered vibration signal; The first vibration signal is subjected to a bandpass filter to obtain the filtered second vibration signal; The second vibration signal is subjected to a second bandpass filter to obtain the filtered third vibration signal; The third vibration signal is processed by automatic gain control to obtain the final vibration signal.

[0062] Specifically, after determining the phase-locked loop center frequency, the sensor signal enters the analog anti-low-frequency flooding front end. This front end does not perform final diagnostics, but rather suppresses low-frequency amplitude vibrations and confines the signal to... Nearby, prepare suitable inputs for phase-locked demodulation.

[0063] This front-end can be summarized as follows: ; in, This is the transfer function corresponding to the high-pass or low-frequency suppression circuit. This indicates that for known switching harmonics and blade passing frequency... fs Notch filter circuit, For Centered on, with a bandwidth of B Analog band-limiting circuit, This indicates limiting, low-noise amplification, or gain matching, meaning... It is a vibration signal. This is the first vibration signal. This is the second vibration signal. This is the third vibration signal. This is the final vibration signal.

[0064] This simulation processing is crucial. If the original vibration signal... Directly inputting the signal to the analog-to-digital converter (ADC) would result in low-frequency, large-amplitude rotor vibrations dominating the quantization range, with weak impacts only corresponding to very small changes in digital code values. After the above processing, the strong low-frequency background is weakened, while the target high-frequency ringing is preserved, thereby reducing the risk of weak impacts being masked by quantization noise.

[0065] Figure 3 A schematic diagram of a vibration sensor and a simulated low-frequency flooding-resistant front-end circuit is provided for embodiments of this application, as shown below. Figure 3 As shown, the output of the piezoelectric vibration sensor is converted into a differential charge amplifier. The formula for a differential charge amplifier is: ,in, This is the output voltage of the differential charge amplifier. The amount of charge generated by the sensor. For feedback capacitors; furthermore, in the simulated low-frequency flooding front-end stage, also called the simulated low-frequency masking front-end, after high-pass suppression, notch filtering, and surrounding... Bandpass conditioning and automatic gain control (AGC) processing are performed to generate phase-locked input. For high-pass level suppression, the formula is as follows: ,in, This is the output signal of the high-pass filter. Here is the transfer function for high-pass or low-frequency suppression circuits; the formula for a notch filter is as follows: ,in, For the notch filter output signal, To address known switching harmonics and blade passing frequencies fr Notch filter circuit; around The formula for bandpass conditioning is as follows: ,in, The output signal of the bandpass filter. For Centered on, with a bandwidth of B The analog band-limiting circuit; the formulas for limiting / automatic gain control are as follows: ,in, For the final vibration signal, This indicates limiting, low-noise amplification, or gain matching.

[0066] The specific front-end processing provided in this embodiment, which involves high-pass filtering, first-pass band-pass filtering, second-pass band-pass filtering, and automatic gain control in a cascaded manner, suppresses low-frequency large-amplitude components and narrows the target frequency band step by step. While ensuring unsaturation, it improves the resolution of weak impulses, making the signal entering the phase-locked channel more concentrated in the frequency band where the high-frequency impulse ringing occurs, thereby improving the frequency selectivity and extraction signal-to-noise ratio of subsequent synchronous demodulation.

[0067] In some embodiments, the phase difference between the in-phase reference signal and the quadrature reference signal is 90°, and the in-phase reference signal and the quadrature reference signal are square wave switching signals that correspond to the phase-locked loop center frequency and switch between +1 and -1, or sine reference signal and cosine reference signal that correspond to the phase-locked loop center frequency.

[0068] Specifically, the reference signal can be a square wave switching signal or a sine / cosine reference signal. Square wave switching signals are easy to implement with low-cost hardware; sine / cosine reference signals help reduce harmonic effects. If a square wave switching reference is used, the influence of odd harmonics of the square wave should be considered, and a harmonic signal can be added before phase-locked loop (PLL). The reference waveform is either a bandpass filter centered on the phase-locked loop or a harmonic suppression reference waveform. Regardless of the method used, both the in-phase and quadrature reference signals are related to the phase-locked loop center frequency. They correspond and are 90° out of phase.

[0069] The two reference signals with a 90° phase difference provided in this embodiment, and their optional implementation as square waves or sine / cosine waves, can achieve low-cost, easy-to-implement synchronous demodulation in real time using square wave switching, and can reduce the influence of reference harmonics using sine / cosine waves. Thus, while ensuring the orthogonality of the in-phase and quadrature paths, it balances implementation cost and harmonic suppression, and improves the phase robustness and accuracy of envelope extraction.

[0070] In some embodiments, switch-type synchronous demodulation is achieved by in-phase switch phase-locked channels and quadrature switch phase-locked channels, each of which includes an analog switch, a resistor network, and a low-pass filter. The control terminal of the analog switch is connected to the corresponding reference signal, the input terminal of the analog switch is connected to the vibration signal, and the output terminal of the analog switch is connected to the low-pass filter through a resistor network. The low-pass filter outputs the corresponding in-phase or quadrature components.

[0071] Specifically, Figure 4 A schematic diagram of a switch-type two-phase phase-locked loop demodulation circuit provided in this application embodiment is shown below. Figure 4 As shown, the in-phase switch phase-locked loop (PLL) and quadrature switch PLL can be implemented using analog switches, resistor networks, and low-pass filters. Taking the in-phase channel as an example, the control terminal of the analog switch is connected to an in-phase reference switch signal, and the input terminal is connected to a pre-processed vibration signal. Under the control of the reference signal, the analog switch performs in-phase switching chopping on the input signal. Its output is then fed into a low-pass filter after passing through a resistor network, and the low-pass filter outputs the in-phase component. The orthogonal circuit is similar; its analog switch control terminal is connected to an orthogonal reference switch signal, and the orthogonal component is ultimately output by a low-pass filter. This structure can also be implemented using analog-to-digital converters, digital multipliers, digital low-pass filters, and embedded processors.

[0072] The switch-locked phase-locked channel provided in this embodiment, consisting of an analog switch, a resistor network, and a low-pass filter, provides a specific implementation of switch-type two-phase phase-locked demodulation in hardware circuitry. Switch-type multiplication, low-pass integration, and envelope calculation can all be implemented by analog circuits, digital circuits, microcontrollers, or field-programmable gate arrays, facilitating embedded online monitoring and improving the real-time hardware implementation and engineering feasibility of the solution.

[0073] In some embodiments, the vibration signal output by the vibration sensor is subjected to switching synchronous demodulation with both an in-phase reference signal and a quadrature reference signal to obtain in-phase and quadrature components, including: The vibration signal is multiplied with the in-phase reference signal and the quadrature reference signal respectively, or switched and chopped to obtain the processed first signal and second signal. The first and second signals are filtered to obtain the corresponding in-phase and quadrature components.

[0074] Specifically, the analog-switched dual-phase lock-in demodulator is the core of the proposed method. It consists of a reference signal generator, a 90° phase shifter, an in-phase / quadrature analog switch multiplier, and two low-pass integrators, and is located before the ADC to coherently extract the frequency band of weak impact targets.

[0075] In-phase and quadrature reference signals are defined as follows: , ; The two reference paths are 90° out of phase. After synchronous chopping and low-pass integration, the analog baseband output is: ; The above formula describes both in-phase and orthogonal paths. Only with... Only by referencing coherent high-frequency ringing components nearby can a stable baseband output be formed after multiplication and low-pass integration. The first signal and the second signal are obtained by multiplying them sequentially from top to bottom; the rotor fundamental frequency, low-order harmonics, incoherent high-order harmonics and broadband noise will be suppressed by average.

[0076] This coherent selectivity is key to improving the signal-to-noise ratio in two-phase switch demodulation. Conventional bandpass filtering relies solely on the passband width; if rotor harmonics fall into the passband, they will still be included in the envelope analysis. In contrast, phase-locked loop demodulation requires not only close frequency but also... It also requires that the reference remain coherent during the integration time.

[0077] The conditioned signal and the in-phase and quadrature switch references and Synchronous multiplication, followed by low-pass integration to obtain the baseband output. and .

[0078] Figure 5A structural diagram of an impact vibration signal extraction device provided in an embodiment of this application is shown below. Figure 5 The diagram illustrates the connections between the vibration sensor, front-end processing module, reference signal generation module, dual-phase lock-in module, low-pass integration module, and envelope recognition module.

[0079] Figure 6 A timing diagram for a two-phase switch phase-locked loop provided in an embodiment of this application is shown below. Figure 6 As shown in the figure, the input high-frequency ringing is shown from top to bottom. In-phase reference Orthogonal reference In-phase low-pass output Orthogonal low-pass output and the final impact envelope The in-phase reference and the quadrature reference are 90° out of phase. Whenever high-frequency ringing occurs at the input, the two low-pass outputs form a low-frequency envelope pulse, which is finally synthesized to obtain a phase-robust impulse envelope output. Figure 6 This is used to illustrate the timing relationship of two-phase switch phase-locked demodulation and the formation process of the impulse envelope.

[0080] The multiplication or switching chopper plus filtering processing method provided in this embodiment downconverts the target high-frequency ringing to a low-frequency baseband by multiplying or switching chopper at the same frequency. Then, it retains the narrowband information near the reference frequency and suppresses other frequency components by low-pass filtering, which is equivalent to forming a narrowband synchronous detection around the phase-locked loop center frequency. This allows for the stable extraction of in-phase and quadrature components while suppressing low-frequency large-amplitude vibrations, noise and broadband interference, thereby improving the selectivity of the target frequency band.

[0081] In some embodiments, determining the impact envelope based on the in-phase component and the quadrature component to output an impact vibration signal includes: The impact envelope is calculated based on the in-phase and quadrature components. Fault analysis and processing are performed on the impact envelope to output the impact vibration signal.

[0082] Specifically, first, the in-phase components... and orthogonal components Synthesized to obtain the impact envelope Compared to directly observing the original signal, the envelope more closely approximates the time distribution of the impact. Further fault analysis processing of this impact envelope, such as smoothing, threshold discrimination, impact counting, impact interval statistics, envelope spectrum analysis, or fault classification, can then output impact vibration signals indicating the presence of weak impacts, the time of impact occurrence, impact repetition frequency, or equipment fault status.

[0083] The method provided in this embodiment, which calculates the impact envelope before performing fault analysis, uses the phase-robust impact envelope as the input for fault analysis. This allows subsequent analysis to be built on the enhanced and denoised low-frequency envelope, facilitating the formation of interpretable diagnostic features such as the impact occurrence time, impact repetition frequency, envelope spectrum peak value, and energy index, thereby improving the accuracy and interpretability of early fault identification.

[0084] In some embodiments, calculating the impact envelope based on the in-phase component and the quadrature component includes: The in-phase component and the quadrature component are squared respectively to obtain the first squared component and the second squared component. The third square component is obtained by summing the first square component and the second square component. The impact envelope is obtained by taking the square root of the third square component.

[0085] Specifically, the formula is as follows: ;in, The first square component, The second square component. The third square component, This is the impact envelope. This envelope is related to the high-frequency impact ringing energy and does not depend on the specific phase when the impact arrives.

[0086] The impact envelope calculation method provided in this embodiment, which uses square, summation, and square root, synthesizes the amplitude envelope from in-phase and quadrature components. This allows the envelope result to utilize the energy of both in-phase and quadrature channels, thereby reducing the detection error caused by the uncertainty of the arrival phase of the impact signal. It also ensures that the envelope amplitude extracted from impacts of the same intensity at different phases remains consistent, improving phase robustness and detection stability.

[0087] In some embodiments, the fault analysis processing method includes at least one or more of the following processing methods: smoothing processing, threshold detection processing, impact counting processing, impact interval statistical processing, and fault classification processing.

[0088] Specifically, smoothing is used to suppress random fluctuations in the envelope; threshold detection is used to determine the presence of an impact; impact counting is used to count the number of impacts per unit time; impact interval statistics are used to estimate the impact repetition period; and fault classification is used to determine the fault type by combining impact characteristics. These processing methods can be used individually or in combination according to diagnostic needs.

[0089] The various optional fault analysis and processing methods provided in this embodiment can be flexibly combined according to different diagnostic objectives to extract multi-dimensional features such as the presence of impact, number of impacts, impact interval and fault type from the impact envelope. This facilitates integration with traditional fault diagnosis rules and improves the comprehensiveness and adaptability of fault identification.

[0090] In some embodiments, fault analysis processing is performed on the impact envelope to output an impact vibration signal, including: Envelope spectrum analysis was performed on the impact envelope to obtain the envelope spectrum of the impact envelope; The fault type is determined based on the spectral peak characteristic parameters corresponding to the fault characteristic frequency in the envelope spectrum, so as to output the impact vibration signal.

[0091] Specifically, digital envelope synthesis has two advantages. First, it avoids the nonlinear errors and temperature drift caused by analog squaring, summing, and square root circuits. Second, it preserves the original... I and Q Baseband information facilitates amplitude, phase, envelope spectrum, and multi-condition analysis.

[0092] like Nearby narrowband impact ringing is approximately Then the two-phase outputs are approximately proportional to each other. and Therefore, the aforementioned envelope pairs phase. It is insensitive and can stably reflect changes in impact energy.

[0093] The failure frequency was finally estimated by the mean-free digital envelope spectrum. ; ;in, To estimate the failure frequency, To impact the envelope, Represents the envelope sequence The average value, T For a period of time, t For the current time, For Fourier transform, Fault frequency search set / search interval, if If the nearby envelope spectral peaks are significantly higher than the local background, it is identified as a local fault in the corresponding bearing.

[0094] The envelope spectrum analysis and peak matching method provided in this embodiment makes the components near the fault frequency more interpretable because the envelope comes from the synchronous demodulation output of the frequency response amplification region. It can form identifiable peaks in the envelope spectrum, which can be used to identify the type of weak impact faults such as early bearing failure, slight blade chipping or local wear, and improve the reliability of fault location.

[0095] In some embodiments, the frequency response curve corresponding to the high-frequency amplification region comes from the inherent frequency response of the vibration sensor itself, or from the combined frequency response formed by the vibration sensor's mounting base, local resonance of the mechanical structure, and additional sensitive structures.

[0096] Specifically, the frequency response amplification region can originate from the sensor's inherent frequency response, i.e., the resonant region formed by the internal mass block and elastic structure of the sensor; or it can originate from the combined frequency response formed by the sensor mounting base, local resonance of the mechanical structure, or additional sensitive structures, i.e., the sensitivity enhancement region exhibited in a certain high-frequency band after the sensor is coupled with the mounting structure. Both types of sources can serve as target frequency bands for enhancing weak impact ringing.

[0097] The diversity of frequency response amplification sources provided in this embodiment allows the solution to go beyond the inherent frequency response of the sensor itself. It can also make full use of the combined frequency response formed by the mounting base, local resonance of the mechanical structure, or additional sensitive structures, thereby expanding the selection range of the available high-frequency amplification region. This makes it easier to find the operating frequency band sensitive to impact ringing under different installation conditions and improves the engineering adaptability of the solution.

[0098] In some embodiments, the bandwidth of the bandpass filter in the anti-low-frequency flooding front-end processing method covers the preset energy value of the impact ringing, and the bandwidth is selected as a preset bandwidth when it covers the preset energy value of the impact ringing.

[0099] Specifically, front-end bandwidth B The main energy of the impact ringing should be covered, that is, the preset energy value of the impact ringing should be covered, so as to ensure that the target impact energy enters the phase-locked channel; at the same time, while meeting the requirement of covering the preset energy value, the bandwidth should be as narrow as possible, that is, the preset bandwidth should be selected as small as possible, so as to improve the equivalent quality factor and interference suppression capability.

[0100] The bandpass bandwidth selection principle provided in this embodiment achieves a balance between covering the main energy of impact ringing and narrowing the passband, so that the bandwidth can retain the impact energy while being as narrow as possible. This reduces the entry of frequency conversion harmonics, structural vibration and broadband noise into the phase-locked channel while effectively forming a high-quality factor narrowband extraction, thereby improving the selectivity of the target frequency band and the extraction signal-to-noise ratio.

[0101] In some embodiments, the front-end gain of the automatic gain control processing of the anti-low-frequency flooding front-end processing method is obtained by matching the low-frequency vibration amplitude and the impact amplitude range.

[0102] Specifically, the front-end gain should be matched according to the amplitude range of low-frequency vibrations and impacts, prioritizing unsaturation and then pursuing resolution for weak impacts. That is, the front-end gain should be automatically adjusted based on the amplitude of the current low-frequency large-amplitude vibrations and the expected amplitude range of weak impacts, so that the signal does not exceed the dynamic range of the back-end and saturate, while maximizing the effective resolution of weak impacts.

[0103] The front-end gain matching method provided in this embodiment matches the front-end gain of automatic gain control with the range of low-frequency vibration amplitude and impact amplitude. This avoids front-end saturation while making full use of the effective number of bits of the analog-to-digital converter, improving the quantization resolution and dynamic range utilization of weak impacts, and further reducing the risk of weak impacts being compressed or submerged in the acquisition stage.

[0104] In some embodiments, it also includes: Acquire the rotational speed signal of the rotating equipment; If the rate of change of the rotation speed signal is greater than the preset rate of change, the impact envelope is resampled according to the rotation speed signal; or, the low-pass bandwidth corresponding to the in-phase component and the quadrature component, as well as the diagnostic threshold used when outputting the impact vibration signal, are adjusted according to the operating conditions of the rotating equipment.

[0105] Specifically, if the equipment speed changes significantly, i.e. the rate of change of the speed signal is greater than the preset rate of change, the impact envelope can be resampled in order based on the speed signal to align the speed-related fault features to the order domain; or the low-pass bandwidth corresponding to the in-phase and quadrature components and the diagnostic threshold used when outputting the impact vibration signal can be dynamically adjusted according to the equipment's operating conditions to make the extraction and discrimination parameters adapt to the operating conditions.

[0106] The speed adaptive processing method provided in this embodiment enables the solution to stably extract speed-related fault features even under variable operating conditions with large speed fluctuations by resampling order or dynamically adjusting the low-pass bandwidth and diagnostic threshold according to the operating conditions. This reduces feature frequency drift and misjudgment caused by speed changes, and improves the robustness and detection stability of the solution under variable operating conditions.

[0107] In some embodiments, the cutoff frequency of the low-pass filter is greater than the impact repetition frequency and fault characteristic frequency corresponding to the impact vibration signal, and less than the phase-locked loop center frequency.

[0108] Specifically, the low-pass filter cutoff frequency should be higher than the expected impulse repetition frequency and fault characteristic frequency to ensure that the impulse repetition information and fault characteristic information in the impulse envelope can be completely preserved; at the same time, this cutoff frequency should be much lower than the phase-locked loop center frequency. To fully filter out the high-frequency components remaining after synchronous demodulation and those related to... Unrelated interfering components.

[0109] The low-pass filter cutoff frequency conditions provided in this embodiment, while preserving the impact repetition frequency and fault characteristic frequency, fully suppress high-frequency residues and irrelevant interference, so that the in-phase and quadrature components can more accurately reflect the energy changes of the target impact, thereby improving the extraction quality of the impact envelope and the accuracy of subsequent fault analysis.

[0110] Figure 7This is a schematic diagram of another impact vibration signal extraction process provided in an embodiment of this application, as shown below. Figure 7 As shown, during frequency response calibration, the frequency response curve of the sensor is acquired, and then the sensor is selected within the frequency response amplification region to avoid strong low-frequency interference. Further steps include anti-low-frequency submersion front-end processing, dual-phase lock-in demodulation, envelope extraction, and fault output.

[0111] The foregoing has described in detail various embodiments of the method for extracting impact vibration signals. Based on this, this application also discloses an apparatus for extracting impact vibration signals corresponding to the above-described method. Figure 8 This is a structural diagram of an impact vibration signal extraction device provided in an embodiment of this application. Figure 8 As shown, the device for extracting impact vibration signals includes: The acquisition module 11 is used to acquire the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor; Generation module 12 is used to generate in-phase reference signal and quadrature reference signal corresponding to the phase-locked center frequency; The demodulation module 13 is used to perform on-off synchronous demodulation of the vibration signal output by the vibration sensor with the in-phase reference signal and the quadrature reference signal, respectively, to obtain the in-phase component and the quadrature component; Output module 14 is used to determine the impact envelope based on the in-phase component and the quadrature component in order to output the impact vibration signal.

[0112] Since the embodiments of the device part correspond to the embodiments described above, please refer to the embodiments of the method part for the description of the device part, and will not be repeated here.

[0113] For a description of the impact vibration signal extraction device provided in this application, please refer to the above method embodiments. This application will not repeat the description here, but it has the same beneficial effects as the above impact vibration signal extraction method.

[0114] Figure 9 A structural diagram of an impact vibration signal extraction device provided in an embodiment of this application is shown below. Figure 9 As shown, the device includes: Memory 21 is used to store computer programs; Processor 22 is used to implement the steps of the method for extracting impact vibration signals when executing a computer program.

[0115] The impact vibration signal extraction device provided in this embodiment may include, but is not limited to, tablet computers, laptop computers, or desktop computers.

[0116] The processor 22 may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor 22 may be implemented using at least one of the following hardware forms: Digital Signal Processor (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 22 may also include a main processor and a coprocessor. The main processor, also known as the Central Processing Unit (CPU), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 22 may integrate a Graphics Processing Unit (GPU), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor 22 may also include an Artificial Intelligence (AI) processor, which handles computational operations related to machine learning.

[0117] The memory 21 may include one or more computer-readable storage media, which may be non-transitory. The memory 21 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In this embodiment, the memory 21 is used to store at least the following computer program 211, which, after being loaded and executed by the processor 22, is capable of implementing the relevant steps of the impact vibration signal extraction method disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 21 may also include an operating system 212 and data 213, etc., and the storage method may be temporary storage or permanent storage. The operating system 212 may include Windows, Unix, Linux, etc. The data 213 may include, but is not limited to, the data involved in the impact vibration signal extraction method, etc.

[0118] In some embodiments, the shock vibration signal extraction device may further include a display screen 23, an input / output interface 24, a communication interface 25, a power supply 26, and a communication bus 27.

[0119] Those skilled in the field can understand, Figure 9 The structure shown does not constitute a limitation on the device for extracting impact vibration signals and may include more or fewer components than shown.

[0120] The processor 22 implements the method for extracting impact vibration signals provided in any of the above embodiments by calling instructions stored in the memory 21.

[0121] For a description of the impact vibration signal extraction device provided in this application, please refer to the above method embodiments. This application will not repeat the description here, but it has the same beneficial effects as the above impact vibration signal extraction method.

[0122] Furthermore, this application also provides a computer-readable storage medium storing a computer program, which, when executed by processor 22, implements the steps of the above-described method for extracting impact vibration signals.

[0123] It is understood that if the methods in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and executes all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] For a description of the computer-readable storage medium provided in this application, please refer to the above method embodiments. This application will not repeat the description here, but it has the same beneficial effects as the above-described method for extracting impact vibration signals.

[0125] The simulation uses a sampling rate of 204800Hz. The low-frequency large-amplitude vibration is composed of rotor / blade related components of 96Hz, 192Hz, 384Hz, 576Hz and 864Hz, superimposed with a small amount of background noise. The fault impact repetition frequency is set to 270Hz. The high-frequency impact ringing frequency is located near the frequency response amplification region of the sensor. The sensor frequency response parameters include f1=5Hz, f2=20Hz, f3=12kHz, f4=32kHz, and a resonant frequency of approximately 40kHz. The phase-locked operating frequency is currently selected as 38kHz, which is automatically selected based on the frequency response amplification region and envelope spectrum response. Figure 10 A sensor frequency response curve and schematic diagram indicating f1, f2, f3, f4, resonant frequency, and phase-locked loop operating frequency are provided for embodiments of this application, as shown below. Figure 10 As shown, conventional vibration measurements typically emphasize flat frequency bands, while this scheme intentionally utilizes the high-frequency response amplification region to enhance the damped ringing excited by weak impacts. Figure 11 This application provides a schematic diagram of a local magnification of the resonant region of the frequency response curve and the selection of the phase-locked loop operating frequency, as shown in the embodiment. Figure 11 As shown, the frequency response curve further magnifies the region near the resonance, highlighting the relationship between the resonant frequency and the selected phase-locked loop (PLL) operating frequency. This figure supports the following technical characteristics: the PLL center frequency is not arbitrarily set, but determined based on the frequency response amplification range of the sensor or sensor-mounting structure combination, and a frequency point more sensitive to impact ringing is selected while avoiding excessive nonlinearity or saturation risks. It should be noted that... Figure 10 and Figure 11 The horizontal axis represents frequency (Hz), and the vertical axis represents relative gain (dB). Additionally, in... Figure 10 The frequencies marked f1 and f2 cannot be displayed in the graph because the horizontal axis range is in kHz.

[0126] Figure 12 This application provides a schematic diagram illustrating the time-domain superposition relationship between low-frequency amplitude vibration and weak impact response, as shown in the embodiment. Figure 12 As shown, the horizontal axis represents time (s) and the vertical axis represents amplitude (g), displaying the amplitude difference between the related vibrations and weak high-frequency effects of large rotors / blades within the same time window. Low-frequency large-amplitude vibrations dominate the amplitude range, and weak impacts are not easily observed directly in ordinary time-domain waveforms. This is precisely why a low-frequency flooding front-end and phase-locked loop extraction are needed. Figure 13 This application provides a schematic diagram illustrating the magnitude difference between low-frequency amplitude vibration and weak impact response under a dual-coordinate system, as shown in the embodiment of the present application. Figure 13 As shown, the horizontal axis represents time in seconds (s), the left vertical axis represents the amplitude of strong low-frequency vibration in g, and the right vertical axis represents the amplitude of the weak impact response. Placing the strong low-frequency vibration and the weak impact on the left and right axes respectively emphasizes the difference in their magnitudes. This figure illustrates that if the raw signal is directly acquired at full range, the low-frequency components will significantly occupy the dynamic range of the sensor and analog-to-digital converter, reducing the resolution of the weak impact. Figure 14 This application provides a schematic diagram illustrating the frequency domain behavior of a low-frequency large-amplitude vibration after incorporating a weak impact, as shown in the embodiment of this application. Figure 14 As shown, the horizontal axis represents frequency (Hz), and the vertical axis represents amplitude, displaying the frequency domain distribution after the superposition of strong low-frequency vibration and weak impact. The low-frequency rotor-related components dominate in the low-frequency range, while the high-frequency ringing energy corresponding to the weak impact requires a combination of sensor frequency response amplification range and narrowband synchronous detection for stable extraction. Figure 15 This is a schematic diagram illustrating how a weak impact in the original sensor output signal is masked by a low-frequency large-amplitude vibration, as provided in an embodiment of this application. Figure 15As shown, the horizontal axis represents time in seconds, and the vertical axis represents amplitude, presenting the local time-domain waveform of the sensor output signal. It can be seen that in the presence of low-frequency large-amplitude vibrations and background noise, weak impacts do not necessarily form obvious spikes in the original waveform; therefore, simple time-domain threshold detection is prone to missing detections.

[0127] Figure 16 A schematic diagram of spectrum comparison before low-frequency flooding front-end processing is provided for an embodiment of this application, as shown below. Figure 16 As shown in the figure, the spectrum is the sensor output spectrum. Figure 17 A schematic diagram showing a spectrum comparison after anti-low-frequency flooding front-end processing is provided for an embodiment of this application, as shown below. Figure 17 As shown in the figure, the spectrum is the spectrum after low-frequency front-end processing. Figure 16 , 17 The x-axis represents frequency in Hz, and the y-axis represents amplitude, comparing the sensor output spectrum and the spectrum after low-frequency flooding front-end processing. Front-end processing reduces the impact of low-frequency amplitude components on subsequent phase-locked demodulation channels through high-pass / band-limiting methods, making target ringing components near the high-frequency amplification region more suitable for entering the synchronous detection module.

[0128] Figure 18 This application provides an embodiment of a switch-type two-phase phase-locked loop demodulated in-phase circuit. I Channels and Orthogonals Q Channel baseband output schematic diagram, as shown Figure 18 As shown, the horizontal axis represents time in seconds, and the vertical axis represents amplitude, displaying the baseband output of the in-phase and positive-mode traffic channels. Both signals were obtained by synchronous demodulating the same high-frequency amplification region signal with a reference signal 90° out of phase, providing a basis for subsequent calculations of the phase-robust impact envelope. Figure 19 This application provides a schematic diagram of the time-domain result of the impact envelope obtained by synthesizing in-phase / positive traffic channels, as shown in the embodiments of this application. Figure 19 As shown, the horizontal axis represents time in seconds, and the vertical axis represents the envelope amplitude, displaying the result of... and The synthesized impact envelope. Compared with direct observation of the original signal, the envelope more closely approximates the time distribution of the impact, making it suitable for further impact counting, interval statistics, and envelope spectrum analysis. Figure 20 This application provides a schematic diagram of the frequency domain results of a switch-type dual-phase lock-in extraction envelope and a magnified display of the 270Hz fault frequency, as shown in the embodiment of this application. Figure 20 As shown, the horizontal axis represents frequency in Hz, and the vertical axis represents amplitude, revealing a significant peak around 270Hz in the phase-locked envelope spectrum. A magnified view illustrates the prominence of this fault frequency peak relative to the adjacent background spectral lines, demonstrating that this method can recover the fault impact repetition frequency from a low-frequency, large-amplitude vibration background.

[0129] Figure 21 A schematic diagram comparing the envelope spectrum of a switch-type dual-phase lock-in circuit with that of a conventional Hilbert envelope spectrum is provided for an embodiment of this application, as shown below. Figure 21 As shown, the horizontal axis represents frequency (Hz), and the vertical axis represents amplitude. The phase-locked envelope spectrum of this scheme is compared with the ordinary Hilbert envelope spectrum. This figure illustrates that this scheme is not a simple back-end envelope processing, but rather enhances the target frequency band through frequency response amplification region selection and phase-locked synchronization detection, making the components near the fault frequency more interpretable. Figure 22 This is a time-domain comparison diagram of a phase-locked envelope and a normal Hilbert envelope provided in an embodiment of this application, as shown below. Figure 22 As shown, the horizontal axis represents time in seconds, and the vertical axis represents the envelope amplitude, further comparing the time-domain performance of the two envelope results. The phase-locked envelope, derived from the synchronous demodulation output of the selected frequency response amplification region, can more directly reflect the energy changes of the target's high-frequency ringing; the ordinary envelope, on the other hand, relies more on the passband selection of the bandpass filter. Figure 23 A frequency domain detail comparison diagram in the 0-2000Hz range of phase-locked envelope and ordinary Hilbert envelope is provided for embodiments of this application, such as... Figure 23 As shown, the two envelope spectra are displayed in two separate figures, with the horizontal axis representing frequency in Hz and the vertical axis representing amplitude. The frequency on the horizontal axis is extended to 2000 Hz to observe the frequency domain characteristics near the fault frequency and its harmonics.

[0130] Figure 24 A time-domain waveform diagram of a weak impulse response provided in an embodiment of this application, such as... Figure 24 As shown, the horizontal axis represents time in seconds, and the vertical axis represents amplitude. This figure separately illustrates the time-domain characteristics of a weak impact response, which is short in duration, small in amplitude, and exhibits high-frequency decaying ringing features. This demonstrates that weak impacts are more suitable for detection through high-frequency ringing envelopes, rather than relying solely on the original time-domain peak values ​​against a low-frequency, large-amplitude vibration background. Figure 25 A weak impact response spectrum and a locally magnified view near the impact ringing frequency are provided for embodiments of this application, as shown in the figure. Figure 25 As shown, the horizontal axis represents frequency (Hz), and the vertical axis represents amplitude, indicating that the high-frequency ringing energy of the weak impulse response is concentrated near the sensor's frequency response amplification region. This result is consistent with... Figure 10 , Figure 11 The frequency response amplification regions in the two circuits correspond to each other, indicating that the selection of the phase-locked loop operating frequency has a physical basis.

[0131] The above provides a detailed description of the method, apparatus, device, and medium for extracting impact vibration signals provided in this application. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

[0132] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

Claims

1. A method for extracting impact vibration signals, characterized in that, include: Obtain the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor; Generate in-phase and quadrature reference signals corresponding to the phase-locked center frequency; The vibration signal output by the vibration sensor is simultaneously demodulated with the in-phase reference signal and the quadrature reference signal in a switching manner to obtain the in-phase component and the quadrature component. The impact envelope is determined based on the in-phase component and the quadrature component to output the impact vibration signal.

2. The method for extracting impact vibration signals according to claim 1, characterized in that, The high-frequency amplification region is the high-frequency band in the frequency response curve where the output sensitivity is higher than the average sensitivity of the flat frequency band. The phase-locked loop center frequency is located within the high-frequency amplification region, and at least one of the following conditions is met during the determination of the phase-locked loop center frequency: The sensitivity corresponding to the phase-locked center frequency is greater than the average sensitivity corresponding to the flat frequency band within the frequency response curve. The frequency difference between the low-frequency converter and the corresponding preset harmonic is greater than the preset frequency difference; It is confirmed that the vibration sensor and its corresponding front-end circuit do not saturate; High-frequency ringing caused by covering impact vibration signals.

3. The method for extracting impact vibration signals according to claim 2, characterized in that, The process of determining the center frequency of the phase-locked loop includes: Obtain the flat frequency band and corresponding average sensitivity of the vibration sensor; The first boundary of the high-frequency amplification region is defined in advance; The high-frequency amplification region is determined based on the first boundary, the average sensitivity, and the resonant frequency of the vibration sensor; The candidate frequency evaluation function is determined based on the safety penalty item, impact ringing energy, rotor interference energy, and noise energy. Within the high-frequency amplification region, the phase-locked center frequency is determined based on the candidate frequency evaluation function.

4. The method for extracting impact vibration signals according to claim 1, characterized in that, Before performing on / off synchronous demodulation with the in-phase reference signal and the quadrature reference signal, the method further includes: The vibration signal output by the vibration sensor is subjected to anti-low-frequency submersion front-end processing to obtain the final vibration signal; The final vibration signal is used as the input signal and then subjected to a switch-type synchronous demodulation step with the in-phase reference signal and the quadrature reference signal, respectively.

5. The method for extracting impact vibration signals according to claim 4, characterized in that, The anti-low-frequency flooding front-end processing includes at least one or more of the following: high-pass filtering, band-pass filtering, notch filtering, limiting, automatic gain control, differential input, isolation amplification, and anti-aliasing filtering.

6. The method for extracting impact vibration signals according to claim 5, characterized in that, The specific process of the anti-low-frequency flooding front-end processing includes: The vibration signal is subjected to high-pass filtering to obtain the filtered first vibration signal; The first vibration signal is subjected to a bandpass filter to obtain the filtered second vibration signal; The second vibration signal is subjected to a second bandpass filter to obtain the filtered third vibration signal; The third vibration signal is processed by automatic gain control to obtain the final vibration signal.

7. The method for extracting impact vibration signals according to claim 1, characterized in that, The in-phase reference signal and the quadrature reference signal are 90° out of phase, and the in-phase reference signal and the quadrature reference signal are square wave switching signals that correspond to the phase-locked loop center frequency and switch between +1 and -1, or sine reference signal and cosine reference signal that correspond to the phase-locked loop center frequency.

8. The method for extracting impact vibration signals according to claim 7, characterized in that, The switch-type synchronous demodulation is achieved by in-phase switch phase-locked channels and quadrature switch phase-locked channels. Each of the switch phase-locked channels includes an analog switch, a resistor network, and a low-pass filter. The control terminal of the analog switch is connected to the corresponding reference signal, the input terminal of the analog switch is connected to the vibration signal, and the output terminal of the analog switch is connected to the low-pass filter through the resistor network, so that the low-pass filter outputs the corresponding in-phase component or quadrature component.

9. The method for extracting impact vibration signals according to claim 8, characterized in that, The vibration signal output by the vibration sensor is simultaneously demodulated with the in-phase reference signal and the quadrature reference signal using a switching method to obtain the in-phase component and the quadrature component, including: The vibration signal is multiplied or chopper-processed with the in-phase reference signal and the quadrature reference signal respectively to obtain the processed first signal and second signal. The first signal and the second signal are filtered to obtain the corresponding in-phase component and quadrature component.

10. The method for extracting impact vibration signals according to claim 1, characterized in that, Determining the impact envelope based on the in-phase component and the quadrature component to output the impact vibration signal includes: The impact envelope is calculated based on the in-phase component and the quadrature component. The impact envelope is subjected to fault analysis processing to output the impact vibration signal.

11. The method for extracting impact vibration signals according to claim 10, characterized in that, Calculating the impact envelope based on the in-phase component and the quadrature component includes: The in-phase component and the quadrature component are squared respectively to obtain the first squared component and the second squared component. The first squared component and the second squared component are summed to obtain the third squared component; The impact envelope is obtained by taking the square root of the third square component.

12. The method for extracting impact vibration signals according to claim 10, characterized in that, The fault analysis and processing methods include at least one or more of the following: smoothing processing, threshold detection processing, impact counting processing, impact interval statistical processing, and fault classification processing.

13. The method for extracting impact vibration signals according to claim 12, characterized in that, Fault analysis processing is performed on the impact envelope to output the impact vibration signal, including: Envelope spectrum analysis was performed on the impact envelope to obtain the envelope spectrum of the impact envelope; The fault type is determined based on the spectral peak characteristic parameters corresponding to the fault characteristic frequency in the envelope spectrum, so as to output the impact vibration signal.

14. The method for extracting impact vibration signals according to claim 1, characterized in that, The frequency response curve corresponding to the high-frequency amplification region comes from the inherent frequency response of the vibration sensor itself, or from the combined frequency response formed by the mounting base of the vibration sensor, local resonance of the mechanical structure, and additional sensitive structures.

15. The method for extracting impact vibration signals according to claim 5, characterized in that, The bandwidth of the bandpass filter in the anti-low-frequency flooding front-end processing method covers the preset energy value of the impact ringing, and the bandwidth is selected based on the preset bandwidth while covering the preset energy value of the impact ringing.

16. The method for extracting impact vibration signals according to claim 5, characterized in that, The front-end gain of the automatic gain control processing of the anti-low-frequency flooding front-end processing method is obtained by matching the range of low-frequency vibration amplitude and impact amplitude.

17. The method for extracting impact vibration signals according to claim 1, characterized in that, Also includes: Acquire the rotational speed signal of the rotating equipment; If the rate of change of the rotational speed signal is greater than a preset rate of change, the impact envelope is resampled according to the rotational speed signal; or, the low-pass bandwidth corresponding to the in-phase component and the quadrature component, as well as the diagnostic threshold used when outputting the impact vibration signal, are adjusted according to the operating conditions of the rotating equipment.

18. The method for extracting impact vibration signals according to claim 8, characterized in that, The cutoff frequency of the low-pass filter is greater than the impact repetition frequency and fault characteristic frequency corresponding to the impact vibration signal, and less than the phase-locked center frequency.

19. A device for extracting impact vibration signals, characterized in that, include: The acquisition module is used to acquire the phase-locked center frequency of the high-frequency amplification region within the frequency response curve of the vibration sensor; The generation module is used to generate in-phase reference signals and quadrature reference signals corresponding to the phase-locked center frequency; The demodulation module is used to perform on / off synchronous demodulation of the vibration signal output by the vibration sensor with the in-phase reference signal and the quadrature reference signal, respectively, to obtain the in-phase component and the quadrature component; The output module is used to determine the impact envelope based on the in-phase component and the quadrature component, so as to output the impact vibration signal.

20. A device for extracting impact vibration signals, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the method for extracting impact vibration signals as described in any one of claims 1 to 18.

21. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method for extracting impact vibration signals as described in any one of claims 1 to 18.

Images