A vibration spectrum-based coupler mechanical state monitoring system

Abstract

The application relates to the technical field of laser medical equipment, and discloses a coupler mechanical state monitoring system based on vibration spectrum, which comprises the following steps: a vibration acquisition unit acquires a three-dimensional mechanical vibration original signal; a signal processing unit carries out multi-component decomposition on the three-dimensional mechanical vibration original signal; a cause analysis unit reconstructs a vibration energy transmission path, determines a mechanical structure part of a vibration source, eliminates environmental common mode noise, and distinguishes low-frequency resonance characteristics from high-frequency impact characteristics; a state discrimination unit corrects reference characteristic threshold values according to real-time working condition parameters, and compares and generates a mechanical looseness recognition signal or a mechanical wear recognition signal; and a safety linkage unit carries out hierarchical protection, adjusts laser input power, or triggers a light path cutting mechanism. The application realizes comprehensive determination of vibration energy transmission characteristics, determines mechanical structure parts corresponding to vibration abnormalities, and realizes classification recognition of mechanical looseness and mechanical wear.

Description

Technical Field

[0001] This invention relates to the field of laser medical equipment technology, and more specifically, to a coupler mechanical condition monitoring system based on vibration spectrum. Background Technology

[0002] Couplers, as key connecting components in laser transmission systems, are widely used in high-power laser equipment, fiber optic transmission devices, and precision optical systems. Their main function is to achieve stable connections between mechanical structures and reliable transmission of optical signals or laser energy. During long-term operation, couplers are typically subjected to continuous vibration loads, thermal stress changes, and periodic mechanical impacts, which can easily lead to structural abnormalities such as loosening of connection points and wear at contact interfaces.

[0003] In existing technologies, monitoring methods for the mechanical condition of couplers mainly include manual periodic maintenance, single-point vibration amplitude monitoring, or alarm mechanisms based on simple spectral thresholds. These monitoring methods typically only assess the overall amplitude of the vibration signal or the energy of a specific frequency band, lacking the ability to analyze the spatial propagation characteristics of vibration and making it difficult to accurately distinguish between local structural anomalies and overall resonance phenomena. Furthermore, in complex operating environments, vibrations of the equipment base, external environmental disturbances, or overall machine resonance often superimpose on the acquired signals, forming common-mode interference, leading to false alarms or missed alarms.

[0004] Therefore, it is necessary to design a coupler mechanical condition monitoring system based on vibration spectrum to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a coupler mechanical condition monitoring system based on vibration spectrum, which aims to solve the problem of lack of analytical ability on the propagation characteristics of vibration in the spatial dimension, making it difficult to accurately distinguish between local structural anomalies and overall resonance phenomena, leading to false alarms or missed alarms.

[0006] This invention proposes a coupler mechanical condition monitoring system based on vibration spectrum, comprising: Vibration acquisition unit is used to acquire raw three-dimensional mechanical vibration signals; The signal processing unit is used to perform multi-component decomposition on the original three-dimensional mechanical vibration signal to generate several axial vibration component signals and energy component signals of different frequency bands. The causal analysis unit is used to determine the phase synchronization and amplitude coupling correlation between different axial vibration components, reconstruct the vibration energy transfer path based on the phase synchronization and amplitude coupling correlation, determine the mechanical structure part of the vibration source, eliminate environmental common mode noise according to the vibration modal characteristics corresponding to the mechanical structure part, and distinguish low-frequency resonance characteristics and high-frequency impact characteristics according to the correspondence between the vibration source part and the frequency band, and generate frequency domain energy distribution characteristics. The state discrimination unit is used to correct the reference feature threshold according to the real-time operating parameters; compare the frequency domain energy distribution feature with the corrected reference feature threshold, and generate a mechanical loosening identification signal or a mechanical wear identification signal according to the comparison result; The safety linkage unit is used to perform graded protection based on the mechanical loosening identification signal or mechanical wear identification signal, and to adjust the laser input power or trigger the optical path cut-off mechanism.

[0007] Furthermore, the vibration acquisition unit includes at least three triaxial MEMS accelerometers arranged in a spatially distributed array, which are respectively installed on the main load-bearing surface of the coupler housing, the connecting flange, and the optical device mounting position. The triaxial MEMS accelerometers are connected to each other via a clock synchronization signal line.

[0008] Furthermore, when the signal processing unit performs multi-component decomposition on the original three-dimensional mechanical vibration signal to generate several axial vibration component signals and energy component signals of different frequency bands, it includes: The signal processing unit filters out the DC bias and noise exceeding the effective frequency response range of the original three-dimensional mechanical vibration signal, maps the time domain signal to the frequency domain, projects the frequency domain signal onto an orthogonal coordinate system according to the vibration direction, generates axial vibration component signals representing different axial directions, and energy component signals of different frequency bands divided according to octave bands.

[0009] Furthermore, when the causal analysis unit determines the phase synchronization and amplitude coupling correlation between different axial vibration components, it includes: The causal analysis unit determines the phase synchronization based on the arrival time difference of the corresponding time-domain features in the different axial vibration component signals, and determines the amplitude coupling correlation based on the amplitude ratio of the different axial vibration component signals at the same time.

[0010] Furthermore, when the cause analysis unit determines the mechanical structural location of the vibration source, it includes: The cause analysis unit determines the propagation order of vibration energy between the vibration acquisition units based on the arrival time difference of different axial vibration component signals in the phase synchronization; it determines the energy intensity distribution relationship of vibration energy between the vibration acquisition units based on the amplitude ratio relationship of different axial vibration component signals in the amplitude coupling correlation; it matches the propagation order and energy intensity distribution relationship with the structural layout of the coupler housing to determine the central region of energy divergence, maps the central region to the physical structure of the coupler, and locates the mechanical structural part of the vibration source.

[0011] Furthermore, when the causal analysis unit eliminates environmental common-mode noise based on the vibration modal characteristics corresponding to the mechanical structure part, it includes: The causal analysis unit compares all the original three-dimensional mechanical vibration signals, identifies global vibration components with consistent phase and fixed amplitude ratio, determines the global vibration components as environmental common-mode noise, constructs a cancellation signal with opposite phase and equal amplitude to the global vibration components, and eliminates the cancellation signal from the original three-dimensional mechanical vibration signals.

[0012] Furthermore, when the causal analysis unit distinguishes between low-frequency resonance characteristics and high-frequency impact characteristics, it includes: The causal analysis unit marks the signals of the mechanical structural parts of the vibration source that are connection structures and whose energy is concentrated in a preset low-frequency range as the low-frequency resonance feature; it marks the signals of the mechanical structural parts of the vibration source that are contact interfaces and whose energy is concentrated in a preset high-frequency range as the high-frequency impact feature; and it combines the marked low-frequency resonance feature and high-frequency impact feature to generate the frequency domain energy distribution feature.

[0013] Furthermore, when the state discrimination unit corrects the reference feature threshold based on real-time operating parameters, it includes: The state discrimination unit acquires the coupler's operating temperature, drive current, and operating load rate as the real-time operating parameters; the reference feature threshold includes a low-frequency reference threshold and a high-frequency reference threshold; based on a multi-dimensional mapping relationship, a low-frequency correction coefficient and a high-frequency correction coefficient matching the real-time operating parameters are determined; a weighted operation is performed on the reference low-frequency reference threshold and the low-frequency correction coefficient to generate a corrected low-frequency reference threshold; a weighted operation is performed on the reference high-frequency reference threshold and the high-frequency correction coefficient to generate a corrected high-frequency reference threshold.

[0014] Furthermore, when the state discrimination unit generates a mechanical loosening identification signal or a mechanical wear identification signal based on the comparison result, it includes: The state discrimination unit extracts low-frequency and high-frequency energy values ​​from the frequency domain energy distribution characteristics; compares the low-frequency energy value with the corrected low-frequency reference threshold, and generates the mechanical loosening identification signal when the low-frequency energy value is greater than the corrected low-frequency reference threshold; compares the high-frequency energy value with the corrected high-frequency reference threshold, and generates the mechanical wear identification signal when the high-frequency energy value is greater than the corrected high-frequency reference threshold; and generates both the mechanical loosening identification signal and the mechanical wear identification signal simultaneously when both the low-frequency energy value and the high-frequency energy value are greater than the corrected high-frequency reference threshold.

[0015] Furthermore, when the safety linkage unit performs graded protection based on the mechanical loosening identification signal or the mechanical wear identification signal, it includes: When only the mechanical loosening identification signal is generated, the safety linkage unit performs a first-level protection action, issues an alarm signal, and limits the laser input power to a first preset ratio; when only the mechanical wear identification signal is generated, the safety linkage unit performs a second-level protection action, cuts off the laser drive power supply, and triggers the optical path cut-off mechanism; when both the mechanical loosening identification signal and the mechanical wear identification signal are generated simultaneously, the safety linkage unit performs the second-level protection action.

[0016] Compared with existing technologies, the advantages of this invention are as follows: By performing multi-component decomposition on the original three-dimensional mechanical vibration signal, extracting the axial vibration component and energy component signals of different frequency bands, and analyzing the phase synchronization and amplitude coupling correlation between each axial vibration component, a comprehensive judgment of the vibration energy transfer characteristics is achieved, thereby identifying the mechanical structural parts corresponding to vibration anomalies. Combining the vibration modal characteristics of the mechanical structural parts to eliminate environmental common-mode noise improves the signal-to-noise ratio and the stability of anomaly identification. By distinguishing between low-frequency resonance characteristics and high-frequency impact characteristics, the classification and identification of mechanical loosening and mechanical wear are achieved, and the reference feature threshold is dynamically corrected according to real-time operating parameters, enhancing the adaptability to different operating states. After identifying an abnormal state, adjusting the laser input power or triggering the optical path cutoff mechanism through a graded protection mechanism can promptly reduce equipment risks and improve the operational safety and reliability of the coupler and related optical systems. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1A functional block diagram of a coupler mechanical condition monitoring system based on vibration spectrum provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the laser therapy device coupler provided in an embodiment of the present invention.

[0018] Among them: 1. Coupler; 2. Main load-bearing surface; 3. Connecting flange; 4. Position of optical component mounting base. Detailed Implementation

[0019] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] In some embodiments of this application, see Figure 1 As shown, a coupler mechanical condition monitoring system based on vibration spectrum is proposed, comprising: Vibration acquisition unit is used to acquire raw three-dimensional mechanical vibration signals; The signal processing unit is used to perform multi-component decomposition on the original three-dimensional mechanical vibration signal to generate several axial vibration component signals and energy component signals of different frequency bands. The causal analysis unit is used to determine the phase synchronization and amplitude coupling correlation between different axial vibration components, reconstruct the vibration energy transfer path based on the phase synchronization and amplitude coupling correlation, and determine the mechanical structure part of the vibration source; eliminate environmental common mode noise according to the vibration modal characteristics corresponding to the mechanical structure part; and distinguish between low-frequency resonance characteristics and high-frequency impact characteristics according to the correspondence between the vibration source part and the frequency band, and generate frequency domain energy distribution characteristics. The state discrimination unit is used to correct the reference feature threshold according to the real-time operating parameters; compare the frequency domain energy distribution characteristics with the corrected reference feature threshold, and generate a mechanical loosening identification signal or a mechanical wear identification signal based on the comparison result; The safety linkage unit is used to perform graded protection based on mechanical loosening identification signals or mechanical wear identification signals, and to adjust the laser input power or trigger the optical path cut-off mechanism.

[0021] Specifically, see Figure 2The laser therapy device coupler shown, during actual use, typically connects the laser generating module and the optical transmission components. Under conditions of high-power continuous output, drive motor operation, and equipment handling or vibration, it is susceptible to cyclic stress from thermal expansion and contraction and periodic mechanical impacts. This causes micro-gaps or surface wear to gradually form on the main load-bearing surface 2 of the coupler housing, the connecting flange 3, and the optical component mounting position 4, resulting in abnormal phenomena such as enhanced low-frequency resonance or increased high-frequency impact vibration. If these abnormalities are not identified in time, they may lead to optical path deviation, decreased laser coupling efficiency, unstable output power, and in severe cases, even damage to optical components or equipment shutdown. Existing technologies typically employ single-point vibration amplitude monitoring or fixed-frequency threshold alarm methods, lacking the comprehensive analytical capabilities for vibration propagation characteristics in space and structural correlation. This makes it difficult to accurately distinguish between different fault types such as mechanical loosening and mechanical wear, and is prone to misjudgment or missed judgment under different operating conditions. This technology uses multi-axis vibration signal acquisition and spectral decomposition, combined with phase synchronization and amplitude coupling correlation analysis to determine the abnormal vibration location, and distinguishes between low-frequency resonance and high-frequency impact based on the vibration source location and frequency band characteristics. At the same time, it introduces operating parameters to dynamically correct the discrimination threshold, and links with laser power control and optical path cutting mechanism to achieve refined identification and graded safety protection of the mechanical state of coupler 1.

[0022] Specifically, the vibration acquisition unit includes several triaxial MEMS accelerometers (preferably at least 3) arranged in a spatially distributed array. These sensors are respectively installed on the main load-bearing surface of the coupler, the connecting flange, and the optical device mounting position. They are synchronized with the central acquisition module via a clock synchronization line or a synchronization bus. The sampling rate is preferably 10kHz (which can be adjusted according to the frequency response of the monitored structure).

[0023] Specifically, the signal processing unit first performs hardware anti-aliasing low-pass filtering and DC bias removal on the acquired three-dimensional mechanical vibration raw time-domain signal, and then performs digital preprocessing frame by frame (window function such as Hann window, frame length 2048 points, 50% overlap is preferred), and maps the time-domain signal to the frequency domain (FFT) to obtain the amplitude spectrum and phase spectrum; the signal processing unit further projects the frequency domain signal onto an orthogonal coordinate system according to the vibration direction to generate axial vibration component signals representing different axial directions, and calculates the energy component signal and energy spectral density of each frequency band according to octave / octave band or preset frequency band (for example, the low frequency band 0–500Hz is used for resonance analysis, and the high frequency band 1kHz–10kHz is used for impact / wear analysis, and the specific boundary can be adjusted according to the characteristics of the equipment).

[0024] Specifically, the causal analysis unit calculates the phase difference parameter (e.g., using the phase angle of the cross spectrum or the arrival time delay estimation based on the time-domain cross-correlation function) and the amplitude coupling correlation (e.g., the Pearson correlation coefficient of the amplitude envelope or the coherence function of the cross spectrum) based on the complex spectrum of each axial vibration component, and combines the phase / amplitude indices into a coupling feature matrix. Based on this coupling feature matrix, a simple inversion strategy of time difference-energy weight sorting is adopted (arriving order of arrival time of each sensor feature and combining amplitude proportion to determine the order and intensity distribution of energy propagation), and it is matched with the structural layout model of the coupler (a pre-established structural node-sensor mapping table) to determine the high-probability central region of vibration energy divergence (i.e., locating it to a certain structure). (High-probability indication of location, rather than absolute coordinates); For components with consistent global phase and stable amplitude ratios, they can be identified as environmental common-mode noise and suppressed using adaptive filters or by constructing offset components with opposite phases but adjusted according to actual amplitude ratios (implementation methods include minimum mean square error adaptive noise reduction or common-mode component projection removal based on subspace decomposition) to improve the signal-to-noise ratio of local abnormal features; then, based on the structural type and frequency band energy concentration of the vibration source, low-frequency resonance features and high-frequency impact features are distinguished (e.g., when the energy is mainly concentrated in the preset low-frequency range and the corresponding location is a connecting structure, it is marked as "mechanical loosening" tendency; when the energy is significantly in the high-frequency band and the corresponding location is a contact interface, it is marked as "mechanical wear / impact" tendency).

[0025] Specifically, the state discrimination unit extracts low-frequency and high-frequency energy values ​​from the frequency domain energy distribution characteristics. Simultaneously, the state discrimination unit collects real-time operating parameters related to the system (operating temperature, drive current, operating load rate, etc.), calculates correction coefficients for low and high frequencies using pre-established multi-dimensional mapping relationships or calibration curves, and compares the corrected thresholds with the real-time extracted energy characteristics: when the low-frequency energy value exceeds the corrected low-frequency threshold, a mechanical loosening identification signal is triggered; when the high-frequency energy value exceeds the corrected high-frequency threshold, a mechanical wear identification signal is triggered; if both exceed the limits simultaneously, both types of identification signals are triggered simultaneously.

[0026] Specifically, the safety linkage unit executes a graded protection strategy based on the identification results (preferred execution strategy: Level 1 protection - when only a mechanical loosening identification signal is detected, the system sends a power reduction command to the laser control unit to limit the laser input power to a first preset ratio, such as 50% to 70%; Level 2 protection - when a mechanical wear identification signal is detected or both types of identification signals occur simultaneously, the laser drive power supply is cut off and the optical path mechanical / electrical cut-off mechanism is triggered to interrupt the output).

[0027] Understandably, by synchronously acquiring three-dimensional vibration signals from the coupler and combining this with multi-component decomposition, phase synchronization analysis, and amplitude coupling correlation calculation, the vibration energy transfer path was reconstructed and the vibration source structural components were located. This allowed for the differentiation of different causes, such as mechanical loosening and mechanical wear, at the structural correlation level. By introducing vibration modal characteristics to eliminate environmental common-mode noise and dynamically correcting the characteristic thresholds using real-time operating parameters, the accuracy and anti-interference capabilities of anomaly identification were improved, reducing the risk of false alarms and missed alarms. Linking the identification results with laser power adjustment and optical path interruption mechanisms enabled graded safety protection, allowing for early intervention in the fault stage and preventing consequences such as optical path deviation, power instability, and damage to optical components.

[0028] In some embodiments of this application, the vibration acquisition unit includes a plurality of triaxial MEMS accelerometers arranged in a spatially distributed array, which are respectively installed on the main load-bearing surface of the coupler housing, the connecting flange, and the optical device mounting position, and the triaxial MEMS accelerometers are connected to each other through a clock synchronization signal line.

[0029] Specifically, the vibration acquisition unit includes multiple triaxial MEMS accelerometers arranged in a spatially distributed array, preferably no fewer than three, respectively mounted on the main load-bearing surface of the coupler housing, the connecting flange, and the optical component mounting position. The main load-bearing surface is used to acquire the main vibration response signal of the overall structure under stress. The connecting flange position is used to monitor low-frequency resonance characteristics caused by loosening or gap changes at the connection interface. The optical component mounting position is used to detect high-frequency vibration characteristics caused by micro-displacement or impact wear of the optical components. Each triaxial MEMS accelerometer is used to synchronously acquire vibration acceleration signals in the X, Y, and Z axes to achieve complete acquisition of spatial vibration vector information. Preferably, each sensor is connected to the central control module through an independently deployed clock synchronization signal line or a shared synchronization bus, enabling all sampling channels to sample synchronously under a unified time reference, thereby eliminating time deviations between different acquisition nodes and ensuring the accuracy of phase analysis and coupling calculation. Each sensor can be rigidly fixed with screws or bonded with high-strength structural adhesive to the surface of the coupler structure, and a vibration guiding medium layer is set between the mounting surface and the structural surface to improve the vibration signal transmission efficiency and reduce interface attenuation error.

[0030] Understandably, by deploying multi-point triaxial MEMS accelerometers at key stress and connection points of the coupler and achieving clock-synchronized sampling, complete spatial vibration propagation information can be obtained, ensuring phase consistency and time alignment accuracy between signals at each measurement point, thereby improving the accuracy of vibration source localization and structural cause analysis. The multi-location acquisition method can distinguish between overall common-mode vibration and local abnormal vibration, enhancing the ability to identify early fault characteristics such as mechanical loosening and wear.

[0031] In some embodiments of this application, when the signal processing unit performs multi-component decomposition on the original three-dimensional mechanical vibration signal to generate several axial vibration component signals and energy component signals of different frequency bands, it includes: The signal processing unit filters out the DC bias and noise beyond the effective frequency response range of the original three-dimensional mechanical vibration signal, maps the time domain signal to the frequency domain, projects the frequency domain signal onto the orthogonal coordinate system according to the vibration direction, generates axial vibration component signals representing different axial directions, and energy component signals divided into different frequency bands according to octave bands.

[0032] Specifically, the signal processing unit uses a digital high-pass filtering algorithm to filter out the DC bias component introduced by sensor zero drift or installation error, and simultaneously uses low-pass or band-pass filtering to remove high-frequency noise or external electromagnetic interference signals that exceed the effective frequency response range of the coupler structure, thereby improving the signal-to-noise ratio. Preferably, a finite impulse response (FIR) filter or an infinite impulse response (IIR) filter can be used to achieve frequency shaping.

[0033] Specifically, after time-domain signal purification, the signal processing unit divides the processed time-domain vibration signal into frames according to a preset sampling window length, and maps the time-domain signal to the frequency domain using Fast Fourier Transform (FFT) to obtain the corresponding amplitude and phase spectrum information. Subsequently, the frequency-domain signal is processed by vector decomposition and orthogonal projection according to the three axes acquired by the sensor, and the vibration vector components are projected onto the pre-set orthogonal coordinate system of the structure to generate axial vibration component signals representing the X-axis, Y-axis, and Z-axis directions, thereby realizing the separation and quantization of vibration in different spatial directions.

[0034] Specifically, the signal processing unit divides the frequency domain signal into frequency bands according to a preset octave band division rule (e.g., 1 / 1 octave, 1 / 3 octave, or a characteristic frequency band set according to the coupler's structural characteristics), and calculates the energy value or power spectral density integral value within each frequency band to generate corresponding energy component signals for different frequency bands. These energy component signals are used to characterize the energy distribution characteristics of vibrations in different frequency ranges.

[0035] It is understandable that by applying DC bias and filtering out invalid frequency band noise from the original three-dimensional mechanical vibration signal, and combining frequency domain mapping, orthogonal projection decomposition and octave band energy partitioning, the vibration characteristics can be accurately characterized in both spatial direction and frequency range, thereby improving the signal-to-noise ratio and feature resolution, which helps to accurately identify structural low-frequency resonance and local high-frequency impact components.

[0036] In some embodiments of this application, when the causal analysis unit determines the phase synchronization and amplitude coupling correlation between different axial vibration components, it includes: The causal analysis unit determines phase synchronization based on the arrival time difference of the corresponding time-domain features in different axial vibration component signals, and determines amplitude coupling correlation based on the amplitude ratio of different axial vibration component signals at the same time.

[0037] Specifically, the causal analysis unit first extracts representative time-domain feature points from each axial vibration component signal, including but not limited to peak points, zero-crossing points, envelope extrema points, or energy mutation points, and performs time alignment processing on each axial vibration component signal based on a unified time reference. Subsequently, by calculating the arrival time difference of corresponding time-domain feature points in different axial vibration component signals, the propagation time delay relationship between each axial vibration component is determined; when the arrival time difference is less than a preset synchronization threshold, it is determined to be a high phase synchronization state; when the arrival time difference is greater than the preset threshold, it is determined that there is a propagation delay or phase difference, thus forming a phase synchronization index.

[0038] Specifically, based on the established time synchronization relationship, the causal analysis unit extracts the instantaneous amplitude or envelope amplitude of different axial vibration component signals at the same time or within the same sampling window, and calculates the amplitude ratio between each axial vibration component. Preferably, the coupling degree parameter of vibration energy in different directions can be obtained by constructing an amplitude ratio matrix or calculating the proportional coefficient after amplitude normalization. When the amplitude ratio between different axial vibration components remains relatively stable and exhibits a fixed distribution pattern, it is determined to be a characteristic of the overall structural response; when the amplitude of a certain axial vibration component deviates significantly from the proportional relationship, it is determined to be a characteristic of local structural abnormal response. Through the above-mentioned phase synchronization analysis based on time difference and coupling correlation analysis based on amplitude ratio, multidimensional vibration correlation characteristic parameters are formed.

[0039] Understandably, by utilizing the time difference of arrival in the time domain to determine phase synchronization and by determining the amplitude coupling correlation through the amplitude ratio at the same moment, the correlation between vibrations in different directions can be comprehensively characterized from the time dimension and the energy distribution dimension. This distinguishes between overall structural resonance and local abnormal vibration sources, which not only improves the accuracy of vibration source location analysis but also enhances the ability to identify differences in vibration propagation under complex working conditions.

[0040] In some embodiments of this application, when the causal analysis unit determines the mechanical structural location of the vibration source, it includes: The cause analysis unit determines the propagation sequence of vibration energy between vibration acquisition units based on the arrival time difference of different axial vibration component signals in phase synchronization; it determines the energy intensity distribution relationship between vibration acquisition units based on the amplitude ratio of different axial vibration component signals in amplitude coupling correlation; it matches the propagation sequence and energy intensity distribution relationship with the structural layout of the coupler shell to determine the central region of energy divergence, maps the central region to the physical structure of the coupler, and locates the mechanical structural part of the vibration source.

[0041] Specifically, the cause analysis unit first extracts the characteristic arrival time difference of the corresponding axial vibration component signals at different vibration acquisition units based on the phase synchronization analysis results, and sorts them according to the order of arrival time, thereby determining the propagation path and propagation order of vibration energy among each vibration acquisition unit; preferably, the vibration acquisition unit that first detects the characteristic response can be used as the initial response node, and a vibration propagation time sequence diagram can be constructed by combining the response time difference of adjacent acquisition units.

[0042] Specifically, the causal analysis unit extracts the amplitude ratio or energy proportion parameter of the corresponding axial vibration component signal at different vibration acquisition units based on the amplitude coupling correlation parameter, and determines the intensity distribution relationship of vibration energy among the acquisition units. Preferably, the energy value of each acquisition unit can be normalized to form an energy distribution matrix or weight vector to reflect the relative intensity of vibration response at different locations.

[0043] Specifically, after obtaining the propagation sequence and energy intensity distribution, the causal analysis unit matches this information with a pre-established coupler shell structural layout model. The structural layout model may include the spatial correspondence between key structural components of the coupler (such as the main load-bearing surface, connecting flanges, and optical component mounting positions) and the installation position of the vibration acquisition unit. By overlaying the propagation time sequence diagram and the energy weight distribution diagram onto the structural layout model, the central region of vibration energy dispersion is determined; preferably, the spatial location of the energy concentration region can be calculated using a weighted center algorithm or a minimum propagation path analysis method. Subsequently, the central region is mapped onto the actual physical structure of the coupler, thereby locating the mechanical structural component corresponding to the vibration source and achieving reverse inference from the signal layer to the structural layer.

[0044] Understandably, by comprehensively utilizing the temporal sequence of vibration propagation and the relationship between energy intensity distribution, and matching and locating it with the coupler structure layout, it is possible to accurately identify the mechanical structural parts of the vibration source, avoid the misjudgment problem caused by judging based on a single point amplitude, distinguish abnormal vibrations generated at different structural locations, and improve the positioning accuracy and diagnostic reliability of the vibration source.

[0045] In some embodiments of this application, when the causal analysis unit eliminates environmental common-mode noise based on the vibration modal characteristics corresponding to the mechanical structure parts, it includes: The causal analysis unit compares all the original three-dimensional mechanical vibration signals, identifies global vibration components with consistent phase and fixed amplitude ratio, determines the global vibration components as environmental common-mode noise, constructs a cancellation signal with opposite phase and equal amplitude to the global vibration components, and eliminates the cancellation signal from the original three-dimensional mechanical vibration signals.

[0046] Specifically, the causal analysis unit performs a comprehensive comparative analysis of the three-dimensional mechanical vibration raw signals output by all vibration acquisition units. Under a unified time reference, it extracts the corresponding frequency domain or time domain characteristic components of each acquisition point and calculates the phase difference and amplitude ratio between each acquisition point. When a vibration component is detected to have basically consistent phase and no significant difference in arrival time in all vibration acquisition units, and the amplitude ratio between each acquisition point remains fixed within a preset tolerance range, then the vibration component is determined to be a global vibration component introduced by external environmental vibration sources (such as equipment base vibration, external impact, or overall body resonance), i.e., environmental common-mode noise.

[0047] Specifically, after common-mode component identification, the causal analysis unit constructs a set of cancellation signals with opposite phase and equal amplitude to the global vibration component based on the amplitude and phase parameters of the global vibration component. Preferably, the cancellation signals can be generated by amplitude replication and phase inversion of the original vibration signal, or synthesized in real time using a digital signal processing algorithm. Subsequently, the cancellation signals are superimposed on the original three-dimensional mechanical vibration signal to achieve phase cancellation of the global vibration component, thereby eliminating the influence of environmental common-mode noise in each acquisition channel.

[0048] Understandably, by identifying global vibration components with consistent phase and fixed amplitude ratios and constructing cancellation signals with opposite phases to eliminate them, common-mode noise introduced by external environment or overall body resonance is eliminated, improving the signal-to-noise ratio and feature purity of vibration signals, avoiding the influence of environmental interference on the judgment of local fault characteristics, and improving the accuracy and stability of vibration source location and status identification.

[0049] In some embodiments of this application, the causal analysis unit distinguishes between low-frequency resonance characteristics and high-frequency impact characteristics, including: The causal analysis unit marks signals whose mechanical structural parts of the vibration source are connection structures and whose energy is concentrated in a preset low-frequency range as low-frequency resonance characteristics; it marks signals whose mechanical structural parts of the vibration source are contact interfaces and whose energy is concentrated in a preset high-frequency range as high-frequency impact characteristics; and it combines the marked low-frequency resonance characteristics and high-frequency impact characteristics to generate frequency domain energy distribution characteristics.

[0050] Specifically, the causal analysis unit first determines the type of mechanical structure corresponding to the vibration source based on the aforementioned vibration source location results. It defines the connecting flanges, threaded fasteners, and locating pin connection surfaces in the coupler—areas with constraint and load-bearing functions—as connecting structures. It defines structural interfaces with relative contact, compression, or sliding relationships, such as axial pressing surfaces and optical component fixing plate contact surfaces, as contact interfaces. Subsequently, the causal analysis unit reads the frequency domain energy component signal corresponding to the vibration source and analyzes its energy distribution within a preset frequency range. The preset low-frequency range is preferably the modal frequency range of the coupler's overall structure or the concentrated range of low-order natural frequencies, used to reflect the overall resonance enhancement phenomenon caused by structural loosening or stiffness changes. The preset high-frequency range is preferably the high-frequency response range generated by material contact impact, micro-slip friction, or localized wear. When the mechanical structure corresponding to the vibration source is a connecting structure, and its energy peak or energy integral value is mainly concentrated within the preset low-frequency range, the causal analysis unit marks the signal as a low-frequency resonance characteristic. When the mechanical structure corresponding to the vibration source is a contact interface, and its energy is mainly concentrated within the preset high-frequency range, the signal is marked as a high-frequency impact characteristic.

[0051] Specifically, after completing the above classification and labeling, the causal analysis unit combines the low-frequency resonance characteristics and high-frequency impact characteristics, and constructs a unified frequency domain energy distribution feature vector or feature matrix according to preset weights or proportion parameters, which includes at least low-frequency energy indicators, high-frequency energy indicators and their relative proportion parameters.

[0052] Understandably, by associating the structural type of the vibration source with the frequency band of energy concentration, it is possible to distinguish between low-frequency resonance and high-frequency impact signals in both structural properties and frequency characteristics, thereby improving the ability to identify different types of abnormalities such as mechanical loosening and contact wear.

[0053] In some embodiments of this application, when the state discrimination unit corrects the reference feature threshold based on real-time operating condition parameters, it includes: The state discrimination unit acquires the coupler's operating temperature, drive current, and operating load rate as real-time operating parameters; the reference characteristic thresholds include low-frequency reference thresholds and high-frequency reference thresholds; based on multi-dimensional mapping relationships, it determines low-frequency correction coefficients and high-frequency correction coefficients that match the real-time operating parameters; it performs a weighted operation on the reference low-frequency reference thresholds and low-frequency correction coefficients to generate corrected low-frequency reference thresholds; and it performs a weighted operation on the reference high-frequency reference thresholds and high-frequency correction coefficients to generate corrected high-frequency reference thresholds.

[0054] Specifically, the state discrimination unit acquires the coupler's operating temperature, drive current, and operating load rate in real time as real-time operating condition parameters. The operating temperature is used to reflect the influence of thermal expansion and material stiffness changes on vibration characteristics, the drive current is used to reflect changes in drive intensity and excitation level, and the operating load rate is used to reflect the stress state of the mechanical structure and energy input level.

[0055] Specifically, the reference characteristic thresholds include a baseline low-frequency reference threshold and a baseline high-frequency reference threshold, which are used to determine anomalies in corresponding low-frequency resonance characteristics and high-frequency impact characteristics, respectively. During the system's factory delivery or calibration phase, a multi-dimensional mapping relationship model is established based on normal operating data collected under different temperature, current, and load conditions. This model can be constructed using a lookup table method, a function fitting method, or a multivariate linear / nonlinear regression model. The state discrimination unit matches the current real-time operating parameters within the multi-dimensional mapping relationship model to determine the low-frequency correction coefficient and high-frequency correction coefficient corresponding to the current operating condition.

[0056] Specifically, the state discrimination unit generates a corrected low-frequency reference threshold by weighting a baseline low-frequency reference threshold with a low-frequency correction coefficient; and generates a corrected high-frequency reference threshold by weighting a baseline high-frequency reference threshold with a high-frequency correction coefficient. The weighting operation preferably employs a proportional amplification or proportional correction method, such as multiplying the threshold by the correction coefficient or adjusting based on the deviation, allowing the threshold to dynamically adjust with changes in operating conditions. The corrected low-frequency and high-frequency reference thresholds are then compared with the real-time extracted frequency domain energy distribution features to accurately identify mechanical loosening or wear conditions.

[0057] Understandably, by introducing real-time operating parameters such as operating temperature, drive current, and operating load rate to dynamically correct the reference threshold, the interference of natural fluctuations in vibration amplitude caused by changes in operating conditions on the judgment results is eliminated, false alarms are avoided under high load or high temperature conditions, and the sensitivity of abnormal identification under low load conditions is improved.

[0058] In some embodiments of this application, when the state discrimination unit generates a mechanical loosening identification signal or a mechanical wear identification signal based on the comparison result, it includes: The state discrimination unit extracts low-frequency and high-frequency energy values ​​from the frequency domain energy distribution characteristics; it compares the low-frequency energy value with a corrected low-frequency reference threshold, and generates a mechanical loosening identification signal when the low-frequency energy value is greater than the corrected low-frequency reference threshold; it compares the high-frequency energy value with a corrected high-frequency reference threshold, and generates a mechanical wear identification signal when the high-frequency energy value is greater than the corrected high-frequency reference threshold; and it generates both a mechanical loosening identification signal and a mechanical wear identification signal when both the low-frequency energy value and the high-frequency energy value are greater than the corrected high-frequency reference threshold.

[0059] Specifically, the state discrimination unit first extracts low-frequency and high-frequency energy values ​​from the frequency domain energy distribution characteristics output by the causal analysis unit. The low-frequency energy value is the integral or weighted average of the energy across all frequency bands within a preset low-frequency range, while the high-frequency energy value is the integral or weighted average of the energy across all frequency bands within a preset high-frequency range. Subsequently, the state discrimination unit compares the low-frequency energy value with a corrected low-frequency reference threshold. When the low-frequency energy value exceeds the corrected low-frequency reference threshold, it is determined to be a low-frequency abnormal enhancement state, generating a mechanical loosening identification signal. This identification signal indicates structural resonance enhancement caused by a decrease in the stiffness of the connection structure or abnormal fastening conditions. Simultaneously, the state discrimination unit compares the high-frequency energy value with a corrected high-frequency reference threshold. When the high-frequency energy value exceeds the corrected high-frequency reference threshold, it is determined to be a high-frequency impact enhancement state, generating a mechanical wear identification signal. This identification signal indicates abnormal impact, micro-slip friction, or surface wear at the contact interface. Furthermore, when the low-frequency energy value is greater than the corrected low-frequency reference threshold and the high-frequency energy value is greater than the corrected high-frequency reference threshold, the state discrimination unit determines that the coupler may have a combined abnormal state of loose connection and contact wear at the same time, and generates mechanical looseness identification signal and mechanical wear identification signal at the same time.

[0060] Understandably, by independently comparing the low-frequency energy value and the high-frequency energy value, and outputting a composite abnormal signal when both thresholds exceed the limit, the system distinguishes between mechanical loosening, mechanical wear, and the complex state of both coexisting, thereby improving the pertinence and accuracy of abnormal classification.

[0061] In some embodiments of this application, when the safety linkage unit performs graded protection based on mechanical loosening identification signals or mechanical wear identification signals, it includes: When only a mechanical loosening identification signal is generated, the safety linkage unit performs a first-level protection action, issues an alarm signal, and limits the laser input power to a first preset ratio; when only a mechanical wear identification signal is generated, the safety linkage unit performs a second-level protection action, cuts off the laser drive power supply, and triggers the optical path cut-off mechanism; when both mechanical loosening identification signals and mechanical wear identification signals are generated simultaneously, the safety linkage unit performs a second-level protection action.

[0062] Specifically, when the status judgment unit only generates a mechanical loosening identification signal, the safety linkage unit determines that the coupler is in an early structural abnormality state, and at this time, it executes the first-level protection action. The first-level protection action includes: sending an audible and visual alarm signal or an interface prompt signal to the equipment control system, and simultaneously sending a power limiting command to the laser control module to limit the laser input power to a first preset ratio (e.g., 50% to 70% of the rated power, the specific ratio can be set according to the equipment design), so as to reduce the structural load and vibration excitation intensity, thereby preventing the loosening from further expanding, and providing maintenance personnel with early warning time.

[0063] Specifically, when the status determination unit only generates a mechanical wear identification signal, the safety linkage unit determines that the coupler has a relatively serious risk of contact impact or wear. Such an anomaly may cause optical element misalignment or local overheating, thus executing a secondary protection action. The secondary protection action includes: immediately cutting off the laser drive power supply, stopping the laser output, and simultaneously triggering an optical path cutoff mechanism, such as controlling the closing of the electric shutter, mechanical baffle, or safety gate, to physically isolate the laser channel and prevent the laser from continuing to transmit.

[0064] Specifically, when the status judgment unit generates both a mechanical loosening identification signal and a mechanical wear identification signal, the safety linkage unit determines that the coupler is in a combined abnormal or serious failure risk state. In order to avoid equipment damage or safety accidents, it prioritizes the execution of secondary protection actions, namely cutting off the drive power supply and triggering the optical path cut-off mechanism.

[0065] Understandably, by setting up a graded protection mechanism, mechanical loosening and mechanical wear are associated with different safety response levels. This allows for power limiting measures to control risks in the early stages of anomalies, and timely cut-off of laser output in cases of severe or combined anomalies, preventing the escalation of the fault and damage to optical components. This graded linkage method takes into account both the continuous operation requirements of the equipment and ensures operational safety.

[0066] In summary, by performing multi-component decomposition on the original three-dimensional mechanical vibration signal, extracting the axial vibration component and energy components in different frequency bands, and analyzing the phase synchronization and amplitude coupling correlation between each axial vibration component, a comprehensive judgment of the vibration energy transfer characteristics was achieved, thereby identifying the mechanical structural parts corresponding to vibration anomalies. Combining the vibration modal characteristics of the mechanical structural parts to eliminate environmental common-mode noise improves the signal-to-noise ratio and the stability of anomaly identification. By distinguishing between low-frequency resonance characteristics and high-frequency impact characteristics, mechanical loosening and mechanical wear can be classified and identified. Furthermore, dynamic correction of the reference feature threshold based on real-time operating parameters enhances adaptability to different operating states. After identifying an abnormal state, adjusting the laser input power or triggering the optical path cutoff mechanism through a graded protection mechanism can promptly reduce equipment risks and improve the operational safety and reliability of the coupler and related optical systems.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A coupler mechanical condition monitoring system based on vibration spectrum, characterized in that, include: Vibration acquisition unit is used to acquire raw three-dimensional mechanical vibration signals; The signal processing unit is used to perform multi-component decomposition on the original three-dimensional mechanical vibration signal to generate several axial vibration component signals and energy component signals of different frequency bands. The causal analysis unit is used to determine the phase synchronization and amplitude coupling correlation between different axial vibration components, reconstruct the vibration energy transfer path based on the phase synchronization and amplitude coupling correlation, determine the mechanical structure part of the vibration source, eliminate environmental common mode noise according to the vibration modal characteristics corresponding to the mechanical structure part, and distinguish low-frequency resonance characteristics and high-frequency impact characteristics according to the correspondence between the vibration source part and the frequency band, and generate frequency domain energy distribution characteristics. The state discrimination unit is used to correct the reference feature threshold according to the real-time operating parameters; compare the frequency domain energy distribution feature with the corrected reference feature threshold, and generate a mechanical loosening identification signal or a mechanical wear identification signal according to the comparison result; The safety linkage unit is used to perform graded protection based on the mechanical loosening identification signal or the mechanical wear identification signal, and to adjust the laser input power or trigger the optical path cutoff mechanism. The vibration acquisition unit includes several triaxial MEMS accelerometers arranged in a spatially distributed array, which are installed at least on the main load-bearing surface of the coupler housing, the connecting flange, and the optical device mounting position, and the triaxial MEMS accelerometers are connected to each other through a clock synchronization signal line. When the causal analysis unit distinguishes between low-frequency resonance characteristics and high-frequency impact characteristics, it includes: The causal analysis unit marks the signals of the mechanical structure of the vibration source that are connection structures and whose energy is concentrated in a preset low-frequency range as the low-frequency resonance feature; it marks the signals of the mechanical structure of the vibration source that are contact interfaces and whose energy is concentrated in a preset high-frequency range as the high-frequency impact feature; and it combines the marked low-frequency resonance feature and high-frequency impact feature to generate the frequency domain energy distribution feature. When the safety linkage unit performs graded protection based on the mechanical loosening identification signal or the mechanical wear identification signal, it includes: When only the mechanical loosening identification signal is generated, the safety linkage unit performs a first-level protection action, issues an alarm signal, and limits the laser input power to a first preset ratio; when only the mechanical wear identification signal is generated, the safety linkage unit performs a second-level protection action, cuts off the laser drive power supply, and triggers the optical path cut-off mechanism; when both the mechanical loosening identification signal and the mechanical wear identification signal are generated simultaneously, the safety linkage unit performs the second-level protection action.

2. The coupler mechanical condition monitoring system based on vibration spectrum according to claim 1, characterized in that, When the signal processing unit performs multi-component decomposition on the original three-dimensional mechanical vibration signal to generate several axial vibration component signals and energy component signals of different frequency bands, it includes: The signal processing unit filters out the DC bias and noise exceeding the effective frequency response range of the original three-dimensional mechanical vibration signal, maps the time domain signal to the frequency domain, projects the frequency domain signal onto an orthogonal coordinate system according to the vibration direction, generates axial vibration component signals representing different axial directions, and energy component signals of different frequency bands divided according to octave bands.

3. The coupler mechanical condition monitoring system based on vibration spectrum according to claim 1, characterized in that, When the causal analysis unit determines the phase synchronization and amplitude coupling correlation between different axial vibration components, it includes: The causal analysis unit determines the phase synchronization based on the arrival time difference of the corresponding time-domain features in the different axial vibration component signals, and determines the amplitude coupling correlation based on the amplitude ratio of the different axial vibration component signals at the same time.

4. The coupler mechanical condition monitoring system based on vibration spectrum according to claim 3, characterized in that, When the causal analysis unit determines the mechanical structural location of the vibration source, it includes: The cause analysis unit determines the propagation order of vibration energy between the vibration acquisition units based on the arrival time difference of different axial vibration component signals in the phase synchronization; it determines the energy intensity distribution relationship of vibration energy between the vibration acquisition units based on the amplitude ratio relationship of different axial vibration component signals in the amplitude coupling correlation; it matches the propagation order and energy intensity distribution relationship with the structural layout of the coupler housing to determine the central region of energy divergence, maps the central region to the physical structure of the coupler, and locates the mechanical structural part of the vibration source.

5. The coupler mechanical condition monitoring system based on vibration spectrum according to claim 4, characterized in that, When the causal analysis unit eliminates environmental common-mode noise based on the vibration modal characteristics corresponding to the mechanical structure parts, it includes: The causal analysis unit compares all the original three-dimensional mechanical vibration signals, identifies global vibration components with consistent phase and fixed amplitude ratio, determines the global vibration components as environmental common-mode noise, constructs a cancellation signal with opposite phase and equal amplitude to the global vibration components, and eliminates the environmental common-mode noise from the original three-dimensional mechanical vibration signals.

6. The coupler mechanical condition monitoring system based on vibration spectrum according to claim 1, characterized in that, When the state discrimination unit corrects the reference feature threshold based on real-time operating parameters, it includes: The state discrimination unit acquires the coupler's operating temperature, drive current, and operating load rate as the real-time operating parameters; the reference feature threshold includes a low-frequency reference threshold and a high-frequency reference threshold; based on a multi-dimensional mapping relationship, a low-frequency correction coefficient and a high-frequency correction coefficient matching the real-time operating parameters are determined; a weighted operation is performed on the reference low-frequency reference threshold and the low-frequency correction coefficient to generate a corrected low-frequency reference threshold; a weighted operation is performed on the reference high-frequency reference threshold and the high-frequency correction coefficient to generate a corrected high-frequency reference threshold.

7. The coupler mechanical condition monitoring system based on vibration spectrum according to claim 6, characterized in that, When the state discrimination unit generates a mechanical loosening identification signal or a mechanical wear identification signal based on the comparison result, it includes: The state discrimination unit extracts low-frequency and high-frequency energy values ​​from the frequency domain energy distribution characteristics; compares the low-frequency energy value with the corrected low-frequency reference threshold, and generates the mechanical loosening identification signal when the low-frequency energy value is greater than the corrected low-frequency reference threshold; compares the high-frequency energy value with the corrected high-frequency reference threshold, and generates the mechanical wear identification signal when the high-frequency energy value is greater than the corrected high-frequency reference threshold; and generates both the mechanical loosening identification signal and the mechanical wear identification signal simultaneously when both the low-frequency energy value and the high-frequency energy value are greater than the corrected high-frequency reference threshold.

Images