Order spectrum generation method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides an order spectrum generation method and device, electronic equipment and a storage medium, and relates to the technical field of industrial equipment monitoring. The method comprises the following steps: performing short-time Fourier transform on an acquired original vibration signal of the rotating machine to obtain a time-frequency spectrogram of the rotating machine, extracting a plurality of continuous candidate frequency ridge lines from the time-frequency spectrogram based on an image processing technology, screening one of the candidate frequency ridge lines as a reference harmonic ridge line, and according to the reference harmonic ridge line, obtaining a harmonic signal of the rotating machine; calculating to obtain an instantaneous estimation phase of the main shaft, demodulating the original vibration signal based on the instantaneous estimation phase to obtain a demodulated vibration signal, performing equal-angle resampling on the original vibration signal according to the demodulated vibration signal to obtain an angle domain vibration signal with an equally-spaced angle domain, and performing Fourier transform on the angle domain vibration signal to obtain an angle domain vibration signal with an equally-spaced angle domain; and generating an order spectrum. Therefore, the state of the rotating machine can be accurately monitored by using the order spectrum in a speed change scene of the rotating machine.

Description

Technical Field

[0001] This invention relates to the field of industrial equipment monitoring technology, and more specifically, to a method, apparatus, electronic device, and storage medium for generating order spectra. Background Technology

[0002] In condition monitoring of rotating machinery (such as wind turbines and rolling mills), order analysis is a key technology for fault diagnosis. When the equipment speed is constant, traditional Fourier transform can effectively extract characteristic frequencies (such as gear meshing frequencies and bearing fault frequencies) from vibration signals, achieving accurate diagnosis. However, in actual operation, the speed of many devices changes continuously, causing modulation of the frequency components of the vibration signal. Traditional spectrum analysis suffers from energy diffusion and spectrum ambiguity, making it difficult to identify key fault characteristics.

[0003] To address this, order tracking technology was proposed. Its core principle is to convert the time-domain signal into a stationary angle-domain signal through angle resampling, thereby eliminating the influence of speed variations. However, existing methods heavily rely on hardware speed sensors to provide the key phase signal. These sensors suffer from problems such as susceptibility to damage, installation difficulties, high costs, and signal distortion in industrial settings, limiting their reliability. Summary of the Invention

[0004] In view of this, the object of the embodiments of the present invention is to provide a method, apparatus, electronic device and storage medium for generating order spectra to at least partially improve the above-mentioned problems.

[0005] To achieve the above objectives, the technical solutions adopted in the embodiments of the present invention are as follows: In a first aspect, embodiments of the present invention provide a method for generating order spectra, comprising: A short-time Fourier transform is performed on the original vibration signal of the rotating machinery to obtain the time-frequency spectrum of the rotating machinery. Based on image processing technology, multiple continuous candidate frequency ridges are extracted from the time-spectrum graph; wherein each frequency ridge represents the instantaneous frequency trajectory of a vibrational harmonic component. From the candidate frequency ridges, one is selected as the reference harmonic ridge. Based on the reference harmonic ridge, the instantaneous estimated phase of the main shaft of the rotating machinery is calculated, and the original vibration signal is demodulated based on the instantaneous estimated phase to obtain the demodulated vibration signal; Based on the demodulated vibration signal, the original vibration signal is resampled at equal angles to obtain an angle domain vibration signal with equal intervals in the angle domain; The angular domain vibration signal is subjected to Fourier transform to generate an order spectrum for use in the condition monitoring and fault diagnosis of the rotating machinery.

[0006] Optionally, the step of extracting multiple consecutive candidate frequency ridges from the time-spectrum graph based on image processing technology includes: The time-spectrum is binarized to obtain a binarized image; Connectivity analysis is performed on the binarized image to obtain multiple initial frequency line segments composed of pixels; For each initial frequency line segment, its geometric features are extracted; the geometric features include the starting point coordinates and the ending point coordinates. Based on the endpoint coordinates of each initial frequency line segment, the starting coordinates of other initial frequency line segments that meet the preset conditions are found in the preset neighborhood window downstream, and potential connection relationships are established between the corresponding line segments. A directed graph is constructed based on the potential connections described above. The directed graph is then traversed using a graph traversal algorithm to merge the interconnected initial frequency segments into complete candidate frequency ridges.

[0007] Optionally, the step of using the endpoint coordinates of each initial frequency line segment as a reference, searching for the starting coordinates of other initial frequency line segments that meet preset conditions within a preset neighborhood window downstream of them, and establishing potential connection relationships between the corresponding line segments, includes: Each initial frequency line segment is taken as a source line segment, and the end coordinates of the source line segment are taken as the search origin. A neighborhood window is set downstream of it; the neighborhood window includes a preset height and a preset width. Scan all other initial frequency segments and select other initial frequency segments whose starting coordinates are within the neighborhood window as candidate connection segments; Based on preset connection criteria, a target connection segment is determined from the candidate connection segments; A directed potential connection relationship is established between the source line segment and the target connecting line segment.

[0008] Optionally, the geometric features further include segment length, head direction vector, tail direction vector, and tail direction stability; the step of determining the target connecting segment from the candidate connecting segments based on a preset connection criterion includes: For each of the candidate connecting segments, calculate the Euclidean distance between the endpoint coordinates of the source segment and the starting coordinates of each candidate connecting segment; Determine if there are any neighboring candidate line segments whose row difference between the endpoint coordinates and the starting coordinates is less than or equal to a preset row difference threshold and whose column difference is less than or equal to a preset column difference threshold; If it exists, the nearest candidate line segment with the smallest Euclidean distance will be taken as the target connecting line segment; If it does not exist, calculate the first angle between the tail direction vector of the source line segment and the head direction vector of each candidate connecting line segment; Calculate the second angle between the tail direction vector of the source line segment and the connecting vector; the connecting vector is the connecting direction vector from the end point of the source line segment to the starting point of the candidate connecting line segment; If the tail direction stability of the source line segment is stable and the line segment length is greater than or equal to the preset length, then determine whether there are candidate line segments with a first included angle less than the first preset angle and a second included angle less than the second preset angle. If it exists, the candidate line segment with the smallest first included angle is taken as the target connecting line segment.

[0009] Optionally, selecting one from the candidate frequency ridges as a reference harmonic ridge includes: Based on the number of teeth and rotational speed range of the rotating machinery, the theoretical range of meshing frequency is calculated; From the candidate frequency ridges, select the first candidate ridge with a frequency within the theoretical value range; Select a predetermined number of second candidate ridges with high energy from each of the first candidate ridges; For each of the second candidate ridges, calculate the theoretical turning frequency of the second candidate ridge and determine whether there is a ridge with a frequency equal to the theoretical turning frequency among the candidate frequency ridges. If it exists, the second candidate ridge line is used as the reference harmonic ridge line; If not, determine whether there is a ridge among the candidate frequency ridges whose frequency is an integer multiple of the frequency of the second candidate ridge; If it exists, the second candidate ridge line is used as the reference harmonic ridge line; If none exists, the smoothest and most continuous ridge among the second candidate ridges is taken as the reference harmonic ridge.

[0010] Optionally, the step of calculating the instantaneous estimated phase of the principal axis based on the reference harmonic ridge, and demodulating the original vibration signal based on the instantaneous estimated phase to obtain a demodulated vibration signal, includes: Integrating the reference harmonic ridge line yields the instantaneous phase of the reference harmonic ridge line; The instantaneous estimated phase of the principal axis is calculated based on the instantaneous phase and the order of the reference harmonic ridge. Based on the instantaneously estimated phase, a phase demodulation operator is constructed; The demodulated vibration signal is obtained by multiplying the phase demodulation operator with the original vibration signal.

[0011] Optionally, the step of resampling the original vibration signal at equal angles based on the demodulated vibration signal to obtain equally spaced angle-domain vibration signals includes: The demodulated vibration signal is subjected to low-pass filtering to remove high-frequency components, thereby obtaining the baseband signal; Perform an inverse Fourier transform on the baseband signal to obtain a complex signal; The phase information of the complex signal is extracted to obtain the instantaneous precise phase of the main shaft; Construct the target phase sequence based on the preset number of sampling points per revolution; Based on the instantaneous precise phase, calculate the acquisition time series corresponding to the target phase sequence; Using an interpolation algorithm, the amplitude of each time point in the acquired time series in the original vibration signal is determined, and an angular domain vibration signal with equal intervals in the angular domain is obtained.

[0012] Secondly, embodiments of the present invention provide an order spectrum generation apparatus, comprising: The time-spectrum acquisition unit is used to perform a short-time Fourier transform on the acquired original vibration signal of the rotating machinery to obtain the time-spectrum of the rotating machinery. The ridge extraction unit is used to extract multiple consecutive candidate frequency ridges from the time-spectrum graph based on image processing technology; wherein each frequency ridge represents the instantaneous frequency trajectory of a vibrational harmonic component. The reference harmonic ridge line determination unit is used to select one as a reference harmonic ridge line from each of the candidate frequency ridge lines. The demodulated vibration signal calculation unit is used to calculate the instantaneous estimated phase of the main shaft based on the reference harmonic ridge, and demodulate the original vibration signal based on the instantaneous estimated phase to obtain the demodulated vibration signal; An angle domain vibration signal calculation unit is used to perform equal-angle resampling on the original vibration signal based on the demodulated vibration signal to obtain an angle domain vibration signal with equal intervals in the angle domain. The order spectrum generation unit is used to perform Fourier transform on the angular domain vibration signal to generate an order spectrum for use in the condition monitoring and fault diagnosis of the rotating machinery.

[0013] Thirdly, embodiments of the present invention provide an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the method described in any of the above-mentioned embodiments.

[0014] Fourthly, embodiments of the present invention provide a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in any of the preceding claims.

[0015] The present invention provides an order spectrum generation method, apparatus, electronic device and storage medium that, by relying solely on the original vibration signal of rotating machinery and combining time-frequency analysis, image processing and generalized demodulation technology, achieves high-precision order tracking without the need for hardware speed sensors, thereby generating a clear and stable order spectrum for equipment condition monitoring and fault diagnosis.

[0016] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A schematic structural block diagram of an electronic device provided in an embodiment of the present invention; Figure 2 A flowchart illustrating a method for generating an order spectrum according to an embodiment of the present invention; Figure 3 This is another flowchart illustrating a method for generating order spectra according to an embodiment of the present invention; Figure 4 A flowchart illustrating step S230 provided in an embodiment of the present invention; Figure 5 A flowchart illustrating step S240 provided in an embodiment of the present invention; Figure 6 A flowchart illustrating step S250 provided in an embodiment of the present invention; Figure 7 This is a schematic structural block diagram of an order spectrum generation device provided in an embodiment of the present invention.

[0019] Icons: 100 - Electronic device; 101 - Memory; 102 - Communication interface; 103 - Processor; 104 - Communication bus; 300 - Order spectrum generation device; 310 - Time-spectrum acquisition unit; 320 - Ridge extraction unit; 330 - Reference harmonic ridge determination unit; 340 - Demodulated vibration signal calculation unit; 350 - Angular domain vibration signal calculation unit; 360 - Order spectrum generation unit. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] It should be noted that, in this document, 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 said element.

[0024] Rotating machinery is a core piece of equipment in industrial production. Monitoring its condition and diagnosing its faults are crucial for ensuring production safety and improving equipment reliability. Spectrum analysis is the foundation of fault diagnosis. At constant rotational speed, Fourier transform analysis of vibration signals can clearly identify characteristics such as gear meshing frequencies and bearing failure frequencies, thereby determining the equipment's health status.

[0025] However, in actual operating conditions, the rotational speed of many rotating machines is constantly changing. This causes the characteristic frequency of the vibration signal to change accordingly, resulting in frequency modulation. In this case, the traditional Fourier transform will disperse the energy across a continuous frequency band, causing spectral ambiguity, making it difficult to identify key fault characteristics, and leading to diagnostic failure.

[0026] To address the issue of variable speed, order tracking technology was developed. Its core idea is to convert non-stationary signals acquired in the time domain into stationary signals in the angle domain through resampling based on angle, thereby eliminating the impact of speed variations on spectral analysis and allowing traditional methods such as Fourier transform to continue to be used.

[0027] The key and prerequisite for achieving order tracking is to obtain accurate instantaneous rotational speed or instantaneous phase information. However, existing methods rely heavily on hardware speed sensors to provide key phase signals. Such sensors are prone to damage, difficult to install, costly, and suffer from signal distortion in industrial settings, which limits the reliability of the application.

[0028] Based on the above, embodiments of the present invention provide a method, apparatus, electronic device, and storage medium for generating order spectra. By relying solely on the original vibration signals of rotating machinery and combining time-frequency analysis, image processing, and generalized demodulation techniques, high-precision order tracking without the need for hardware speed sensors is achieved, thereby generating clear and stable order spectra for equipment condition monitoring and fault diagnosis.

[0029] To implement the process steps and functions of the various examples of this invention, please refer to [link / reference]. Figure 1 , Figure 1 This is a schematic structural block diagram of an electronic device provided in an embodiment of the present invention. The electronic device 100 includes a memory 101 and a processor 103, which are electrically connected directly or indirectly to each other to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses 104 or signal lines. The memory 101 can be used to store software programs and modules, and the processor 103 executes the software programs and modules stored in the memory 101, thereby performing various functional applications and data processing.

[0030] Electronic device 100 can be, but is not limited to, a personal computer (PC), a server, a distributed computer, etc. It is understood that electronic device 100 is not limited to a physical server, but can also be a virtual machine on a physical server, a virtual machine built on a cloud platform, or any other computer that can provide the same functionality as the server or virtual machine. The operating system of electronic device 100 can be, but is not limited to, Windows, Linux, etc.

[0031] The memory 101 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0032] The communication connection between the electronic device 100 and external devices is achieved through at least one communication interface 102 (which can be wired or wireless).

[0033] Processor 103 may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of this embodiment can be completed by integrated logic circuits in the hardware of processor 103 or by instructions in software form. Processor 103 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0034] Understandable. Figure 1 The structure shown is for illustrative purposes only; the electronic device 100 may also include components that are more advanced than those shown. Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown. Figure 1 The components shown can be implemented using hardware, software, or a combination thereof.

[0035] The following is an exemplary description of the order spectrum generation method provided by this invention. See also... Figure 2 The subject executing this method can be one of the above. Figure 1 The electronic device 100 shown, the method includes as follows Figure 2 The following steps are described: S210: Perform a short-time Fourier transform on the acquired raw vibration signal of the rotating machinery to obtain the time spectrum of the rotating machinery.

[0036] S220: Based on image processing technology, multiple continuous candidate frequency ridges are extracted from the time-spectrum graph; each frequency ridge represents the instantaneous frequency trajectory of a vibrational harmonic component.

[0037] S230: Select one from the candidate frequency ridges as the reference harmonic ridge.

[0038] S240: Based on the reference harmonic ridge, the instantaneous estimated phase of the main shaft of the rotating machinery is calculated, and the original vibration signal is demodulated based on the instantaneous estimated phase to obtain the demodulated vibration signal.

[0039] S250: Based on the demodulated vibration signal, the original vibration signal is resampled at equal angles to obtain an angle-domain vibration signal with equal intervals in the angle domain.

[0040] S260: Performs Fourier transform on the angular domain vibration signal to generate an order spectrum for use in condition monitoring and fault diagnosis of rotating machinery.

[0041] The original vibration signal x(t) is acquired by an accelerometer installed on a key part of rotating machinery (such as the gearbox housing). A short-time Fourier transform (STFT) is performed on this signal to obtain its two-dimensional time-frequency representation, which is the time-frequency spectrum. This spectrum reflects the distribution of vibration energy over time and frequency, providing a visual basis for subsequent identification of key vibration components.

[0042] The time-spectrum graph is treated as a grayscale image. Image processing techniques are used to detect and extract several energy-concentrated, time-extending curves, termed "frequency ridges." Each ridge corresponds to a harmonic component caused by a major vibration source (such as spindle rotation frequency, gear meshing, etc.), and the trajectory of its frequency coordinate change over time represents the instantaneous frequency variation of that harmonic. Through this step, a set of candidate frequency ridges reflecting the dynamic characteristics of the equipment is obtained.

[0043] Among all extracted candidate frequency ridges, the ridge that best represents the rotational state of the equipment spindle is selected as the reference harmonic ridge. This ridge typically corresponds to the main excitation source that has a clear multiple relationship with the spindle speed, such as the meshing frequency of a gearbox or the pole pass frequency of a motor. The screening process can comprehensively consider the energy level, continuity, physical meaning, and prior knowledge of the system of the ridge to ensure that the selected ridge has high reliability and representativeness.

[0044] Based on the instantaneous frequency information represented by the selected reference harmonic ridge, the cumulative phase of the harmonic component can be derived through integration. Combined with its order relative to the spindle rotation frequency, the instantaneous estimated phase of the spindle can be further calculated. A demodulation function is then constructed based on this phase to perform phase compensation processing on the original vibration signal, thereby eliminating the frequency modulation effect caused by rotational speed fluctuations and obtaining a relatively stable demodulated vibration signal.

[0045] By utilizing the phase information contained in the demodulated vibration signal, a series of target sampling times with equal angular intervals are determined. Based on this, an interpolation method can be used to extract the amplitude at the corresponding time points from the original vibration signal, thereby generating a discrete signal sequence uniformly distributed across the rotation angle, i.e., an angular domain vibration signal. This signal achieves the conversion from the time domain to the angular domain, eliminating the sampling unevenness problem caused by changes in rotational speed.

[0046] Finally, a standard Fourier transform is performed on the obtained angular domain vibration signal to obtain its frequency domain expression. Here, the horizontal axis represents the "order," indicating the number of vibrations occurring within a complete rotation cycle. The peak positions and amplitudes in the order spectrum directly correspond to the fault characteristics of different mechanical components and can be used to identify typical faults such as gear wear, broken teeth, and bearing damage, achieving efficient and accurate condition monitoring and fault diagnosis.

[0047] This method can perform high-quality order analysis without relying on external speed sensors, significantly improving its applicability and reliability in complex industrial environments, and has important engineering application value.

[0048] In an alternative implementation, in step S220 above, multiple consecutive candidate frequency ridges are extracted from the time-spectrum graph based on image processing techniques, see [link to relevant documentation]. Figure 3 This can be further included in the following sub-steps to achieve effective reconstruction of broken, incomplete frequency trajectories: S221: Binarize the time-spectrum graph to obtain a binary image.

[0049] To highlight significant vibrational components and suppress background noise interference, the time-spectrum image obtained through short-time Fourier transform is first subjected to amplitude thresholding. Optionally, an energy threshold is set (which can be adaptively adjusted according to the signal-to-noise ratio), and all pixels in the spectrum below the threshold are set to 0 (black), while pixels above or equal to the threshold are set to 1 (white), thus generating a binarized image. This image preserves the energy concentration areas of the main vibrational harmonics, providing a clear data foundation for subsequent geometric structure analysis.

[0050] S222: Perform connected component analysis on the binarized image to obtain multiple initial frequency line segments composed of pixels.

[0051] In a binarized image, spatially adjacent points with pixel values ​​all equal to 1 are grouped into the same connected region. By scanning and labeling the entire image, all independent connected regions are identified. Each connected region corresponds to a locally continuous energy trajectory in both time and frequency directions. Further skeleton extraction processing is performed on this connected region, simplifying it into a single-pixel-width centerline structure, forming an initial frequency line segment composed of ordered pixels. Each initial frequency line segment has a definite start and end point; however, due to factors such as rapid changes in rotational speed, noise interference, or transient signal impacts, these segments often exhibit discontinuity, branching, or overlap, and cannot directly represent the complete frequency evolution trajectory.

[0052] S223: For each initial frequency line segment, extract its geometric features; the geometric features include the starting point coordinates and the ending point coordinates.

[0053] To support subsequent correlation determination between line segments, geometric attribute modeling is required for each initial frequency line segment. This includes the time-frequency coordinates of the start and end points. The location information of the start and end points is the key basis for determining whether line segments can be connected, and is used to limit the search range and establish the connection direction.

[0054] S224: Using the endpoint coordinates of each initial frequency line segment as a reference, find the starting coordinates of other initial frequency line segments that meet the preset conditions within the preset neighborhood window downstream, and establish potential connection relationships between the corresponding line segments.

[0055] Using the endpoint of a given initial frequency segment as the reference origin, a two-dimensional neighborhood window (e.g., a rectangular area with fixed row height and column width) is defined downstream of its time axis. Within this area, a search is conducted to determine if the starting points of other initial frequency segments fall within it. If so, a potential physical continuation relationship is considered, and this pairing is recorded as a potential connection. This process iterates through the endpoints of all initial frequency segments, attempting to establish candidate connections one by one, thus initially constructing a topological network of connections between the segments.

[0056] In one alternative implementation, step S224 may include the following sub-steps: S2241: Take each initial frequency line segment as the source line segment, take the end coordinates of the source line segment as the search origin, and set a neighborhood window downstream of it; the neighborhood window includes a preset height and a preset width.

[0057] For the initial frequency segment to be processed as the source segment, a rectangular neighborhood window is constructed in the positive direction of the time axis (i.e., future time) using its endpoint position (r_e, c_e) in the time spectrum as the reference point. The preset height (frequency direction) and preset width (time direction) can be preset according to the signal sampling rate, frequency resolution, and typical frequency change rate. For example, the preset height is 50 pixels and the preset width is 20 pixels.

[0058] S2242: Scan all other initial frequency segments and select other initial frequency segments whose starting coordinates are within the neighborhood window as candidate connection segments.

[0059] Iterate through all extracted initial frequency segments except the source segment, checking if their starting coordinates fall within the defined neighborhood window. If the starting point (r_s, c_s) of an initial frequency segment satisfies: r_s≥r_e, |r_s - r_e| ≤ Δr_max, |c_s - c_e| ≤ Δc_max Wherein, Δr_max and Δc_max are the preset height and preset width, respectively.

[0060] Then it is included in the candidate connecting segment set of the current source segment.

[0061] S2243: Based on preset connection criteria, determine the target connection segment from the candidate connection segments.

[0062] After obtaining a set of candidate connecting segments, it is necessary to further evaluate their continuity probability with the source segment based on multi-dimensional criteria, thereby selecting the optimal matching object. To this end, embodiments of the present invention provide a hierarchical decision-making mechanism that prioritizes spatial compactness and introduces directional consistency judgment when indeterminacy is not possible.

[0063] Therefore, the geometric features extracted in step S223 may also include line segment length, head direction vector, tail direction vector, and tail direction stability.

[0064] In one alternative implementation, step S2243 may include the following sub-steps: S22431: For each candidate connecting line segment, calculate the Euclidean distance between the endpoint coordinates of the source line segment and the starting coordinates of each candidate connecting line segment.

[0065] Calculate the two-dimensional planar distance from the end point of the source segment to the start point of each candidate connecting segment:

[0066] This distance reflects the spatial proximity of the two in the time-frequency plane and is a fundamental indicator for measuring connection priority.

[0067] S22432: Determine whether there are any neighboring candidate line segments whose row difference between the endpoint coordinates and the starting coordinates is less than or equal to a preset row difference threshold and whose column difference is less than or equal to a preset column difference threshold.

[0068] First, determine if there are candidate connecting segments that satisfy stricter local proximity conditions, for example: a preset row difference threshold of 2 pixels and a preset column difference threshold of 3 pixels. The determination condition would then be:

[0069] Such "nearby candidate segments" indicate that their starting point is immediately adjacent to the end point of the source segment, and they have a high probability of physical continuity.

[0070] S22433: If it exists, take the nearest candidate line segment with the smallest Euclidean distance as the target connecting line segment.

[0071] When there is at least one candidate line segment that satisfies the local proximity condition, the line segment with the shortest Euclidean distance is selected as the final target line segment to ensure the smoothest connection path and the least jump.

[0072] S22434: If it does not exist, calculate the first angle between the tail direction vector of the source line segment and the head direction vector of each candidate connecting line segment.

[0073] When no candidate line segment enters the strictly adjacent region, it indicates that the break spacing is large or there is a transient disturbance. At this time, the system switches to direction guidance mode. The first angle between the tail direction vector of the source line segment and the head direction vector of the candidate connecting line segment is calculated.

[0074] S22435: Calculate the second angle between the tail direction vector of the source line segment and the connecting vector; the connecting vector is the connecting direction vector from the end point of the source line segment to the starting point of the candidate connecting line segment.

[0075] Define the link vector v_link = (r_s - r_e, c_s - c_e), and calculate the angle between it and the direction vector at the end of the source segment. This angle reflects the consistency between the trend extension direction of the source segment and the actual jump direction. If the second angle is small, it indicates that the candidate segment is located on the natural extension path of the source segment, supporting the connection.

[0076] S22436: If the tail direction stability of the source line segment is stable and the line segment length is greater than or equal to the preset length, then determine whether there is an angle candidate line segment with a first included angle less than the first preset angle and a second included angle less than the second preset angle.

[0077] S22437: If it exists, take the candidate line segment with the smallest first included angle as the target connecting line segment.

[0078] A direction-driven connection strategy is only employed when the source line segment itself has high reliability. If the line segment is long, for example, ≥20, and its tail direction has a low rate of change (i.e., the tail direction is stable), then its direction information is considered reliable. Under this premise, candidate line segments with angles where the first included angle is less than a first preset angle and the second included angle is less than a second preset angle are further filtered. For example, the first preset angle is 20° and the second preset angle is 30°.

[0079] If there are candidate line segments with angles, select the candidate line segment with the smallest first included angle (i.e. the best directional consistency) as the target connecting line segment to complete the optimal matching.

[0080] If it does not exist, then the action of connecting the source segment to the next line ends.

[0081] S2244: Establish a directed potential connection between the source line segment and the target connecting line segment.

[0082] Once the target connecting segment is determined, a directed potential connection is established between the source segment and the target connecting segment, indicating that the former may extend to the latter in time evolution. This potential connection will be used to subsequently build a complete directed graph model, supporting global path reconstruction.

[0083] S225: Construct a directed graph based on each potential connection relationship, and traverse the directed graph using a graph traversal algorithm to merge the interrelated initial frequency segments into complete candidate frequency ridges.

[0084] All initial frequency segments are treated as nodes in a directed graph, and each potential connection is considered as a directed edge pointing from the source segment to the target segment, thus constructing a directed graph model. A graph traversal algorithm (such as Depth-First Search (DFS) or Breadth-First Search (BFS)) is then used to traverse the graph, identifying all sequences of nodes that can be connected by edges. Each maximum connected path corresponds to a set of segments that extend continuously in space and time. By concatenating all segments in this set in chronological order, a complete and continuous candidate frequency ridge can be reconstructed. Finally, all such reconstructed ridges are output as the base dataset for subsequent selection of reference harmonics.

[0085] In one alternative implementation, in step S230 above, one frequency ridge is selected from the candidate frequency ridges as a reference harmonic ridge, see [link to relevant documentation]. Figure 4 This may include the following sub-steps S231-S238 to ensure that the selected ridge has high reliability and physical interpretability, thereby providing an accurate benchmark for subsequent phase estimation and order analysis.

[0086] S231: Calculate the theoretical range of meshing frequency based on the number of teeth and rotational speed range of the rotating machinery.

[0087] For gear-driven equipment, one of the main sources of vibration excitation is the gear meshing process. If the number of teeth Z of the driving gear and the possible operating speed range [n_min, n_max] (unit: rpm) of the equipment are known, the approximate range of the theoretical meshing frequency can be derived: f_mesh = (n / 60) × Z.

[0088] Therefore, the theoretical range of meshing frequency is: [(n_min × Z) / 60, (n_max × Z) / 60].

[0089] This range forms the basis for subsequent initial screening, effectively narrowing the search space and avoiding the misselection of unrelated harmonics as reference signals.

[0090] S232: Select the first candidate ridge line whose frequency is within the theoretical range from all candidate frequency ridge lines.

[0091] Traverse all complete candidate frequency ridges obtained through image processing and reconstruction, and check whether their average or typical frequency over the entire time axis falls within the theoretical meshing frequency band calculated above. If the frequency trajectory of a ridge is within this interval, then include it in the "first candidate ridge" set.

[0092] S233: Select a preset number of second candidate ridges with high energy from each of the first candidate ridges.

[0093] Within the first candidate ridge set, ridges are further sorted according to the vibration energy level they represent. The intensity of each ridge can be quantified by calculating the average squared amplitude or logarithmic energy of its duration. The top K ridges with the highest energy (e.g., K=1 or K=5) are selected for the next stage of analysis, forming the "second candidate ridge" set. This strategy prioritizes retaining significant vibration components with high signal-to-noise ratios, improving the reliability of subsequent judgments.

[0094] S234: For each second candidate ridge, calculate the theoretical turning frequency of the second candidate ridge and determine whether there is a ridge with a frequency equal to the theoretical turning frequency among the candidate frequency ridges.

[0095] Suppose the frequency of a certain second candidate ridge is f_k(t). If it is the gear meshing frequency, then the corresponding spindle rotation frequency should be: f (t) = f k (t) / Z.

[0096] Subsequently, the system searches among all the original candidate frequency ridges for another ridge whose frequency trajectory is highly consistent with f_r(t). If it exists, it indicates that the system has simultaneously captured the meshing frequency and its fundamental frequency component, which together form a complete physical harmonic chain, greatly enhancing the credibility of the candidate ridge as the true meshing frequency.

[0097] S235: If it exists, then the second candidate ridge line is used as the reference harmonic ridge line.

[0098] Once it is confirmed that there is an actual ridge line that matches the theoretical frequency, it can be determined that the second candidate ridge line is highly likely to be the true main meshing frequency, and it can be directly selected as the reference harmonic ridge line for subsequent phase estimation and demodulation processing.

[0099] S236: If not, determine whether there exists a ridge among the candidate frequency ridges whose frequency is an integer multiple of the frequency of the second candidate ridge. If yes, proceed to step S235; if no, proceed to step S237.

[0100] When a clear fundamental frequency ridge cannot be found, it can be verified in reverse whether it is a lower-order component of a higher-order harmonic. That is, check whether there are other ridge frequencies approximately n×f_k(t) (n = 2, 3...).

[0101] If it is found to participate in forming a set of integer multiple harmonic relationships (such as f_k, 2f_k, 3f_k), it indicates that the frequency has good periodicity and structure, supporting it as the core reference frequency, and therefore it can still be identified as the reference harmonic ridge.

[0102] S237: If it does not exist, the most continuous and smoothest ridge among the second candidate ridges shall be used as the reference harmonic ridge.

[0103] If neither of the above two physical relationships is satisfied, the strategy reverts to a fallback approach: selecting the ridge with the most continuous frequency trajectory and the smallest fluctuation from the second candidate ridges as the final reference ridge. The smoothness of each ridge can be measured by calculating indicators such as the standard deviation of its frequency change rate, the number of breaks, or the integral of curvature.

[0104] Through the above-mentioned multi-level, multi-criteria fusion screening process, the embodiments of the present invention can still robustly identify the most representative reference harmonic ridges even in the absence of ideal operating conditions.

[0105] However, the instantaneous phase obtained by directly integrating the frequency trajectory of the reference harmonic ridge may still have a slight deviation, especially under long-term operation or drastic changes in operating conditions. Numerical integration is prone to cumulative errors, and if it is directly used for equal-angle resampling, it will lead to broadening or even misalignment of the order spectrum.

[0106] Therefore, this embodiment of the invention does not directly rely on the integral phase for resampling. Instead, it further utilizes this phase to construct a phase demodulation operator, performing "frequency straightening" processing on the original vibration signal in the sense of generalized Fourier transform. Through this demodulation operation, the frequency modulation effect caused by the rotational speed change can be effectively eliminated, generating a demodulated signal containing a pure periodic component, from which higher-precision instantaneous phase information can be extracted. See also Figure 5 Step S240 may include the following sub-steps: S241: Integrate the reference harmonic ridge line to obtain the instantaneous phase of the reference harmonic ridge line.

[0107] S242: Calculate the instantaneous estimated phase of the principal axis based on the instantaneous phase and the order of the reference harmonic ridge.

[0108] S243: Construct a phase demodulation operator based on the instantaneous estimated phase.

[0109] S244: Multiply the phase demodulation operator with the original vibration signal to obtain the demodulated vibration signal.

[0110] The instantaneous frequency trajectory characterized by the selected reference harmonic ridge is denoted as... By integrating this frequency function in the time domain, its corresponding instantaneous phase can be obtained. Since the reference harmonic ridge has undergone multiple screenings and optimizations in the preceding steps, its frequency trajectory has high continuity and low noise interference. Therefore, the phase drift and cumulative error generated by the integration process are significantly reduced, and the obtained instantaneous phase has high accuracy.

[0111] The order of the reference harmonic ridge pair is (For example, if the meshing frequency is Z times the rotational frequency, then...) =Z), the instantaneous estimated phase of the principal axis is then calculated as follows: = .

[0112] Then, based on the idea of ​​generalized Fourier transform, a complex-valued demodulation function (i.e., phase demodulation operator) is constructed: .

[0113] Where j is the imaginary unit. This operator "straightens" the periodic components in the original signal that bend with rotational speed, making them appear as a constant frequency line in the frequency domain. By introducing this nonlinear phase compensation mechanism, the problem of failure of traditional fixed filtering or simple integration methods under wide-range variable speeds is effectively overcome.

[0114] The acquired raw vibration signal x(t) is then compared with the aforementioned phase demodulation operator. By multiplying point by point, the demodulated vibration signal x1(t) is obtained, which contains the vibration information after phase compensation. At this point, the components of each order in the original signal that are synchronized with the rotation of the main shaft have been converted into approximately stationary narrowband signals, and the frequency modulation phenomenon is greatly suppressed, laying a good foundation for subsequent accurate extraction of instantaneous phase and realization of equal-angle resampling.

[0115] To construct a more accurate angular domain vibration signal, in one alternative implementation, see [link to relevant documentation]. Figure 6 Step S250 may include the following sub-steps: S251: Perform low-pass filtering on the demodulated vibration signal to remove high-frequency components and obtain the baseband signal.

[0116] S252: Perform inverse Fourier transform on the baseband signal to obtain a complex signal.

[0117] S253: Extract the phase information of the complex signal to obtain the instantaneous precise phase of the spindle.

[0118] S254: Construct the target phase sequence based on the preset number of sampling points per revolution.

[0119] S255: Calculate the acquisition time series corresponding to the target phase sequence based on the instantaneous accurate phase.

[0120] S256: Using an interpolation algorithm, the amplitude of each time point in the acquired time series in the original vibration signal is determined, and the angular domain vibration signal with equal intervals in the angular domain is obtained.

[0121] First, the demodulated vibration signal x1(t) is input to a low-pass filter, the cutoff frequency of which is set according to the bandwidth of the lowest-order component of the system (e.g., 0~50 Hz). This operation aims to preserve the slowly varying periodic components synchronized with the spindle rotation, while suppressing high-frequency noise, uncorrelated harmonics, and residual modulation components. After filtering, the signal energy is concentrated near zero frequency, forming a stable "baseband signal," providing a clean data foundation for subsequent phase extraction.

[0122] An inverse Fourier transform is performed on the filtered baseband signal to convert it back from the frequency domain to the time domain, resulting in a new complex time series z(t). Since the rotational speed-related vibration components in the original signal have been effectively "straightened out," this complex signal exhibits approximately constant amplitude and frequency characteristics, with a smooth envelope and continuous phase, making it very suitable for high-precision phase tracking.

[0123] Next, the arctangent function is applied to the complex signal z(t) to calculate the instantaneous precise phase of its principal axis: = arg(z(t))=atan2(b(t),a(t)) Where z(t) = a(t) + j b(t) is a complex signal, a(t) is the real part of the signal, b(t) is the imaginary part of the signal, and atan2() is the arctangent function.

[0124] Then, set the number of sampling points N required for each rotation (e.g., N=128), then the angular interval between two adjacent points is... Therefore, an equal-angle target phase sequence is constructed:

[0125] This sequence represents the sampling positions that are expected to be uniformly distributed in the angular domain, and serves as the target reference for subsequent interpolation.

[0126] according to and It is possible to find the target phase The corresponding time How much, i.e., what to solve This allows us to construct a set of corresponding target sampling times, i.e., the acquisition time series.

[0127] On the original vibration signal x(t), for each found time point Interpolation operations (such as linear interpolation, cubic spline interpolation, or Lagrange interpolation) are performed to obtain the corresponding vibration amplitude. Finally, a discrete sequence of length N is formed: x(θ) = [x(t0), x(t1), x(t2), ...], which yields a discrete signal sequence x(θ) uniformly distributed over the rotation angle, which is the angular domain equally spaced vibration signal.

[0128] Furthermore, embodiments of the present invention provide an order spectrum generation device, see [link to related document]. Figure 7 The order spectrum generation device 300 includes: The time-spectrum acquisition unit 310 is used to perform a short-time Fourier transform on the acquired original vibration signal of the rotating machinery to obtain the time-spectrum of the rotating machinery.

[0129] The ridge extraction unit 320 is used to extract multiple continuous candidate frequency ridges in the time-spectrum graph based on image processing technology; wherein each frequency ridge represents the instantaneous frequency trajectory of a vibrational harmonic component.

[0130] The reference harmonic ridge determination unit 330 is used to select one as a reference harmonic ridge from each candidate frequency ridge.

[0131] The demodulated vibration signal calculation unit 340 is used to calculate the instantaneous estimated phase of the main shaft based on the reference harmonic ridge, and demodulate the original vibration signal based on the instantaneous estimated phase to obtain the demodulated vibration signal.

[0132] The angle domain vibration signal calculation unit 350 is used to perform equal-angle resampling on the original vibration signal based on the demodulated vibration signal to obtain an angle domain vibration signal with equal intervals.

[0133] The order spectrum generation unit 360 is used to perform Fourier transform on the angular domain vibration signal to generate an order spectrum for condition monitoring and fault diagnosis of rotating machinery.

[0134] In summary, the order spectrum generation method, apparatus, electronic device, and storage medium provided by this invention, by combining time-frequency analysis, image processing, and generalized demodulation techniques, achieve adaptive order tracking under variable speed conditions in rotating machinery. First, by performing a short-time Fourier transform on the original vibration signal and extracting candidate frequency ridges using image processing techniques, continuous harmonic frequency trajectories can be automatically identified and reconstructed from complex backgrounds, improving the robustness and accuracy of key feature extraction. Second, by determining the optimal reference harmonic ridge through a multi-level screening strategy (based on theoretical frequency range, energy level, and physical harmonic relationships), the reliability and physical interpretability of the reference signal are ensured. Furthermore, by calculating the instantaneous estimated phase of the main shaft based on the reference ridge and constructing a demodulation operator, phase compensation is performed on the original signal, effectively suppressing the frequency modulation effect caused by speed changes; and by extracting the instantaneous precise phase in reverse from the demodulated signal, high-precision equal-angle resampling is achieved, significantly reducing the accumulated error caused by traditional integration methods. The resulting order spectrum is characterized by concentrated energy and clear spectral lines, and can accurately reflect the vibration characteristics of each order of the equipment under the condition of no bond phase signal, thereby realizing effective condition monitoring and fault diagnosis of key components such as gears and bearings.

[0135] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0136] In addition, the functional modules in the various embodiments of the present invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0137] If the functionality is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a computer-readable storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. 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.

[0138] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0139] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for generating order spectra, characterized in that, include: A short-time Fourier transform is performed on the original vibration signal of the rotating machinery to obtain the time-frequency spectrum of the rotating machinery. Based on image processing technology, multiple continuous candidate frequency ridges are extracted from the time-spectrum graph; wherein each frequency ridge represents the instantaneous frequency trajectory of a vibrational harmonic component. From the candidate frequency ridges, one is selected as the reference harmonic ridge. Based on the reference harmonic ridge, the instantaneous estimated phase of the main shaft of the rotating machinery is calculated, and the original vibration signal is demodulated based on the instantaneous estimated phase to obtain the demodulated vibration signal; Based on the demodulated vibration signal, the original vibration signal is resampled at equal angles to obtain an angle domain vibration signal with equal intervals in the angle domain; The angular domain vibration signal is subjected to Fourier transform to generate an order spectrum for use in the condition monitoring and fault diagnosis of the rotating machinery.

2. The method according to claim 1, characterized in that, The image processing technology is used to extract multiple consecutive candidate frequency ridges from the time-spectrum graph, including: The time-spectrum is binarized to obtain a binarized image; Connectivity analysis is performed on the binarized image to obtain multiple initial frequency line segments composed of pixels; For each initial frequency line segment, its geometric features are extracted; the geometric features include the starting point coordinates and the ending point coordinates. Based on the endpoint coordinates of each initial frequency line segment, the starting coordinates of other initial frequency line segments that meet the preset conditions are found in the preset neighborhood window downstream, and potential connection relationships are established between the corresponding line segments. A directed graph is constructed based on the potential connections described above. The directed graph is then traversed using a graph traversal algorithm to merge the interconnected initial frequency segments into complete candidate frequency ridges.

3. The method according to claim 2, characterized in that, The step of using the endpoint coordinates of each initial frequency line segment as a reference, searching for the starting coordinates of other initial frequency line segments that meet preset conditions within a preset neighborhood window downstream, and establishing potential connection relationships between the corresponding line segments includes: Each initial frequency line segment is taken as a source line segment, and the end coordinates of the source line segment are taken as the search origin. A neighborhood window is set downstream of it; the neighborhood window includes a preset height and a preset width. Scan all other initial frequency segments and select other initial frequency segments whose starting coordinates are within the neighborhood window as candidate connection segments; Based on preset connection criteria, a target connection segment is determined from the candidate connection segments; A directed potential connection relationship is established between the source line segment and the target connecting line segment.

4. The method according to claim 3, characterized in that, The geometric features also include segment length, head direction vector, tail direction vector, and tail direction stability; the step of determining the target connecting segment from the candidate connecting segments based on preset connection criteria includes: For each of the candidate connecting segments, calculate the Euclidean distance between the endpoint coordinates of the source segment and the starting coordinates of each candidate connecting segment; Determine if there are any neighboring candidate line segments whose row difference between the endpoint coordinates and the starting coordinates is less than or equal to a preset row difference threshold and whose column difference is less than or equal to a preset column difference threshold; If it exists, the nearest candidate line segment with the smallest Euclidean distance will be taken as the target connecting line segment; If it does not exist, calculate the first angle between the tail direction vector of the source line segment and the head direction vector of each candidate connecting line segment; Calculate the second angle between the tail direction vector of the source line segment and the connecting vector; the connecting vector is the connecting direction vector from the end point of the source line segment to the starting point of the candidate connecting line segment; If the tail direction stability of the source line segment is stable and the line segment length is greater than or equal to the preset length, then determine whether there are candidate line segments with a first included angle less than the first preset angle and a second included angle less than the second preset angle. If it exists, the candidate line segment with the smallest first included angle is taken as the target connecting line segment.

5. The method according to claim 1, characterized in that, The step of selecting one as a reference harmonic ridge from the candidate frequency ridges includes: Based on the number of teeth and rotational speed range of the rotating machinery, the theoretical range of meshing frequency is calculated; From the candidate frequency ridges, select the first candidate ridge with a frequency within the theoretical value range; Select a predetermined number of second candidate ridges with high energy from each of the first candidate ridges; For each of the second candidate ridges, calculate the theoretical turning frequency of the second candidate ridge and determine whether there is a ridge with a frequency equal to the theoretical turning frequency among the candidate frequency ridges. If it exists, the second candidate ridge line is used as the reference harmonic ridge line; If not, determine whether there is a ridge among the candidate frequency ridges whose frequency is an integer multiple of the frequency of the second candidate ridge; If it exists, the second candidate ridge line is used as the reference harmonic ridge line; If none exists, the smoothest and most continuous ridge among the second candidate ridges is taken as the reference harmonic ridge.

6. The method according to claim 1, characterized in that, The step of calculating the instantaneous estimated phase of the principal axis based on the reference harmonic ridge, and demodulating the original vibration signal based on the instantaneous estimated phase to obtain a demodulated vibration signal includes: Integrating the reference harmonic ridge line yields the instantaneous phase of the reference harmonic ridge line; The instantaneous estimated phase of the principal axis is calculated based on the instantaneous phase and the order of the reference harmonic ridge. Based on the instantaneously estimated phase, a phase demodulation operator is constructed; The demodulated vibration signal is obtained by multiplying the phase demodulation operator with the original vibration signal.

7. The method according to claim 1 or 6, characterized in that, The step of resampling the original vibration signal at equal angles based on the demodulated vibration signal to obtain equally spaced angle-domain vibration signals includes: The demodulated vibration signal is subjected to low-pass filtering to remove high-frequency components, thereby obtaining the baseband signal; Perform an inverse Fourier transform on the baseband signal to obtain a complex signal; The phase information of the complex signal is extracted to obtain the instantaneous precise phase of the main shaft; Construct the target phase sequence based on the preset number of sampling points per revolution; Based on the instantaneous precise phase, calculate the acquisition time series corresponding to the target phase sequence; Using an interpolation algorithm, the amplitude of each time point in the acquired time series in the original vibration signal is determined, and an angular domain vibration signal with equal intervals in the angular domain is obtained.

8. An order spectrum generation device, characterized in that, include: The time-spectrum acquisition unit is used to perform a short-time Fourier transform on the acquired original vibration signal of the rotating machinery to obtain the time-spectrum of the rotating machinery. The ridge extraction unit is used to extract multiple consecutive candidate frequency ridges from the time-spectrum graph based on image processing technology; wherein each frequency ridge represents the instantaneous frequency trajectory of a vibrational harmonic component. The reference harmonic ridge line determination unit is used to select one as a reference harmonic ridge line from each of the candidate frequency ridge lines. The demodulated vibration signal calculation unit is used to calculate the instantaneous estimated phase of the main shaft of the rotating machinery based on the reference harmonic ridge, and demodulate the original vibration signal based on the instantaneous estimated phase to obtain the demodulated vibration signal. An angle domain vibration signal calculation unit is used to perform equal-angle resampling on the original vibration signal based on the demodulated vibration signal to obtain an angle domain vibration signal with equal intervals in the angle domain. The order spectrum generation unit is used to perform Fourier transform on the angular domain vibration signal to generate an order spectrum for use in the condition monitoring and fault diagnosis of the rotating machinery.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 7.

Images