A method and system for quantifying the degradation state of a rotating machine

Abstract

The application belongs to the technical field of rotating machinery state monitoring. A rotating machinery degradation state quantification method and system are provided. In view of the problems that the traditional Renyi entropy is susceptible to noise and working condition fluctuation interference, and the degradation trend characterization stability is insufficient, the application first acquires the original vibration signal of the rotating machinery operation and constructs an envelope energy sequence, extracts a candidate period set and calculates a period consistency strength sequence, combines zero-crossing interval expansion to modify the envelope energy sequence, calculates the period consistency modified Renyi entropy of the modified envelope energy sequence, unifies the direction of the period consistency modified Renyi entropy, and obtains the rotating machinery degradation state quantification index. The application can improve the monotonicity and trend of the degradation index, and can provide reliable support for rotating machinery degradation state evaluation and early fault warning.

Description

Technical Field

[0001] This invention relates to the field of rotating machinery condition monitoring technology, and specifically to a method and system for quantifying the degradation state of rotating machinery. Background Technology

[0002] Rotating machinery is widely used in industries such as energy, manufacturing, rail transportation, and metallurgy. Key components such as bearings, gears, and shafts operate under high-speed, heavy-load, and complex conditions for extended periods, making them prone to degradation faults such as wear, pitting, cracking, and spalling. Degradation state quantification technology can construct health indicators from monitoring signals to characterize the equipment performance degradation process, providing support for equipment operation and maintenance, fault early warning, and remaining life prediction. Vibration signals are sensitive to fault characteristics such as mechanical shock and friction, making them a mainstream data source for degradation state monitoring. Entropy-based indicators based on vibration signals are easy to calculate. Among them, Renyi entropy, as a generalized form of Shannon entropy, can characterize the concentration and uncertainty of the signal energy probability distribution, exhibiting a certain sensitivity to impact-related faults and is commonly used for the quantitative analysis of the degradation state of rotating machinery.

[0003] Existing methods for quantifying the degradation of rotating machinery based on Reynolds entropy only calculate indicators based on the statistical distribution of signal amplitude or energy probability, without incorporating the inherent periodic consistency characteristics of rotating machinery fault impacts. In the early stages of equipment degradation, fault impacts are typically weak and easily mixed with background noise, random impacts, and operating condition fluctuations. Fault-related periodic impacts and irregular interference components are included equally in the statistical calculations, leading to significant fluctuations in the obtained Reynolds entropy. The monotonicity and consistency of the degradation trend are insufficient, making it difficult to stably reflect the true performance degradation process of the equipment, reducing the reliability of degradation state assessment, and hindering subsequent fault early warning and life prediction work. Summary of the Invention

[0004] To address the shortcomings of existing methods for quantifying the degradation state of rotating machinery based on Raney entropy, such as failure to consider fault cycle characteristics, susceptibility to interference, and insufficient stability in characterizing degradation trends, this invention provides a method and system for quantifying the degradation state of rotating machinery. By constructing an envelope energy sequence, extracting a candidate period set, calculating the period consistency strength, and correcting the envelope energy sequence, the method calculates the period consistency correction Raney entropy based on the corrected energy sequence and achieves directional unification. This enables quantitative characterization of the degradation state of rotating machinery and improves the monotonicity, trend, and anti-interference capability of the degradation state quantification indicators.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for quantifying the degradation state of rotating machinery.

[0006] A method for quantifying the degradation state of rotating machinery includes the following steps: Obtain the original vibration signals during the operation of rotating machinery, and construct an envelope energy sequence based on the original vibration signals; A candidate period set is extracted based on the envelope sequence corresponding to the envelope energy sequence, and a period consistency strength sequence is calculated based on the candidate period set and the envelope sequence. The periodic uniformity intensity sequence is extended based on the zero-crossing interval of the original vibration signal, and the envelope energy sequence is corrected based on the extended periodic uniformity intensity sequence to obtain the corrected envelope energy sequence. The periodic consistency corrected Raney entropy is calculated from the corrected envelope energy sequence, and the direction of the periodic consistency corrected Raney entropy is unified to obtain a quantitative index of the degradation state of rotating machinery.

[0007] In one implementation of the first aspect of the present invention, constructing an envelope energy sequence based on the original vibration signal includes: After preprocessing the original vibration signal, a Hilbert transform is performed to obtain an analytic signal. The magnitude of the analytic signal is taken to obtain the envelope sequence. The envelope sequence is then squared to construct the envelope energy sequence.

[0008] In one implementation of the first aspect of the present invention, extracting a candidate period set based on the envelope sequence includes: Calculate the normalized autocorrelation coefficients of the envelope sequence under different delays, combine them to obtain a normalized autocorrelation sequence, perform a Fourier transform on the normalized autocorrelation sequence to obtain a candidate period spectrum, search for peak points in the candidate period spectrum, and calculate the candidate period set based on the ratio of the frequency corresponding to the peak point to the sampling frequency.

[0009] In one implementation of the first aspect of the present invention, calculating the periodic consistency strength sequence based on the candidate period set and the envelope sequence includes: Construct a sliding window matrix based on the envelope sequence, and calculate the correlation of the corresponding period offset position of each sliding window for each candidate period in the candidate period set. For each sliding window, the maximum value of the correlation under all candidate periods is taken as the periodic consistency strength of the corresponding sliding window, and the combination is used to obtain the periodic consistency strength sequence.

[0010] As a further limitation of the first aspect of the present invention, calculating the correlation of the corresponding period offset positions of each sliding window includes: Perform mean removal and norm normalization on each window vector in the sliding window matrix; For each candidate period, construct a multi-order set of positive and negative offsets, calculate the correlation between the current window vector and the window vector at the corresponding period offset position, and take the average of the effective correlations under the same candidate period to obtain the period consistency strength of the current sliding window under the corresponding candidate period.

[0011] In one implementation of the first aspect of the present invention, the periodic uniformity intensity sequence is extended according to the zero-crossing interval of the original vibration signal, including: The set of zero-crossing points is determined based on the sign change of the original vibration signal, and the oscillation interval is constructed from adjacent zero-crossing points as the zero-crossing interval; The maximum value of the periodic consistency intensity within each zero-crossing interval is taken as the periodic consistency intensity of all sampling points within the interval, thus obtaining the extended periodic consistency intensity sequence.

[0012] In one implementation of the first aspect of the present invention, the envelope energy sequence is modified based on the extended periodic consistency intensity sequence, including: Determine the periodicity threshold and compare the intensity value of each sampling point in the expanded periodicity intensity sequence with the periodicity threshold. The envelope energy values ​​corresponding to sampling points with intensity values ​​not less than the periodicity consistency threshold are retained, while the envelope energy values ​​corresponding to sampling points with intensity values ​​less than the periodicity consistency threshold are set to zero, thus obtaining the corrected envelope energy sequence.

[0013] As a further limitation of the first aspect of the present invention, determining the periodic consistency threshold includes: Set an initial periodicity consistency threshold and count the number of sampling points in the current sample whose periodicity consistency strength is greater than the initial periodicity consistency threshold; If the number of sampling points in the current sample is not less than the maximum number of sampling points in the historical samples whose periodic consistency strength is greater than the initial periodic consistency threshold, then the initial periodic consistency threshold is used as the periodic consistency threshold. If the number of sampling points in the current sample is less than the maximum number of sampling points in the historical samples whose periodic consistency strength is greater than the initial periodic consistency threshold, then the periodic consistency strength of the current sample is sorted in descending order, and the strength value at the corresponding sorting position is selected as the periodic consistency threshold.

[0014] In one implementation of the first aspect of the present invention, calculating the periodic consistency corrected Raney entropy of the corrected envelope energy sequence includes: The corrected envelope energy sequence is normalized to an energy probability distribution; Based on the normalized energy probability distribution, the periodic consistency modified Raney entropy is calculated by substituting it into the Raney entropy calculation formula. When the sum of the corrected envelope energy sequence is zero, a preset value is output to avoid calculation errors. To unify the direction of the periodic consistency correction Reni entropy, including: Based on the changing trend of the periodically consistent modified Reni entropy with the degree of equipment degradation, we can choose to directly use the periodically consistent modified Reni entropy or perform a reverse transformation on the periodically consistent modified Reni entropy to obtain a quantitative index of the degradation state of rotating machinery that changes monotonically with the increase of the degree of equipment degradation.

[0015] Secondly, the present invention provides a system for quantifying the degradation state of rotating machinery.

[0016] A system for quantifying the degradation state of rotating machinery, comprising: The envelope energy construction unit is configured to: acquire the original vibration signal during the operation of the rotating machinery, and construct an envelope energy sequence based on the original vibration signal; The period intensity calculation unit is configured to: extract a candidate period set based on the envelope sequence corresponding to the envelope energy sequence, and calculate a period consistency intensity sequence based on the candidate period set and the envelope sequence; The energy sequence correction unit is configured to: extend the periodic uniformity intensity sequence according to the zero-crossing interval of the original vibration signal, and correct the envelope energy sequence based on the extended periodic uniformity intensity sequence to obtain the corrected envelope energy sequence. The degradation index generation unit is configured to: calculate the periodic consistency modified Raney entropy on the modified envelope energy sequence, and unify the direction of the periodic consistency modified Raney entropy to obtain a quantitative index of the degradation state of rotating machinery.

[0017] Compared with the prior art, the beneficial effects of the present invention are: This invention combines periodic consistency correction with Reni entropy calculation, effectively improving the stability and trend of quantitative results for the degradation state of rotating machinery while retaining the sensitivity of Reni entropy to signal energy probability distribution and impact distribution characteristics. First, an envelope energy sequence is constructed based on the original vibration signal to characterize vibration and impact features in energy form. Then, periodic consistency strength is calculated through candidate period extraction and sliding window correlation analysis. Combined with zero-crossing interval expansion, the window-level strength is mapped to the sampling point level. The periodic consistency strength is used to filter and correct the envelope energy sequence, suppressing the influence of random impacts, background noise, and aperiodic disturbances, while retaining the periodic impact components related to rotational faults. Finally, the periodic consistency-corrected Reni entropy is calculated based on the corrected energy sequence, and its direction is unified to obtain a quantitative index for the degradation state. The obtained index more clearly presents the evolution process of equipment performance degradation, enhances the monotonicity and coherence of the degradation trend, and reduces the impact of operating condition fluctuations and noise interference on state assessment. This provides more reliable quantitative support for early fault identification, degradation state assessment, and operation and maintenance decisions for rotating machinery.

[0018] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0019] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0020] Figure 1 A flowchart illustrating a method for quantifying the degradation state of rotating machinery, provided as an exemplary embodiment of the present invention; Figure 2 A comparison of the time-varying trends of conventional Rényi entropy and PC-Rényi entropy is provided as an exemplary embodiment of the present invention, wherein... Figure 2 In the figure (a), the traditional Rényi entropy changes over time. Figure 2 (b) in the figure represents the trend of PC-Rényi entropy over time; Figure 3 A quantitative comparison of conventional Rényi entropy and PC-Rényi entropy in terms of monotonicity, trend and robustness is provided for an exemplary embodiment of the present invention; Figure 4 A schematic diagram of a rotating machinery degradation state quantification system provided as an exemplary embodiment of the present invention; Figure 5 A schematic diagram of a computer device provided for an exemplary embodiment of the present invention. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0022] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0023] Condition monitoring of rotating machinery is mainly achieved by collecting multi-source monitoring data such as vibration signals, current signals, acoustic emission signals, or temperature signals. Among these, vibration signals are widely used in fault diagnosis and degradation assessment of rotating machinery due to their sensitivity to mechanical shock, friction, loosening, and periodic faults. Existing methods for quantifying the degradation state of vibration signals mainly include time-domain statistical indicators, frequency-domain and time-frequency characteristic indicators, mode decomposition characteristic indicators, entropy-based complexity indicators, and data-driven health indicator construction methods. Among entropy-based and statistical complexity indicators, methods such as Shannon entropy, sample entropy, permutation entropy, spectral entropy, and Rényi entropy can describe the equipment condition from the perspectives of probability distribution, energy concentration, and signal complexity. Rényi entropy, a generalized form of Shannon entropy, adjusts the degree of attention given to high-probability components or energy-concentrated components through an order parameter. When introduced into vibration signal analysis, it can be used to characterize the non-uniformity of signal amplitude, envelope energy, or spectral energy distribution. When rotating machinery suffers localized damage, the periodic fault impact alters the energy distribution and complexity characteristics of the vibration signal, thus Rényi entropy can serve as a quantitative indicator of degradation state. However, in actual industrial environments, early fault impacts are often weak and are superimposed with background noise, random interference, frequency modulation, and operating condition fluctuations, making it difficult for traditional Rényi entropy to stably reflect the true degradation trend.

[0024] The root cause of these shortcomings lies in the fact that most existing degradation quantification indices only focus on changes in the amplitude, energy, frequency, or probability distribution of vibration signals, without fully considering the "periodic consistency" characteristic of rotating machinery fault impacts. For rotating machinery, impacts caused by localized damage typically recur with the rotation cycle; while random noise, accidental impacts, and some operating condition disturbances often lack stable periodic consistency. Therefore, if periodic consistency constraints can be introduced into the Rényi entropy calculation process, the influence of non-periodic disturbances on degradation state quantification can be reduced while retaining the Rényi entropy's sensitivity to the complexity and concentration of energy distribution.

[0025] To address the problems of existing solutions, this invention proposes a method for quantifying the degradation state of rotating machinery based on periodic consistency-corrected Rényi entropy. The purpose of this invention is to improve the stability, monotonicity, and anti-interference ability of traditional Rényi entropy in quantifying the degradation state of rotating machinery, enabling the obtained health indicators to more accurately reflect the evolution of equipment from a normal state to a fault state. Specifically, for example... Figure 1 As shown, it includes the following steps: Step S1: Acquire the raw vibration signals during the operation of the rotating machinery and construct an impact energy characterization sequence for quantifying the degradation state. More specifically, this includes: Step S101: Collect vibration signals during the operation of rotating machinery. Let the first... The original vibration signal corresponding to each sample for: (1); in, , The number of samples collected. The length of a single vibration signal segment. Representing the The vibration amplitude of the first sampling point in each sample. Representing the The vibration amplitude at the second sampling point in the sampled data. Representing the The th sample Vibration amplitude at each sampling point This represents transposition.

[0026] Step S102: Perform finite value screening on the original vibration signal to eliminate non-numerical, infinite, or abnormal sampling points; if necessary, perform mean removal, trend removal, or bandpass filtering on the original vibration signal to obtain the vibration signal to be analyzed.

[0027] Step S103: Perform Hilbert transform on the preprocessed vibration signal to obtain the analytical signal. : (2); in, express Transformation, The imaginary unit, Representing the The th sample Vibration amplitude at each sampling point.

[0028] Obtain the envelope sequence by taking the modulus value of the analytic signal. : (3); Step S104: Construct the envelope energy sequence based on the envelope sequence: (4); in, Indicates the first The sample at the th The impact energy intensity at each sampling point. This envelope energy sequence serves as input for subsequent periodic consistency-corrected Rényi entropy calculations.

[0029] Step S2: Extract a candidate period set based on the envelope correlation features of the vibration signal, and calculate the period consistency intensity sequence through sliding window correlation analysis. Specifically, this includes: Step S201: Calculate the normalized autocorrelation sequence of the envelope sequence, which is used to characterize the repetition similarity of the envelope signal under different delays. For delays... The normalized autocorrelation coefficient can be expressed as: (5); in, , and These are the means of the two envelope sequences involved in the relevant calculations. To prevent extremely small positive numbers with a denominator of zero, normalized autocorrelation sequences are used. It is used to characterize the degree of repetition similarity of envelope signals under different delays.

[0030] Step S202: Perform a Fast Fourier Transform on the normalized autocorrelation sequence to obtain the periodic candidate spectrum: (6); in, Representing the The frequency in each sampled sample is Periodic candidate spectral amplitude at Frequency coordinates for: (7); In the formula, Sampling frequency, To normalize the autocorrelation sequence length, Indicates to Round down. Retain spectral amplitude values ​​within the non-negative frequency range as candidate periodic spectra.

[0031] Step S203: Search for peak points in the periodic candidate spectrum. Preferably, search for peaks only within a frequency range not lower than 10Hz, and select the first 5 peak positions according to their amplitude from largest to smallest. Let the preset lower frequency limit be... Only when the conditions are met Search for peak points within the frequency range to obtain the peak set. : (8); in, This is a local peak value. According to peak amplitude Peak set from largest to smallest Sort, select the first Key peak locations: (9); Candidate periods are calculated based on peak frequencies. : (10); in, Representative to Perform rounding operation.

[0032] This forms a candidate periodicity set. : (11); In the formula, The number of candidate cycles.

[0033] Preferably, the periods corresponding to the top 5 peaks with the largest amplitudes in the candidate periodic spectrum are selected as candidate periods. When no valid peak points are detected, the following preset period set is used as the candidate period set: (12); Step S204: Construct a sliding window matrix based on the envelope sequence. Set the sliding window length. The number of windows is Then the sliding window matrix is: (12); Among them, preferably . No. sliding window vectors for: (13); Step S205: Perform mean removal and norm normalization on each sliding window vector: (14); in, Window vector The mean, It is a vector consisting entirely of 1s.

[0034] Step S206: For candidate periods Construct a multi-order set of positive and negative offsets : (15); In the formula, The maximum period multiple, preferably .

[0035] Step S207: Calculate the correlation between the current window and its periodic offset window: (16); In the formula, ,and Within the valid window range; Representing the Each window is at an offset of The correlation coefficient at that time.

[0036] Step S208: Average the effective correlations under the same candidate period to obtain the periodic consistency strength under that candidate period: (17); Among them, when When the window position is valid, ,otherwise ; Representing the In the nth sample The window in the first Periodic consistency strength under candidate periods; This represents the effective location indicator factor.

[0037] Step S209: Take the maximum value of all candidate periods to obtain the window position. Periodic consistency strength at: (18).

[0038] Step S3: Extend the periodic consistency intensity sequence based on the zero-crossing interval, and correct the impact energy characterization sequence according to the periodic consistency intensity. Specifically, this includes: Step S301: Determine the set of zero-crossing points based on the sign change of the vibration signal to be analyzed. : (19); in, The sign function is defined, and the oscillation interval is constructed from adjacent zero crossover points. ; Representing the The position number of each zero-crossing point.

[0039] Step S302: Extend the window-level periodicity intensity to the sampling point-level periodicity intensity. For those located in the same oscillation interval For each sampling point within a given interval, the maximum periodicity strength within that interval is taken as the periodicity value of all sampling points within that interval. (twenty one); in, This represents the point-by-point periodic consistency strength after interval expansion.

[0040] Step S303: Set the periodic consistency threshold The periodic consistency threshold can be a fixed threshold or a threshold dynamically adjusted based on historical samples. Let the initial periodic consistency threshold be... In this embodiment, the peak statistical results in the normalized cross-correlation plot of the healthy stage samples are used to determine the optimal value. Then the first The number of sampling points in a sample whose periodicity strength is greater than the initial threshold for: (twenty two); in, For indicator functions: (twenty three); when Or the current sample satisfies: (twenty four); in, Representing the The total number of sampling points in a sample whose periodic consistency strength is greater than the initial threshold.

[0041] The initial threshold is used as the periodic consistency threshold for the current sample: ; when And the current sample satisfies: (25); Sort the current sample's periodic consistency strength sequence in descending order: (26); make: (27); And select the first The periodic consistency strength value is used as the periodic consistency threshold for the current sample: (28); Step S304: Based on the periodic consistency threshold Correct the envelope energy sequence: (29); in, This is the envelope energy value after periodic consistency correction. This step can suppress random shocks, background noise, and non-degradational perturbations with weak periodic consistency, while retaining degradation-related shock components with stable and repetitive characteristics.

[0042] Step S4: Calculate the normalized Rényi entropy on the impact energy characterization sequence after periodic consistency correction, and unify the degradation direction to obtain the PC-Rényi entropy, a quantitative index of the degradation state of rotating machinery. Specifically, this includes: Step S401: Normalize the envelope energy sequence after periodic consistency correction into a probability distribution: (30); in, Indicates the first The sample at the th Normalized energy probability at each sampling point To prevent extremely small positive numbers with a denominator of zero.

[0043] Step S402: Calculate the periodic consistency correction Rényi entropy: (31); in, For the first The periodic consistency correction Rényi entropy corresponding to each sample Let be the Rényi entropy order. Preferably, take . To enhance the indicator's response to periodic shock energies with higher probability weights; As a normalization factor, the resulting entropy value is in principle located within the range of... Interval. When (32); At that time, it can be Set to a preset value to avoid calculation errors when there is no effective energy component. Step S403: Based on the application direction of the degradation index, unify the direction of the periodic consistency correction Rényi entropy. Preferably, when... As the overall value increases with the severity of degradation in rotating machinery, the following quantitative index of degradation state is constructed: (33); when When the overall size decreases as the degree of degradation of rotating machinery increases, a reverse transformation can be performed: (34); in, For the first The periodic consistency correction Rényi entropy health index corresponds to each sample. Arranging the indices corresponding to all sampled samples in chronological order yields the degradation state quantification sequence: (35) In the formula, This represents the number of samples collected. This sequence can be used for tracking the degradation trend of rotating machinery, classifying degradation stages, assessing condition, and issuing early warnings.

[0044] like Figure 2 (a) and Figure 2 As shown in (b) in the figure, Figure 2 As shown, the lifetime trend of the traditional Rényi index after unifying the direction is compared with that of the PC-Rényi entropy obtained in this invention as a function of sampling time. To ensure that both indices change in the direction that "a larger value indicates a more severe degree of degradation," the traditional normalized Rényi entropy is adjusted. To perform a direction change, in order to As a comparative indicator with traditional methods, this transformation only changes the direction of change of the traditional Rényi entropy, without altering its original fluctuation characteristics and degradation information. The traditional Rényi index remains at a low level for a long time during the healthy and early degradation stages, with small numerical fluctuations, and its response to the initial degradation process is not obvious. As the degradation worsens, the index generally increases, but multiple local peaks and large fluctuations appear in the middle and late stages of degradation, and it decreases significantly in the near-failure stage, thus still failing to continuously and stably reflect the equipment performance degradation process. In contrast, the PC-Rényi entropy obtained by this invention has a low and relatively stable initial baseline in the healthy stage, gradually increases in the early degradation stage, then shows a clearer multi-stage step increase with the severity of degradation, and further increases rapidly in the near-failure stage. The overall direction of change is consistent with the process of rotating machinery evolving from a healthy state to a faulty state. The results demonstrate that the periodic consistency correction, by reducing the direct involvement of aperiodic random shocks and irrelevant energy components in the Rényi entropy calculation, highlights degradation-related shock energies with stable and repetitive characteristics, thereby altering the effective energy probability distribution. This allows the resulting health indicators to more clearly reflect the performance degradation process of rotating machinery. Therefore, compared to traditional Rényi entropy, the PC-Rényi entropy of this invention exhibits a clearer positive degradation trend, better degradation stage differentiation, and a more pronounced early degradation response.

[0045] like Figure 3As shown, a quantitative comparison is made between the traditional Rényi entropy and the PC-Rényi entropy of this invention in terms of robustness, trend, and monotonicity. Referring to the health indicator quality evaluation method, monotonicity characterizes the ability of an indicator to maintain a consistent direction of change during the degradation process, calculated by statistically analyzing the differences in the number of positive and negative changes between adjacent differences in the health indicator sequence; trend characterizes the correlation between the indicator and the running time, calculated by the absolute value of the Pearson correlation coefficient between the health indicator sequence and the collection sequence number, and the direction of change of the indicator is determined based on the signed correlation coefficient; robustness characterizes the ability of an indicator to resist local fluctuations and maintain a stable degradation trend, calculated by the degree of deviation of the health indicator from its smooth trend term. Specifically, the robustness of the traditional Rényi entropy is 0.997, slightly higher than the PC-Rényi entropy's 0.993, but both maintain high robustness. The trend and monotonicity of the PC-Rényi entropy reach 0.947 and 0.134 respectively, both significantly higher than the traditional Rényi entropy's 0.812 and 0.006, indicating that the periodic consistency correction can highlight the degradation-related components with stable and repetitive characteristics, making the obtained health indicators more correlated with the equipment degradation process and better monotonic consistency. In summary, although the robustness evaluation value of the PC-Rényi entropy is slightly lower, its trend and monotonicity are more significantly improved, and the overall evaluation result is better than that of the traditional Rényi entropy, verifying the effectiveness of the method of the present invention in tracking the degradation trend of rotating machinery, early fault warning, and condition assessment.

[0046] In summary, this invention, through candidate period extraction, sliding window period consistency analysis, zero-crossing interval expansion, and period consistency threshold correction, highlights the contribution of degradation-related periodic impact components to the energy probability distribution before Rényi entropy calculation, and reduces the influence of random impacts and non-degradational perturbations on entropy calculation. Compared with traditional Rényi entropy, the PC-Rényi entropy constructed in this invention maintains higher robustness while exhibiting better trend characteristics, monotonic consistency, and interpretability of degradation direction, and can more clearly characterize the evolution process of rotating machinery from a healthy state to a degraded state.

[0047] This invention, based on the traditional Rényi entropy measurement of signal energy probability distribution characteristics, introduces a periodic consistency correction mechanism to highlight the periodic impact component related to the fault mechanism of rotating machinery, enabling health indicators to more accurately characterize the equipment degradation process. This invention weakens the influence of random impacts, background noise, and non-fault components on indicator calculations through periodic consistency screening, avoiding the problem of traditional entropy indicators including both accidental large-amplitude interference and fault periodic impacts in the statistics, thus improving the reliability of degradation state quantification results under complex operating conditions. The PC-Rényi entropy index constructed by this invention enhances the regularity of health indicator changes with the degradation process, reduces the impact of local fluctuations on degradation trend judgment, and makes the overall direction of index change more consistent with the evolution of equipment from a normal state to a fault state, making it more suitable for subsequent tasks such as fault early warning and remaining life prediction. This invention does not rely on complex deep learning training processes or require a large number of fault label samples, and can be applied to the quantitative analysis of degradation state of vibration signals from various types of rotating machinery such as bearings, gearboxes, motors, and fans.

[0048] Figure 4 A system for quantifying the degradation state of rotating machinery is shown, comprising: Envelope energy construction unit 401 is configured to: acquire the original vibration signal during the operation of rotating machinery, and construct an envelope energy sequence based on the original vibration signal; The period intensity calculation unit 402 is configured to: extract a candidate period set based on the envelope sequence corresponding to the envelope energy sequence, and calculate a period consistency intensity sequence based on the candidate period set and the envelope sequence; The energy sequence correction unit 403 is configured to: extend the periodic uniformity intensity sequence according to the zero-crossing interval of the original vibration signal, and correct the envelope energy sequence based on the extended periodic uniformity intensity sequence to obtain the corrected envelope energy sequence. The degradation index generation unit 404 is configured to: calculate the periodic consistency modified Raney entropy on the modified envelope energy sequence, and unify the direction of the periodic consistency modified Raney entropy to obtain a quantitative index of the degradation state of rotating machinery.

[0049] It is understood that the aforementioned units can be individually or entirely merged into one or more other units, or some of the units can be further divided into multiple functionally smaller units. This achieves the same operation without affecting the technical effects of the embodiments of the present invention. The aforementioned units are based on logical functional division. In practical applications, the function of one unit can be implemented by multiple units, or the function of multiple units can be implemented by one unit. In other embodiments of the present invention, the system may also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented collaboratively by multiple units.

[0050] According to another embodiment of the present invention, the system of this embodiment can be constructed by running a computer program (including program code) capable of performing the steps involved in the corresponding method of the present invention on a general-purpose computing device, such as a computer, which includes processing elements and storage elements such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM). The computer program can be recorded on, for example, a computer-readable recording medium, loaded into the aforementioned computing device through the computer-readable recording medium, and run therein.

[0051] Figure 5 A computer device is shown, which includes a processor 501, a communication interface 502, and a computer-readable storage medium 503. The processor 501, communication interface 502, and computer-readable storage medium 503 can be connected via a bus or other means.

[0052] The communication interface 502 is used to receive and send data. The computer-readable storage medium 503 can be stored in the memory of the electronic device. The computer-readable storage medium 503 is used to store computer programs, which include program instructions. The processor 501 is used to execute the program instructions stored in the computer-readable storage medium 503.

[0053] The processor 501 is the computing and control core of the electronic device. It is suitable for implementing one or more instructions, specifically for loading and executing one or more instructions to achieve the corresponding method flow or corresponding function.

[0054] Processor 501 is configured to perform the following procedure: Obtain the original vibration signals during the operation of rotating machinery, and construct an envelope energy sequence based on the original vibration signals; A candidate period set is extracted based on the envelope sequence corresponding to the envelope energy sequence, and a period consistency strength sequence is calculated based on the candidate period set and the envelope sequence. The periodic uniformity intensity sequence is extended based on the zero-crossing interval of the original vibration signal, and the envelope energy sequence is corrected based on the extended periodic uniformity intensity sequence to obtain the corrected envelope energy sequence. The periodic consistency corrected Raney entropy is calculated from the corrected envelope energy sequence, and the direction of the periodic consistency corrected Raney entropy is unified to obtain a quantitative index of the degradation state of rotating machinery.

[0055] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can implement the described functions using different methods for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0056] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic cable, digital cable) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data processing device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0057] The above description is merely a preferred embodiment of the present invention and is not intended to limit the 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.

Claims

1. A method for quantifying the degradation state of rotating machinery, characterized in that, Includes the following processes: Obtain the original vibration signals during the operation of rotating machinery, and construct an envelope energy sequence based on the original vibration signals; A candidate period set is extracted based on the envelope sequence corresponding to the envelope energy sequence, and a period consistency strength sequence is calculated based on the candidate period set and the envelope sequence. The periodic uniformity intensity sequence is extended based on the zero-crossing interval of the original vibration signal, and the envelope energy sequence is corrected based on the extended periodic uniformity intensity sequence to obtain the corrected envelope energy sequence. The periodic consistency corrected Raney entropy is calculated from the corrected envelope energy sequence, and the direction of the periodic consistency corrected Raney entropy is unified to obtain a quantitative index of the degradation state of rotating machinery.

2. The method for quantifying the degradation state of rotating machinery as described in claim 1, characterized in that, The envelope energy sequence is constructed based on the original vibration signal, including: After preprocessing the original vibration signal, a Hilbert transform is performed to obtain an analytic signal. The magnitude of the analytic signal is taken to obtain the envelope sequence. The envelope sequence is then squared to construct the envelope energy sequence.

3. The method for quantifying the degradation state of rotating machinery as described in claim 1, characterized in that, Candidate period sets are extracted based on envelope sequences, including: Calculate the normalized autocorrelation coefficients of the envelope sequence under different delays, combine them to obtain a normalized autocorrelation sequence, perform a Fourier transform on the normalized autocorrelation sequence to obtain a candidate period spectrum, search for peak points in the candidate period spectrum, and calculate the candidate period set based on the ratio of the frequency corresponding to the peak point to the sampling frequency.

4. The method for quantifying the degradation state of rotating machinery as described in claim 1, characterized in that, The periodic consistency strength sequence is calculated based on the candidate periodic set and the envelope sequence, including: Construct a sliding window matrix based on the envelope sequence, and calculate the correlation of the corresponding period offset position of each sliding window for each candidate period in the candidate period set. For each sliding window, the maximum value of the correlation under all candidate periods is taken as the periodic consistency strength of the corresponding sliding window, and the combination is used to obtain the periodic consistency strength sequence.

5. The method for quantifying the degradation state of rotating machinery as described in claim 4, characterized in that, Calculate the correlation of the corresponding period offset positions for each sliding window, including: Perform mean removal and norm normalization on each window vector in the sliding window matrix; For each candidate period, construct a multi-order set of positive and negative offsets, calculate the correlation between the current window vector and the window vector at the corresponding period offset position, and take the average of the effective correlations under the same candidate period to obtain the period consistency strength of the current sliding window under the corresponding candidate period.

6. The method for quantifying the degradation state of rotating machinery as described in claim 1, characterized in that, The periodic uniformity intensity sequence is extended based on the zero-crossing intervals of the original vibration signal, including: The set of zero-crossing points is determined based on the sign change of the original vibration signal, and the oscillation interval is constructed from adjacent zero-crossing points as the zero-crossing interval; The maximum value of the periodic consistency intensity within each zero-crossing interval is taken as the periodic consistency intensity of all sampling points within the interval, thus obtaining the extended periodic consistency intensity sequence.

7. The method for quantifying the degradation state of rotating machinery as described in claim 1, characterized in that, The envelope energy sequence is modified based on the extended periodic consistency strength sequence, including: Determine the periodicity threshold and compare the intensity value of each sampling point in the expanded periodicity intensity sequence with the periodicity threshold. The envelope energy values ​​corresponding to sampling points with intensity values ​​not less than the periodicity consistency threshold are retained, while the envelope energy values ​​corresponding to sampling points with intensity values ​​less than the periodicity consistency threshold are set to zero, thus obtaining the corrected envelope energy sequence.

8. The method for quantifying the degradation state of rotating machinery as described in claim 7, characterized in that, Determining the periodic consistency threshold includes: Set an initial periodicity consistency threshold and count the number of sampling points in the current sample whose periodicity consistency strength is greater than the initial periodicity consistency threshold; If the number of sampling points in the current sample is not less than the maximum number of sampling points in the historical samples whose periodic consistency strength is greater than the initial periodic consistency threshold, then the initial periodic consistency threshold is used as the periodic consistency threshold. If the number of sampling points in the current sample is less than the maximum number of sampling points in the historical samples whose periodic consistency strength is greater than the initial periodic consistency threshold, then the periodic consistency strength of the current sample is sorted in descending order, and the strength value at the corresponding sorting position is selected as the periodic consistency threshold.

9. The method for quantifying the degradation state of rotating machinery as described in claim 1, characterized in that, The periodic consistency corrected Reni entropy is calculated for the corrected envelope energy sequence, including: The modified envelope energy sequence is normalized to an energy probability distribution; Based on the normalized energy probability distribution, the periodic consistency modified Raney entropy is calculated by substituting it into the Raney entropy calculation formula. When the sum of the corrected envelope energy sequence is zero, a preset value is output to avoid calculation errors. To unify the direction of the periodic consistency correction Reni entropy, including: Based on the changing trend of the periodically consistent modified Reni entropy with the degree of equipment degradation, we can choose to directly use the periodically consistent modified Reni entropy or perform a reverse transformation on the periodically consistent modified Reni entropy to obtain a quantitative index of the degradation state of rotating machinery that changes monotonically with the increase of the degree of equipment degradation.

10. A system for quantifying the degradation state of rotating machinery, characterized in that, include: The envelope energy construction unit is configured to: acquire the original vibration signal during the operation of the rotating machinery, and construct an envelope energy sequence based on the original vibration signal; The period intensity calculation unit is configured to: extract a candidate period set based on the envelope sequence corresponding to the envelope energy sequence, and calculate a period consistency intensity sequence based on the candidate period set and the envelope sequence; The energy sequence correction unit is configured to: extend the periodic uniformity intensity sequence according to the zero-crossing interval of the original vibration signal, and correct the envelope energy sequence based on the extended periodic uniformity intensity sequence to obtain the corrected envelope energy sequence. The degradation index generation unit is configured to: calculate the periodic consistency modified Raney entropy on the modified envelope energy sequence, and unify the direction of the periodic consistency modified Raney entropy to obtain a quantitative index of the degradation state of rotating machinery.

Images