Gear health assessment method and system based on magnetic memory signal analysis
Through the method based on magnetic memory signal analysis, combined with historical working condition parameters and time-frequency characteristic analysis, the gear health status is evaluated, and the accuracy of magnetic memory signal acquisition and analysis under complex working conditions is solved, achieving higher evaluation accuracy and reliability.
Patent Information
- Application Number
- CN202411196177.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-29
- Publication Date
- 2024-12-06
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing gear health assessment methods are difficult to achieve real-time and accurate magnetic memory signal acquisition and analysis under complex operating conditions, and are greatly affected by factors such as temperature, speed, and magnetic interference.
Using a method based on magnetic memory signal analysis, the historical working conditions parameters of the target gear are obtained, and the magnetic memory signal data is obtained by using the control variable method. Combined with time and frequency domain feature analysis, the acquisition accuracy and reliability are evaluated, and the health assessment time is determined. Window, obtain health assessment signal data in real time, and perform health status assessment through decision tree algorithm.
It improves the accuracy and reliability of gear health assessment, reduces the impact of environmental interference factors, and provides more accurate health status assessment results.
Smart Images

Figure CN119089162A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of magnetic memory signal analysis, and in particular to a gear health assessment method and system based on magnetic memory signal analysis. Background Art
[0002] Gears are key components in mechanical transmission systems, and their health status directly affects the performance and service life of the entire system. With the improvement of the automation level of industrial equipment, the importance of equipment operation status monitoring and fault diagnosis technology has become increasingly prominent. Traditional gear health assessment methods mainly rely on technologies such as vibration analysis, acoustic emission detection, and oil analysis. However, these methods are often affected by complex working conditions in practical applications, making it difficult to conduct real-time and accurate assessment of the health status of gears.
[0003] In recent years, magnetic memory detection technology has gradually become an emerging method for gear health assessment. The magnetic memory effect refers to the phenomenon that when a ferromagnetic material is subjected to force, its internal magnetic domain structure changes and an abnormal magnetic field is generated on its surface. By detecting the magnetic memory signal on the surface of the gear, the stress distribution and damage of the gear during operation can be obtained. Compared with traditional detection methods, magnetic memory signal detection has the advantages of non-contact and real-time monitoring, and can evaluate the health status of the gear without affecting its normal operation.
[0004] However, the application of magnetic memory signal detection technology in gear health assessment still faces some challenges. Since the magnetic memory signals of gears under different working conditions are greatly affected by factors such as temperature, rotation speed, and magnetic interference, how to accurately collect and analyze these signals and then evaluate the health status of the gears has become the focus and difficulty of current research. Therefore, a gear health assessment method based on magnetic memory signal analysis is proposed, which can improve the accuracy and reliability of magnetic memory signal data acquisition under complex working conditions, thereby providing effective support for gear health status assessment, which has important theoretical and practical significance. Summary of the invention
[0005] In order to solve at least one of the above technical problems, the present invention proposes a gear health assessment method and system based on magnetic memory signal analysis.
[0006] A first aspect of the present invention provides a gear health assessment method based on magnetic memory signal analysis, comprising:
[0007] Obtain historical operating parameters of the target gear under working conditions, obtain magnetic memory signal data of the target gear under each operating parameter based on a control variable method, and calibrate them as first magnetic memory signal data;
[0008] Determine the acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear according to the first magnetic memory signal data, and obtain acquisition accuracy data;
[0009] Performing a magnetic memory signal data acquisition reliability assessment on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability assessment result;
[0010] Determine a health assessment time window for the target gear according to the collected reliability assessment result, and acquire health assessment magnetic memory signal data for the target gear in real time according to the health assessment time window;
[0011] A health assessment is performed on the target gear according to the health assessment magnetic memory signal data to obtain a health assessment result of the target gear.
[0012] In this solution, the historical operating parameters of the target gear under the working state are obtained, and the magnetic memory signal data of the target gear under each operating parameter is obtained based on the control variable method, which is calibrated as the first magnetic memory signal data, specifically:
[0013] Acquire historical operating parameters of the target gear under working conditions, wherein the operating parameters include gear temperature, gear speed, and gear magnetic interference intensity;
[0014] Determine the historical parameters of each operating condition parameter item according to the historical operating condition parameters, and construct a parameter matrix with the historical parameters of each operating condition parameter item;
[0015] The magnetic memory signal data of each historical parameter of the target gear in each operating parameter item is obtained by the control variable method according to the parameter matrix and calibrated as the first magnetic memory signal data.
[0016] In this solution, the acquisition accuracy of the magnetic memory signal data is determined based on the first magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear, and the acquisition accuracy data is obtained, specifically:
[0017] Acquire a healthy target gear sample, and acquire magnetic memory signal data of the target gear sample under a preset working condition, and calibrate it as second magnetic memory signal data;
[0018] Sorting the first magnetic memory signal data and the second magnetic memory signal data in time series to obtain first time domain signal data and second time domain signal data;
[0019] Extracting time domain features of the first time domain signal data and the second time domain signal data, the time domain features including signal mean, variance, root mean square value, and kurtosis;
[0020] The first time domain signal data and the second time domain signal data are converted into frequency domain signals based on discrete Fourier transform, and frequency domain features of the first magnetic memory signal data and the second magnetic memory signal data are extracted, wherein the frequency domain features include spectrum frequency, frequency center, spectrum energy, and spectrum bandwidth;
[0021] Calculating the Manhattan distance between the time domain characteristics and frequency domain characteristics of the first magnetic memory signal data and the time domain characteristics and frequency domain characteristics of the second magnetic memory signal data, and determining the influence weight of the time domain characteristics and frequency domain characteristics on the difference evaluation of the magnetic memory signal data according to the information entropy weight method;
[0022] The difference between the first magnetic memory signal data and the second magnetic memory signal data is determined according to the Manhattan distance and the influence weight, and the acquisition accuracy of the magnetic memory signal data for the operating parameters of different operating items under the target gear working state is evaluated according to the difference to obtain the acquisition accuracy data.
[0023] In this solution, the magnetic memory signal data acquisition reliability evaluation is performed on the target gear's operating parameter change data within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability evaluation result, which is specifically:
[0024] Based on the K-means clustering algorithm, K data points are randomly selected from the collection accuracy data as the initial cluster centers. For each data point in the collection accuracy data, the Euclidean distance between each data point and each initial cluster center is calculated, and the data point is assigned to the cluster with the closest initial cluster center. The mean of all data points in the cluster after the data point is assigned is calculated, and the cluster center is updated according to the mean.
[0025] Iteratively calculate the Euclidean distance of each data point to the new cluster center and assign the data point to the cluster where the nearest cluster center is located, and update the cluster center until the cluster center no longer changes, thus obtaining the data clustering result of collection accuracy;
[0026] According to the acquisition accuracy data clustering results, a magnetic memory signal data acquisition reliability level evaluation is performed on each clustering result to determine the operating parameter range of each operating condition item in different magnetic memory signal data acquisition reliability levels;
[0027] Constructing a membership function according to the operating condition parameter range of each operating condition item in different magnetic memory signal data acquisition reliability levels, and constructing a magnetic memory signal data acquisition reliability evaluation model based on the membership function;
[0028] The operating parameter change data of the target gear within a preset time period is obtained, the operating parameter change data is imported into the reliability assessment model, the reliability of the magnetic memory signal data acquisition within the preset time period is evaluated, and the collection reliability assessment result of the magnetic memory signal data within the preset time period is obtained.
[0029] In this solution, the health assessment time window of the target gear is determined according to the collected reliability assessment result, and the health assessment magnetic memory signal data of the target gear is acquired in real time according to the health assessment time window, specifically:
[0030] Sorting the acquisition reliability evaluation results according to the time series to obtain the acquisition reliability time series change data of the magnetic memory signal data within a preset time period;
[0031] Building an acquisition reliability prediction model based on the LSTM algorithm, obtaining historical acquisition reliability time series change data and importing it into the model for model training, importing the acquisition reliability time series change data into the acquisition reliability prediction model after training, and predicting the acquisition reliability prediction data for magnetic memory signal data collection of the target gear in the future preset time period;
[0032] Determine a high reliability time window for magnetic memory signal data acquisition within a future preset time period according to the acquisition reliability prediction data, and mark it as a health assessment time window for the target gear;
[0033] The health assessment magnetic memory signal data of the target gear is acquired in real time in the health assessment time window.
[0034] In this solution, the health assessment of the target gear is performed according to the health assessment magnetic memory signal data to obtain the health assessment result of the target gear, which is specifically:
[0035] Obtaining magnetic memory signal data of target gears with different quality problems, extracting time domain features and frequency domain features of the magnetic memory signal data, and labeling the extracted time domain features and frequency domain features with the types of quality problems to obtain labeled data;
[0036] Building a target gear health status assessment model based on a decision tree algorithm, and importing the labeled data into the target gear health status assessment model for training;
[0037] Extracting the time domain features and frequency domain features of the health assessment magnetic memory signal data and marking them as current feature data;
[0038] Marking the time domain features and frequency domain features of the second magnetic memory signal data as reference data, comparing the current feature data with the reference data, calculating the Euclidean distance between the current feature data and the reference data, and evaluating the difference between the current feature data and the reference data according to the Euclidean distance;
[0039] A difference threshold is preset, and if the difference between the current characteristic data and the reference data is less than the difference threshold, the target gear is calibrated as a healthy state;
[0040] If the difference between the current feature data and the benchmark data is greater than the difference threshold, the current feature data is imported into the target gear health status assessment model to identify quality problems, and the health status of the target gear is assessed based on the identified quality problems to obtain a health assessment result of the target gear.
[0041] The second aspect of the present invention further provides a gear health assessment system based on magnetic memory signal analysis, the system comprising: a memory, a processor, the memory comprising a gear health assessment method program based on magnetic memory signal analysis, and when the gear health assessment method program based on magnetic memory signal analysis is executed by the processor, the following steps are implemented:
[0042] Obtain historical operating parameters of the target gear under working conditions, obtain magnetic memory signal data of the target gear under each operating parameter based on a control variable method, and calibrate them as first magnetic memory signal data;
[0043] Determine the acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear according to the first magnetic memory signal data, and obtain acquisition accuracy data;
[0044] Performing a magnetic memory signal data acquisition reliability assessment on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability assessment result;
[0045] Determine a health assessment time window for the target gear according to the collected reliability assessment result, and acquire health assessment magnetic memory signal data for the target gear in real time according to the health assessment time window;
[0046] A health assessment is performed on the target gear according to the health assessment magnetic memory signal data to obtain a health assessment result of the target gear.
[0047] The present invention discloses a gear health assessment method and system based on magnetic memory signal analysis. The method obtains the magnetic memory signal data of the target gear under working conditions and combines the historical working condition parameters to evaluate the health status of the gear. First, the magnetic memory signal data of the target gear under different working conditions is obtained, the accuracy of signal acquisition under each working condition is determined, and the acquisition accuracy data is obtained; then, the signal acquisition reliability of the gear within a preset time period is evaluated based on the acquisition accuracy data to obtain an evaluation result; then, the health assessment time window is determined, and the magnetic memory signal data is acquired in real time within this window; finally, a health assessment is performed based on the signal data to obtain the health status of the gear. This method improves the accuracy and reliability of gear health assessment. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Figure 1A flow chart of a gear health assessment method based on magnetic memory signal analysis of the present invention is shown;
[0049] Figure 2 A flow chart showing the method of determining the health assessment time window of a target gear according to the present invention is shown;
[0050] Figure 3 A flow chart showing the health assessment result of a target gear obtained by the present invention;
[0051] Figure 4 A block diagram of a gear health assessment system based on magnetic memory signal analysis of the present invention is shown. DETAILED DESCRIPTION
[0052] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.
[0053] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.
[0054] Figure 1 A flow chart of a gear health assessment method based on magnetic memory signal analysis of the present invention is shown.
[0055] like Figure 1 As shown, the first aspect of the present invention provides a gear health assessment method based on magnetic memory signal analysis, comprising:
[0056] S102, obtaining historical operating parameters of the target gear under the working state, obtaining magnetic memory signal data of the target gear under each operating parameter based on the control variable method, and calibrating them as first magnetic memory signal data;
[0057] S104, determining the acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear according to the first magnetic memory signal data, and obtaining acquisition accuracy data;
[0058] S106, performing a magnetic memory signal data acquisition reliability assessment on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data, and obtaining an acquisition reliability assessment result;
[0059] S108, determining a health assessment time window for a target gear according to the collected reliability assessment result, and acquiring health assessment magnetic memory signal data for the target gear in real time according to the health assessment time window;
[0060] S110, performing a health assessment on the target gear according to the health assessment magnetic memory signal data to obtain a health assessment result of the target gear.
[0061] It should be noted that, when the target gear is running, the magnetic memory signal data obtained under different working parameters of the target gear will deviate, so the obtained magnetic memory signal data cannot accurately evaluate the health status of the target gear; therefore, by obtaining the magnetic memory signal data of the target gear under different working parameters (such as temperature, rotation speed, magnetic interference, etc.) based on the control variable method, and analyzing these signal data, the influence of each working parameter on the accuracy of the magnetic memory signal acquisition data can be accurately evaluated, and the working parameter change data within a preset time period can be evaluated for acquisition reliability according to the acquisition accuracy data, and the time window for health evaluation of the target gear is determined according to the evaluation result, and the magnetic memory signal data is obtained within the time window to evaluate the health status of the target gear. By determining the optimal time window for health evaluation, it is ensured that the data is collected during the period with the highest accuracy of magnetic memory signal data acquisition and the smallest change of working parameters. In this way, the influence of environmental interference factors can be minimized, more accurate magnetic memory signal data can be obtained, and the accuracy of the health status evaluation results of the target gear can be improved.
[0062] According to an embodiment of the present invention, the historical operating parameters of the target gear under the working state are obtained, and the magnetic memory signal data of the target gear under each operating parameter is obtained based on the control variable method, and is calibrated as the first magnetic memory signal data, specifically:
[0063] Acquire historical operating parameters of the target gear under working conditions, wherein the operating parameters include gear temperature, gear speed, and gear magnetic interference intensity;
[0064] Determine the historical parameters of each operating condition parameter item according to the historical operating condition parameters, and construct a parameter matrix with the historical parameters of each operating condition parameter item;
[0065] The magnetic memory signal data of each historical parameter of the target gear in each operating parameter item is obtained by the control variable method according to the parameter matrix and calibrated as the first magnetic memory signal data.
[0066] It should be noted that there are multiple historical parameters. For example, the target gear has experienced operating conditions of 50, 60, 70 degrees and so on during its historical working process, then these temperatures are all historical parameters of the gear temperature operating parameter item; the method of obtaining the magnetic memory signal data of each historical parameter of the target gear in each operating parameter item through the control variable method according to the parameter matrix is specifically: taking an A operating parameter item as a variable and other operating parameter items as constants, obtaining the magnetic memory signal data of the A operating parameter item under each historical parameter. For example, when the gear temperature is a variable, the gear speed and the gear magnetic interference intensity are set to 0, and the target gear is assigned each historical temperature parameter value in the parameter matrix in turn, and the magnetic memory signal data under each temperature parameter is obtained. This step is to assign temperature or speed or magnetic interference intensity to the target gear through experiments.
[0067] According to an embodiment of the present invention, the acquisition accuracy of the magnetic memory signal data is determined based on the first magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear to obtain the acquisition accuracy data, specifically:
[0068] Acquire a healthy target gear sample, and acquire magnetic memory signal data of the target gear sample under a preset working condition, and calibrate it as second magnetic memory signal data;
[0069] Sorting the first magnetic memory signal data and the second magnetic memory signal data in time series to obtain first time domain signal data and second time domain signal data;
[0070] Extracting time domain features of the first time domain signal data and the second time domain signal data, the time domain features including signal mean, variance, root mean square value, and kurtosis;
[0071] The first time domain signal data and the second time domain signal data are converted into frequency domain signals based on discrete Fourier transform, and frequency domain features of the first magnetic memory signal data and the second magnetic memory signal data are extracted, wherein the frequency domain features include spectrum frequency, frequency center, spectrum energy, and spectrum bandwidth;
[0072] Calculating the Manhattan distance between the time domain characteristics and frequency domain characteristics of the first magnetic memory signal data and the time domain characteristics and frequency domain characteristics of the second magnetic memory signal data, and determining the influence weight of the time domain characteristics and frequency domain characteristics on the difference evaluation of the magnetic memory signal data according to the information entropy weight method;
[0073] The difference between the first magnetic memory signal data and the second magnetic memory signal data is determined according to the Manhattan distance and the influence weight, and the acquisition accuracy of the magnetic memory signal data for the operating parameters of different operating items under the target gear working state is evaluated according to the difference to obtain the acquisition accuracy data.
[0074] It should be noted that the magnetic memory signal data was comprehensively analyzed by extracting time domain features (such as signal mean, variance, root mean square value, kurtosis) and frequency domain features (such as spectrum frequency, frequency center, spectrum energy, spectrum bandwidth). The time domain features can reflect the changing trend of the signal in time, while the frequency domain features reveal the distribution of the signal in frequency. The combination of the two can more comprehensively capture the information in the magnetic memory signal and effectively improve the accuracy of signal analysis; by calculating the Manhattan distance between the magnetic memory signal features of the target gear and the healthy sample, and combining the information entropy weight method to evaluate the impact of each feature on the difference, it can more sensitively capture the impact of the operating parameters on the accuracy of magnetic memory signal data acquisition; frequency domain analysis converts the time domain signal into the frequency domain signal through discrete Fourier transform, and can accurately decompose the various frequency components in the signal; the different operating conditions are gear temperature, gear speed, and gear magnetic interference intensity, and the operating parameters of different operating conditions are the specific parameters of each operating condition.
[0075] According to an embodiment of the present invention, the magnetic memory signal data acquisition reliability evaluation is performed on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability evaluation result, which is specifically:
[0076] Based on the K-means clustering algorithm, K data points are randomly selected from the collection accuracy data as the initial cluster centers. For each data point in the collection accuracy data, the Euclidean distance between each data point and each initial cluster center is calculated, and the data point is assigned to the cluster with the closest initial cluster center. The mean of all data points in the cluster after the data point is assigned is calculated, and the cluster center is updated according to the mean.
[0077] Iteratively calculate the Euclidean distance of each data point to the new cluster center and assign the data point to the cluster where the nearest cluster center is located, and update the cluster center until the cluster center no longer changes, thus obtaining the data clustering result of collection accuracy;
[0078] According to the acquisition accuracy data clustering results, a magnetic memory signal data acquisition reliability level evaluation is performed on each clustering result to determine the operating parameter range of each operating condition item in different magnetic memory signal data acquisition reliability levels;
[0079] Constructing a membership function according to the operating condition parameter range of each operating condition item in different magnetic memory signal data acquisition reliability levels, and constructing a magnetic memory signal data acquisition reliability evaluation model based on the membership function;
[0080] The operating parameter change data of the target gear within a preset time period is obtained, the operating parameter change data is imported into the reliability assessment model, the reliability of the magnetic memory signal data acquisition within the preset time period is evaluated, and the collection reliability assessment result of the magnetic memory signal data within the preset time period is obtained.
[0081] It should be noted that by further analyzing the acquisition accuracy data and using the clustering algorithm to classify the data, the reliability of magnetic memory signal acquisition under different operating parameters can be quantified. This process can identify which parameters have a greater impact on signal acquisition under different working conditions, and by building a reliability evaluation model, evaluate the impact of changes in these parameters in different time periods on signal acquisition reliability; clustering the acquisition accuracy data based on the K-means clustering algorithm can automatically group large-scale data by similarity. This automated clustering processing method can reduce possible errors caused by human intervention and improve the efficiency and consistency of data processing; by constructing a membership function based on the clustering results and using it to establish a magnetic memory signal data acquisition reliability evaluation model, a quantitative analysis of acquisition reliability can be achieved. This model based on the membership function can comprehensively consider the impact of multiple operating parameters on data acquisition reliability, thereby more accurately reflecting the reliability change trend of data acquisition and improving the accuracy and reliability of the evaluation model.
[0082] Figure 2 A flow chart of determining a health assessment time window for a target gear according to the present invention is shown.
[0083] According to an embodiment of the present invention, determining the health assessment time window of the target gear according to the collected reliability assessment result, and acquiring the health assessment magnetic memory signal data of the target gear in real time according to the health assessment time window, specifically, is as follows:
[0084] S202, sorting the acquisition reliability evaluation results according to the time series to obtain acquisition reliability time series change data of the magnetic memory signal data within a preset time period;
[0085] S204, constructing an acquisition reliability prediction model based on the LSTM algorithm, obtaining historical acquisition reliability time series change data and importing it into the model for model training, importing the acquisition reliability time series change data into the acquired reliability prediction model after training, and predicting the acquisition reliability prediction data for magnetic memory signal data acquisition of the target gear in a preset time period in the future;
[0086] S206, determining a high reliability time window for magnetic memory signal data acquisition within a preset time period in the future according to the acquisition reliability prediction data, and marking it as a health assessment time window for the target gear;
[0087] S208, acquiring health assessment magnetic memory signal data of the target gear in real time in the health assessment time window.
[0088] It should be noted that the LSTM (Long Short-Term Memory Network) algorithm has a strong ability to process time series data and can capture long-term dependencies and trend changes. The prediction model based on the LSTM algorithm can more accurately analyze and predict the time series changes of acquisition reliability, thereby improving the prediction accuracy of future acquisition reliability; predicting future acquisition reliability in advance allows health assessments to be performed during the high reliability window period, thereby identifying potential gear problems in advance. This early warning capability helps to formulate timely maintenance plans and strategies, reduce the risk of equipment failure, and extend the service life of equipment; determine the high reliability time window, and collect data during this time period to reduce the interference of low reliability data.
[0089] Figure 3 A flow chart of obtaining health assessment results of a target gear according to the present invention is shown.
[0090] According to an embodiment of the present invention, the health assessment of the target gear is performed according to the health assessment magnetic memory signal data to obtain the health assessment result of the target gear, specifically:
[0091] S302, obtaining magnetic memory signal data of target gears with different quality problems, extracting time domain features and frequency domain features of the magnetic memory signal data, and labeling the extracted time domain features and frequency domain features with quality problem types to obtain labeled data;
[0092] S304, constructing a target gear health status assessment model based on a decision tree algorithm, and importing the labeled data into the target gear health status assessment model for training;
[0093] S306, extracting the time domain features and frequency domain features of the health assessment magnetic memory signal data, and marking them as current feature data;
[0094] S308, calibrating the time domain features and frequency domain features of the second magnetic memory signal data as reference data, comparing the current feature data with the reference data, calculating the Euclidean distance between the current feature data and the reference data, and evaluating the difference between the current feature data and the reference data according to the Euclidean distance;
[0095] S310, presetting a difference threshold, if the difference between the current feature data and the reference data is less than the difference threshold, calibrating the target gear to be in a healthy state;
[0096] S312, if the difference between the current feature data and the reference data is greater than the difference threshold, the current feature data is imported into the target gear health status assessment model to identify quality problems, and the health status of the target gear is assessed based on the identified quality problems to obtain a health assessment result of the target gear.
[0097] It should be noted that the decision tree algorithm can generate a tree model based on the features and labels in the training data, clarify the features and decision rules of various health states; the decision tree model effectively classifies and distinguishes the time domain features and frequency domain features. By automatically processing these feature data, the model can quickly determine the health status of the gear, reduce the workload and time of manual analysis, and improve evaluation efficiency; by comparing the Euclidean distance between the current feature data and the benchmark data, the difference between the current data and the standard state can be accurately quantified. If the difference is less than the preset threshold, it can be determined that the target gear is in a healthy state, thereby reducing false alarms and ensuring the accuracy of the health status judgment; when the target gear is in a healthy state, no subsequent health status detection is performed, thereby improving the efficiency of health status detection and reducing the data processing pressure of the system; if the difference between the current feature data and the benchmark data is greater than the difference threshold, it is considered that the target gear has a quality problem, so a health status detection is required, and the specific type of quality problem is identified based on the magnetic memory signal data. Through detailed quality problem identification, clear fault information can be provided to maintenance personnel to help them formulate more targeted repair and maintenance strategies, thereby improving maintenance efficiency and effectiveness; the different quality problems include degree of wear, internal cracks, gear defects, material fatigue, and uneven stress.
[0098] According to an embodiment of the present invention, it also includes:
[0099] Obtaining a structural diagram of a target machine, obtaining position information of a target gear in the target machine, and determining information on the number of gears within a preset range of the position of the target gear according to the position information and the structural diagram of the target machine;
[0100] If the number of gears is greater than a preset value, obtain the multi-gear hybrid magnetic memory signal data within the preset range where the target gear is located, remove noise from the multi-gear hybrid magnetic memory signal data based on a low-pass filter and align the signal data according to a time series to obtain pre-processed signal data;
[0101] The independent component analysis method is introduced to construct a data matrix with the preprocessed signal data, where each column represents a mixed signal and each row represents the signal value at a time point;
[0102] The data matrix is centrally processed, a covariance matrix of the data matrix after the centralization processing is calculated, the covariance matrix is eigenvalue decomposed to obtain eigenvalues and eigenvectors, and a unit matrix of the covariance matrix is calculated based on the eigenvalues and eigenvectors;
[0103] The unit matrix is processed based on an independent component analysis method to separate independent component signals of multi-gear mixed magnetic memory signal data, and an independent component matrix and an unmixing matrix are constructed;
[0104] The independent component signals are restored to magnetic memory signal data of each gear according to the independent component matrix and the unmixing matrix.
[0105] It should be noted that in a multi-gear system, magnetic memory signals may be mixed together due to the mutual influence between multiple gears. This mixed signal makes it difficult to separate the individual signals of each gear, affecting the accurate assessment of the health status of each gear; therefore, the independent component analysis (ICA) method can extract each independent source signal from multiple mixed signals; the ICA algorithm can automatically distinguish between signals and noise in the process of separating signals, thereby improving the quality and accuracy of signals; ICA achieves accurate separation of mixed signals through mathematical models and algorithms. This makes the extraction of time domain and frequency domain features more accurate during data processing; ICA algorithms usually use some optimization techniques to find the optimal separation matrix. These optimization techniques include maximizing non-Gaussian metrics (such as negative entropy), minimizing certain cost functions, or achieving this through contrast measures (such as Kurtos is). By continuously adjusting the separation matrix, the ICA algorithm can make the separated signals as close to independent components as possible. Through independent component analysis, the algorithm constructs an unmixing matrix that can decompose mixed signals into independent source signals; the data centering process is to subtract the mean of each column in the data matrix from its mean so that the mean of each column is zero.
[0106] Figure 4 A block diagram of a gear health assessment system based on magnetic memory signal analysis of the present invention is shown.
[0107] The second aspect of the present invention further provides a gear health assessment system 4 based on magnetic memory signal analysis, the system comprising: a memory 41, a processor 42, the memory comprising a gear health assessment method program based on magnetic memory signal analysis, the gear health assessment method program based on magnetic memory signal analysis when executed by the processor, implements the following steps:
[0108] Obtain historical operating parameters of the target gear under working conditions, obtain magnetic memory signal data of the target gear under each operating parameter based on a control variable method, and calibrate them as first magnetic memory signal data;
[0109] Determine the acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear according to the first magnetic memory signal data, and obtain acquisition accuracy data;
[0110] Performing a magnetic memory signal data acquisition reliability assessment on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability assessment result;
[0111] Determine a health assessment time window for the target gear according to the collected reliability assessment result, and acquire health assessment magnetic memory signal data for the target gear in real time according to the health assessment time window;
[0112] A health assessment is performed on the target gear according to the health assessment magnetic memory signal data to obtain a health assessment result of the target gear.
[0113] The present invention discloses a gear health assessment method and system based on magnetic memory signal analysis. The method obtains the magnetic memory signal data of the target gear under working conditions and combines the historical working condition parameters to evaluate the health status of the gear. First, the magnetic memory signal data of the target gear under different working conditions is obtained, the accuracy of signal acquisition under each working condition is determined, and the acquisition accuracy data is obtained; then, the signal acquisition reliability of the gear within a preset time period is evaluated based on the acquisition accuracy data to obtain an evaluation result; then, the health assessment time window is determined, and the magnetic memory signal data is acquired in real time within this window; finally, a health assessment is performed based on the signal data to obtain the health status of the gear. This method improves the accuracy and reliability of gear health assessment.
[0114] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.
[0115] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0116] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.
[0117] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical disks, and other media that can store program codes.
[0118] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.
[0119] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.
Claims
1. A gear health assessment method based on magnetic memory signal analysis, characterized in that: The following steps are involved: Obtain historical operating parameters of the target gear under working conditions, obtain magnetic memory signal data of the target gear under each operating parameter based on a control variable method, and calibrate them as first magnetic memory signal data; Determine the acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear according to the first magnetic memory signal data, and obtain acquisition accuracy data; Performing a magnetic memory signal data acquisition reliability assessment on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability assessment result; Determine a health assessment time window for the target gear according to the collected reliability assessment result, and acquire health assessment magnetic memory signal data for the target gear in real time according to the health assessment time window; A health assessment is performed on the target gear according to the health assessment magnetic memory signal data to obtain a health assessment result of the target gear.
2. A gear health assessment method based on magnetic memory signal analysis according to claim 1, characterized in that: The historical working condition parameters of the target gear under the working state are obtained, and the magnetic memory signal data of the target gear under each working condition parameter is obtained based on the control variable method, and calibrated as the first magnetic memory signal data, specifically: Acquire historical operating parameters of the target gear under working conditions, wherein the operating parameters include gear temperature, gear speed, and gear magnetic interference intensity; Determine the historical parameters of each operating condition parameter item according to the historical operating condition parameters, and construct a parameter matrix with the historical parameters of each operating condition parameter item; The magnetic memory signal data of each historical parameter of the target gear in each operating parameter item is obtained by the control variable method according to the parameter matrix and calibrated as the first magnetic memory signal data.
3. The gear health assessment method based on magnetic memory signal analysis according to claim 1 is characterized in that: The acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear is determined according to the first magnetic memory signal data to obtain the acquisition accuracy data, specifically: Acquire a healthy target gear sample, and acquire magnetic memory signal data of the target gear sample under a preset working condition, and calibrate it as second magnetic memory signal data; Sorting the first magnetic memory signal data and the second magnetic memory signal data in time series to obtain first time domain signal data and second time domain signal data; Extracting time domain features of the first time domain signal data and the second time domain signal data, the time domain features including signal mean, variance, root mean square value, and kurtosis; The first time domain signal data and the second time domain signal data are converted into frequency domain signals based on discrete Fourier transform, and frequency domain features of the first magnetic memory signal data and the second magnetic memory signal data are extracted, wherein the frequency domain features include spectrum frequency, frequency center, spectrum energy, and spectrum bandwidth; Calculating the Manhattan distance between the time domain characteristics and frequency domain characteristics of the first magnetic memory signal data and the time domain characteristics and frequency domain characteristics of the second magnetic memory signal data, and determining the influence weight of the time domain characteristics and frequency domain characteristics on the difference evaluation of the magnetic memory signal data according to the information entropy weight method; The difference between the first magnetic memory signal data and the second magnetic memory signal data is determined according to the Manhattan distance and the influence weight, and the acquisition accuracy of the magnetic memory signal data for the operating parameters of different operating items under the target gear working state is evaluated according to the difference to obtain the acquisition accuracy data.
4. The gear health assessment method based on magnetic memory signal analysis according to claim 1 is characterized in that: The reliability evaluation of magnetic memory signal data acquisition is performed on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability evaluation result, which is specifically: Based on the K-means clustering algorithm, K data points are randomly selected from the collection accuracy data as the initial cluster centers. For each data point in the collection accuracy data, the Euclidean distance between each data point and each initial cluster center is calculated, and the data point is assigned to the cluster with the closest initial cluster center. The mean of all data points in the cluster after the data point is assigned is calculated, and the cluster center is updated according to the mean. Iteratively calculate the Euclidean distance of each data point to the new cluster center and assign the data point to the cluster where the nearest cluster center is located, and update the cluster center until the cluster center no longer changes, thus obtaining the data clustering result of collection accuracy; According to the acquisition accuracy data clustering results, a magnetic memory signal data acquisition reliability level evaluation is performed on each clustering result to determine the operating parameter range of each operating condition item in different magnetic memory signal data acquisition reliability levels; Constructing a membership function according to the operating condition parameter range of each operating condition item in different magnetic memory signal data acquisition reliability levels, and constructing a magnetic memory signal data acquisition reliability evaluation model based on the membership function; The operating parameter change data of the target gear within a preset time period is obtained, the operating parameter change data is imported into the reliability assessment model, the reliability of the magnetic memory signal data acquisition within the preset time period is evaluated, and the collection reliability assessment result of the magnetic memory signal data within the preset time period is obtained.
5. The gear health assessment method based on magnetic memory signal analysis according to claim 1 is characterized in that: The health assessment time window of the target gear is determined according to the collected reliability assessment result, and the health assessment magnetic memory signal data of the target gear is acquired in real time according to the health assessment time window, specifically: Sorting the acquisition reliability evaluation results according to the time series to obtain the acquisition reliability time series change data of the magnetic memory signal data within a preset time period; Building an acquisition reliability prediction model based on the LSTM algorithm, obtaining historical acquisition reliability time series change data and importing it into the model for model training, importing the acquisition reliability time series change data into the acquisition reliability prediction model after training, and predicting the acquisition reliability prediction data for magnetic memory signal data collection of the target gear in the future preset time period; Determine a high reliability time window for magnetic memory signal data acquisition within a future preset time period according to the acquisition reliability prediction data, and mark it as a health assessment time window for the target gear; The health assessment magnetic memory signal data of the target gear is acquired in real time in the health assessment time window.
6. The gear health assessment method based on magnetic memory signal analysis according to claim 1 is characterized in that: The health assessment of the target gear is performed according to the health assessment magnetic memory signal data to obtain the health assessment result of the target gear, specifically: Obtaining magnetic memory signal data of target gears with different quality problems, extracting time domain features and frequency domain features of the magnetic memory signal data, and labeling the extracted time domain features and frequency domain features with the types of quality problems to obtain labeled data; Building a target gear health status assessment model based on a decision tree algorithm, and importing the labeled data into the target gear health status assessment model for training; Extracting the time domain features and frequency domain features of the health assessment magnetic memory signal data and marking them as current feature data; Marking the time domain features and frequency domain features of the second magnetic memory signal data as reference data, comparing the current feature data with the reference data, calculating the Euclidean distance between the current feature data and the reference data, and evaluating the difference between the current feature data and the reference data according to the Euclidean distance; A difference threshold is preset, and if the difference between the current characteristic data and the reference data is less than the difference threshold, the target gear is calibrated as a healthy state; If the difference between the current feature data and the benchmark data is greater than the difference threshold, the current feature data is imported into the target gear health status assessment model to identify quality problems, and the health status of the target gear is assessed based on the identified quality problems to obtain a health assessment result of the target gear.
7. A gear health assessment system based on magnetic memory signal analysis, characterized in that: The gear health assessment system based on magnetic memory signal analysis includes a storage device and a processor. The storage device includes a gear health assessment method program based on magnetic memory signal analysis. When the gear health assessment method program based on magnetic memory signal analysis is executed by the processor, the following steps are implemented: Obtain historical operating parameters of the target gear under working conditions, obtain magnetic memory signal data of the target gear under each operating parameter based on a control variable method, and calibrate them as first magnetic memory signal data; Determine the acquisition accuracy of the magnetic memory signal data for the working condition parameters of different working condition items under the working state of the target gear according to the first magnetic memory signal data, and obtain acquisition accuracy data; Performing a magnetic memory signal data acquisition reliability assessment on the operating parameter change data of the target gear within a preset time period according to the acquisition accuracy data to obtain an acquisition reliability assessment result; Determine a health assessment time window for the target gear according to the collected reliability assessment result, and acquire health assessment magnetic memory signal data for the target gear in real time according to the health assessment time window; A health assessment is performed on the target gear according to the health assessment magnetic memory signal data to obtain a health assessment result of the target gear.
