A gear defect detection system and method based on gear pump machining

Abstract

The application discloses a gear defect detection system and method based on gear pump processing, and relates to the technical field of nondestructive testing.The gear defect detection system comprises a gear defect detection platform, which is communicatively connected with a magnetization preprocessing module, a magnetic signal acquisition module, a magnetic signal processing module, a gear defect analysis module and a defect visualization labeling module, wherein electrical signals are connected between the modules; the magnetization preprocessing module is used for customizing a magnetization scheme according to the material and structural characteristics of the gear. The application adopts signal separation and feature extraction technology, can effectively extract independent features of different defect types from the magnetic signals, construct a defect classification model, accurately distinguish surface damage, geometric deformation and fatigue cracks, and quantitatively analyze the severity of the defects, thereby improving the accuracy of defect classification and avoiding classification errors caused by overlapping magnetic signals in traditional methods.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nondestructive testing, in particular to a gear defect detection system and method based on gear pump processing. BACKGROUND

[0002] Gear pump is a common mechanical element in hydraulic system, widely used in automobile, engineering machinery, aerospace, ship, energy and other fields, its main function is to realize the delivery and pressurization of liquid through the meshing and rotation of gear, and the quality and performance of gear have a crucial influence on the overall performance of gear pump, if the gear has defects such as crack, wear, broken tooth, etc., it will directly affect the delivery efficiency, sealing performance and service life of gear pump, therefore, defect detection of gear is an important link to ensure the quality of gear pump.

[0003] For example, the gear defect intelligent detection method and system based on magnetic memory technology in Chinese patent publication No. CN119246666A can detect, locate and identify the stress defects of gear based on magnetic memory technology, so as to provide reliable fault detection basis for gear maintenance and repair, and improve the accuracy and sustainability of gear use and maintenance.

[0004] In the prior art, the stress defects of gear are detected, located and identified based on magnetic memory technology, which solves the problem that it is difficult to express and capture the potential defects and types of gear through the changes of magnetization intensity and magnetic domain structure, but due to the differences in material, structure and working environment of gear, the magnetization characteristics and defect performance are quite different, and the coincidence degree of magnetization intensity change is compared to identify the defect type, the magnetic signals of different defects exist overlapping, which leads to the problem of inaccurate defect classification, therefore, the present application proposes a gear defect detection system and method based on gear pump processing to solve the above problems. SUMMARY

[0005] The present application aims to provide a gear defect detection system and method based on gear pump processing to solve the problems raised in the background.

[0006] To solve the above technical problems, the technical solution adopted by the present application is:

[0007] In a first aspect, a gear defect detection system based on gear pump processing includes a gear defect detection platform, the gear defect detection platform is communicatively connected with a magnetization preprocessing module, a magnetic signal acquisition module, a magnetic signal processing module, a gear defect analysis module and a defect visualization labeling module, wherein the modules are electrically connected;

[0008] The magnetization preprocessing module is used to customize the magnetization scheme according to the material and structural characteristics of the gear;

[0009] The magnetic signal acquisition module is used for acquiring magnetic signals during gear detection by using a high-resolution and high-sensitivity magnetic sensor array, including magnetic signal data of the gear surface and the inside, and improving the detection sensitivity by capturing the magnetic signal changes caused by the defects on the gear surface and the inside.

[0010] The magnetic signal processing module is used for applying signal processing techniques to process different types of noise in the magnetic signal, solving the overlapping problem of the magnetized signal, separating different defect signals, improving the signal-to-noise ratio of the magnetic signal, reducing the influence of noise on the detection result, and improving the clarity and accuracy of the magnetic signal.

[0011] The gear defect analysis module is used for performing feature analysis on the processed magnetic signal, extracting defect features, and separating different defect signals, and then constructing a defect classification model to identify the gear defect type and analyze the defect severity to determine the usability of the gear.

[0012] The defect visualization labeling module is used for visualizing the detected defect position, size, type and usability information, providing a user-friendly interface, and facilitating the user to view, analyze and label the detection result, making the detection result more intuitive and easy to understand, and improving the user experience.

[0013] The further improvement of the technical scheme of the present application is that the gear defect analysis module includes a defect signal separation unit, a defect classification identification unit and a defect severity analysis unit.

[0014] The defect signal separation unit is used for extracting defect features from the processed magnetic signal, and distinguishing different defect types of the magnetic signal through signal analysis, and then separating different defect signals, effectively extracting independent features of different defects from the overlapping magnetic signals, reducing the classification error caused by defect type overlap, and improving the detection accuracy of the system.

[0015] The defect classification identification unit is used for combining the separated defect signal and defect feature to construct a defect classification model, identify the defect type of the gear, and accurately distinguish different types of defects.

[0016] The defect severity analysis unit is used for further analyzing the severity of the defect according to the characteristics and classification results of the defect, evaluating the influence of the defect on the gear of the gear pump, judging the usability of the gear, and identifying the defects that have a greater impact on the performance of the gear pump by quantitatively evaluating the defects of the gear.

[0017] The further improvement of the technical scheme of the present application is that the magnetization preprocessing module specifically includes:

[0018] According to the working characteristics (flow, pressure, speed, etc.) of the gear pump and the expected use environment (temperature, humidity, corrosive medium, etc.), the gear parameters are determined, the gear parameters are input into the system, and the material and structural characteristics of the gear are analyzed;

[0019] Based on the material and structural characteristics of the gear, the magnetization characteristics are analyzed, the required magnetization strength, magnetization direction and pretreatment steps are determined, and the magnetization scheme is customized;

[0020] The demagnetization equipment is used for demagnetizing the gear, and after demagnetization, the magnetization equipment is used for pre-magnetizing the gear, so that the gear has a consistent magnetization state;

[0021] The magnetization state of the pre-magnetized gear is verified, whether the magnetization state of the gear meets the expected requirements is verified, the comparability and stability of the magnetic signal are ensured, and the parameters (magnetization strength, magnetization direction, equipment type, etc.) of the magnetization process are recorded and stored in the system database, and a unique magnetization state identifier is generated for each gear, which is convenient for subsequent detection process tracking and comparison.

[0022] The further improvement of the technical scheme of the application is that the magnetic signal acquisition module specifically comprises:

[0023] According to the size and shape of the gear, the magnetic sensor array is configured, a plurality of magnetic sensors are arranged in an array to ensure that the surface and possible internal defect areas of the gear can be fully covered, and a multi-axis detection function is adopted to capture magnetic field changes in different directions, wherein the magnetic sensor is a tunnel magnetoresistance sensor or a magnetic flux gate sensor;

[0024] Based on the customized magnetization scheme, the gear is magnetized, and the magnetized gear is placed on the detection platform, so that the relative position between the gear and the magnetic sensor array is fixed;

[0025] The magnetic signal generated on the surface and inside of the gear is captured by the magnetic sensor array, wherein the magnetic signal includes the magnetic field generated by the normal structure of the gear and the magnetic field change caused by the defect.

[0026] The further improvement of the technical scheme of the application is that the magnetic signal processing module specifically comprises:

[0027] The original magnetic signal collected is amplified by a differential amplifier to improve the detectability of the signal, and a lock-in amplifier is used to amplify the frequency of the magnetic signal to further enhance the signal strength and improve the detectability of the signal;

[0028] The magnetic signal is subjected to DC offset processing to eliminate baseline drift caused by equipment or environmental factors, the DC offset can be realized by a high-pass filter, the stability of the signal is ensured, and the preprocessed signal is subjected to spectrum analysis, the main noise type and frequency range in the signal are identified, the characteristics of the noise are analyzed, whether it is white noise, periodic noise or random noise is judged, and then according to the result of noise analysis, a corresponding filter is designed to process different types of noise, wherein for periodic noise, a notch filter or band-stop filter is used for suppression, and for random noise, an adaptive filter or Kalman filter method is used for noise reduction;

[0029] The signal separation technology of independent component analysis is applied to separate the overlapping magnetization signals, the statistical independence between signals is utilized to separate different defect signals from the mixed signals, and the separated signals are subjected to feature extraction, the peak value, mean value, variance and frequency statistical characteristics of the signals are calculated;

[0030] A 24-bit high-precision analog-to-digital converter (ADC) is used to convert the amplified and filtered analog signals into digital signals to ensure high resolution of the data, and the collected digital signals are transmitted to the host computer through a communication module for subsequent analysis, and then a unique detection identifier is generated for each gear, the magnetic signal data of the gear is stored, and the identification information, detection parameters (magnetization intensity and direction) and acquisition time stamp of the gear are attached, so as to facilitate subsequent tracing.

[0031] The further improvement of the technical scheme of the application is that the defect signal separation unit specifically comprises:

[0032] According to the working characteristics of the gear pump and the expected use environment, the defect types of the gear are determined, including surface damage defects, geometric deformation defects and fatigue crack defects, and the corresponding defect features are extracted for each type of defect, and the detection reference values of the defect features are determined, wherein the detection reference values are determined according to the design requirements of the gear, the use environment and the industry standards, and are used to evaluate whether the defects exceed the acceptable range;

[0033] For the surface damage defect, the defect features are wear features, pitting features and bonding features, for the geometric deformation defect, the defect features are tooth profile error, axial / radial runout and center distance deviation, and for the fatigue crack defect, the defect features are crack length, crack depth and crack number, and then the detection reference values of the defect features are integrated to obtain a standard defect feature sequence;

[0034] The defect signal separation unit receives the processed magnetic signal data, and performs feature analysis on the magnetic signal data, identifies the frequency components in the signal, distinguishes different types of defect features, and then uses independent component analysis technology to separate different defect signals, decomposes the mixed signal into independent defect signals through the statistical independence between signals, and extracts the independent features of each defect from the separated signals;

[0035] The separated defect features are compared with the standard defect feature sequence, classified into known defect types (surface damage, geometric deformation or fatigue crack), and the separated defect signals are output to the defect classification and recognition unit.

[0036] The further improvement of the technical scheme of the application is that the defect classification and recognition unit specifically comprises:

[0037] Collecting relevant data of gear pump gear defect detection, including different types of defect signals and corresponding defect features, and then extracting gear defect sample data from the gear pump gear defect detection database, and marking the defect features contained in the surface damage defect, geometric deformation defect and fatigue crack defect in the sample data to form a defect data set, which is divided into a training set and a test set, to ensure that the training set and the test set contain various types of defect samples to ensure the generalization ability of the model;

[0038] According to the data characteristics of the gear defect detection, a recurrent neural network is selected as the model architecture of the defect classification model, and the neural network model is trained using the labeled training set data, in the training process, according to the model training performance and verification result, adjust the parameters of the model (learning rate, batch size, network layer number, etc., use regularization technology (Dropout, L2 regularization, etc.) to prevent overfitting, use early stopping method to speed up the training process and improve the model performance, and then use the test set to test the model, evaluate the accuracy and generalization ability of the model, according to the evaluation result, adjust the model structure or parameter, further optimize the model performance;

[0039] The defect classification model identifies and classifies the defect type (surface damage, geometric deformation or fatigue crack) according to the extracted defect features, and then outputs the identified defect type and its position information to the defect severity analysis unit.

[0040] The further improvement of the technical scheme of the application is that the process of identifying and classifying the defect type comprises:

[0041] Integrate the extracted defect features to form a feature vector as the input of the defect classification model, and normalize the input feature vector to ensure that the feature values are within the same dimension range, check the integrity and consistency of the feature vector to ensure that there is no missing value or abnormal value;

[0042] Load the trained defect classification model, initialize the input interface of the model, ensure that the format of the input feature vector is consistent with the model training, and input the constructed feature vector into the defect classification model, the model calculates the probability of the input feature belonging to different defect types according to the input feature vector, and the probability of the input feature belonging to surface damage, geometric deformation or fatigue crack is calculated through the internal neural network structure step by step.

[0043] According to the probability value output by the model, the defect type with the highest probability is selected as the final recognition result, and then the recognized defect type and its related information are arranged into a structured output format and output to the defect severity analysis unit for subsequent defect severity analysis. The output information includes defect type, defect feature specific value, defect position and probability value output by the model.

[0044] The further improvement of the technical scheme of the application is that the defect severity analysis unit specifically comprises:

[0045] Obtain the defect type, defect feature specific value, defect position and probability value output by the defect classification and recognition unit;

[0046] According to the defect type and the specific value of the defect feature, the severity of the defect is quantitatively evaluated, wherein for surface damage defect, the wear depth or area, the area or number of pitting and the size or depth of the bonding area are calculated, for geometric deformation defect, the tooth profile deviation value, the deviation value of the jump and the deviation between the actual value and the design value of the center distance are measured, for fatigue crack defect, the extension length of the crack is measured, the depth of the crack is measured and the number of cracks is counted.

[0047] Compare the quantified defect features with the pre-set defect allowable value to preliminarily judge the severity of the defect, and then comprehensively consider the ratio of all defect features to the defect allowable value to calculate the defect severity coefficient to evaluate the comprehensive severity of the gear defect.

[0048] According to the defect severity analysis result of the gear and the defect size and position, the usability of the gear is judged, and the gears with defects are divided into three types of usable, repairable and unusable, wherein usable means that the defect has little effect on the performance of the gear and does not affect normal operation, repairable means that the defect has a certain effect on the performance of the gear and can be used after repair, and unusable means that the defect is serious and may cause gear failure.

[0049] The further improvement of the technical scheme of the present application is that the defect visualization labeling module specifically comprises:

[0050] The detection result data received from the defect severity analysis unit includes defect position (addendum, dedendum, gear face, etc.), defect size (crack length, wear depth, pitting area, etc.), defect type (surface damage, geometric deformation, fatigue crack), defect severity coefficient, and gear availability judgment (available, needs repair, unavailable);

[0051] A visualization interface is created to display the three-dimensional model of the gear, and the defect position is labeled on the three-dimensional model, with different colors used to distinguish defect types and severity. Red dots are used to represent surface damage defects, blue rectangles are used to represent geometric deformation defects, and green triangles are used to represent fatigue crack defects.

[0052] Magnification, reduction, and rotation tools are provided to allow users to view various parts of the gear. Cross-sectional views are provided to show internal defect conditions of the gear. Users are allowed to label detection results, add annotations, record specific defect conditions or repair suggestions, and save and export functions are provided to save the labeled results as pictures or report files.

[0053] In a second aspect, a gear defect detection method based on gear pump machining is implemented based on the gear defect detection system based on gear pump machining described above, comprising the following steps:

[0054] Step one, customize magnetization scheme according to gear material and structure, perform demagnetization and pre-magnetization to ensure consistent magnetization state;

[0055] Step two, use a high-resolution magnetic sensor array to collect gear surface and internal magnetic signals and capture signal changes caused by defects;

[0056] Step three, apply signal processing techniques to pre-process the collected magnetic signals and separate different defect signals to improve signal-to-noise ratio and clarity;

[0057] Step four, extract defect features from the processed magnetic signals, build a defect classification model, and identify defect types, including surface damage defects, geometric deformation defects, and fatigue crack defects;

[0058] Step five, combine defect features and positions to calculate defect severity coefficients and judge gear availability, which are available, need repair, and unavailable, respectively;

[0059] Step six, create a three-dimensional model of the gear to label defect positions, types, and severity, and provide a user interface to support labeling and report export.

[0060] Due to the adoption of the above technical scheme, the present application has the following technical progress compared with the prior art:

[0061] 1. The present application provides a gear defect detection system and method based on gear pump machining, which adopts signal separation and feature extraction technology, can effectively extract independent features of different defect types from magnetic signals, and construct a defect classification model, thereby accurately distinguishing surface damage, geometric deformation and fatigue cracks, and quantitatively analyzing the severity of defects, not only improving the accuracy of defect classification, but also avoiding classification errors caused by overlapping magnetic signals in traditional methods.

[0062] 2. The present application provides a gear defect detection system and method based on gear pump machining, which can provide consistent and stable magnetization state for gears of different materials and structures through customized magnetization preprocessing scheme, significantly reducing the problem of uneven magnetization caused by material differences, and combining with high-resolution and high-sensitivity magnetic sensor array, the system can accurately capture the tiny defect signal on the surface and inside of the gear, greatly improving the accuracy and reliability of the detection, in addition, through signal processing technology to customize the removal of noise in the magnetic signal, further improving the clarity and accuracy of the magnetic signal.

[0063] 3. The present application provides a gear defect detection system and method based on gear pump machining, which quantitatively evaluates the severity of defects according to the characteristics and classification results of defects, and through the calculation of the size, position and development trend of defects, can quickly judge the usability of the gear, and divide the gear into three types of usable, repairable and unusable, not only improving the evaluation efficiency, but also providing a scientific basis for the machining and production of gear pump, which helps to ensure the service life and maintenance cost of the gear pump. BRIEF DESCRIPTION OF DRAWINGS

[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0065] Figure 1 The system function module diagram of the present application;

[0066] Figure 2 The working flowchart of the defect severity analysis unit of the present application;

[0067] Figure 3 The method flowchart of the present application. DETAILED DESCRIPTION

[0068] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the protection scope of the present application.

[0069] As shown in Embodiment 1, Figure 1 The present application provides a gear defect detection system based on gear pump processing, comprising a gear defect detection platform, the gear defect detection platform is communicatively connected with a magnetization preprocessing module, a magnetic signal acquisition module, a magnetic signal processing module, a gear defect analysis module and a defect visualization labeling module, wherein the modules are electrically connected.

[0070] The magnetization pretreatment module is used to customize magnetization schemes based on the material and structural characteristics of the gears. This includes selecting appropriate magnetization equipment, magnetization intensity, and magnetization direction. Through demagnetization and pre-magnetization steps, it reduces the influence of residual magnetic fields within the material on the test results, ensuring the gears have a consistent magnetization state before testing. This improves the comparability and stability of magnetic signals, reduces magnetization inhomogeneity caused by material differences, and provides a reliable foundation for subsequent testing. The module determines gear parameters based on the operating characteristics of the gear pump (flow rate, pressure, speed, etc.) and the expected operating environment (temperature, humidity, corrosive media, etc.), inputs these parameters into the system, and analyzes the material and structural characteristics of the gears. The gear materials include both ferromagnetic and non-ferromagnetic materials. Gears made of different materials respond differently to magnetization. Structural characteristics include size, shape, number of teeth, tooth pitch, and tooth thickness. Ferromagnetic materials (such as carbon steel and alloy steel) have high permeability and hysteresis characteristics, resulting in a strong response to magnetic fields. Non-ferromagnetic materials (such as stainless steel and aluminum alloys) have weaker magnetization capabilities. Based on the material and structural characteristics of the gears, their magnetization characteristics are analyzed to determine the required magnetization intensity, magnetization direction, and pretreatment steps to customize a magnetization scheme. The selection of magnetization intensity and direction depends on the gear material and the required detection sensitivity. Higher magnetization intensity results in higher detection sensitivity, but the gear's load-bearing capacity and response to magnetization must also be considered. The magnetization direction should be perpendicular to the expected direction of the defect to improve detection accuracy. To ensure accuracy, a demagnetizing device is used to demagnetize the gears. After demagnetization, a magnetizing device is used to pre-magnetize the gears to ensure a consistent magnetization state. The demagnetizing process removes residual magnetic fields from the gears, preventing interference with subsequent magnetization and inspection processes. The gears are placed in the demagnetizing device, starting with a high magnetic field strength and gradually decreasing it to zero while slowly rotating the gear to ensure the magnetic field dissipates evenly. This process is repeated multiple times to minimize the residual magnetic field inside the gears. The demagnetizing device is a coil device with an adjustable magnetic field strength. The pre-magnetizing process applies an initial, uniform magnetization state to the gears, providing a consistent background signal for subsequent defect detection. The specific magnetization process depends on the gear's material and structure. Based on structural characteristics, select appropriate magnetization equipment (AC magnetization equipment, DC magnetization equipment, or pulse magnetization equipment), set magnetization intensity and direction, use circumferential magnetization for axial defect detection and longitudinal magnetization for radial defect detection, and then pre-magnetize the gear to ensure uniform and stable magnetization process. Multiple magnetization and verification are required to ensure consistency of magnetization state. Verify the magnetization state of the gear after pre-magnetization to verify whether the magnetization state of the gear meets the expected requirements, ensure the comparability and stability of magnetic signals, and record and store the magnetization parameters (magnetization intensity, magnetization direction, equipment type, etc.) in the system database. At the same time, generate a unique magnetization state identifier for each gear to facilitate traceability and comparison in subsequent detection processes.

[0071] The magnetic signal acquisition module is used to acquire the magnetic signal of the gear detection by using a high-resolution and high-sensitivity magnetic sensor array, including the magnetic signal data of the gear surface and the internal part, and the sensitivity of the detection is improved by capturing the magnetic signal changes caused by the defects on the gear surface and the internal part. According to the size and shape of the gear, the magnetic sensor array is configured, and multiple magnetic sensors are arranged to form an array to ensure that the surface of the gear and the possible internal defect area can be fully covered, and a multi-axis detection function is adopted to capture the magnetic field changes in different directions. The magnetic sensor is a tunnel magnetoresistance sensor or a fluxgate sensor. The gear is magnetized based on a customized magnetization scheme, and the magnetized gear is placed on the detection platform to fix the relative position between the gear and the magnetic sensor array. The magnetic signal generated by the gear surface and the internal part is captured by the magnetic sensor array. The magnetic signal includes the magnetic field generated by the normal structure of the gear and the magnetic field changes caused by the defects.

[0072] The magnetic signal processing module is used to apply signal processing techniques to customize the processing of different types of noise in the magnetic signal to solve the problem of overlapping magnetized signals, separate different defect signals, improve the signal-to-noise ratio of the magnetic signal, reduce the influence of noise on the detection result, improve the clarity and accuracy of the magnetic signal, and amplify the original magnetic signal collected by a differential amplifier to improve the detectability of the signal. A lock-in amplifier is used to amplify the frequency of the magnetic signal to further enhance the signal strength and improve the detectability of the signal. The magnetic signal is subjected to a direct current offset removal process to eliminate the baseline drift caused by equipment or environmental factors. The direct current offset removal can be achieved by a high-pass filter to ensure the stability of the signal. The preprocessed signal is subjected to frequency spectrum analysis to identify the main noise types and frequency ranges in the signal, analyze the characteristics of the noise, and determine whether it is white noise, periodic noise, or random noise. Based on the results of the noise analysis, a corresponding filter is designed to process different types of noise. For periodic noise, a notch filter or band-stop filter is used for suppression. For random noise, an adaptive filter or Kalman filter method is used for noise reduction. The independent component analysis signal separation technique is applied to separate the overlapping magnetized signals. The statistical independence between signals is used to separate different defect signals from the mixed signals. The separated signals are subjected to feature extraction to calculate the statistical characteristics of the peak value, mean value, variance, and frequency of the signal. A 24-bit high-precision analog-to-digital converter (ADC) is used to convert the amplified and filtered analog signal into a digital signal to ensure high resolution of the data. The collected digital signal is transmitted to the host computer through a communication module for subsequent analysis. A unique detection identifier is generated for each gear. The magnetic signal data of the gear is stored along with the identification information of the gear, the detection parameters (magnetization strength and direction), and the acquisition timestamp for subsequent tracing.

[0073] The gear defect analysis module is used to perform feature analysis on the signal after magnetic signal processing, extract defect features, separate different defect signals, and then build a defect classification model to identify gear defect types. At the same time, it analyzes the severity of defects to determine the usability of the gear.

[0074] The defect visualization and annotation module is used to visualize the location, size, type, and availability information of detected defects. It provides a user-friendly interface, making it easy for users to view, analyze, and annotate the detection results, making the results more intuitive and easy to understand, and improving the user experience.

[0075] Example 2, as Figure 2 As shown, based on Embodiment 1, the present invention provides a technical solution: preferably, the gear defect analysis module includes a defect signal separation unit, a defect classification and identification unit, and a defect severity analysis unit;

[0076] The defect signal separation unit is used to extract defect features from the processed magnetic signals, and to distinguish the magnetic signals of different defect types through signal analysis, so as to separate different defect signals, effectively extract independent features of different defects from the overlapping magnetic signals, reduce the classification error caused by the overlap of defect types, improve the detection accuracy of the system, and determine the defect types of the gear according to the working characteristics of the gear pump and the expected use environment, including surface damage defects, geometric deformation defects and fatigue crack defects, and extract the corresponding defect features for each type of defect, while determining the detection reference value of each defect feature, wherein the detection reference value is determined according to the design requirements of the gear, the use environment and the industry standard, and is used to evaluate whether the defect exceeds the acceptable range. For surface damage defects, the defect features are wear features, pitting features and gluing features. For geometric deformation defects, the defect features are tooth profile error, axial / radial runout and center distance deviation. For fatigue crack defects, the defect features are crack length, crack depth and crack number. Then, the detection reference values of the defect features are integrated to obtain a standard defect feature sequence, wherein the wear feature reflects the wear degree of the gear surface material, the pitting feature represents the small hole-shaped damage on the gear surface caused by fatigue or corrosion, the gluing feature reflects the adhesion damage on the gear surface caused by poor lubrication or overload, the tooth profile error represents the shape deviation of the gear tooth profile, the axial / radial runout represents the position deviation of the gear in the axial or radial direction, the center distance deviation represents the deviation between the actual value and the design value of the gear center distance, the crack length represents the extension length of the crack, the crack depth represents the depth of the crack, and the crack number represents the number of cracks on the gear. The defect signal separation unit receives the processed magnetic signal data and performs feature analysis on the magnetic signal data to identify the frequency components in the signal and distinguish different types of defect features. Then, the independent component analysis technology is used to separate different defect signals, and the mixed signals are decomposed into independent defect signals through the statistical independence between signals. The independent features of each defect are extracted from the separated signals to ensure that the features of different defect signals are clear and distinguishable. Specifically, the implementation steps of the independent component analysis technology are as follows: the collected magnetic signals are subjected to mean removal and whitening processing to make the signals have unit variance and be unrelated to each other; the FastICA (Fast Independent Component Analysis) algorithm is used for fixed-point iteration to maximize the non-Gaussianity of the signal and separate the independent components; statistical features such as peak value, mean value, variance, frequency feature and kurtosis are extracted from the separated independent components; the extracted features are compared with the standard defect feature sequence to be classified into known defect types and the separation result is output, wherein the wear feature extracts the amplitude change of the signal to reflect the wear degree of the surface material, the pitting feature extracts the high-frequency component of the signal to reflect the existence of the small hole-shaped damage, the gluing feature extracts the low-frequency component of the signal to reflect the characteristics of the adhesion damage, and the tooth profile error extracts the shape deviation feature of the signal.The position deviation feature of the axial / radial runout extraction signal, the center distance deviation feature of the center distance deviation extraction signal, the crack feature of the crack length, depth and number extraction signal, such as the magnetic field distortion caused by the crack, the separated defect feature is compared with the standard defect feature sequence, classified into known defect types (surface damage, geometric deformation or fatigue crack), and the separated defect signal is output to the defect classification and identification unit;

[0077] The defect classification and identification unit is used to combine the separated defect signal and the defect feature, construct a defect classification model, identify the defect type of the gear, accurately distinguish different types of defects, collect related data of gear pump gear defect detection, including different types of defect signals and corresponding defect features, and then extract gear defect sample data from the gear pump gear defect detection database, and mark the defect features contained in the surface damage defect, geometric deformation defect and fatigue crack defect in the sample data, form a defect data set, divide it into a training set and a test set, ensure that the training set and the test set contain various types of defect samples, to ensure the generalization ability of the model, according to the data characteristics of the gear defect detection, select the recurrent neural network as the model architecture of the defect classification model, and use the labeled training set data to train the neural network model, in the training process, according to the model training performance and verification result, adjust the parameters of the model (learning rate, batch size, network layer number, etc.), use regularization techniques (Dropout, L2 regularization, etc.) to prevent overfitting, use early stopping method to accelerate the training process and improve the model performance, then use the test set to test the model, evaluate the accuracy and generalization ability of the model, according to the evaluation result, adjust the model structure or parameter, further optimize the model performance, input the defect features contained in the current separated defect signal to the trained defect classification model, the defect classification model identifies and classifies the defect type (surface damage, geometric deformation or fatigue crack) according to the extracted defect feature, and then outputs the identified defect type and its position information to the defect severity analysis unit;

[0078] In addition, the process of identifying and classifying the defect type includes:

[0079] The extracted defect features are integrated to form a feature vector, which is used as the input of the defect classification model. The input feature vector is normalized to ensure that the feature values are within the same dimension range. The integrity and consistency of the feature vector are checked to ensure that there are no missing values or outliers. A trained defect classification model is loaded, and the input interface of the model is initialized to ensure that the format of the input feature vector is consistent with the model training. The constructed feature vector is input into the defect classification model. The model calculates the probability of the input feature belonging to different defect types by gradually extracting and integrating features through its internal neural network structure. The probability of the input feature belonging to surface damage, geometric deformation, or fatigue crack is calculated. Based on the probability value output by the model, the defect type with the highest probability is selected as the final recognition result. The recognized defect type and related information are then organized into a structured output format and output to the defect severity analysis unit for subsequent defect severity analysis. The output information includes the defect type, specific values of the defect features, defect location, and probability value output by the model.

[0080] The expression of the probability of the input feature belonging to different defect types is as follows:

[0081] ;

[0082] ;

[0083] In the formula, is the probability of the input feature belonging to different defect types, is the input feature vector, which contains the extracted defect features, , is the defect type (surface damage, geometric deformation, or fatigue crack), is the index, which represents the traversal of all defect types in the summation function, is the score function of the model for the th defect type, which represents the matching degree of the input feature vector and the th defect type, is the total number of defect types, is the weight coefficient of the th defect type, which represents the importance of the defect type, is the detection reference value of the th feature, which is used to normalize the feature value , is the bias term of the th defect type, which is used to adjust the baseline of the model, is the dimension (number of features) of the input feature vector.

[0084] ;

[0085] wherein represents the probability value of the model output for each defect type , and the defect type with the highest probability is selected as the recognition result;

[0086] a defect severity analysis unit configured to further analyze the severity of the defect based on the characteristics and classification result of the defect, evaluate the influence of the defect on the gear of the gear pump, and determine the usability of the gear by quantitatively evaluating the defect of the gear, identifying the defect that has a greater influence on the performance of the gear pump, obtaining the defect type, specific numerical value of the defect characteristics, defect position, and probability value output by the defect classification and recognition unit from the defect classification and recognition unit, quantitatively evaluating the severity of the defect based on the defect type and specific numerical value of the defect characteristics, wherein for surface damage defects, calculating the wear depth or area, the area or number of pitting, and the size or depth of the bonding area, for geometric deformation defects, measuring the deviation value of the tooth profile, the deviation value of the runout, and the deviation between the actual value and the design value of the center distance, for fatigue crack defects, measuring the extension length of the crack, measuring the depth of the crack, and counting the number of cracks, comparing the quantified defect characteristics with the pre-set defect allowable value to preliminarily determine the severity of the defect, and then comprehensively calculating the defect severity coefficient by comprehensively evaluating the overall severity of the gear defect based on the ratio of all defect characteristics to the defect allowable value, determining the usability of the gear based on the defect severity analysis result and the size and position of the defect, and dividing the gear with defects into three types: usable, repairable, and unusable, wherein usable means that the defect has little influence on the performance of the gear and does not affect normal operation, repairable means that the defect has some influence on the performance of the gear and can be used after repair, and unusable means that the defect is severe and may cause gear failure;

[0087] The expression of the defect severity coefficient is:

[0088] ;

[0089] wherein, is the defect severity coefficient for evaluating the overall severity of the gear defect, is the quantitative value of the th defect characteristic, is the allowable value of the th defect characteristic, is the total number of defect characteristics, is an adjustment coefficient for adjusting the sensitivity of the defect severity coefficient, is a natural exponential function for mapping the weighted sum of defect characteristics to the range (0, 1), the value range of is (0, 1), and when Close to 0, indicating that the defect severity is lower, the gear state is better, when Close to 1, indicating that the defect severity is higher, the gear state is worse, when all are much smaller than their allowable values , Close to 0, resulting in Close to 0, when some Close to or exceed their allowable values , Increase, resulting in Increase, the value of the adjustment coefficient The greater the defect severity coefficient is more sensitive to the change of the defect characteristic value, which can be used: the defect severity coefficient is lower, , the defect position is not in the key area (such as the dedendum, addendum, etc.), the defect size is much smaller than the allowable value ( ), need to be repaired: the defect severity coefficient is moderate, , the defect position is in the non-key area, but the defect size is close to the allowable value ( ), or the defect position is in the key area, but the defect size does not exceed the allowable value ( ), not available: the defect severity coefficient is higher, , the defect position is in the key area (such as the dedendum, addendum, etc.), the defect size exceeds the allowable value ( );

[0090] The defect visualization labeling module specifically includes:

[0091] Receive detection result data from the defect severity analysis unit, including: defect position (addendum, dedendum, gear face, etc.), defect size (crack length, wear depth, pitting area, etc.), defect type (surface damage, geometric deformation, fatigue crack), defect severity coefficient, and gear availability judgment (available, need to be repaired, not available), create a visualization interface, display the three-dimensional model of the gear, label the defect position on the three-dimensional model, use different colors to distinguish defect types and severity, wherein red dots represent surface damage defects, blue rectangles represent geometric deformation defects, and green triangles represent fatigue crack defects, provide zoom in, zoom out and rotation tools to allow users to view various parts of the gear, provide a cross-sectional view to show the internal defect situation of the gear, allow users to label the detection results, add annotations, record the specific situation or repair suggestions of the defect, and provide save and export functions to save the labeled results as pictures or report files.

[0092] As shown in Figure 3 embodiment 3, based on embodiments 1-2, the application also provides a gear defect detection method based on gear pump machining, which is realized based on the above-mentioned gear defect detection system based on gear pump machining, including the following steps:

[0093] Step one, customize magnetization scheme according to gear material and structure, demagnetization and pre-magnetization, ensure consistent magnetization state;

[0094] Step two, use high-resolution magnetic sensor array to collect gear surface and internal magnetic signal, capture signal changes caused by defects;

[0095] Step three, apply signal processing technology to preprocess the collected magnetic signal, separate different defect signals, and improve signal-to-noise ratio and clarity;

[0096] Step four, extract defect features from processed magnetic signals, build defect classification model, identify defect types, including surface damage defects, geometric deformation defects and fatigue crack defects;

[0097] Step five, combine defect features and location, calculate defect severity coefficient, and judge gear usability, respectively available, need to repair and unavailable;

[0098] Step six, create a three-dimensional model of the gear to mark the location, type and severity of the defect, provide a user interface, support labeling and report export.

[0099] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this, any skilled person in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A gear defect detection system based on gear pump machining, comprising a gear defect detection platform, characterized in that: The gear defect detection platform is communicatively connected with a magnetization preprocessing module, a magnetic signal acquisition module, a magnetic signal processing module, a gear defect analysis module, and a defect visualization labeling module, wherein the modules are electrically connected; The magnetization preprocessing module is configured to customize a magnetization scheme according to the material and structural characteristics of the gear; The magnetic signal acquisition module is configured to acquire magnetic signals of the gear during detection by using a magnetic sensor array, including magnetic signal data of the gear surface and inside; The magnetic signal processing module is configured to apply signal processing techniques to customize the processing of different types of noise in the magnetic signals; The gear defect analysis module is configured to analyze the features of the processed magnetic signals, extract defect features, separate different defect signals, and further construct a defect classification model to identify the gear defect type and analyze the defect severity to determine the usability of the gear. The defect signal separation unit is configured to extract defect features from the processed magnetic signals and distinguish different defect types by signal analysis, and further separate different defect signals. The defect classification identification unit is configured to construct a defect classification model by combining the separated defect signals and defect features to identify the defect type of the gear. The defect severity analysis unit is configured to further analyze the severity of the defect based on the defect features and classification results, evaluate the impact of the defect on the gear of the gear pump, and determine the usability of the gear. The defect visualization labeling module is configured to visually display the detected defect position, size, type, and usability information. The magnetization preprocessing module specifically includes: Determine the gear parameters based on the working characteristics of the gear pump and the expected use environment, input the gear parameters into the system, and analyze the material and structural characteristics of the gear; Based on the material and structural characteristics of the gear, analyze its magnetization characteristics, determine the required magnetization strength, magnetization direction, and preprocessing steps to customize the magnetization scheme; Use demagnetization equipment to demagnetize the gear, and after demagnetization, use magnetization equipment to pre-magnetize the gear to have a consistent magnetization state; Verify the magnetization state of the pre-magnetized gear to determine whether the magnetization state meets the expected requirements, record and store the magnetization parameters in the system database, and generate a unique magnetization state identifier for each gear.

2. A gear defect detection system based on gear pump machining according to claim 1, characterized in that: The magnetic signal acquisition module specifically includes: According to the size and shape of the gear, configure a magnetic sensor array, layout multiple magnetic sensors to form an array, and use multi-axis detection function to capture magnetic field changes in different directions, wherein the magnetic sensor is a tunnel magnetoresistance sensor or a magnetic flux gate sensor; Based on the customized magnetization scheme, magnetize the gear and place the magnetized gear on the detection platform to fix the relative position between the gear and the magnetic sensor array; Use the magnetic sensor array to capture the magnetic signals generated on the surface and inside of the gear, wherein the magnetic signals include the magnetic field generated by the normal structure of the gear and the magnetic field changes caused by defects.

3. A gear defect detection system based on gear pump machining according to claim 2, characterized in that: The magnetic signal processing module specifically comprises: The collected original magnetic signal is amplified by a differential amplifier, and a phase-locked amplifier is used for frequency modulation amplification processing of the magnetic signal, further enhancing the signal strength; The magnetic signal is subjected to direct current offset processing to eliminate baseline drift caused by equipment or environmental factors, and the preprocessed signal is subjected to frequency spectrum analysis to identify the main noise type and frequency range in the signal, analyze the characteristics of the noise, determine whether it is white noise, periodic noise or random noise, and then design a corresponding filter to process different types of noise according to the results of the noise analysis; The signal separation technology of independent component analysis is applied to separate the overlapping magnetization signals, and the statistical independence between signals is used to separate different defect signals from the mixed signals, and the separated signals are subjected to feature extraction to calculate the statistical characteristics of the peak value, mean value, variance and frequency of the signals; A 24-bit high-precision analog-to-digital converter is used to convert the amplified and filtered analog signal into a digital signal, and the collected digital signal is transmitted to the upper computer through a communication module, and then a unique detection identifier is generated for each gear, the magnetic signal data of the gear is stored, and the identification information, detection parameters and collection time stamp of the gear are attached.

4. The gear defect detection system based on gear pump machining according to claim 1, characterized in that: The defect signal separation unit specifically comprises: According to the working characteristics of the gear pump and the expected use environment, the defect types of the gear are determined, including surface damage defects, geometric deformation defects and fatigue crack defects, and the corresponding defect features are extracted for each type of defect, and the detection reference values of the defect features are determined; For surface damage defects, the defect features are wear features, pitting features and bonding features, for geometric deformation defects, the defect features are tooth profile error, axial / radial runout and center distance deviation, and for fatigue crack defects, the defect features are crack length, crack depth and crack number, and then the detection reference values of the defect features are integrated to obtain a standard defect feature sequence; The defect signal separation unit receives the processed magnetic signal data and performs feature analysis on the magnetic signal data to identify the frequency components in the signal and distinguish different types of defect features, and then uses independent component analysis technology to separate different defect signals and decompose the mixed signal into independent defect signals, and extracts the independent features of each defect from the separated signals; The separated defect features are compared with the standard defect feature sequence, classified into known defect types, and the separated defect signals are output to the defect classification and recognition unit.

5. A gear defect detection system based on gear pump machining as claimed in claim 4, wherein: The defect classification and recognition unit specifically comprises: Collecting relevant data for gear pump gear defect detection, including different types of defect signals and corresponding defect features, and then extracting gear defect sample data from the gear pump gear defect detection database and marking the defect features contained in the surface damage defects, geometric deformation defects and fatigue crack defects in the sample data to form a defect data set, which is divided into a training set and a test set; According to the data characteristics of gear defect detection, a recurrent neural network is selected as the model architecture of the defect classification model, and the neural network model is trained using the labeled training set data. During the training process, the model parameters are adjusted according to the model training performance and validation results. Then, the model is tested using the test set to further optimize the model performance. The defect features contained in the current separated defect signal are input into the trained defect classification model. The defect classification model identifies and classifies the defect type based on the extracted defect features, and then outputs the identified defect type and its location information to the defect severity analysis unit.

6. A gear defect detection system based on gear pump machining as claimed in claim 5, wherein: The process of identifying and classifying the defect type includes: Integrate the extracted defect features to form a feature vector as the input of the defect classification model, and perform normalization processing on the input feature vector. Load the trained defect classification model, initialize the input interface of the model, and input the constructed feature vector into the defect classification model. The model calculates the probability of the input feature belonging to different defect types by gradually extracting and integrating features through its internal neural network structure. The probability represents the probability of the input feature belonging to surface damage, geometric deformation, or fatigue crack. According to the probability value output by the model, select the defect type with the highest probability as the final recognition result. Then, the identified defect type and its related information are arranged into a structured output format and output to the defect severity analysis unit for subsequent defect severity analysis. The output information includes defect type, specific values of defect features, defect location, and probability value output by the model.

7. A gear defect detection system based on gear pump machining according to claim 6, characterized in that: The defect severity analysis unit specifically includes: Obtain the defect type, specific values of defect features, defect location, and probability value output by the defect classification model from the defect classification and recognition unit. According to the defect type and specific values of defect features, quantitatively evaluate the severity of the defect. For surface damage defects, calculate the wear depth or area, the area or number of pitting, and the size or depth of the bonding area. For geometric deformation defects, measure the tooth profile deviation value, the deviation value of the jump, and the deviation between the actual value and the design value of the center distance. For fatigue crack defects, measure the crack extension length, measure the crack depth, and count the number of cracks. Compare the quantified defect features with the pre-set defect allowable values to preliminarily judge the severity of the defect. Then, integrate the ratios of all defect features and defect allowable values to calculate the defect severity coefficient, and evaluate the comprehensive severity of the gear defect. According to the results of gear defect severity analysis and defect size and location, judge the usability of the gear, and divide the gear with defects into three types: usable, repairable, and unusable.

8. A gear defect detection method based on gear pump machining, realized based on the gear defect detection system based on gear pump machining according to any one of claims 1-7, characterized in that, The steps include: Step 1: Customize the magnetization scheme according to the gear material and structure, and perform demagnetization and pre-magnetization. Step 2: Use a high-resolution magnetic sensor array to collect the magnetic signals on the surface and inside of the gear, and capture the signal changes caused by defects. Step 3: Apply signal processing techniques to preprocess the collected magnetic signals and separate different defect signals. Step four, extract defect features from the processed magnetic signals, build a defect classification model, identify the defect type, including surface damage defects, geometric deformation defects and fatigue crack defects; Step five, combine defect features and positions to calculate defect severity coefficients, and judge the gear usability, which are available, need to be repaired and unavailable respectively; Step six, create a three-dimensional model of the gear to mark the defect position, type and severity, and provide a user interface to support labeling and report export.

Images