A method and system for detecting fatigue damage of precision hardware

Abstract

The application discloses a kind of precision hardware fatigue damage detection method and system, it is related to precision manufacturing technical field, method includes: obtaining the theoretical magnetic memory signal of reference block;The reference block is measured to magnetic memory signal, and reference measurement signal is obtained;According to the reference measurement signal and the theoretical magnetic memory signal, adjustment factor reflecting overall measurement deviation is calculated;According to the adjustment factor, initial magnetic memory signal is corrected, and target magnetic memory signal is obtained, and the initial magnetic memory signal is measured to obtain by magnetic memory signal to the precision hardware of test;According to the target magnetic memory signal, fatigue damage assessment processing is carried out to the precision hardware of test, and fatigue damage assessment result is obtained.The application realizes fatigue damage detection, and improves accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of precision manufacturing, and particularly relates to a precision hardware fatigue damage detection method and system. BACKGROUND

[0002] When a precision hardware part bears alternating loads, fatigue damage will accumulate in the internal material, which may lead to structural failure. As a non-contact and non-destructive detection method, the magnetic memory fatigue damage detection method is widely used to evaluate the service state of these hardware parts. This method captures the weak magnetic leakage signals caused by changes in the internal microstructure of the material to determine the degree of fatigue damage. However, in actual industrial applications, the detection system may face electrical interference from the internal power supply unit. These interference signals sometimes show similarities in waveform with the real fatigue damage signals, leading to system misjudgment and low detection accuracy.

[0003] Therefore, the technical problems in the related art need to be improved. SUMMARY

[0004] The main purpose of the embodiments of the present application is to provide a precision hardware fatigue damage detection method and system, which can correct the magnetic memory signal by adjusting the factor to realize fatigue damage detection and improve the accuracy.

[0005] In one aspect, the embodiments of the present application provide a precision hardware fatigue damage detection method, comprising the following steps:

[0006] obtaining a theoretical magnetic memory signal of a reference block;

[0007] measuring the magnetic memory signal of the reference block to obtain a reference measurement signal;

[0008] calculating an adjustment factor reflecting the overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal;

[0009] correcting an initial magnetic memory signal according to the adjustment factor to obtain a target magnetic memory signal, wherein the initial magnetic memory signal is obtained by measuring the magnetic memory signal of a precision hardware part to be measured;

[0010] performing fatigue damage evaluation processing on the precision hardware part to be measured according to the target magnetic memory signal to obtain a fatigue damage evaluation result.

[0011] In some embodiments, the adjustment factor reflecting the overall measurement deviation is calculated according to the reference measurement signal and the theoretical magnetic memory signal, comprising:

[0012] extracting a first spatial distribution of the theoretical magnetic memory signal and a second spatial distribution of the reference measurement signal;

[0013] comparing the first spatial distribution and the second spatial distribution to determine a spatial deviation;

[0014] spatially correcting the reference measurement signal according to the spatial deviation;

[0015] calculating the adjustment factor according to a theoretical magnetic memory signal and the spatially corrected reference measurement signal.

[0016] In some embodiments, the fatigue damage assessment processing of the precision hardware under test according to the target magnetic memory signal comprises:

[0017] generating a load application instruction for applying a local gradually changing mechanical load to an evaluation region of the precision hardware under test;

[0018] after applying the local gradually changing mechanical load, collecting a magnetic memory signal of the evaluation region to obtain initial load and signal response data;

[0019] performing feature analysis on the initial load and signal response data to obtain a response feature representing a material fatigue accumulation pattern;

[0020] performing evaluation on the precision hardware under test according to the target magnetic memory signal, the response feature, and a preset judgment criterion to obtain the fatigue damage assessment result.

[0021] In some embodiments, the feature analysis on the initial load and signal response data to obtain a response feature representing a material fatigue accumulation pattern comprises:

[0022] after the local gradually changing mechanical load is removed, collecting a magnetic memory signal of the evaluation region to obtain residual magnetic memory signal data;

[0023] generating target load and signal response data according to the residual magnetic memory signal data and the initial load and signal response data;

[0024] extracting the response feature from the target load and signal response data.

[0025] In some embodiments, the generation of target load and signal response data according to the residual magnetic memory signal data and the initial load and signal response data comprises:

[0026] performing component analysis on the residual magnetic memory signal data to identify interference components introduced by transient mechanical vibration or local magnetic field disturbance, the interference components including non-fatigue noise or interference signals;

[0027] According to the interference component, the initial load and signal response data are deinterfered to obtain target load and signal response data.

[0028] In some embodiments, the component analysis on the residual magnetic memory signal data to identify the interference component introduced by transient mechanical vibration or local magnetic field disturbance comprises:

[0029] A reference region signal data is obtained by collecting magnetic memory signals in a reference region adjacent to the to-be-evaluated region and free of fatigue damage;

[0030] According to the reference region signal data, an interference characteristic parameter is determined;

[0031] The interference characteristic parameter is compared with the residual magnetic memory signal data to obtain a first comparison result;

[0032] According to the first comparison result, the interference component is identified from the residual magnetic memory signal data.

[0033] In some embodiments, the interference component is identified from the residual magnetic memory signal data according to the first comparison result, comprising:

[0034] A vibration application instruction is generated, which is used to apply mechanical vibration of a preset amplitude and frequency to the to-be-evaluated region;

[0035] After applying mechanical vibration, a dynamic response magnetic memory signal data is obtained by collecting magnetic memory signals in the to-be-evaluated region;

[0036] The dynamic response magnetic memory signal data is processed by feature analysis to obtain a magnetic memory signal modulation feature synchronized with the mechanical vibration, which is used to represent the force-magnetic coupling response of the material;

[0037] According to the first comparison result and the magnetic memory signal modulation feature, the interference component is identified from the residual magnetic memory signal data.

[0038] In some embodiments, the dynamic response magnetic memory signal data is processed by feature analysis to obtain a magnetic memory signal modulation feature synchronized with the mechanical vibration, comprising:

[0039] The dynamic response magnetic memory signal data is analyzed by response analysis to obtain a nonlinear response under different mechanical vibration intensities;

[0040] According to the nonlinear response, the dynamic response magnetic memory signal data is analyzed by frequency spectrum analysis to extract a target harmonic component of the mechanical vibration frequency, the target harmonic component comprising a second harmonic component or a third harmonic component;

[0041] modulation feature of the magnetic memory signal.

[0042] In some embodiments, the identifying the interference component from the residual magnetic memory signal data according to the first comparison result and the modulation feature of the magnetic memory signal comprises:

[0043] extracting a fatigue damage feature from the residual magnetic memory signal data according to the first comparison result;

[0044] comparing the fatigue damage feature with the modulation feature of the magnetic memory signal to obtain a second comparison result;

[0045] according to the second comparison result, taking a signal component without the modulation feature of the magnetic memory signal in the residual magnetic memory signal data as the interference component.

[0046] In another aspect, an embodiment of the present application provides a fatigue damage detection system for precision hardware, comprising:

[0047] a reference theoretical feature acquisition module configured to acquire a theoretical magnetic memory signal of a reference block;

[0048] a reference measurement module configured to perform magnetic memory signal measurement on the reference block to obtain a reference measurement signal;

[0049] a deviation calculation module configured to calculate an adjustment factor reflecting overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal;

[0050] a signal correction module configured to correct an initial magnetic memory signal according to the adjustment factor to obtain a target magnetic memory signal, the initial magnetic memory signal being obtained by performing magnetic memory signal measurement on a to-be-tested precision hardware;

[0051] a fatigue damage evaluation module configured to perform fatigue damage evaluation processing on the to-be-tested precision hardware according to the target magnetic memory signal to obtain a fatigue damage evaluation result.

[0052] The embodiments of the present application have at least the following beneficial effects: the embodiments of the present application first acquire a theoretical magnetic memory signal of a reference block, perform magnetic memory signal measurement on the reference block to obtain a reference measurement signal, then calculate an adjustment factor reflecting overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal, correct an initial magnetic memory signal according to the adjustment factor to obtain a target magnetic memory signal, and finally perform fatigue damage evaluation processing on a to-be-tested precision hardware according to the target magnetic memory signal to obtain a fatigue damage evaluation result, so that the fatigue damage detection can be realized by correcting the magnetic memory signal through the adjustment factor, and the accuracy is improved.

[0053] Other features and advantages of the present application will be set forth in the description that follows, and in part will be apparent from the description, or can be learned by practice of the application. The purposes and other advantages of the present application will be realized and attained by the structure particularly pointed out in the description and appended claims. BRIEF DESCRIPTION OF DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

[0055] Figure 1 A flow chart of a precision hardware fatigue damage detection method according to an embodiment of the present application;

[0056] Figure 2 A structural schematic diagram of a precision hardware fatigue damage detection system according to an embodiment of the present application. DETAILED DESCRIPTION

[0057] In order to make the purposes, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated.

[0058] Before the embodiments of the present application are described in detail, first, some nouns and terms involved in the embodiments of the present application will be described, and the nouns and terms involved in the embodiments of the present application are applicable to the following explanations.

[0059] Fatigue damage: refers to damage accumulation in the process of cyclic loading. Generally divided into high cycle fatigue and low cycle fatigue. In low cycle fatigue, the plastic strain is large enough to measure the stress corresponding to the yield limit, and the cycle number is ≤10000 times. High cycle fatigue refers to the unrecoverable strain in the fatigue process, which is only composed of micro-plasticity, the stress is low, generally lower than the yield limit, but the cycle number is high, generally ≥100000 times.

[0060] In the related art, the traditional existing precise hardware magnetic memory fatigue damage detection system has the problem of misjudgment when detecting, because the electrical interference signal generated by the internal power supply unit is similar to the real fatigue damage signal waveform. Further, when the system triggers the operation process due to misjudgment and unexpectedly increases the load of the internal components, it may cause a chain failure, such as burning of the power supply unit, and further cause hidden damage to the key calibration data of the system, so that the subsequent detection results have persistent system deviation, and finally the real fatigue damage cannot be accurately identified.

[0061] For example, assuming that in an industrial detection workshop, a magnetic memory fatigue damage detection system for detecting aircraft engine blade connecting pieces is running. The system is powered by a separate power supply unit, and is equipped with a magnetic field sensor and a data acquisition card, wherein the data acquisition card internally stores a calibration parameter table. During a detection process, the cooling fan inside the power supply unit has a reduced speed due to bearing grease deterioration, causing the power supply unit to be in an overheated state for a long time. This overheating causes the output voltage stability of the voltage stabilizing circuit to decrease, generating a small voltage ripple of a specific frequency. This ripple is conducted to the high-gain amplification circuit at the front end of the magnetic field sensor through the connecting cable, and after being amplified, it forms a stable internal interference signal. The frequency of the interference signal is just outside the main frequency band of the existing signal filtering rules of the system, so it enters the subsequent data processing link without being filtered. Since the form of the interference signal in the time domain is similar to the magnetic signal mutation characteristics generated by the early micro-crack tip stress concentration area of the material, the feature extraction logic module mistakenly identifies it as a fatigue damage point and issues a damage warning. After receiving the warning, the user manually increases the excitation current of the magnetizing coil according to the operation procedure for review. This operation unexpectedly increases the load of the power supply unit, causing its internal components to burn due to overheating, and generating a high-voltage voltage spike pulse at the moment of burning. This voltage spike impacts the data acquisition card, causing some data bits in the storage unit used to store the calibration parameter table to flip, and the calibration parameter table is damaged. After the system is replaced and restarted, the damaged calibration parameter table is loaded, causing all subsequent collected measurement data to be introduced into a fixed, nonlinear systematic deviation. This deviation continuously masks the weak magnetic memory fatigue damage signal caused by alternating load, making the system unable to effectively identify the early deterioration of the internal material of the hardware.

[0062] If the above problems are not solved, the early fatigue damage of precision hardware parts cannot be accurately identified and evaluated. This will directly lead to the loss of the core function of the detection system, and the reliable basis for the service state of the key components cannot be provided. In the long run, this may cause the hardware parts with potential fatigue damage to continue to serve, thereby increasing the risk of structural failure, and having a negative impact on the safety and economic benefits of equipment operation. In addition, due to the concealment of systematic bias, users may unknowingly rely on inaccurate detection results, further exacerbating the risk.

[0063] When facing the above problems, the first thought of the present application is to remove the internal interference signal by improving the signal filtering algorithm, and to regularly calibrate the system to correct the errors of the calibration parameter table. However, the waveform of the internal interference signal is similar to the real damage signal, and its frequency may avoid the conventional filtering frequency band, so that simple filtering is difficult to distinguish and effectively remove. At the same time, the damage of the calibration parameter table is hidden, and may cause systematic bias. Regular calibration may not completely eliminate this bias, and even may introduce new errors in the calibration process, because the calibration itself also depends on the measurement ability of the system. For this, the present application further thinks that the system has a difficult-to-detect, continuous measurement bias, a reference with a known state can be introduced, and the bias can be quantified and corrected in real time or quasi-real time by measuring the reference. This idea can avoid direct repair or rely on possibly damaged calibration data, but correct the bias of the whole measurement link through external reference. By actually measuring the reference block of the known theoretical signal, and comparing it with the theoretical signal, an adjustment factor reflecting the overall measurement bias of the current system can be calculated. This adjustment factor can comprehensively reflect all systematic biases including calibration data damage, internal interference and other environmental factors. Then, the adjustment factor is applied to the initial measurement signal of the hardware part to be measured, so as to obtain a corrected target magnetic memory signal closer to the real state, and then accurate fatigue damage evaluation is carried out.

[0064] The embodiments of the present application will be specifically explained in combination with the drawings:

[0065] Figure 1 is an optional flowchart of a precision hardware fatigue damage detection method provided by an embodiment of the present application, Figure 1 The method in can include but is not limited to steps S101 to S105.

[0066] Step S101, acquiring a theoretical magnetic memory signal of a reference block;

[0067] Step S102, measuring the magnetic memory signal of the reference block to obtain a reference measurement signal;

[0068] Step S103, calculating an adjustment factor reflecting overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal;

[0069] Step S104, correcting the initial magnetic memory signal to obtain a target magnetic memory signal according to the adjustment factor, the initial magnetic memory signal being obtained by measuring the magnetic memory signal of the precision hardware piece to be measured;

[0070] Step S105, performing fatigue damage evaluation processing on the precision hardware piece to be measured according to the target magnetic memory signal to obtain a fatigue damage evaluation result.

[0071] The steps S101 to S105 shown in the embodiments of the present application can correct the magnetic memory signal by the adjustment factor to realize fatigue damage detection, thereby improving the accuracy.

[0072] In some embodiments, steps S101 to S105 introduce a reference block as a calibration reference, and based on the difference between the theoretical magnetic memory signal and the actual measurement signal, dynamically calculate and apply the adjustment factor to real-time correct the initial magnetic memory signal of the precision hardware piece to be measured, effectively overcoming the measurement deviation caused by internal interference or damaged calibration data of the existing system, and significantly improving the accuracy and reliability of fatigue damage evaluation.

[0073] The theoretical magnetic memory signal of the reference block can be obtained first, which serves as an ideal, unbiased reference basis for subsequent deviation identification. The magnetic memory signal of the reference block is measured to obtain a reference measurement signal, which contains all systematic deviations and environmental influences that the current detection system may introduce in actual operation. Then, an adjustment factor reflecting overall measurement deviation is calculated according to the reference measurement signal and the theoretical magnetic memory signal. The adjustment factor can quantify and characterize the current systematic error of the detection system, such as the deviation caused by internal interference or damaged calibration data. Then, the initial magnetic memory signal is corrected according to the adjustment factor to obtain a target magnetic memory signal, thereby effectively eliminating the systematic error, so that the target magnetic memory signal can more truly reflect the actual magnetic memory state of the precision hardware piece to be measured. The initial magnetic memory signal is obtained by measuring the magnetic memory signal of the precision hardware piece to be measured. Finally, the precision hardware piece to be measured is subjected to fatigue damage evaluation processing according to the target magnetic memory signal to obtain a fatigue damage evaluation result, thereby obtaining an accurate and reliable fatigue damage evaluation result. This dynamic calibration process ensures that even in the case of damaged system calibration data or internal interference, high-precision detection data can be continuously provided, thereby effectively identifying early fatigue damage.

[0074] It can be understood that the reference block refers to a reference object with known and stable magnetic properties, which is usually made of the same or similar material as the precision hardware to be tested and is in a non-damaged state, and its purpose is to provide a stable and repeatable calibration reference standard. The adjustment factor refers to one or a set of parameters calculated based on the difference between the theoretical magnetic memory signal and the reference measurement signal, which quantifies the overall measurement deviation of the detection system, which can be a simple scaling factor, offset or more complex conversion function, and its purpose is to correct subsequent measurement data.

[0075] In order to more clearly illustrate the technical scheme, specific examples are used in the following explanation. First, the theoretical magnetic memory signal of a known non-damaged reference block can be calculated in advance by a high-precision simulation model or measured in a strictly controlled laboratory environment, thereby obtaining the theoretical magnetic memory signal and storing it in the database of the detection system. For example, the reference block can be a non-damaged standard part similar in material and geometry to the precision hardware to be tested. Before actual detection or during periodic calibration, the reference block is placed under the sensor of the detection system for magnetic memory signal measurement to obtain the reference measurement signal. Then, the data processing unit, such as a high-performance industrial computer, receives the theoretical magnetic memory signal and the reference measurement signal. The data processing unit can run a pre-set algorithm, such as a least squares method or a neural network model, to calculate an adjustment factor reflecting the overall measurement deviation of the current system. This adjustment factor can be a linear or nonlinear mapping function. Subsequently, when the precision hardware to be tested is measured for magnetic memory signal, the obtained initial magnetic memory signal is transmitted to the data processing unit. The data processing unit uses the previously calculated adjustment factor to real-time correct the initial magnetic memory signal to generate the target magnetic memory signal. Finally, the target magnetic memory signal is input to the fatigue damage assessment module, which can integrate various assessment algorithms, such as threshold comparison, pattern recognition or machine learning classifier, to output the fatigue damage assessment result of the precision hardware to be tested, such as "no damage", "minor damage" or "severe damage".

[0076] Through the above technical scheme, the present embodiment can effectively overcome the measurement deviation problem of the existing precision hardware fatigue damage detection system in complex industrial environments due to internal interference or damaged calibration data. By introducing the reference block for dynamic calibration and calculating the adjustment factor to correct the initial magnetic memory signal, it ensures that even in the case of damaged system calibration parameters or unknown systematic deviation, accurate and reliable target magnetic memory signals can be obtained. This significantly improves the accuracy and reliability of fatigue damage assessment, enabling the system to more effectively identify the early fatigue damage state of precision hardware, avoiding the chain failure that may be caused by misjudgment, thereby ensuring the authenticity and effectiveness of the detection results.

[0077] In some embodiments, in step S103, calculating the adjustment factor reflecting the overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal can include but is not limited to the following steps:

[0078] extracting a first spatial distribution of the theoretical magnetic memory signal and a second spatial distribution of the reference measurement signal;

[0079] comparing the first spatial distribution and the second spatial distribution to determine the spatial deviation;

[0080] spatially correcting the reference measurement signal according to the spatial deviation;

[0081] calculating the adjustment factor according to the theoretical magnetic memory signal and the spatially corrected reference measurement signal.

[0082] In some embodiments, due to the influence of the measurement environment and the device itself, the reference measurement signal may have a spatial deviation. If this spatial deviation is not corrected, directly calculating the adjustment factor will result in the adjustment factor failing to accurately reflect the overall measurement deviation, thereby affecting the accuracy of fatigue damage assessment. Therefore, the first spatial distribution of the theoretical magnetic memory signal and the second spatial distribution of the reference measurement signal can be extracted first, and then the first spatial distribution and the second spatial distribution are compared to determine the spatial deviation of the reference measurement signal relative to the theoretical magnetic memory signal. This deviation may be caused by sensor positioning errors, inconsistent scanning paths, or reference block placement deviations, etc. Then, according to the spatial deviation, the reference measurement signal is spatially corrected so that the corrected reference measurement signal is highly aligned with the theoretical magnetic memory signal in space. Finally, according to the theoretical magnetic memory signal and the spatially corrected reference measurement signal, the adjustment factor is calculated, which can more accurately reflect the overall deviation of the measurement system in the non-spatial dimension, such as sensor sensitivity drift, environmental magnetic field fluctuation, etc. This spatially corrected adjustment factor can provide a more accurate correction amount when applied to correct the initial magnetic memory signal.

[0083] It can be understood that the spatial deviation refers to the inconsistency of the first spatial distribution and the second spatial distribution in spatial position, shape or scale, which can specifically manifest as translation, rotation, scaling or more complex nonlinear deformation, and its purpose is to quantify the geometric distortion of the reference measurement signal relative to the theoretical magnetic memory signal. Spatial correction refers to geometric transformation processing of the reference measurement signal according to the determined spatial deviation, so that its spatial distribution is aligned or matched with the spatial distribution of the theoretical magnetic memory signal, which can specifically be achieved by using affine transformation, perspective transformation, non-rigid registration algorithm or interpolation resampling technology, and its purpose is to eliminate or reduce the spatial error in the reference measurement signal, so that it more accurately reflects the true magnetic memory characteristics of the reference block.

[0084] To make the technical solution clearer, specific examples are used for explanation below. In extracting the first spatial distribution of the theoretical magnetic memory signal and the second spatial distribution of the reference measurement signal, a high-precision three-dimensional scanning device can be used to scan the reference block to obtain its geometric model, and combined with magnetic field simulation software to generate three-dimensional spatial distribution data of the theoretical magnetic memory signal on the surface of the reference block, forming a dense point cloud or grid data. At the same time, a multi-channel magnetic field sensor array is used to actually measure the reference block, and the sensor array scans at a preset path and interval to collect the values of the reference measurement signal at the corresponding spatial positions, forming another spatial distribution data. In comparing the first spatial distribution and the second spatial distribution to determine the spatial deviation, an iterative closest point (ICP) algorithm or a registration algorithm based on feature points can be used. For example, representative feature points can be extracted from the two spatial distributions, and then an optimal rigid transformation is calculated by minimizing the distance between these feature points to determine the spatial deviation of the reference measurement signal relative to the theoretical magnetic memory signal. In spatially correcting the reference measurement signal according to the spatial deviation, the rigid transformation parameters determined in the previous step can be used to transform the original spatial coordinates of the reference measurement signal to align it with the spatial coordinate system of the theoretical magnetic memory signal. Subsequently, bilinear interpolation or cubic spline interpolation methods can be used to resample the corrected reference measurement signal to the same spatial grid or resolution as the theoretical magnetic memory signal, ensuring that the two have a one-to-one correspondence in space. In calculating the adjustment factor based on the theoretical magnetic memory signal and the spatially corrected reference measurement signal, the spatially aligned theoretical magnetic memory signal and reference measurement signal can be calculated point by point to obtain a spatial distribution of adjustment factor matrix. Alternatively, linear regression analysis can be performed on the two to obtain a global scaling factor as the adjustment factor. For example, the ratio of the corrected reference measurement signal to the theoretical magnetic memory signal at each corresponding spatial point can be calculated, and then the ratios are averaged or weighted to obtain an adjustment factor that reflects the overall measurement deviation.

[0085] By the technical solution, the embodiment can effectively solve the problem that the adjustment factor is inaccurate due to the spatial deviation of the reference measurement signal. By extracting and comparing the spatial distribution of the theoretical magnetic memory signal and the reference measurement signal, the spatial deviation in the reference measurement signal can be accurately identified and quantified. Then, the reference measurement signal is spatially corrected, eliminating the geometric distortion introduced in the measurement process, so that the corrected reference measurement signal can more truly reflect the magnetic memory characteristics of the reference block. The adjustment factor calculated on this basis can more accurately reflect the overall deviation of the measurement system, thereby ensuring that the correction of the initial magnetic memory signal is more accurate. Finally, this makes the generated target magnetic memory signal have higher reliability, significantly improves the accuracy of the fatigue damage evaluation result of the precision hardware, and provides more reliable data support for industrial detection.

[0086] In some embodiments, in step S105, the fatigue damage evaluation process of the to-be-tested precision hardware is performed according to the target magnetic memory signal, and a fatigue damage evaluation result is obtained, which can include but is not limited to the following steps:

[0087] Step S201, a load application instruction is generated, and the load application instruction is used to apply a local gradually changing mechanical load to a to-be-evaluated region of the to-be-tested precision hardware;

[0088] Step S202, after the local gradually changing mechanical load is applied, the to-be-evaluated region is subjected to magnetic memory signal collection, and initial load and signal response data are obtained;

[0089] Step S203, feature analysis is performed on the initial load and signal response data, and a response feature representing a material fatigue accumulation mode is obtained;

[0090] Step S204, the to-be-tested precision hardware is evaluated according to the target magnetic memory signal, the response feature, and a preset judgment criterion, and a fatigue damage evaluation result is obtained.

[0091] In some embodiments, due to the reliance on only the target magnetic memory signal for evaluation, its accuracy can be insufficient, because the fatigue damage of the material is a dynamic accumulation process, which is affected by multiple factors, such as load type, load size, loading method, etc. A single magnetic memory signal is difficult to fully reflect the fatigue state of the material, thereby reducing the accuracy of the evaluation results. To this end, a load application instruction can be generated, which is used to apply a local gradual mechanical load to the evaluation region of the precision hardware under test, aiming to simulate the dynamic stress environment that the material is subjected to in actual working conditions, thereby triggering the potential fatigue damage inside the material, making it show a response on the magnetic memory signal. After applying the local gradual mechanical load, the magnetic memory signal of the evaluation region is collected to obtain initial load and signal response data. These data contain the force-magnetic coupling information of the material under dynamic load, which can reflect the changes of the microstructure of the material during the loading process.

[0092] Then the initial load and signal response data are analyzed to obtain response characteristics representing the fatigue accumulation pattern of the material, so that key indicators directly related to fatigue damage, such as magnetic field change rate, magnetic hysteresis characteristics, etc., can be extracted from complex dynamic signals, which can finely reveal the fatigue accumulation state of the material. Then, according to the target magnetic memory signal, the response characteristics, and the preset judgment criteria, the precision hardware under test is evaluated to obtain the fatigue damage evaluation results. The target magnetic memory signal provides the baseline information of the macro fatigue damage of the material, and the response characteristics supplement the details of the micro damage accumulation of the material under dynamic stress. By comprehensively considering the static target magnetic memory signal (which reflects the overall fatigue state of the material) and the dynamic response characteristics (which reflect the local damage situation of the material during loading), the evaluation is no longer limited to single-dimensional information, but can capture the dynamic evolution process of the fatigue damage of the material and the damage characteristics of the local sensitive region.

[0093] It can be understood that the local gradual mechanical load refers to a mechanical force applied on the evaluation region, whose size or direction gradually changes over time or spatial position, which can be specifically achieved by controlling the output of the loading device so that the load gradually increases from zero to a certain peak value, or different intensities of load are applied at different positions, aiming to simulate the stress state of the precision hardware during actual service, thereby effectively triggering the fatigue damage response inside the material. The response characteristics refer to key parameters or indicators extracted from the initial load and signal response data, which can reflect the degree and pattern of fatigue damage accumulation of the material, which can specifically include magnetic field strength change rate, magnetic hysteresis loop shape parameter, magnetic domain structure change characteristic, magnetoelastic effect parameter, or magnetic impedance spectrum change, etc., aiming to identify the information directly related to the fatigue state of the material from complex signal data, which has diagnostic significance.

[0094] To make the technical solution clearer, specific examples are used for explanation. First, a servo-controlled loading device, such as a piezoelectric actuator or a hydraulic loading system, can be used to generate a load application command and apply a local and gradual mechanical load to a specific evaluation area of the precision hardware under test. The load can be set to gradually increase from zero to a preset maximum value, or cyclically loaded with a specific frequency and amplitude to simulate the stress cycle under actual working conditions. During the load application process, or after the load application is completed, a high-sensitivity magnetic field sensor array, such as a Hall sensor array or a giant magnetoresistance sensor array, can be used to collect the magnetic memory signal of the evaluation area in real time or quasi-real time. These sensors can be arranged on or near the surface of the evaluation area to capture the magnetic field changes caused by the load, thereby obtaining the initial load and signal response data. These data can include load size, loading time point, and corresponding magnetic memory signal strength, gradient, or waveform information at that time. Further, the obtained initial load and signal response data are analyzed for features. This can include time domain, frequency domain, or time-frequency domain analysis of the signal, such as calculating the slope, peak, valley, hysteresis loop area, or specific frequency components in the Fourier transform spectrum of the magnetic memory signal during the load change process. In addition, statistical methods such as mean, variance, skewness, kurtosis, etc. can also be used to quantify the characteristics of the signal. These analyses aim to extract response features that can represent the material's fatigue accumulation pattern, such as identifying abnormal patterns of magnetic signals related to micro-crack initiation or propagation. Finally, the corrected target magnetic memory signal, which reflects the overall fatigue state of the material, is combined with the response features obtained through the above analysis and a preset judgment criterion to evaluate the precision hardware under test.

[0095] Through the above technical solution, the present embodiment can combine the static magnetic memory signal of the material with the dynamic load response, thereby providing a fatigue damage evaluation method with a wide coverage and improved information depth. By applying a local and gradual mechanical load and collecting the corresponding magnetic memory signal response, the fatigue damage characteristics within the material can be effectively excited and captured, especially those early-stage damages that are not obvious or difficult to identify under static conditions. Feature analysis of the initial load and signal response data can extract key response features that represent the material's fatigue accumulation pattern, which can finely reflect the fatigue state of the material. Finally, combining the target magnetic memory signal, response features, and a preset judgment criterion for evaluation improves the accuracy and reliability of fatigue damage evaluation, solves the problem of insufficient accuracy of single magnetic memory signal evaluation, and improves the accuracy of fatigue damage state judgment for precision hardware.

[0096] In some embodiments, the feature analysis of the initial load and signal response data in step S203 to obtain response features representing the material's fatigue accumulation pattern can include, but is not limited to, the following steps:

[0097] Step S301, after the local gradually changing mechanical load is removed, the magnetic memory signal of the to-be-evaluated region is collected to obtain residual magnetic memory signal data;

[0098] Step S302, target load and signal response data is generated according to the residual magnetic memory signal data and the initial load and signal response data;

[0099] Step S303, response features are extracted from the target load and signal response data.

[0100] In some embodiments, since the initial load and signal response data can be affected by factors such as transient mechanical vibration or local magnetic field disturbance, these factors can introduce non-fatigue noise or interference signals, thereby affecting the accuracy of the response features, and further leading to deviation of the fatigue damage evaluation results. In order to accurately extract the response features, the magnetic memory signal of the to-be-evaluated region can be collected after the local gradually changing mechanical load is removed to obtain residual magnetic memory signal data. The residual magnetic memory signal data can reflect the accumulation of material fatigue damage, and it is less affected by non-fatigue factors such as transient mechanical vibration or local magnetic field disturbance. Then, target load and signal response data is generated according to the residual magnetic memory signal data and the initial load and signal response data. By effectively combining and processing the two kinds of data, the non-fatigue noise or interference signals introduced by transient mechanical vibration or local magnetic field disturbance in the initial load and signal response data can be identified and eliminated or significantly weakened. For example, the residual magnetic memory signal data can be used as a kind of reference or correction factor to filter out the transient fluctuations in the initial data that are irrelevant to fatigue damage. Then, response features are extracted from the target load and signal response data. Since the target load and signal response data has excluded most interference factors, the response features extracted therefrom can more accurately represent the fatigue accumulation pattern of the material, avoiding misjudgment caused by noise interference.

[0101] It can be understood that the residual magnetic memory signal data refers to the magnetic signal data reflecting the internal stress state and microstructure changes of the material obtained by magnetic memory signal collection after the local gradually changing mechanical load is applied and removed. It can be measured non-contact by a high-sensitivity magnetic sensor array, and its purpose is to capture the stable magnetic memory effect formed after the material fatigue damage accumulates, to provide damage information irrelevant to transient load.

[0102] To make the technical solution clearer, specific examples are used for explanation below. After applying a local ramp mechanical load to the region to be evaluated of the precision hardware and completing initial load and signal response data collection, a high-sensitivity magnetic sensor array, such as a linear array composed of multiple Hall sensors, can be used to perform a non-contact scan of the region to be evaluated after the mechanical load is completely removed to obtain residual magnetic memory signal data. These sensors can collect data at a preset scan speed and sampling frequency to ensure that stable magnetic signals after stress redistribution inside the material are captured. Subsequently, target load and signal response data can be generated according to the collected residual magnetic memory signal data and the previously obtained initial load and signal response data. Specifically, a signal processing unit, such as an embedded digital signal processor, can be used to fuse process the two sets of data. One implementation is to use the residual magnetic memory signal data as a reference to identify and remove non-fatigue noise components introduced by transient mechanical vibration or local magnetic field disturbance from the initial load and signal response data through an adaptive filtering algorithm, such as the least mean square algorithm. Another implementation is to perform wavelet transform on the two sets of data, identify and retain low-frequency components related to fatigue damage, while filtering out high-frequency noise components, and then reconstruct the processed signals to obtain more pure target load and signal response data. Finally, response features are extracted from the generated target load and signal response data. This can be done by a data analysis module, which can run on a high-performance processor. For example, the magnetic field gradient change rate of the target load and signal response data in a specific region, the peak value, the root mean square value of the signal, or frequency spectrum analysis to extract specific frequency components related to fatigue accumulation can be calculated, and these features can be used as response features representing the fatigue accumulation pattern of the material for subsequent fatigue damage evaluation.

[0103] Through the above technical solution, the embodiment can effectively overcome the influence of non-fatigue noise or interference signals introduced by factors such as transient mechanical vibration or local magnetic field disturbance. By collecting residual magnetic memory signal data after the local ramp mechanical load is removed and combining it with the initial load and signal response data to generate target load and signal response data, the interference components in the data can be significantly reduced, so that the response features extracted from the target load and signal response data can more accurately represent the fatigue accumulation pattern of the material, thereby improving the accuracy and reliability of the fatigue damage evaluation results and avoiding false judgments caused by noise.

[0104] In some embodiments, in step S302, generating target load and signal response data according to residual magnetic memory signal data and initial load and signal response data can include but is not limited to the following steps:

[0105] In step S401, component analysis is performed on the residual magnetic memory signal data to identify interference components introduced by transient mechanical vibration or local magnetic field disturbance, including non-fatigue noise or interference signals.

[0106] In step S402, according to the interference components, the initial load and signal response data are deinterference processed to obtain target load and signal response data.

[0107] In some embodiments, since the residual magnetic memory signal data can be affected by transient mechanical vibration or local magnetic field disturbance, non-fatigue noise or interference signals and other interference components are introduced, which will affect the accuracy of the target load and signal response data, and further affect the reliability of the fatigue damage assessment. Therefore, the residual magnetic memory signal data can be analyzed first to identify the interference components introduced by transient mechanical vibration or local magnetic field disturbance, so that the system can decompose the signal into different components to accurately identify which are the effective signals related to fatigue damage, wherein the interference components include non-fatigue noise or interference signals. Then, according to the interference components, the initial load and signal response data are deinterference processed to obtain target load and signal response data, aiming to accurately eliminate or significantly weaken the influence of these interference components on the original data, so as to ensure that the generated target load and signal response data can more truly reflect the actual fatigue state of the precision hardware under test.

[0108] It can be understood that the non-fatigue noise or interference signal refers to a random or systematic signal that is irrelevant to material fatigue damage and has a negative impact on the magnetic memory signal, which can be represented by Gaussian white noise, power frequency interference, power supply ripple or sensor drift signal, and its purpose is to specify the type of interference components.

[0109] To more clearly illustrate the technical solutions, specific examples are used in the following for explanation. Firstly, the residual magnetic memory signal data collected after the local gradual mechanical load removal of the to-be-evaluated region can be subjected to component analysis by using a wavelet decomposition method. Specifically, the residual magnetic memory signal data can be decomposed into wavelet coefficients of different frequency scales, and by analyzing the energy distribution and time-frequency characteristics of these wavelet coefficients, interference components introduced by transient mechanical vibration or local magnetic field disturbance can be identified. For example, if it is detected that there is a signal with energy concentration in a specific high frequency band and synchronized with the mechanical vibration frequency, or a transient spike appears when the magnetic field disturbance occurs, these can be identified as non-fatigue noise or interference signals. Subsequently, according to the identified interference components, the initial load and signal response data are subjected to de-interference processing. Specifically, an adaptive filtering algorithm, such as the least mean square algorithm or the recursive least squares algorithm, can be used. In the adaptive filtering process, the identified interference components can be taken as the reference input, and the initial load and signal response data can be taken as the expected input, and by dynamically adjusting the filter coefficients through the algorithm, the interference components can be effectively removed or suppressed from the initial load and signal response data. For example, if the power supply ripple is identified as an interference component, the adaptive filter can learn and offset its influence on the signal. After such de-interference processing, pure and accurate target load and signal response data can be obtained, which can more accurately reflect the fatigue damage state of the material.

[0110] Through the above technical solutions, the embodiment can effectively solve the problem of the influence of interference components existing in the residual magnetic memory signal data on the accuracy of fatigue damage evaluation. By performing component analysis on the residual magnetic memory signal data, non-fatigue noise or interference signals introduced by transient mechanical vibration or local magnetic field disturbance can be accurately identified and distinguished, and the initial load and signal response data can be subjected to de-interference processing accordingly, which can significantly improve the purity and accuracy of the target load and signal response data. This enables the subsequent fatigue damage evaluation to be based on more reliable data, thereby avoiding false judgments caused by interference signals and improving the reliability of the fatigue damage evaluation results.

[0111] In some embodiments, in step S401, the component analysis on the residual magnetic memory signal data to identify interference components introduced by transient mechanical vibration or local magnetic field disturbance can include but is not limited to the following steps:

[0112] Step S501, magnetic memory signal collection is performed on the reference region to obtain reference region signal data, the reference region is adjacent to the to-be-evaluated region and does not have fatigue damage;

[0113] Step S502, determining interference characteristic parameters according to the reference region signal data;

[0114] In step S503, the interference characteristic parameter is compared with the residual magnetic memory signal data to obtain a first comparison result.

[0115] In step S504, the interference component is identified from the residual magnetic memory signal data according to the first comparison result.

[0116] In some embodiments, the identification of the interference component in the residual magnetic memory signal is still challenging because the interference component can be mixed with the real fatigue damage signal and is difficult to distinguish, directly affecting the effect of deinterference and the accuracy of fatigue damage assessment. To accurately identify the interference component, the magnetic memory signal of the reference region can be collected first to obtain reference region signal data, wherein the reference region is adjacent to the to-be-evaluated region and does not have fatigue damage. The reference region signal data mainly reflects environmental interference and material background magnetic field, thereby providing a pure interference reference. Then, the interference characteristic parameter is determined according to the reference region signal data, which is an inherent attribute of the interference signal, such as its frequency, amplitude or waveform mode, which serves as a basis for identifying interference and effectively eliminates the interference of other factors. Then, the interference characteristic parameter is compared with the residual magnetic memory signal data to obtain a first comparison result, which can preliminarily reveal whether there is an interference component similar to the reference region in the residual magnetic memory signal. Finally, the interference component is identified from the residual magnetic memory signal data according to the first comparison result, so that the system can use the interference characteristics of the known undamaged region to accurately separate the interference caused by non-fatigue damage from the complex signal of the to-be-evaluated region.

[0117] It can be understood that the interference characteristic parameter refers to a parameter extracted from the reference region signal data, which can characterize the characteristics of transient mechanical vibration or local magnetic field disturbance, which can be realized by the frequency, amplitude, phase, waveform mode or statistical characteristics of the signal, and the purpose is to provide a basis for subsequent identification of interference in the residual magnetic memory signal. The reference region refers to a region adjacent to the to-be-evaluated region and not having fatigue damage, which can be realized by selecting a known undamaged material region near the to-be-evaluated region, and the purpose is to provide a pure reference signal containing only environmental interference and material background magnetic field information.

[0118] To make the technical solution clearer, specific examples are used for explanation below. First, in the vicinity of the area to be evaluated, a region known to be free of fatigue damage is selected as a reference region. The reference region is scanned using a magnetic memory signal sensor to collect its magnetic memory signal, thereby obtaining reference region signal data. For example, a Hall sensor array can be used to non-contact scan the surface of the reference region at a preset scanning speed and interval, recording the magnetic field strength data at each position. Next, the collected reference region signal data is processed to determine the interference characteristic parameters. For example, the reference region signal data can be subjected to fast Fourier transform (FFT) to analyze its frequency spectrum and identify specific frequency peaks caused by environmental vibration or power fluctuations and their corresponding amplitudes, which can be combined as interference characteristic parameters. Alternatively, the transient characteristics of the signal can also be extracted through wavelet analysis as interference characteristic parameters. Then, the determined interference characteristic parameters are compared with the residual magnetic memory signal data of the area to be evaluated. For example, the residual magnetic memory signal data can also be subjected to frequency spectrum analysis, and then the interference frequency peaks of the reference region are compared with the corresponding frequency peaks in the residual magnetic memory signal data to calculate their similarity or correlation coefficient, obtaining a first comparison result. If the similarity exceeds a preset threshold, it indicates that there may be interference in the residual magnetic memory signal. Finally, according to the first comparison result, the interference components are identified from the residual magnetic memory signal data. For example, if the comparison result shows that there are signal components in the residual magnetic memory signal data that highly match the interference characteristic parameters of the reference region, adaptive filtering, notch filtering or independent component analysis (ICA) and other signal processing techniques can be used to separate these matching signal components from the residual magnetic memory signal data, determine them as interference components and remove them, thereby obtaining purer magnetic memory signals reflecting the true fatigue damage.

[0119] Through the above technical solution, the magnetic memory signal data of the reference region is used as a reference to accurately determine the interference characteristic parameters introduced by transient mechanical vibration or local magnetic field disturbance. Comparing these interference characteristic parameters with the residual magnetic memory signal data of the area to be evaluated can effectively identify the interference components in the signal that are unrelated to fatigue damage, avoiding the problem of interference signal and true fatigue damage signal aliasing. This enables the subsequent interference removal process to be more accurate, thereby obtaining purer target load and signal response data, significantly improving the accuracy of fatigue damage evaluation.

[0120] In some embodiments, in step S504, identifying the interference components from the residual magnetic memory signal data according to the first comparison result can include but is not limited to the following steps:

[0121] In step S601, a vibration application instruction is generated, and the vibration application instruction is used to apply mechanical vibration of a preset amplitude and frequency to the to-be-evaluated region.

[0122] In step S602, after the mechanical vibration is applied, the to-be-evaluated region is subjected to magnetic memory signal acquisition, and dynamic response magnetic memory signal data is obtained.

[0123] In step S603, the dynamic response magnetic memory signal data is subjected to feature analysis processing, and magnetic memory signal modulation features synchronized with the mechanical vibration are obtained, and the magnetic memory signal modulation features are used to represent the force-magnetic coupling response of the material.

[0124] In step S604, according to the first comparison result and the magnetic memory signal modulation features, interference components are identified from the residual magnetic memory signal data.

[0125] In some embodiments, since the comparison is only performed by the static reference region signal, it is difficult to distinguish the interference components introduced by the transient mechanical vibration or local magnetic field disturbance in the residual magnetic memory signal, because these interference components may exhibit similarities with the fatigue damage features on some static features, thereby causing misjudgment. Therefore, a vibration application instruction can be generated first, and the vibration application instruction is used to apply mechanical vibration of a preset amplitude and frequency to the to-be-evaluated region, aiming to introduce a controllable dynamic disturbance to observe the response of the material to the disturbance. By applying mechanical vibration with a preset amplitude and frequency, the magnetic domain structure inside the material can be dynamically changed, thereby generating vibration-related magnetic memory signal modulation. After the mechanical vibration is applied, the to-be-evaluated region is subjected to magnetic memory signal acquisition, and dynamic response magnetic memory signal data is obtained, aiming to capture the magnetic memory signal changes of the material when subjected to mechanical vibration. These data contain the response information of the material to the mechanical vibration. The dynamic response magnetic memory signal data is subjected to feature analysis processing, and magnetic memory signal modulation features synchronized with the mechanical vibration are obtained, wherein the magnetic memory signal modulation features are used to represent the force-magnetic coupling response of the material, aiming to extract feature information related to the mechanical vibration, such as signal amplitude, frequency or phase modulation. These modulation features reflect the force-magnetic coupling effect of the material, that is, the characteristics of the change of the magnetic performance of the material with mechanical stress.

[0126] Then, according to the first comparison result and the magnetic memory signal modulation feature, the interference component is identified from the residual magnetic memory signal data. The embodiment can verify and supplement the first comparison result by introducing the dynamic response magnetic memory signal data and its modulation feature. Specifically, if a certain component in the residual magnetic memory signal data is determined as a potential interference in the first comparison result, but it does not show the magnetic memory signal modulation feature synchronized with the mechanical vibration in the dynamic response magnetic memory signal data, it can be confirmed as an interference component and removed. On the contrary, if a certain signal component shows the magnetic memory signal modulation feature synchronized with the mechanical vibration, it indicates that it is a manifestation of the material force-magnetic coupling response, and it is more likely to be a fatigue damage signal. This method of combining static comparison and dynamic response analysis can distinguish between interference components introduced by transient mechanical vibration or local magnetic field disturbance and fatigue damage features, avoid misjudgment, and thus ensure the reliability of subsequent fatigue damage assessment.

[0127] It can be understood that the magnetic memory signal modulation feature refers to the change pattern of the magnetic memory signal of the material when subjected to external mechanical vibration, which can be obtained by analyzing the change law of the amplitude, frequency or phase of the signal with vibration, and its purpose is to characterize the dynamic response of the internal magnetic domain structure of the material to mechanical stress. The force-magnetic coupling response refers to the mutual influence between the magnetic properties of the material and the mechanical stress or strain, which can be manifested as the change of the orientation of the internal magnetic domain, the movement of the magnetic wall and the magnetization state of the material when subjected to mechanical force, thereby causing the change of the magnetic memory signal, and its purpose is to reveal the magnetic behavior of the material under dynamic load, and to provide a basis for distinguishing fatigue damage and transient interference.

[0128] In order to more clearly illustrate the technical scheme, specific examples are used for explanation. First, the vibration application instruction can be generated by the control unit, which is output to the mechanical vibration generator, such as a piezoelectric ceramic driver or an electromagnetic exciter, through a digital signal. The vibration generator is configured to apply mechanical vibration of a predetermined amplitude and frequency, such as sinusoidal vibration, to the region to be evaluated, and the frequency can be set to a non-harmonic frequency of the material inherent frequency to avoid resonance effect, while ensuring the excitation of the force-magnetic coupling response of the material.

[0129] During or immediately after the mechanical vibration application process, the magnetic memory signal of the region to be evaluated can be collected using magnetic field sensors, such as Hall sensor arrays or fluxgate sensors. These sensors can be scanned or fixedly arranged along the surface of the region to be evaluated to obtain dynamic response magnetic memory signal data, which is represented as a sequence of magnetic field strength varying with time.

[0130] Subsequently, the dynamic response magnetic memory signal data can be subjected to feature analysis processing. This can be accomplished by a digital signal processor or embedded system, by performing Fourier transform or wavelet analysis on the collected time-domain signals to extract magnetic memory signal modulation features synchronized with the mechanical vibration frequency. For example, the amplitude or phase variation of the signal at the vibration frequency and its multiples can be analyzed, or the modulation depth of the signal can be calculated. These modulation features can serve as quantitative indicators of the material's magneto-mechanical coupling response.

[0131] Finally, according to the first comparison result and the magnetic memory signal modulation features, the interference components are identified from the residual magnetic memory signal data. Specifically, a decision logic can be constructed. If a certain abnormal signal component in the residual magnetic memory signal data is determined to be a potential interference in the first comparison result after comparing with the reference region signal data, and this component does not exhibit a magnetic memory signal modulation feature synchronized with the applied mechanical vibration in the dynamic response magnetic memory signal data, then this signal component can be confirmed as an interference component and removed. Conversely, if the abnormal signal component exhibits a magnetic memory signal modulation feature synchronized with the mechanical vibration, it can be judged as a fatigue damage signal, as it reflects the material's magnetic response under dynamic stress. This method can improve the accuracy of interference identification, thereby ensuring the reliability of fatigue damage assessment.

[0132] Through the above technical solutions, the present embodiment introduces a mechanism of applying mechanical vibration and analyzing the material's dynamic magneto-mechanical coupling response, enabling the system to distinguish fatigue damage signals from interference components introduced by transient mechanical vibration or local magnetic field disturbance using the material's magnetic response to dynamic load. By observing whether the signal has a modulation feature synchronized with the mechanical vibration, signals that are similar to fatigue damage signals in static features but are essentially interference can be identified, thus avoiding misjudgment. This improves the accuracy of identifying interference components from residual magnetic memory signal data, ensuring the reliability of subsequent fatigue damage assessment, and thus improving the precision of fatigue damage detection of precision hardware.

[0133] In some embodiments, in step S603, the dynamic response magnetic memory signal data is subjected to feature analysis processing to obtain magnetic memory signal modulation features synchronized with the mechanical vibration, which can include but is not limited to the following steps:

[0134] The dynamic response magnetic memory signal data is subjected to response analysis to obtain nonlinear responses under different mechanical vibration intensities;

[0135] According to the nonlinear responses, the dynamic response magnetic memory signal data is subjected to frequency spectrum analysis to extract target harmonic components of the mechanical vibration frequency, the target harmonic components including second harmonic components or third harmonic components;

[0136] The target harmonic components are used as magnetic memory signal modulation features.

[0137] In some embodiments, since the dynamic response magnetic memory signal data can contain complex nonlinear responses, simply performing feature analysis can not accurately extract the magnetic memory signal modulation features synchronized with mechanical vibration, thereby affecting the accuracy of interference component identification. For this purpose, the dynamic response magnetic memory signal data can be analyzed for response first to obtain nonlinear responses under different mechanical vibration intensities. When a material is subjected to mechanical vibration, its magnetic memory signal response is often not a simple linear relationship, but exhibits complex nonlinear characteristics. This nonlinear characteristic is closely related to the microstructure state of the material, especially the accumulation of fatigue damage. According to the nonlinear response, the dynamic response magnetic memory signal data is analyzed for frequency spectrum to extract the target harmonic component of the mechanical vibration frequency, and the target harmonic component is taken as the magnetic memory signal modulation feature. The target harmonic component includes a second harmonic component or a third harmonic component. Traditional frequency spectrum analysis can only focus on the fundamental frequency component. However, due to the nonlinear characteristics of the material, the magnetic memory signal often contains rich harmonic components, which can more sensitively reflect the microstructure changes and fatigue damage state of the material. Because fatigue damage can cause the nonlinear characteristics of the material to increase, thereby increasing the amplitude of the harmonic component.

[0138] It can be understood that the nonlinear response refers to the behavior that the output of the system is not in a simple linear proportion with the input, which can specifically manifest as the amplitude, phase or waveform of the magnetic memory signal changing non-proportionally with the increase of the mechanical vibration intensity. The purpose is to capture the complex characteristics of the microstructure changes of the material during force application.

[0139] In order to more clearly illustrate the technical scheme, specific examples are used for explanation below. First, a piezoelectric ceramic vibrator or an electromagnetic exciter can be used to apply a series of mechanical vibrations with preset frequencies and amplitudes to the evaluation region of the precision hardware, for example, the amplitude can be increased step by step from a low amplitude, and a high-sensitivity magnetic field sensor is used to collect the magnetic memory signal data of the evaluation region in real time. After each change in vibration intensity, the steady-state response or transient response of the magnetic memory signal can be recorded. By analyzing these response data, for example, drawing a curve of the amplitude of the magnetic memory signal versus the vibration intensity, the nonlinear characteristics can be observed, for example, the signal amplitude appears nonlinear growth or saturation phenomenon at a certain vibration intensity, thereby obtaining the nonlinear responses under different mechanical vibration intensities.

[0140] Then, according to the obtained nonlinear response characteristics, the collected dynamic response magnetic memory signal data can be subjected to spectral analysis. For example, the time-domain magnetic memory signal data at each vibration intensity can be subjected to discrete Fourier transform or fast Fourier transform to obtain its frequency-domain spectrum. In the frequency-domain spectrum, in addition to the fundamental frequency component of the mechanical vibration frequency, the harmonic components at integer multiple frequencies of the mechanical vibration frequency can be focused on. Specifically, the second harmonic component and / or the third harmonic component of the mechanical vibration frequency can be identified and extracted, which usually exhibit significant amplitudes when the material undergoes nonlinear force-magnetic coupling effect. For example, a frequency window can be set to accurately locate and extract the amplitude and phase information of these harmonic components.

[0141] Finally, the amplitudes of the extracted target harmonic components, such as the second harmonic component and the third harmonic component, or their combination, are taken as the magnetic memory signal modulation features. These features can be stored or further processed for subsequent interference component identification steps, so as to more accurately characterize the force-magnetic coupling response of the material under mechanical vibration.

[0142] Through the above technical solutions, the embodiment can comprehensively reveal the nonlinear response characteristics of the material under different mechanical vibration intensities by analyzing the dynamic response magnetic memory signal data, which enables more accurate extraction of the target harmonic components, such as the second harmonic component or the third harmonic component, of the mechanical vibration frequency in subsequent spectral analysis. These harmonic components can more sensitively reflect the microstructure changes and fatigue damage state of the material, thereby overcoming the limitation of traditional simple feature analysis that cannot accurately capture complex nonlinear responses. Taking these target harmonic components as the magnetic memory signal modulation features can significantly improve the extraction accuracy of the signal components synchronized with the mechanical vibration, thereby effectively improving the accuracy and reliability of identifying the interference components from the residual magnetic memory signal data, and ensuring the authenticity of the fatigue damage evaluation results.

[0143] In some embodiments, in step S604, identifying the interference components from the residual magnetic memory signal data according to the first comparison result and the magnetic memory signal modulation features can include, but is not limited to, the following steps:

[0144] extracting a fatigue damage feature from the residual magnetic memory signal data according to the first comparison result;

[0145] comparing the fatigue damage feature with the magnetic memory signal modulation features to obtain a second comparison result;

[0146] according to the second comparison result, taking the signal components in the residual magnetic memory signal data that do not have the magnetic memory signal modulation features as the interference components.

[0147] In some embodiments, since only the interference components are directly identified according to the first comparison result, some real fatigue damage features may be misjudged as interference, because the fatigue damage signal itself may have similarities with some interference signals in some features, resulting in low accuracy of identification. Therefore, the fatigue damage features can be extracted from the residual magnetic memory signal data according to the first comparison result, avoiding blind analysis of all signal components, so that the subsequent processing is more focused on the potential damage signal. Then the fatigue damage features are compared with the magnetic memory signal modulation features to obtain a second comparison result. The magnetic memory signal modulation features are the force-magnetic coupling response of the material under the action of mechanical vibration, which has inherent relevance with the real fatigue damage signal, while the interference signal usually does not have such relevance. Due to the difference in this inherent relevance, by comparing the fatigue damage features with the magnetic memory signal modulation features, it can be identified which signal components are truly caused by fatigue damage and which signal components are only interference. Then, according to the second comparison result, the signal components in the residual magnetic memory signal data that do not have the magnetic memory signal modulation features are taken as interference components. This means that only those signal components that meet the preliminary comparison result and are irrelevant to the force-magnetic coupling response will be finally determined as interference. This double verification mechanism, i.e. first screening the potential damage features and then using the force-magnetic coupling response for identification, makes the interference identification process more refined and reliable.

[0148] It can be understood that the fatigue damage features refer to the patterns or properties in the residual magnetic memory signal data that reflect the changes in the microstructure of the material, stress concentration or crack initiation and propagation, which can be characterized by the amplitude, frequency, phase, waveform distortion, energy distribution or specific statistical parameters of the signal, and the purpose is to focus on the signal part directly related to fatigue damage to provide a basis for subsequent accurate identification.

[0149] To make the technical solution clearer, specific examples are used for explanation below. First, according to the first comparison result, fatigue damage features can be extracted from the residual magnetic memory signal data by using signal processing algorithms such as wavelet transform or empirical mode decomposition. For example, specific frequency components, amplitude mutation points or waveform distortion features in the signal that are consistent with known fatigue damage patterns can be identified, which can be quantified as feature vectors. Then, the extracted fatigue damage features are compared with the magnetic memory signal modulation features. The comparison process can use correlation analysis methods, such as calculating the Pearson correlation coefficient between the fatigue damage feature vectors and the magnetic memory signal modulation feature vectors, or using machine learning classifiers such as support vector machines or neural networks for pattern matching, to obtain the second comparison result. For example, if the correlation coefficient is below a preset threshold, or the classifier determines it as a non-fatigue damage pattern, it indicates that the signal component is weakly related to the force-magnetic coupling response. Finally, according to the second comparison result, the signal components in the residual magnetic memory signal data that do not have magnetic memory signal modulation features are regarded as interference components. Specifically, if the second comparison result shows that a certain signal component has very low correlation with the magnetic memory signal modulation features, or is clearly determined by the classifier as a non-fatigue damage signal, then the signal component is identified as interference and removed from the residual magnetic memory signal data to ensure the accuracy of subsequent fatigue damage assessment.

[0150] Through the above technical solution, the embodiment can more accurately identify and distinguish the interference components in the residual magnetic memory signal data from the real fatigue damage signals. By first extracting fatigue damage features from the preliminary comparison result and further comparing them with the magnetic memory signal modulation features representing the material force-magnetic coupling response, it can effectively avoid misjudging real fatigue damage features as interference. This multi-stage identification mechanism makes the identification of interference components more accurate, thereby improving the reliability of fatigue damage assessment and ensuring that the detection system can still accurately identify the early fatigue damage state of precision hardware in a complex interference environment.

[0151] The beneficial effects of implementing the embodiments of the present application include that the embodiments of the present application first acquire a theoretical magnetic memory signal of a reference block, perform magnetic memory signal measurement on the reference block to obtain a reference measurement signal, then calculate an adjustment factor reflecting overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal, correct the initial magnetic memory signal according to the adjustment factor to obtain a target magnetic memory signal, and finally perform fatigue damage assessment processing on the precision hardware under test according to the target magnetic memory signal to obtain a fatigue damage assessment result, so that the fatigue damage detection can be realized by adjusting the magnetic memory signal through the adjustment factor, thereby improving the accuracy.

[0152] As Figure 2As shown, the embodiment of the present application also provides a precise hardware fatigue damage detection system, comprising:

[0153] The reference theoretical feature acquisition module 701 is configured to acquire a theoretical magnetic memory signal of the reference block.

[0154] The reference measurement module 702 is configured to measure the magnetic memory signal of the reference block to obtain a reference measurement signal.

[0155] The deviation calculation module 703 is configured to calculate an adjustment factor reflecting the overall measurement deviation according to the reference measurement signal and the theoretical magnetic memory signal.

[0156] The signal correction module 704 is configured to correct the initial magnetic memory signal to obtain a target magnetic memory signal according to the adjustment factor, wherein the initial magnetic memory signal is obtained by measuring the magnetic memory signal of the precise hardware to be measured.

[0157] The fatigue damage evaluation module 705 is configured to perform fatigue damage evaluation processing on the precise hardware to be measured according to the target magnetic memory signal to obtain a fatigue damage evaluation result.

[0158] The above-mentioned content in the method embodiment is applicable to the system embodiment, the system embodiment specifically implements the same functions as the above-mentioned method embodiment, and achieves the same beneficial effects as the above-mentioned method embodiment.

[0159] The embodiments described in the embodiments of the present application are for more clearly illustrating the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. It is known to those skilled in the art that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

Claims

1. A method for detecting fatigue damage in precision hardware, characterized in that, Includes the following steps: Obtain the theoretical magnetic memory signal of the reference block; The reference block is subjected to magnetic memory signal measurement to obtain the reference measurement signal; Based on the reference measurement signal and the theoretical magnetic memory signal, an adjustment factor reflecting the overall measurement deviation is calculated; The initial magnetic memory signal is corrected according to the adjustment factor to obtain the target magnetic memory signal. The initial magnetic memory signal is obtained by measuring the magnetic memory signal of the precision hardware part under test. Based on the target magnetic memory signal, fatigue damage assessment is performed on the precision hardware part under test to obtain fatigue damage assessment results. The step of calculating the adjustment factor reflecting the overall measurement deviation based on the reference measurement signal and the theoretical magnetic memory signal includes: Extract the first spatial distribution of the theoretical magnetic memory signal and the second spatial distribution of the reference measurement signal; The first spatial distribution and the second spatial distribution are compared to determine the spatial deviation; Based on the spatial deviation, the reference measurement signal is spatially corrected; The adjustment factor is calculated based on the theoretical magnetic memory signal and the spatially corrected reference measurement signal; The step of performing fatigue damage assessment on the precision hardware part under test based on the target magnetic memory signal to obtain fatigue damage assessment results includes: Generate a load application command, which is used to apply a localized, gradual mechanical load to the evaluation area of ​​the precision hardware component under test; After applying a localized gradual mechanical load, magnetic memory signals are acquired in the area to be evaluated to obtain initial load and signal response data. Feature analysis is performed on the initial load and signal response data to obtain response characteristics that characterize the fatigue accumulation mode of the material; The precision hardware component under test is evaluated based on the target magnetic memory signal, the response characteristics, and the preset judgment criteria to obtain the fatigue damage evaluation result. The characteristic analysis of the initial load and signal response data to obtain response characteristics characterizing the material fatigue accumulation mode includes: After the localized gradual mechanical load is removed, magnetic memory signal is acquired in the area to be evaluated to obtain residual magnetic memory signal data. Based on the residual magnetic memory signal data and the initial load and signal response data, target load and signal response data are generated; The response features are extracted from the target load and signal response data.

2. The method according to claim 1, characterized in that, The step of generating target load and signal response data based on the residual magnetic memory signal data and the initial load and signal response data includes: Component analysis is performed on the residual magnetic memory signal data to identify interference components introduced by transient mechanical vibration or local magnetic field disturbance, including non-fatigue noise or interference signals. Based on the interference components, the initial load and signal response data are descrambled to obtain the target load and signal response data.

3. The method according to claim 2, characterized in that, The component analysis of the residual magnetic memory signal data to identify interference components introduced by transient mechanical vibration or local magnetic field disturbance includes: Magnetic memory signal acquisition is performed on a reference region to obtain reference region signal data. The reference region is adjacent to the region to be evaluated and is free from fatigue damage. Based on the signal data of the reference area, determine the interference characteristic parameters; The interference characteristic parameters are compared with the residual magnetic memory signal data to obtain the first comparison result; Based on the first comparison result, the interference component is identified from the residual magnetic memory signal data.

4. The method according to claim 3, characterized in that, The step of identifying the interference component from the residual magnetic memory signal data based on the first comparison result includes: A vibration application command is generated, which is used to apply mechanical vibration with a preset amplitude and frequency to the area to be evaluated; After applying mechanical vibration, magnetic memory signals are acquired in the area to be evaluated to obtain dynamic response magnetic memory signal data. The dynamic response magnetic memory signal data is subjected to feature analysis processing to obtain magnetic memory signal modulation features that are synchronized with the mechanical vibration. The magnetic memory signal modulation features are used to characterize the mechanical-magnetic coupling response of the material. Based on the first comparison result and the modulation characteristics of the magnetic memory signal, the interference component is identified from the residual magnetic memory signal data.

5. The method according to claim 4, characterized in that, The dynamic response magnetic memory signal data is subjected to feature analysis processing to obtain the magnetic memory signal modulation features synchronized with the mechanical vibration, including: Response analysis was performed on the dynamic response magnetic memory signal data to obtain the nonlinear response under different mechanical vibration intensities; Based on the nonlinear response, spectral analysis is performed on the dynamic response magnetic memory signal data to extract the target harmonic components of the mechanical vibration frequency. The target harmonic components include the second harmonic component or the third harmonic component. The target harmonic component is used as the modulation feature of the magnetic memory signal.

6. The method according to claim 4, characterized in that, The step of identifying the interference component from the residual magnetic memory signal data based on the first comparison result and the magnetic memory signal modulation characteristics includes: Based on the first comparison result, fatigue damage features are extracted from the residual magnetic memory signal data; The fatigue damage characteristics are compared with the magnetic memory signal modulation characteristics to obtain a second comparison result; Based on the second comparison result, the signal components in the residual magnetic memory signal data that do not have magnetic memory signal modulation characteristics are taken as the interference components.

7. A precision hardware fatigue damage detection system, characterized in that, include: The reference theory feature acquisition module is used to acquire the theoretical magnetic memory signal of the reference block; A reference measurement module is used to perform magnetic memory signal measurement on the reference block to obtain a reference measurement signal; The deviation calculation module is used to calculate an adjustment factor reflecting the overall measurement deviation based on the reference measurement signal and the theoretical magnetic memory signal. The signal correction module is used to correct the initial magnetic memory signal according to the adjustment factor to obtain the target magnetic memory signal. The initial magnetic memory signal is obtained by measuring the magnetic memory signal of the precision hardware part under test. The fatigue damage assessment module is used to perform fatigue damage assessment processing on the precision hardware part under test based on the target magnetic memory signal, and obtain the fatigue damage assessment result. The step of calculating the adjustment factor reflecting the overall measurement deviation based on the reference measurement signal and the theoretical magnetic memory signal includes: Extract the first spatial distribution of the theoretical magnetic memory signal and the second spatial distribution of the reference measurement signal; The first spatial distribution and the second spatial distribution are compared to determine the spatial deviation; Based on the spatial deviation, the reference measurement signal is spatially corrected; The adjustment factor is calculated based on the theoretical magnetic memory signal and the spatially corrected reference measurement signal; The step of performing fatigue damage assessment on the precision hardware part under test based on the target magnetic memory signal to obtain fatigue damage assessment results includes: Generate a load application command, which is used to apply a localized, gradual mechanical load to the evaluation area of ​​the precision hardware component under test; After applying a localized gradual mechanical load, magnetic memory signals are acquired in the area to be evaluated to obtain initial load and signal response data. Feature analysis is performed on the initial load and signal response data to obtain response characteristics that characterize the fatigue accumulation mode of the material; The precision hardware component under test is evaluated based on the target magnetic memory signal, the response characteristics, and the preset judgment criteria to obtain the fatigue damage evaluation result. The characteristic analysis of the initial load and signal response data to obtain response characteristics characterizing the material fatigue accumulation mode includes: After the localized gradual mechanical load is removed, magnetic memory signal is acquired in the area to be evaluated to obtain residual magnetic memory signal data. Based on the residual magnetic memory signal data and the initial load and signal response data, target load and signal response data are generated; The response features are extracted from the target load and signal response data.

Images