A method and system for multi-point stress synchronous testing of a planetary gearbox

Abstract

The application provides a kind of planetary gearbox multi-point stress synchronous test method and system, it is related to mechanical structure state monitoring and fault diagnosis technical field;The application realizes the three-dimensional space positioning of micro crack source on high-speed rotating planet wheel under the coordinate system of planet wheel body by synchronously collecting acoustic emission waveform and planet carrier motion parameter, constructing dynamic positioning model fusing Doppler compensation;At the same time, by polarizing analysis of acoustic emission longitudinal wave to extract polarized main direction vector, and carrying out vector matching analysis with crack point local geometric coordinate system obtained from three-dimensional digital model, online quantitative discrimination of crack mechanics mode as tensile dominant type or shear dominant type is realized;The application promotes planetary gearbox fault diagnosis from general abnormal alarm to diagnosis level of fixed wheel, fixed point and qualitative diagnosis, and improves the intelligent level and engineering effectiveness of predictive maintenance.

Description

Technical Field

[0001] This invention relates to the field of mechanical structure condition monitoring and fault diagnosis technology, specifically to a method and system for synchronous testing of multi-point stress in a planetary gearbox. Background Technology

[0002] As a core transmission component of high-reliability equipment such as wind power generation and new energy vehicles, early damage monitoring of critical internal components (such as planetary gears) in planetary gearboxes is crucial for preventing catastrophic failures. Acoustic emission technology, due to its extreme sensitivity to transient elastic waves released by localized damage such as micron-level crack propagation, has become a research hotspot for early fault warning.

[0003] However, applying acoustic emission technology to online monitoring of planetary gearboxes presents several challenges. Planetary gears, undergoing both revolution and rotation during operation, are typical sources of moving acoustic emission. When the acoustic emission signal propagates from the moving crack source to the sensor array fixed to the gearbox housing, its propagation path length and direction change in real time, and the Doppler effect is present. Traditional time-difference localization algorithms based on fixed sound speeds and static sensor arrays completely fail, making it impossible to determine which planetary gear the damage occurred on and its specific location. Even if an acoustic emission signal is detected, current technology can only provide scalar parameters such as event counts and energy, failing to determine the mechanical properties of the crack. Is the crack a shear-type crack originating from tooth surface contact fatigue or a tensile crack originating from tooth root bending fatigue? Is its propagation mode open, slip-open, or tear-open? The lack of differentiation between crack orientation and propagation mode prevents accurate root cause analysis and hinders precise guidance for subsequent maintenance decisions, such as whether immediate shutdown is necessary and what repair process to adopt.

[0004] Based on this, the present invention proposes a method and system for synchronous testing of multi-point stress in planetary gearboxes. Summary of the Invention

[0005] The present invention aims to solve the technical problems in the prior art of being unable to spatially locate microcracks on moving planetary gears of planetary gearboxes and being unable to determine the mechanical mode of cracks, thereby providing a method and system for synchronous multi-point stress testing of planetary gearboxes that can realize the integration of dynamic crack location and mechanism determination.

[0006] The technical solution adopted in this invention is as follows:

[0007] A method for synchronous multi-point stress testing of a planetary gearbox, the method comprising:

[0008] S1: Synchronously acquire acoustic emission waveform data, planetary carrier rotation speed signal, and planetary carrier real-time phase angle signal; wherein, the acoustic emission waveform data is acquired by an acoustic emission sensor array arranged on the planetary gearbox housing, the acoustic emission sensor array contains at least four acoustic emission sensors, and at least one of the acoustic emission sensors is a triaxial sensor; the planetary carrier rotation speed signal and the planetary carrier real-time phase angle signal are acquired by a rotation speed and phase acquisition unit;

[0009] S2: When the acoustic emission sensor array captures an acoustic emission event, dynamic positioning calculation is performed based on the time difference data of the acoustic emission event arriving at each acoustic emission sensor in the acoustic emission sensor array, the rotational speed signal of the planetary carrier, and the real-time phase angle signal of the planetary carrier. The dynamic positioning calculation introduces the kinematic model of the planetary gear relative to the housing, performs motion compensation on the time difference data, and outputs the position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gear in which the event occurred.

[0010] S3: Perform polarization analysis on the longitudinal wave component in the acoustic emission waveform data corresponding to the acoustic emission event, and calculate the principal polarization direction vector of the longitudinal wave component;

[0011] S4: Based on the position data output in step S2, determine the local geometric coordinate system corresponding to the position on the planetary gear; perform matching analysis between the polarization principal direction vector obtained in step S3 and the local geometric coordinate system, and generate and output crack diagnosis information based on the matching result; the crack diagnosis information includes at least the identifier of the planetary gear, the position data, and the crack mechanical mode discrimination result.

[0012] Furthermore, the dynamic positioning calculation in step S2 specifically includes:

[0013] S21: Based on the real-time phase angle signal of the planetary carrier, and according to the kinematic model of the planetary gear, calculate the theoretical relative position and theoretical relative velocity vector of the crack source relative to each acoustic emission sensor in the acoustic emission sensor array at the moment the acoustic emission event occurs, for a hypothetical crack source in the coordinate system of the planetary gear where the crack source is located.

[0014] S22: Construct a theoretical time difference function with the coordinates of the assumed crack source in the body coordinate system of the planetary gear as variables; the theoretical time difference function is used to calculate the theoretical propagation time of the acoustic emission event from the assumed crack source to each acoustic emission sensor in the acoustic emission sensor array; the calculation of the theoretical propagation time incorporates the effects of the theoretical relative position and the theoretical relative velocity vector;

[0015] S23: Based on the theoretical time difference function, calculate a set of theoretical propagation times corresponding to the variable, and derive the corresponding theoretical time difference data from the set of theoretical propagation times; iteratively compare the actual time difference data of the acoustic emission event arriving at each acoustic emission sensor with the theoretical time difference data, and with minimizing the error between the two as the optimization objective, solve for the value of the variable that minimizes the error; output the solved value of the variable as the position data.

[0016] Furthermore, step S3 specifically includes:

[0017] S31: Extract the longitudinal wave signal segment that first arrives from the acoustic emission waveform data collected by the triaxial sensor;

[0018] S32: Construct a covariance matrix for the extracted longitudinal wave signal segment, and perform eigenvalue decomposition on the covariance matrix;

[0019] S33: The direction of the eigenvector corresponding to the largest eigenvalue is determined as the principal polarization direction vector.

[0020] Further, in step S4, determining the local geometric coordinate system corresponding to the position on the generating planetary gear specifically involves: calling the pre-stored three-dimensional digital model of the generating planetary gear; inputting the position data output in step S2 into the three-dimensional digital model, querying and obtaining the normal vector and tangential vector of the surface position point corresponding to the position data; and constructing the local geometric coordinate system by the normal vector and the tangential vector.

[0021] Further, in step S4, the matching analysis of the polarization principal direction vector obtained in step S3 with the local geometric coordinate system specifically involves: calculating the first angle between the polarization principal direction vector and the normal vector in the local geometric coordinate system; calculating the second angle between the polarization principal direction vector and the tangential vector in the local geometric coordinate system; and determining whether the crack mechanical mode discrimination result is tensile-dominated or shear-dominated based on the relationship between the first angle and the second angle.

[0022] Furthermore, in step S1, the vibration acceleration signal of the planetary gearbox is also acquired synchronously; the method further includes:

[0023] S5: Based on the location data in the crack diagnosis information output in step S4, extract the sideband energy related to the planetary gear meshing frequency in the vibration acceleration signal within the corresponding time period;

[0024] S6: Compare the sideband energy with a preset energy threshold. If the sideband energy exceeds the energy threshold, add a load impact intensity warning label to the crack diagnosis information.

[0025] Furthermore, the method also includes:

[0026] S7: Continuously record and store the crack diagnosis information output in step S4 multiple times to form a historical diagnosis sequence;

[0027] S8: For crack events at the same location in the historical diagnostic sequence, count their frequency of occurrence and calculate the variance of the change in the direction angle of the polarization principal direction vector;

[0028] S9: If the frequency exceeds a preset frequency threshold and the variance of the direction angle change is less than a preset variance threshold, then generate a crack stability propagation early warning information about the location data.

[0029] Further, in step S22, the theoretical time difference function is constructed in the following way: based on the kinematic model of the planetary gear and the theoretical relative position, the theoretical propagation distance of the sound wave from the motion source point to each acoustic emission sensor in the acoustic emission sensor array is calculated; combined with the component of the theoretical relative velocity vector in the propagation direction, the Doppler effect compensation is performed on the average propagation speed of the sound wave in the material to obtain the equivalent propagation speed; the theoretical propagation distance is divided by the equivalent propagation speed to obtain the theoretical propagation time to each acoustic emission sensor.

[0030] A multi-point stress synchronous testing system for planetary gearboxes, characterized in that it comprises:

[0031] The data acquisition module is used to synchronously acquire acoustic emission waveform data, planetary carrier rotation speed signal, and planetary carrier real-time phase angle signal; wherein, the acoustic emission waveform data is acquired by an acoustic emission sensor array arranged on the planetary gearbox housing, the acoustic emission sensor array includes at least four acoustic emission sensors, and at least one of the acoustic emission sensors is a triaxial sensor; the planetary carrier rotation speed signal and the planetary carrier real-time phase angle signal are acquired by a rotation speed and phase acquisition unit;

[0032] The dynamic positioning processing module is used to perform dynamic positioning calculations when the acoustic emission sensor array captures an acoustic emission event, based on the time difference data of the acoustic emission event arriving at each acoustic emission sensor in the acoustic emission sensor array, the rotational speed signal of the planetary carrier, and the real-time phase angle signal of the planetary carrier. The dynamic positioning calculation introduces the kinematic model of the planetary gears relative to the housing, performs motion compensation on the time difference data, and outputs the position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gear in which the event occurred.

[0033] The polarization analysis processing module is used to perform polarization analysis on the longitudinal wave component in the acoustic emission waveform data corresponding to the acoustic emission event, and calculate the principal polarization direction vector of the longitudinal wave component.

[0034] The diagnostic information generation and output module is used to determine the local geometric coordinate system of the corresponding position on the planetary gear based on the position data output by the dynamic positioning processing module; to perform matching analysis between the polarization principal direction vector obtained by the polarization analysis processing module and the local geometric coordinate system, and to generate and output crack diagnostic information based on the matching result; the crack diagnostic information includes at least the identifier of the planetary gear, the position data, and the crack mechanical mode discrimination result.

[0035] The beneficial effects of this invention are:

[0036] This invention achieves three-dimensional spatial positioning of microcrack sources on high-speed rotating planetary gears in the planetary gear body coordinate system by synchronously acquiring acoustic emission waveforms and planetary carrier motion parameters, and constructing a dynamic positioning model with Doppler compensation. Simultaneously, by performing polarization analysis on the acoustic emission longitudinal waves to extract the principal polarization direction vector, and then performing vector matching analysis with the local geometric coordinate system of the crack point obtained from the three-dimensional digital model, online quantitative discrimination of whether the crack mechanical mode is tensile-dominant or shear-dominant is achieved. This invention elevates planetary gearbox fault diagnosis from general abnormal alarms to a diagnostic level that is gear-specific, location-specific, and qualitative, improving the intelligence level and engineering effectiveness of predictive maintenance. Attached Figure Description

[0037] Figure 1 This is a flowchart of a multi-point stress synchronous testing method for a planetary gearbox according to an embodiment of the present invention;

[0038] Figure 2 This is an overall flowchart of a planetary gearbox multi-point stress synchronous test according to an embodiment of the present invention;

[0039] Figure 3 This is a flowchart of dynamic positioning calculation according to an embodiment of the present invention;

[0040] Figure 4 This is a block diagram of a planetary gearbox multi-point stress synchronous testing system according to an embodiment of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] like Figures 1-3 As shown in the figure, a method for synchronous multi-point stress testing of a planetary gearbox according to an embodiment of the present invention includes the following steps:

[0043] S1: Synchronously acquire acoustic emission waveform data, planetary carrier speed signal and planetary carrier real-time phase angle signal; wherein the acoustic emission waveform data is acquired by an acoustic emission sensor array arranged on the planetary gearbox housing, the acoustic emission sensor array contains at least four acoustic emission sensors, and at least one of the acoustic emission sensors is a triaxial sensor; the planetary carrier speed signal and planetary carrier real-time phase angle signal are acquired by the speed and phase acquisition unit.

[0044] In this embodiment of the invention, acoustic emission waveform data can be acquired by an acoustic emission sensor array arranged on the outer wall of the planetary gearbox housing, wherein the acoustic emission sensor array needs to meet the following requirements: ( The hardware configuration requirements specify the number of acoustic emission sensors, and the array must include at least one triaxial acoustic emission sensor, with the remainder being uniaxial acoustic emission sensors. All sensors are fixedly arranged on the outer wall of the planetary gearbox housing, evenly distributed around the meshing area of ​​the planetary gears, and tightly bonded to the housing surface using a coupling agent to ensure acoustic coupling and prevent acoustic emission signal attenuation. The position vector of each acoustic emission sensor in the fixed coordinate system O-xyz of the housing is denoted as... The values ​​are pre-calibrated. A three-dimensional acoustic emission sensor is used to collect acoustic emission waveform data in three-dimensional space (along the three orthogonal directions x, y, and z). A one-dimensional acoustic emission sensor is used to collect acoustic emission waveform data in one direction, mainly to obtain the time difference data of acoustic emission events arriving at different sensors. It works in conjunction with the three-dimensional acoustic emission sensor to form a multi-point spatial array, providing sufficient spatial calculation data for subsequent dynamic positioning. The planetary carrier speed signal and the planetary carrier real-time phase angle signal are uniformly collected by the speed and phase acquisition unit. The speed and phase acquisition unit is usually composed of a photoelectric encoder, proximity switch, or magnetoelectric speed sensor paired with a professional data acquisition card. This is a commonly used hardware in industry for acquiring the speed and phase of rotating components. This unit needs to be coaxially installed with the planetary carrier rotation shaft of the planetary gearbox, or installed on a transmission component rigidly connected to the planetary carrier, to ensure that the collected signals can truly reflect the actual motion state of the planetary carrier and avoid signal distortion due to transmission backlash. Furthermore, the collected speed and phase signals must be continuous real-time signals, not discrete single-point data, to ensure that the instantaneous occurrence of the acoustic emission event can be captured. Precise motion parameters of the planetary carrier.

[0045] In this invention, in addition to synchronously acquiring acoustic emission waveform data, planetary carrier rotation speed signal, and planetary carrier real-time phase angle signal, the invention also synchronously acquires vibration acceleration signal of the planetary gearbox. Specifically, the vibration acceleration signal can be acquired by a piezoelectric vibration acceleration sensor. This sensor must meet the actual operating conditions of the planetary gearbox, such as temperature resistance and vibration resistance. It is arranged on the same side of the planetary gearbox housing as the acoustic emission sensor array and must be directly facing the planetary gear meshing area to ensure that the acquired vibration signal can truly reflect the load changes during the planetary gear meshing process. The acquisition of this signal is completely synchronized with the basic acquisition object, without any additional hardware layout or acquisition control restrictions.

[0046] It should be noted that, in this embodiment of the invention, synchronous acquisition essentially means that the timestamps of all acquired objects are completely unified. That is, each set of data, including acoustic emission waveform, planetary carrier rotation speed, planetary carrier real-time phase angle, and vibration acceleration, corresponds to the same absolute time point. If the time is not synchronized, the planetary gear motion parameters will not match the acoustic emission event, and subsequent operations such as dynamic positioning and motion compensation based on the planetary gear kinematic model will be completely ineffective. Specifically, synchronous acquisition must meet three core requirements. First, hardware timestamps must be unified. All acquisition hardware (acoustic emission sensor array, rotation speed and phase acquisition unit, vibration acceleration sensor) must be connected to the same high-precision synchronous data acquisition instrument. The acquisition instrument provides a unified system clock to give all acquired data a unique absolute timestamp. The time synchronization accuracy must reach the microsecond level to adapt to the propagation characteristics of high-frequency transient signals of acoustic emission and avoid significant time difference errors caused by millisecond-level synchronization. Second, trigger synchronization must be established. A synchronization marking mechanism based on acoustic emission event triggering is established. When the amplitude of the waveform data acquired by the acoustic emission sensor array exceeds a preset trigger threshold... When the (collection trigger threshold) is reached, the system immediately triggers a synchronization flag and records the timing of the flag. And the actual time of the acoustic emission event The system collects instantaneous values ​​of the planetary carrier rotation speed, real-time phase angle, and vibration acceleration to achieve precise time matching between acoustic emission events and the motion state of the planetary gears. Thirdly, it adapts the sampling frequency. Different acquisition objects are matched with corresponding sampling frequencies according to their own signal characteristics, and all sampling timing is uniformly controlled by the synchronous acquisition instrument. Among them, the acoustic emission waveform data is a high-frequency transient elastic wave, and the sampling frequency is usually 1MHz~10MHz to ensure that the waveform characteristics of the acoustic emission signal can be completely captured. The planetary carrier rotation speed and phase signal are motion parameters of low-speed rotating components, and the sampling frequency is usually 100Hz~1kHz to meet the real-time requirements. The vibration acceleration signal is a medium-low frequency mechanical vibration signal, and the sampling frequency is usually 1kHz~10kHz to ensure that the meshing frequency and sideband characteristics of the planetary gears can be captured.

[0047] In this embodiment of the invention, all acquired data are represented as a continuous time series, wherein the acoustic emission waveform data acquired by the unidirectional sensor is denoted as... ( (Unit: V), the three-dimensional acoustic emission waveform data collected by the three-dimensional sensor (i=1) are denoted as follows: , , (Unit: V), corresponding to the x, y, and z axes of the fixed coordinate system of the box, respectively; the time when the i-th sensor receives a valid acoustic emission event is denoted as . (Unit: s), where the receiving time of the three-dimensional sensor is... As the benchmark for time difference calculation, the actual time difference data of the i-th sensor relative to the three-axis sensor is denoted as . (Unit: s), its calculation formula is: ( All actual time difference data constitute a set. The real-time rotational speed (continuous time series) of the planetary carrier is denoted as... (Unit: r / min), the real-time orbital angular velocity of the planetary carrier obtained from the rotational speed conversion is denoted as... (Unit: rad / s), the formula is: The real-time phase angle of the planetary carrier (continuous time series) is denoted as... (Unit: rad), based on the fixed coordinate system O-xyz of the box; the time of acoustic emission event. The corresponding instantaneous angular velocity and instantaneous phase angle of the planetary carrier are denoted as follows: , This is the core parameter for subsequent kinematic model calculation of the planetary gears. The vibration acceleration signal is represented by both vector and scalar methods; the real-time vibration acceleration vector is denoted as... (Unit: m / s) 2 Arranged along the meshing direction of the planetary gears, the components in the fixed coordinate system of the housing are as follows: The magnitude of the real-time vibration acceleration scalar is the vector magnitude, denoted as . (Unit: m / s) 2 The formula is The moment of acoustic emission event The sequence of vibration acceleration vectors for a duration T before and after the event is denoted as: This data is the core data for subsequent load impact strength analysis.

[0048] After the initial data acquisition is completed, all acquired data needs to undergo lightweight preprocessing. This preprocessing removes only significant environmental noise, retaining the core characteristics of the original signal and avoiding over-processing that could lead to the loss of valuable information. All preprocessed data retains its original timestamp information to ensure synchronization with subsequent steps. The preprocessing operations vary depending on the type of acquired data. Specifically, for acoustic emission waveform data, bandpass filtering can remove low-frequency electromagnetic interference and high-frequency mechanical noise from the environment. The filtering frequency band is typically 100kHz to 1MHz, which is the typical effective frequency band for acoustic emission signals. Simultaneously, the waveform is pre-amplified by 40dB to 60dB to ensure that the weak acoustic emission signal generated by the propagation of micron-sized cracks can be effectively identified. The preprocessed waveform data retains its original characterization symbols. For planetary carrier speed and phase signals, the acquired raw pulse signals can be shaped and de-jittered to remove stray pulses generated by hardware contact or electromagnetic interference. Then, pulse counting and phase calculation convert the raw signal into a continuous digital speed value. and phase angle value For vibration acceleration signals, high-frequency noise can be removed by low-pass filtering, while retaining the mid-to-low frequency vibration characteristics related to planetary gear meshing. At the same time, zero-drift correction can be performed on the signal to eliminate the static offset error of the sensor.

[0049] It is important to note that during this step, to ensure the accuracy, synchronization, and effectiveness of the collected data, the arrangement of the acoustic emission sensor array must avoid acoustic blind spots, and the sensor coverage must include the meshing areas of all planetary gears to ensure that acoustic emission events generated by any planetary gear can be detected at least... Multiple sensors are used to capture data, ensuring the feasibility of three-dimensional spatial positioning. The three-dimensional orthogonal directions of the three-dimensional acoustic emission sensors must be strictly consistent with the x, y, and z axes of the fixed coordinate system O-xyz of the enclosure to avoid directional deviations in subsequent polarization analysis and kinematic calculations for dynamic positioning. The installation of the rotational speed phase acquisition unit must ensure coaxiality; excessive coaxiality error will lead to phase angle acquisition distortion, thus affecting the calculation accuracy of the planetary gear kinematic model in the dynamic positioning step. All acquisition hardware must undergo pre-calibration before implementation, including sensitivity calibration of the acoustic emission sensors, accuracy calibration of the rotational speed phase acquisition unit, and amplitude calibration of the vibration acceleration sensor to ensure the accuracy of the acquired data. The bonding between the acoustic emission sensors and the enclosure surface must use a suitable coupling agent to ensure acoustic coupling between the two and avoid severe attenuation of the acoustic emission signal during propagation, which would affect the subsequent time difference data extraction and polarization analysis. The time synchronization accuracy of the synchronous acquisition system must reach the microsecond level, and it must be consistent with... and Accurate calibration is performed to prevent time base deviations from being transmitted to the dynamic positioning process, which could lead to errors in dynamic positioning calculations; trigger thresholds are also used. It can be precisely calibrated according to the actual operating conditions of the planetary gearbox.

[0050] S2: When the acoustic emission sensor array captures an acoustic emission event, dynamic positioning calculation is performed based on the time difference data of the acoustic emission event arriving at each acoustic emission sensor in the acoustic emission sensor array, the planetary carrier rotation speed signal, and the planetary carrier real-time phase angle signal. The dynamic positioning calculation introduces the kinematic model of the planetary gear relative to the housing, performs motion compensation on the time difference data, and outputs the position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gear in which the event occurred.

[0051] In this embodiment of the invention, the triggering condition for this step is that the amplitude of the acoustic emission waveform data collected by the acoustic emission sensor array exceeds a preset valid event triggering threshold. (Event determination threshold) At this point, it is determined that one valid acoustic emission event caused by the propagation of a microcrack in the planetary gear has been captured, and the occurrence time of this event is recorded as [time value missing]. This time must strictly correspond to the timestamp of the synchronous acquisition of multiple physical quantities in step S1 to ensure that the kinematic parameters match the time of the acoustic emission event. The input data for step S2 includes the actual time difference data of the acoustic emission event arriving at each sensor in the sensor array, Real-time angular velocity of the planetary carrier at any given moment With real-time phase angle Step S0 inputs the planetary gearbox structural parameters (number of planetary gears). Planetary frame radius Planetary gear pitch circle radius (etc.), the kinematic model of the planetary gears based on structural parameters, the three-dimensional digital model of each planetary gear, and the position vector of each sensor in the sensor array in the fixed coordinate system of the housing. The output of step S2 is the three-dimensional position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gear. Furthermore, by comparing and calculating multiple planetary gears, the location of the originating planetary gear can be indirectly determined. This positioning result is not a dynamic position under the fixed coordinate system of the housing, but directly corresponds to the actual physical position of the planetary gear, possessing direct engineering application value. The originating planetary gear refers to the specific planetary gear identified through dynamic positioning calculations as the crack source generating the acoustic emission signal in the analysis of a single acoustic emission event.

[0052] Specifically, dynamic positioning calculation includes:

[0053] S21: Based on the real-time phase angle signal of the planetary carrier, and according to the kinematic model of the planetary gear, calculate the theoretical relative position and theoretical relative velocity vector of the crack source relative to each acoustic emission sensor in the acoustic emission sensor array at the moment the acoustic emission event occurs, for a hypothetical crack source in the body coordinate system of the planetary gear.

[0054] This step is The real-time phase angle of the planetary carrier at time t is a dual reference in both time and space. Combined with the planetary gear kinematic model established in step S0, kinematic calculations are performed on the assumed crack source coordinates to obtain the real-time theoretical motion parameters of the assumed crack source relative to each sensor in the sensor array. Since the actual location of the crack source is unknown, assumptions and calculations need to be performed separately for all planetary gears in the gearbox. For the k-th planetary gear (… ), in its body coordinate system Assume a crack initiation point P with coordinates as follows: This coordinate serves as the core variable for subsequent iterative optimization. Ultimately, the actual location of the planetary gear is determined by the magnitude of the positioning error of each hypothetical source. All position and velocity calculations in this step are based on the box-shaped fixed coordinate system defined in step S0. coordinate system with planetary gear body Complete, ensuring spatial consistency of the kinematic solution. Specifically, first transform the assumed crack initiation point P from the planetary gear body coordinate system to the box-type fixed coordinate system, obtaining... The absolute position vector of point P in the fixed coordinate system of the box and absolute velocity vector This calculation incorporates the combined effects of the planetary gears' revolution and rotation, and the specific solution is based on the planetary gear kinematic equations in step S0, which forms the basis for subsequent relative parameter calculations. Then, using the fixed coordinate system of the box as a reference, the calculation is performed... At any given moment, assume the theoretical relative position vector of the crack initiation point P with respect to the i-th sensor. The formula is: ,in for Let the theoretical relative position vector of the crack initiation point P be assumed to be in meters, and its components be... ; for The absolute position vector of the crack source P in the fixed coordinate system of the box is assumed to be in meters. Let be the position vector of the i-th sensor in the fixed coordinate system of the housing, in meters. The sensor is fixed to the housing, and this vector is a pre-calibrated constant. Simultaneously calculate... At any given moment, assume the crack initiation point P is relative to the i-th sensor. Theoretical relative velocity vector The formula is: ,in for At any given moment, assume the crack source P is relative to the i-th sensor. The theoretical relative velocity vector, in units of m / s, has vector components of... , Let be the velocity vector of the i-th sensor, in m / s. Since the sensor is fixed to the housing, its velocity vector... Since it is a zero vector, the formula can be simplified to: .

[0055] The above calculations need to be performed separately for each sensor in the sensor array, and the final result is... The set of theoretical relative position vectors of the assumed crack source of the k-th planetary gear at time k relative to all sensors. and the theoretical set of relative velocity vectors .

[0056] S22: Construct a theoretical time difference function with the coordinates of the assumed crack source in the body coordinate system of the planetary gear as variables; the theoretical time difference function is used to calculate the theoretical propagation time of the acoustic emission event from the assumed crack source to each acoustic emission sensor in the acoustic emission sensor array; the calculation of the theoretical propagation time incorporates the effects of theoretical relative position and theoretical relative velocity vector.

[0057] The theoretical time difference function is constructed in the following way:

[0058] First, based on the kinematic model and theoretical relative position of the planetary gears, the theoretical propagation distance of the sound wave from the motion source point to each acoustic emission sensor in the acoustic emission sensor array is calculated. Second, by combining the component of the theoretical relative velocity vector in the propagation direction, the Doppler effect compensation is applied to the average propagation speed of the sound wave in the material to obtain the equivalent propagation speed. Finally, the theoretical propagation distance is divided by the equivalent propagation speed to obtain the theoretical propagation time to each acoustic emission sensor.

[0059] In this embodiment of the invention, the theoretical propagation distance is first calculated, and the acoustic emission signal propagates from the moving hypothetical crack source P to the fixed sensor. Its theoretical propagation distance for The magnitude of the theoretical relative position vectors of the two at time t is calculated using the following formula:

[0060] ;

[0061] in for It is assumed that the crack source P propagates to a fixed sensor at all times. The theoretical propagation distance, in meters; Theoretical relative position vector The modulus length; , , for The components of the x, y, and z axes in the fixed coordinate system of the box.

[0062] Simultaneously, a path is defined pointing from the assumed crack source P to the i-th sensor. The propagation direction unit vector Dimensionless, vector components are ,satisfy This unit vector is used for subsequent Doppler effect compensation.

[0063] Next, Doppler effect compensation is performed, and the equivalent propagation velocity is calculated. Specifically, since the crack source is a moving sound source, its velocity component in the direction of acoustic emission signal propagation will cause an apparent change in the propagation velocity of the sound wave received by the sensor (Doppler effect). Therefore, the average propagation velocity in the material needs to be calculated. After compensating for the Doppler effect, the equivalent propagation velocity is obtained, and the formula is:

[0064] ;

[0065] in for The equivalent propagation speed of the acoustic emission signal from the hypothetical crack source to the i-th sensor at any given time is expressed in m / s. Assume the scalar component of the theoretical relative velocity vector of the crack source in the propagation direction, with units of m / s. This component reflects the influence of the Doppler effect: if the crack source moves towards the sensor, this component is positive, and the equivalent propagation velocity increases; if the crack source moves away from the sensor, this component is negative, and the equivalent propagation velocity decreases.

[0066] Next, the theoretical propagation time is calculated, showing the acoustic emission signal propagating from the hypothetical crack source P to the i-th sensor. The theoretical propagation time is the ratio of the theoretical propagation distance to the equivalent propagation speed. The calculation formula is:

[0067] ;

[0068] Finally, a theoretical time difference function is constructed, using the three-dimensional sensor (denoted as i=1) in step S1 as the reference sensor, to calculate the time difference function of the i-th sensor ( The theoretical time difference relative to the reference sensor The formula is:

[0069] ;

[0070] Combined with the actual time difference data extracted in step S1 Constructing the theoretical time difference residual function This function is the residual function between the theoretical time difference and the actual time difference, and the formula is:

[0071] ;

[0072] because Based on the assumed crack initiation coordinates The solution yields the theoretical time difference function. It is a multivariate function with the assumed crack initiation coordinates as the only variable.

[0073] S23: Based on the theoretical time difference function, calculate a set of theoretical propagation times corresponding to the variables, and derive the corresponding theoretical time difference data from the set of theoretical propagation times; iteratively compare the actual time difference data of the acoustic emission event arriving at each acoustic emission sensor with the theoretical time difference data, and with minimizing the error between the two as the optimization objective, solve for the value of the variable that minimizes the error; output the value of the variable obtained by solving as the position data.

[0074] This step aims to minimize the error between the actual time difference data and the theoretical time difference data. It uses a gradient descent iterative algorithm to find the assumed crack source coordinates that minimize the residual. By comparing the calculation results of the multi-planetary gear, the precise location of the crack source and the planetary gear identifier are finally determined.

[0075] Specifically, a least-squares objective function is first constructed. To comprehensively consider the time difference residuals of all sensors, the sum of the squares of the theoretical time difference residual functions of M-1 sensors is used as the least-squares objective function, as shown in the formula:

[0076] ;

[0077] The least squares objective function is dimensionless, and its value is half of the sum of squared time difference residuals of all sensors, reflecting the magnitude of the positioning error of the assumed crack source coordinates; This is to perform a summation operation on all sensors from i=2 to i=M.

[0078] The value of this function reflects the magnitude of the positioning error of the assumed crack initiation coordinates. The optimization objective of this step is to find the crack initiation coordinates that minimize this function. .

[0079] Secondly, the gradient descent method is used for iterative solution. First, the objective function is calculated with respect to the assumed crack initiation coordinates. , , The gradient is calculated using the following formula:

[0080] ;

[0081] in The gradient vector of the objective function is dimensionless and reflects the rate of change and direction of change of the objective function at the current coordinate point. , , For the objective function pair , , The first partial derivative of is dimensionless; , , For theoretical time difference pairs , , The first-order partial derivative, in units of s / m, reflects the degree of influence of the assumed crack source coordinate change on the theoretical time difference.

[0082] Then, iteratively update the coordinates of the assumed crack initiation along the opposite direction of the gradient, using the following update formula:

[0083] ;

[0084] in The learning rate (iteration step size) is dimensionless and its value ranges from 1 to 2. It needs to be calibrated according to the actual working conditions to control the coordinate correction range of each iteration; , , The coordinates of the assumed crack source are given in the nth iteration. , , These are the assumed crack source coordinates for the (n+1)th iteration. , , This represents the value of the first-order partial derivative of the objective function at the nth iteration.

[0085] Simultaneously set the convergence threshold When the absolute value of the difference between the objective function values ​​of two adjacent iterations satisfies ( , When the objective function values ​​are the values ​​at the nth and (n+1th)th iterations, the iteration is considered converged. At this point, the coordinates of the (n+1th)th iteration are... This is the optimal coordinate solution for the assumed crack initiation of the planetary gear. The minimum value of the corresponding objective function is denoted as .

[0086] Finally, the location of the planetary gear and the final crack initiation point were determined. A complete calculation process was performed on all planetary gears within the gearbox, including theoretical relative position and velocity calculations, construction of theoretical time difference functions, and iterative optimization. This yielded the optimal coordinate solution and the corresponding minimum objective function for each planetary gear. All... The planetary gear number corresponding to the minimum value is the planetary gear identifier. ,Right now The optimal coordinate solution corresponding to this planetary gear This refers to the final position data of the crack source corresponding to the acoustic emission event in the coordinate system of the planetary gear body. .

[0087] S3: Perform polarization analysis on the longitudinal wave component in the acoustic emission waveform data corresponding to the acoustic emission event, and calculate the principal polarization direction vector of the longitudinal wave component.

[0088] This step involves polarization analysis of the longitudinal wave component in the acoustic emission signal to calculate the principal polarization vector of the longitudinal wave, providing a unique directional characteristic for crack mechanical mode identification. It should be noted that in the transient elastic wave released by the propagation of a microcrack in the planetary gear, the longitudinal wave is the first wave type to reach the sensor, and its polarization direction is highly consistent with the principal stress direction during crack propagation. By calculating the principal polarization vector of the longitudinal wave, the stress direction characteristics of crack propagation can be indirectly obtained, solving the technical problem of difficulty in identifying crack damage mechanisms in existing technologies.

[0089] Specifically, step S3 includes the following steps:

[0090] S31: Extract the longitudinal wave signal segment that first arrives from the acoustic emission waveform data collected by the three-dimensional sensor.

[0091] This step is the signal preprocessing stage. The propagation speed of acoustic emission longitudinal waves is much greater than that of transverse waves and surface waves. Therefore, the first signal segment to arrive in the acoustic emission signal received by the triaxial sensor is the pure longitudinal wave component. By extracting this signal segment through a limited time window, interference from other wave types can be effectively eliminated, ensuring the accuracy of subsequent polarization analysis.

[0092] Specifically, the effective time interval of the longitudinal wave signal is first determined, based on the timing of the acoustic emission event received by the triaxial sensor. Starting from the point, the duration is... time interval As the effective range of the longitudinal wave signal The longitudinal wave signal window length can be calibrated based on the sound velocity of the enclosure material and the sensor arrangement, with a range of values ​​ranging from [value missing]. .

[0093] Then, the original P-wave signal components are extracted. Within the effective range of the P-wave signal, the extracted amplitude exceeds the trigger threshold. The effective signal is used as the original longitudinal wave signal component, and the portion that does not reach the threshold is set to 0. The specific extraction formula is as follows:

[0094] ;

[0095] in , , These are the original longitudinal wave signal components in the x, y, and z axes, respectively. They are all continuous functions of time t, with units of V, and have non-zero values ​​only within the effective time interval.

[0096] The original P-wave signal contains a DC component introduced by sensor hardware and environmental electromagnetic interference, which can cause deviations in subsequent matrix calculations. Therefore, it is necessary to perform mean-reduction processing on the original P-wave signal components. First, the mean of each original P-wave signal component within the effective time interval is calculated: , , Then, the mean is removed based on the mean value to obtain the mean-removed P-wave signal components. , , :

[0097] ;

[0098] ;

[0099] ;

[0100] The mean of the signal after removing the mean is 0, and its covariance matrix can accurately characterize the correlation and fluctuation characteristics between three-dimensional signals.

[0101] To facilitate subsequent matrix calculations and eigenvalue decomposition, the continuous mean-free longitudinal wave signal is... , , According to the acoustic emission sampling frequency in step S1 Discretize the sample to obtain a discretized signal sequence: ,in denoted as the number of sampling points for the longitudinal wave signal, and j as the sampling point number. The discretized signal completely preserves the polarization characteristics of the original longitudinal wave.

[0102] S32: Construct a covariance matrix for the extracted longitudinal wave signal segments, and perform eigenvalue decomposition on the covariance matrix.

[0103] This step quantifies the linear correlation and energy distribution of the three-dimensional longitudinal wave signal by constructing a covariance matrix, and then extracts the principal energy direction (i.e. the principal polarization direction of the longitudinal wave) from the covariance matrix through eigenvalue decomposition.

[0104] Specifically, for discretized mean-free longitudinal wave signals and ( , First, calculate the covariance between any two signal components in any direction, using the following formula:

[0105] ;

[0106] Where the denominator is taken To ensure unbiased estimation and computational accuracy, when m=n, the covariance is the variance of the signal component, reflecting the signal's own fluctuation energy.

[0107] Based on the above covariance calculation results, a 3×3 order real symmetric positive definite covariance matrix is ​​constructed. The matrix is ​​a real symmetric positive definite matrix (the diagonal elements represent the variances of the components, all of which are positive; the off-diagonal elements are symmetric). The complete form of the matrix is:

[0108] ;

[0109] diagonal elements , , These represent the variances of the longitudinal wave signals along the x, y, and z axes, respectively, characterizing the inherent wave energy of the signal in each direction. A larger value indicates stronger signal energy in that direction. Off-diagonal elements (such as...) The value represents the degree of linear correlation between two signals. A positive value indicates a positive correlation, a negative value indicates a negative correlation, and a value of 0 indicates no linear correlation.

[0110] The principal polarization direction is the propagation direction where the energy of the longitudinal wave signal is most concentrated, and the unit eigenvector corresponding to the largest eigenvalue of the covariance matrix is ​​exactly the principal energy direction of the multidimensional signal. Therefore, for the covariance matrix... By performing eigenvalue decomposition, the principal polarization direction of the longitudinal wave signal can be extracted. The decomposition formula is as follows:

[0111] ;

[0112] in This is an eigenvalue diagonal matrix, with diagonal elements... , , Let be the eigenvalues ​​of the covariance matrix, satisfying The magnitude of the eigenvalue represents the strength of the signal energy in the corresponding direction; The eigenvector matrix and column vectors are... , , These are the unit eigenvectors corresponding to the eigenvalues, and the three eigenvectors are pairwise orthogonal. , ), satisfying the unit vector magnitude requirement ( ).

[0113] It should be noted that, , , The magnitude corresponds to the energy proportion of the longitudinal wave signal in the three orthogonal directions. The maximum eigenvalue corresponds to the direction in which the longitudinal wave signal energy is most concentrated, i.e., the principal polarization direction; , It is a secondary eigenvalue, corresponding to the secondary polarization direction, with a low energy proportion and no effect on the principal polarization direction; , , These are three orthogonal unit vectors, pointing in three orthogonal directions of the longitudinal wave signal energy distribution. Their directions are determined by the direction reference of the box's fixed coordinate system O-xyz, and are completely consistent with the coordinate systems of S1 and S2.

[0114] S33: Determine the direction of the eigenvector corresponding to the largest eigenvalue as the principal polarization direction vector.

[0115] Let the covariance matrix be... The largest eigenvalue is Its corresponding unit eigenvector is ,in , , Let X be the components of the eigenvector along the x, y, and z axes in the fixed coordinate system O-xyz of the box. Then, the principal polarization direction vector of the longitudinal wave component... The expression is:

[0116] ;

[0117] Principal polarization direction vector It is a dimensionless unit vector that satisfies This eliminates the influence of signal amplitude on direction, retaining only the core directional characteristics; the direction of this vector is highly consistent with the principal stress direction during the propagation of microcracks in the planetary gear, and its component form is determined based on the fixed coordinate system of the box, which is completely consistent with the aforementioned coordinate system reference.

[0118] S4: Based on the position data output in step S2, determine the local geometric coordinate system at the corresponding position on the planetary gear; perform matching analysis between the polarization principal direction vector obtained in step S3 and the local geometric coordinate system, and generate and output crack diagnosis information based on the matching results; the crack diagnosis information includes at least the identification of the planetary gear, position data, and crack mechanical mode discrimination results.

[0119] Specifically, determining the local geometric coordinate system at the corresponding position on the generating planetary gear involves: calling the pre-stored three-dimensional digital model of the generating planetary gear; inputting the position data output in step S2 into the three-dimensional digital model, querying and obtaining the normal vector and tangential vector of the surface position point corresponding to the position data; and constructing a local geometric coordinate system from the normal vector and the tangential vector.

[0120] In this embodiment of the invention, before conducting the matching analysis between the polarization principal direction vector and the local geometric coordinate system, it is necessary to first construct a local geometric coordinate system with the crack origin as the origin. This coordinate system is the core geometric reference for vector matching analysis. Its construction is based on the three-dimensional digital model of the planetary gear and consists of three core steps: position mapping, feature vector extraction, and coordinate system construction.

[0121] First, position mapping is performed, and the crack source location data output in step S2 is used... Import the 3D digital model of the planetary gear. Using the model's coordinate retrieval function, accurately locate the physical point corresponding to the planetary gear surface. Then, extract feature vectors: extract the surface normal vector and the surface tangential vector along the meshing motion direction of the crack initiation point from the 3D digital model, and normalize them to unit normal vectors. With tangential unit vector Furthermore, the normal and tangential unit vectors satisfy the requirements of unity and orthogonality, that is... , , The normal unit vector is perpendicular to the planetary gear surface and points outward from the gear body, while the tangential unit vector is along the direction of planetary gear meshing motion; finally, the origin is taken as the crack initiation point. normal unit vector for Axial and tangential unit vectors for The axis is determined by the right-hand screw rule. Axis (secondary tangential), constructing the local geometric coordinate system of the crack initiation. .

[0122] Due to the principal polarization direction vector The coordinate system is defined based on the fixed coordinate system O-xyz of the box body, while the normal and tangential unit vectors are based on the planetary gear body coordinate system. To ensure a consistent spatial reference for subsequent vector angle calculations, the principal polarization direction vector needs to be transformed to the planetary gear body coordinate system. This transformation process is based on the planetary gear kinematic model established in step S0, utilizing the timing of the acoustic emission event. The coordinate transformation matrix is ​​used to achieve this, and the transformation formula is as follows:

[0123] ;

[0124] in The unit vector of the principal polarization direction in the transformed planetary gear body coordinate system still satisfies the unitity requirement. ; for The 3×3 orthogonal coordinate transformation matrix at time step S2 is obtained from step S2. Planetary phase angle at time The solution yields a spatial relationship between the fixed coordinate system of the housing and the coordinate system of the planetary gear body. After coordinate transformation, the unit vectors of the principal polarization direction, normal direction, and tangential direction are all defined based on the coordinate system of the planetary gear body, providing a unified benchmark for subsequent calculations of spatial angles.

[0125] In one embodiment of the present invention, step S4, specifically performing a matching analysis between the polarization principal direction vector obtained in step S3 and the local geometric coordinate system, involves: calculating the first angle between the polarization principal direction vector and the normal vector in the local geometric coordinate system; calculating the second angle between the polarization principal direction vector and the tangential vector in the local geometric coordinate system; and determining whether the crack mechanics mode discrimination result is tensile-dominated or shear-dominated based on the relationship between the magnitude of the first angle and the second angle.

[0126] After completing the coordinate system transformation, a matching analysis between the principal polarization direction vector and the local geometric coordinate system is performed. The core of this analysis is to calculate the spatial angle between the transformed principal polarization direction unit vector and the normal and tangential unit vectors, i.e., the first angle. (The angle between the principal polarization direction and the normal) and the second angle (The angle between the principal polarization direction and the tangent) The angle calculation is based on the correspondence between the dot product of unit vectors and the included angle in spatial analytic geometry. The general principle is: the cosine of the angle between two spatial unit vectors is equal to the dot product of the two vectors. Based on this principle, the first included angle... The calculation formula is:

[0127] ;

[0128] in The unit is ° / rad, and the range of values ​​is [missing information]. The above formula takes the absolute value of the dot product because it only considers the degree of alignment of directions and does not consider the positive or negative direction of the vector. The smaller the value, the closer the principal stress direction of crack propagation is to the normal direction of the planetary gear surface, and the greater the possibility that the crack is dominated by tensile stress.

[0129] Second angle The calculation formula is:

[0130] ;

[0131] in The unit is ° / rad, and the range of values ​​is [missing information]. Similarly, take the absolute value of the dot product, considering only the degree of directional fit; The smaller the value, the closer the principal stress direction of crack propagation is to the tangential direction of the planetary gear surface, and the greater the likelihood that the crack is dominated by shear stress.

[0132] To ensure the accuracy of crack mechanical mode identification, the validity of the angle calculation results must first be determined, excluding cases where the vector direction has no obvious tendency (such as...). The judgment rule is as follows: ,in The threshold for the included angle difference can be determined by engineering experience and testing accuracy, and is usually set to a value of [value missing]. If the judgment rule is met, it indicates that the principal polarization direction has a significant tendency to the normal or tangential direction, and the judgment result is valid; if it is not met, it indicates that the crack is dominated by tensile and shear stresses and needs to be marked as a composite stress crack.

[0133] Based on the validity assessment, the crack mechanical mode is determined according to the comparison of the included angle size. The determination rule is simple, intuitive, and highly interpretable in engineering. Specifically, if... It was determined to be a tensile-dominant crack and marked as... The physical meaning of this is that the principal stress direction of crack propagation coincides with the normal direction of the planetary gear surface, and the crack is caused by normal tensile stress, such as bending tension at the root of the planetary gear teeth or contact tension on the gear surface. This type of crack propagates rapidly and poses a significant threat to equipment safety. It was determined to be a shear-dominant crack and marked as... The physical meaning of this is that the principal stress direction of crack propagation coincides tangentially with the surface of the planetary gear, and the crack is caused by tangential shear stress, such as the sliding shear of the tooth surface during planetary gear meshing, or the shear friction of bearing mating. Such cracks initially propagate slowly and are prone to fatigue cracking; if It was determined to be a composite stress-type crack and marked as... The physical meaning of this is that the crack is caused by a combination of tensile and shear stress, and it often occurs in the stress concentration area of ​​the planetary gear. Its propagation trend needs to be closely monitored.

[0134] The final output of this step is structured and standardized planetary gear crack diagnosis information. This information is a multi-dimensional dataset, which includes at least the planetary gear identification, crack initiation location data, and crack mechanical mode discrimination results. Additional information such as included angle calculation results and stress type descriptions can be added according to engineering operation and maintenance needs. The core output format is... .in, To identify the specific planetary gear where the crack is located, thus achieving "fixed gear" identification; The three-dimensional position data of the crack initiation in the coordinate system of the planetary gear body is used to determine the precise physical location of the crack and achieve "point positioning". This is used to identify the crack mechanics mode, with values ​​of T, S, and C, to clarify the stress-dominant type of the crack and achieve "qualitative" identification. The calculation results for the first and second included angles are used to help illustrate the degree of alignment between the principal polarization direction and the normal and tangential directions, reflecting the strength of stress dominance. Structured crack diagnosis information can be directly mapped onto the physical surface of the planetary gears, providing maintenance personnel with a clear and explicit core basis for developing targeted inspection and maintenance strategies.

[0135] For example, This indicates that a tensile-dominant crack exists at position (15.2, 8.6, 22.5) in the coordinate system of planetary gear No. 2. The principal polarization direction makes an angle of 12° with the normal and 78° with the tangent.

[0136] In one embodiment of the present invention, the method for synchronous multi-point stress testing of planetary gearboxes further includes:

[0137] S5: Based on the location data in the crack diagnosis information output in step S4, extract the sideband energy related to the planetary gear meshing frequency in the vibration acceleration signal within the corresponding time period.

[0138] After a microcrack occurs in a planetary gear, the crack source will generate an additional load impact during meshing. This impact will manifest as an abnormal increase in energy in the meshing frequency and surrounding sidebands of the vibration acceleration signal. By extracting and calculating the energy of this characteristic frequency band, the degree of meshing load impact borne by the crack source location can be quantified, making up for the lack of load impact assessment in simple crack location and mechanical discrimination.

[0139] Specifically, the first step in this process is to define the meshing frequency of the planetary gears and the related sidebands. The core of this process is to calculate the base meshing frequency of the target planetary gears based on the basic parameters of the gearbox transmission and the actual operating parameters of the planetary carrier at the time of the acoustic emission event, and to calibrate the sideband width in conjunction with the actual operating conditions of the equipment, and to delineate the characteristic analysis frequency band.

[0140] The frequency of planetary gear meshing (Unit: Hz) Real-time angular velocity of the planetary carrier (Unit: rad / s) The calculation of gear tooth count parameters accurately matches the actual operating state of the equipment at the time of the acoustic emission event. The calculation formula is as follows:

[0141] ;

[0142] in for Real-time orbital angular velocity of the planetary carrier (unit: rad / s); , , The numbers of teeth (dimensionless) for the sun gear, internal gear ring, and generating planet gears are inherent parameters of the equipment. This formula considers the dual meshing relationship between the planet gears and the sun gear and internal gear ring, and the calculation results are highly consistent with the actual meshing vibration frequency.

[0143] Based on the base engagement frequency, the sideband width is calibrated in conjunction with the equipment operating conditions. Low-speed, heavy-duty planetary gearboxes (such as those used in wind power) exhibit small speed fluctuations. The frequency range is 5~10Hz. High-speed, light-load planetary gearboxes (such as those used in new energy vehicles) experience relatively large speed fluctuations. Select 10~20Hz; based on the fundamental meshing frequency Extending outwards from the center to both sides The range of the sidebands related to the planetary gear meshing frequency is obtained as follows: Subsequent energy calculations will focus on this characteristic frequency band to avoid noise interference from invalid frequency bands.

[0144] For example, if calculation According to the working conditions, the calibration is performed. Then the sideband range is Subsequently, only the vibration energy within this interval will be calculated.

[0145] After defining the characteristic frequency band, characteristic vibration acceleration signals that precisely match the acoustic emission event are extracted to achieve a precise time correspondence between the instant of crack propagation and the impact of the meshing load. Signal extraction is based on the moment of the acoustic emission event. Centered on time, the duration is taken as The time interval is the characteristic vibration signal extraction interval, i.e. ,in It is necessary to ensure that there are at least 3 to 5 planetary gear meshing cycles, while avoiding excessively long cycles that introduce irrelevant vibrations. Typically, the cycle length is 0.1 to 1 second.

[0146] Within this extraction range, the amplitude of the preprocessed vibration acceleration scalar is... In the process, characteristic vibration acceleration scalar signals are extracted. The extraction rule is to directly extract the signal from the corresponding interval without additional filtering (S1 has already performed low-pass filtering and noise reduction). The expression is:

[0147] ;

[0148] In the formula Characteristic vibration signals containing only the impact of meshing loads during crack propagation (unit: m / s) 2 The signal (), which is the original signal for subsequent frequency domain transformation and energy calculation, has a time length of [missing information]. It is completely consistent with the extraction interval.

[0149] To facilitate subsequent frequency domain calculations, the vibration acceleration signal sampling frequency in step S1 is used. (1kHz~10kHz) for continuous characteristic vibration acceleration signals Discretization sampling is performed to obtain a discretized signal sequence. ( ),in The number of sampling points after discretization must be a positive integer. If the calculated result is not an integer, it can be fine-tuned. accomplish, The sampling points are numbered, and the discretized signal completely retains the frequency and energy characteristics of the continuous signal, meeting the accuracy requirements of engineering calculations.

[0150] The total energy of the sidebands is converted from the time domain to the frequency domain using a fast Fourier transform. Then, the frequency domain energy within the characteristic frequency bands is integrated and summed to achieve quantitative quantification of the load impact energy. First, the discretized characteristic vibration signal is analyzed... Perform a Fast Fourier Transform (FFT) to convert the time domain to the frequency domain. The conversion formula is as follows:

[0151] ;

[0152] in The k-th frequency point in the frequency domain Corresponding vibration amplitude (unit: m / s) 2 ), forming the frequency domain amplitude spectrum ; The frequency value at the k-th frequency point; coefficient This is the amplitude normalization coefficient, ensuring the consistency of the amplitude values ​​in the frequency domain and the time domain.

[0153] After FFT transformation, the frequency domain amplitude spectrum of the characteristic vibration signal is obtained. The spectrum reflects the distribution characteristics of vibration amplitude at different frequencies. The abnormal increase in amplitude within the sideband of the meshing frequency indicates the load impact characteristics at the crack initiation location.

[0154] In the frequency domain, the energy density of a vibration signal is the square of its amplitude. (unit: The total energy within a certain frequency range is the integral of the energy density within that range. For the discretized frequency domain amplitude spectrum, summation is used instead of integration to complete the engineering calculation. Combining this with the defined meshing frequency sideband range, the formula for calculating the total energy of the sideband is as follows:

[0155] ;

[0156] in Total energy of the meshing frequency-dependent sidebands (unit: ), which are the core output parameters of step S5; For the k-th frequency point Vibrational energy density; Here, represents the frequency resolution in the frequency domain (in Hz), and represents the distance between two adjacent frequency points. The summation range includes all frequencies falling within the sidebands. Frequency points within This ensures that only the energy related to the impact of the planetary gear meshing load is calculated.

[0157] This formula quantifies the load impact energy by accumulating the energy at all frequency points within the meshing frequency sideband. The larger the value, the stronger the impact of the planetary gear meshing load on the crack initiation location at the moment of acoustic emission, and the more significant the load-driven effect on crack propagation.

[0158] S6: Compare the sideband energy with the preset energy threshold. If the sideband energy exceeds the energy threshold, add a load impact intensity warning label to the crack diagnosis information.

[0159] The core criterion for this step is a preset energy threshold, including the basic energy threshold. With high-risk energy threshold Both require precise calibration based on the actual operating conditions and historical data of the planetary gearbox. There are no uniform fixed values. Calibration must follow four principles: equipment compatibility, operating condition adaptability, historical reference, and margin. At the same time, a standardized engineering method is adopted: First, the vibration acceleration signal of the equipment under normal operation for 30-50 hours is collected. Then, the sideband energy at different times is calculated according to the method in step S5 to obtain the energy statistical sequence. Next, the 95th percentile of the sequence is calculated and used as the basic energy threshold. This excludes accidental energy fluctuations under normal operating conditions; finally, based on the baseline threshold, a factor of 1.5 is applied. As a high-risk energy threshold It can be finely adjusted to 1.2 to 2.0 times depending on the importance of the equipment.

[0160] For example, the 95th percentile of the sideband energy during normal operation of a certain wind turbine planetary gearbox is... Then the calibration , .

[0161] The core judgment process in this step is the quantitative comparison between the total energy of the sideband and the graded threshold. This invention adopts a graded judgment logic, dividing the load impact intensity into three distinct risk levels: no anomaly, general impact anomaly, and high-risk impact anomaly. Each level corresponds to a unique energy comparison condition, and the judgment result is deterministic with no ambiguity. The specific judgment criterion is: when When the load is deemed to be without abnormal impact, the impact energy of the meshing load at the crack initiation location is within the reasonable range of normal equipment operation, and the risk of load-driven crack propagation is low; when When the load impact energy exceeds the normal upper limit but does not reach the high-risk threshold, the risk of crack propagation is increased, and load changes need to be monitored; when When an abnormal impact occurs, it is classified as a high-risk impact anomaly. At this point, the load impact is in a severely abnormal state, and the crack propagation rate will significantly accelerate under strong load impact. This poses a high safety risk to the equipment operation, requiring immediate warning and maintenance measures. During the quantitative comparison process, it is essential to ensure that the threshold values ​​used are completely matched to the current operating conditions of the equipment, and that only energy values ​​are used as the basis for judgment to guarantee the objectivity and consistency of the results.

[0162] After determining the risk level of the load impact intensity, a corresponding load impact intensity warning label must be added to the original crack diagnosis information. The identification can adopt a concise coding rule of single characters and Chinese annotations to achieve rapid identification and display of equipment operation and maintenance systems. The codes strictly correspond one-to-one with risk levels: N (Normal) corresponds to no abnormal load impact (load impact normal, no warning); A (Attention) corresponds to general impact anomaly (load impact slightly abnormal, requires attention); and W (Warning) corresponds to high-risk impact anomaly (load impact severely abnormal, requires warning). During the identification process, the integrity of the original crack diagnosis information must be ensured; no original parameters may be modified or deleted. Simultaneously, the actual total energy of the sidebands must be included. Incorporating these parameters into diagnostic information as supplementary references facilitates subsequent traceability and judgment processes.

[0163] The final output of this step is the structured final crack diagnosis information D' with added warning labels. It integrates the load impact risk level and quantified energy data based on the original diagnosis information, and its structured expression is: .

[0164] This information must meet structured and standardized requirements. All diagnostic information corresponding to acoustic emission events must maintain a consistent order of information items, units, and symbols, including complete crack characteristics such as fixed wheel, fixed point, qualitative, and load risk rating, with no missing information. The final diagnostic information can be directly linked to subsequent steps, serving as the basic data unit for constructing the historical diagnostic sequence in step S7, while also including warning indicators. Total energy of sidebands This provides a key reference for the load impact dimension in the crack propagation trend analysis of steps S8 and S9.

[0165] For example, This indicates that a tensile-dominant crack exists at position (15.2, 8.6, 22.5) in the coordinate system of planetary gear #2. The principal polarization direction forms an angle of 12° with the normal and 78° with the tangent. The meshing load impact at this location is abnormally severe (a warning is needed). The actual total energy in the sidebands is... .

[0166] In one embodiment of the present invention, the method for synchronous multi-point stress testing of planetary gearboxes further includes:

[0167] S7: Continuously record and store the crack diagnosis information output from step S4 multiple times to form a historical diagnosis sequence.

[0168] This step uses planetary gear identifier matching and crack source location distance calculation to filter out all records of the same crack source location from multiple sets of final crack diagnosis information, arrange them in ascending order according to the occurrence time of acoustic emission events, and construct a structured crack history diagnosis sequence.

[0169] Specifically, the first step is to filter the raw data, selecting from the final crack diagnosis information set at multiple time points. Select the same planetary gear identifier from the filter. All records yielded a subset of planetary wheel-level diagnostic information. This achieves precise positioning of the planetary gear dimension. Next, it performs matching of crack origin locations, using the crack origin location of any record within the subset. Using this as a baseline, calculate the spatial Euclidean distance to other records using the following formula:

[0170] ;

[0171] in The spatial distance between the crack initiation locations of events i and j is expressed in mm. If they are the same location of the crack source, then they are determined to be the same location of the crack source.

[0172] After location matching is completed, all diagnostic records for the same crack initiation location are sorted by the time of acoustic emission event occurrence. The records are sorted in ascending order to ensure temporal continuity. Finally, the sorted records are integrated into a structured crack history diagnostic sequence, expressed as:

[0173] ;

[0174] in The number of records in the sequence, i.e., the statistical period. Frequency of acoustic emission events from the same crack source within each group All data includes complete information such as crack location, mechanical mode, load impact, and energy at that point in time, with the statistical period... It can be calibrated according to the operating intensity of the equipment to ensure that the sequence can reflect the continuous trend of crack propagation.

[0175] S8: For crack events at the same location in the historical diagnostic sequence, count their frequency of occurrence and calculate the variance of the direction angle change of the principal polarization direction vector.

[0176] This step is based on the crack history diagnostic sequence constructed in step S7. The frequency statistics of acoustic emission events from the same crack source and the stability calculation of the principal polarization direction vector were completed. Two quantitative indicators were used to reflect the continuity of crack propagation and the stability of the principal stress direction, respectively, providing a dual quantitative basis for subsequent determination of stable propagation trend. The two indicators corroborate each other and avoid misjudgment caused by a single indicator.

[0177] The statistical rule for counting the frequency of acoustic emission events from the same crack source is to directly count the historical diagnostic sequence of cracks. The number of records in the data, that number represents the statistical period. Frequency of acoustic emission events at the same crack source location The core significance of the frequency index lies in determining the continuity of crack propagation. This indicates that repeated acoustic emission events related to crack propagation occur at the same location, suggesting that crack propagation is continuous and not a random, single initiation; if If the result is positive, it is determined to be a random event with no stable expansion trend, and no further vector stability analysis is required; among which... The preset event frequency threshold (usually 3~5) is calibrated based on the frequency of background sound emission events during normal operation of the equipment (the frequency of background events at the same location under normal operating conditions is much lower than this threshold).

[0178] The principal polarization vector reflects the direction of the principal stress during crack propagation. When the crack enters the stable propagation stage, the principal stress direction is highly stable, and the vector dispersion is extremely low. Its stability can be quantified by calculating the overall variance of the vector; the smaller the variance, the more stable the vector and the more fixed the crack propagation direction. Specifically, the vector stability calculation includes: firstly, from the crack history diagnostic sequence... Extract the principal polarization direction vector in the planetary gear body coordinate system corresponding to each set of records. All vectors are unit vectors, satisfying Secondly, through the formula , , Calculate the arithmetic mean of the three components of the vector x, y, and z within the statistical period to obtain the average principal polarization direction vector. This serves as the benchmark for variance calculation. Finally, based on the mean vector, the overall variance of all principal polarization direction vectors relative to the mean vector is calculated. The variance is the arithmetic sum of the variances of the three components, as shown in the formula:

[0179] ;

[0180] in The overall variance of the principal polarization direction vector is a dimensionless value ranging from [0,3]. A value closer to 0 indicates that the principal polarization direction vector of each event is closer to the average vector, and the principal stress direction of crack propagation is more stable. ( To preset the variance threshold, an empirical value is usually taken as...

[0181] If the value is ≤0.05, then the vector height is considered stable and the crack propagation direction is fixed.

[0182] After completing frequency statistics and variance calculation, a preliminary judgment was made using a combination of two indicators. The results were divided into three categories: firstly... and If both benchmarks are satisfied, proceed to step S9 for the final determination of stable extension; secondly... but The frequency met the standard, but the vector was unstable, indicating multi-directional microcrack initiation rather than stable propagation of a single crack; thirdly... The frequency did not meet the standard, and it was determined to be an occasional single crack event with no stable expansion trend.

[0183] S9: If the frequency exceeds the preset frequency threshold and the variance of the direction angle change is less than the preset variance threshold, then generate a crack stability propagation early warning information about the location data.

[0184] This step, based on the preliminary judgment results of the dual indicators in step S8, combined with the changing trend of load impact risk in the crack history diagnosis sequence, completes the final judgment on whether the crack has entered the stable propagation stage, outputs standardized and structured crack propagation early warning information, and provides targeted and implementable engineering operation and maintenance suggestions, realizing the transformation from quantitative analysis to engineering early warning.

[0185] Specifically, the final determination of stable crack propagation adopts a triple verification rule of frequency, variance, and load trend, and is based on the initial judgment that the dual indicators meet the following criteria: and Based on this, further analysis was conducted on the load impact warning signs in the sequence. Total energy of sidebands The changing trend: If the two benchmarks are satisfied, and the sequence... It shows an upgrading trend of N→A→W, or It shows a continuous upward trend (the latter event) If the value is greater than the previous value, the crack is considered to have entered a stable propagation stage, and a warning indicator should be output. If both benchmarks are satisfied, but... Always N and If there is no obvious upward trend, it is determined that the crack propagation is slow and the load driving effect is weak. Therefore, no warning will be issued for the time being, and the statistical cycle needs to be shortened for continuous monitoring.

[0186] This invention employs a simple single-character and Chinese annotation encoding rule to set a crack stability propagation early warning indicator. The coding logic is consistent with that of the preceding load impact warning sign, which facilitates identification and display by the equipment operation and maintenance system. The specific coding rules are as follows: if the crack is determined to have entered the stable expansion stage, the warning sign is W (Warning) (the crack is expanding stably and an immediate warning and operation and maintenance measures should be taken); if the crack is determined to have an unclear expansion trend, the warning sign is M (Monitor) (the crack is expanding slowly and a shorter period of continuous monitoring is required).

[0187] The judgment results, early warning indicators, and core features of historical sequences are integrated into structured crack propagation early warning information. The early warning information should include four parts: core crack characteristics, trend quantification indicators, early warning results, and operation and maintenance suggestions, to ensure the completeness and engineering guidance of the information. The structured expression is as follows:

[0188] .

[0189] in , , To identify the core characteristics of cracks, clarify the fixed wheel, fixed point, and qualitative characteristics of the early warning object; For trend quantification indicators, specific quantitative values ​​for frequency and variance are given as the quantitative basis for judgment; To illustrate the load variation trend, describe the variation trend of load impact indicators and energy. For the early warning results, provide extended early warning indicators; the operation and maintenance suggestions should match the judgment results and be specific and implementable.

[0190] For different warning indicators, standardized engineering operation and maintenance recommendations can be formulated. For warning indicator W, the equipment operating load should be immediately reduced to prevent heavy-load operation from exacerbating crack propagation, the equipment inspection cycle should be shortened, the detection frequency of the crack initiation location should be increased, and shutdown maintenance should be arranged as soon as possible in conjunction with the equipment operation plan, including grinding, welding, or replacing components to repair the crack. For warning indicator M, the crack initiation location should be included in the key monitoring target, the statistical cycle should be shortened by 50% for continuous tracking, and the load impact status should be monitored in real time. If... If an upgrade is initiated, an early warning process will be activated immediately. During routine inspections, the surface condition of the affected area will be carefully checked to observe for the initiation of macroscopic cracks.

[0191] In one embodiment of the present invention, the multi-point stress synchronous testing method for planetary gearbox of the present invention further includes an initialization step before step S1: S0: Input the structural parameters of the planetary gearbox, the structural parameters including the number of planetary gears, the pitch circle radius of the planetary gears, and the radius of the planet carrier; establish the kinematic model of the planetary gears based on the structural parameters, and generate the three-dimensional digital model of each planetary gear.

[0192] S0 is the core initialization step in the multi-point stress synchronous testing method for planetary gearboxes. Its core objective is to accurately acquire all structural parameters of the planetary gearbox, establish a 1:1 matching kinematic model of the planetary gears with the physical equipment, and generate high-precision three-dimensional digital models of each planetary gear. This provides unified, accurate, and traceable model and data support for key aspects such as motion compensation for dynamic positioning calculations, coordinate unification for polarization analysis, and construction of local geometric coordinate systems for crack diagnosis, thereby fundamentally ensuring the calculation accuracy and engineering application effectiveness of the entire testing method.

[0193] Step S0 first involves acquiring and standardizing the structural parameters of the planetary gearbox. As the core data source for modeling, the accuracy of these structural parameters directly determines the reliability of the kinematic and 3D digital models. This step requires the comprehensive acquisition of four categories of inherent characteristic parameters: design geometric parameters, transmission kinematic parameters, material mechanical parameters, and basic operating parameters. These include core geometric parameters such as the number of teeth of the planetary gears, sun gear, and internal gear ring, the pitch circle radius of the planetary gears, and the radius of the planetary carrier; kinematic parameters such as the transmission ratio between the sun gear and the planetary carrier, and between the internal gear ring and the planetary carrier; mechanical parameters such as the elastic modulus, Poisson's ratio, and material sound velocity of core components such as the planetary gears and the gearbox housing; and basic operating parameters such as the rated speed of the planetary carrier and the meshing range of the planetary gears. The parameter acquisition is based on the original equipment manufacturer's design drawings, technical manuals, nameplates, and other materials, supplemented by on-site measurement and verification using high-precision equipment such as laser rangefinders and coordinate measuring machines. The measurement accuracy of core geometric parameters is no less than 0.01mm. When the deviation between the original parameter value and the measured value exceeds 0.1%, the measured value shall prevail. All verified parameters must be listed in a standardized "Planetary Gearbox Basic Parameter Table" according to the parameter name, symbol, value, unit, acquisition method, and verification result. This table serves as the sole data basis for all calculations and modeling and must be updated in a timely manner with equipment maintenance and modification to achieve traceable parameter management.

[0194] Based on the standardized and verified structural parameters, a dedicated kinematic model for the planetary gears needs to be established to address the challenge of dynamic positioning of the acoustic emission source caused by the combined revolution and rotation of the planetary gears. The modeling process first clarifies two coordinate systems: one is the fixed coordinate system O-xyz for the gearbox housing, which is set as the global reference coordinate system, with its origin at the geometric center of the planetary gearbox and its z-axis along the revolution axis of the planetary carrier. The positions of the gearbox housing and all sensors are calibrated based on this coordinate system; the other is the planetary gear body coordinate system. As a local coordinate system for a single planetary gear, with its origin at the geometric center of the planetary gear, this coordinate system undergoes composite motion with the planetary gear, and the position data of the crack source is output based on this coordinate system. Based on the dual coordinate system and combined with the classic kinematic formulas of planetary gear transmission, a 3×3 orthogonal coordinate transformation matrix with the real-time phase angle of the planetary carrier as the variable is derived. The calculation formulas for the revolution angular velocity and rotation angular velocity of the planetary gear under the angular velocity of the planetary carrier are solved, clarifying their proportional relationship. Further, the formulas for calculating the absolute position vector and absolute velocity vector of the assumed crack source from the planetary gear body coordinate system to the fixed coordinate system of the gearbox are derived, forming a complete kinematic model system. After the model is constructed, it is compiled into a "Kinematic Model Manual for Planetary Gearboxes" containing coordinate system definition diagrams, formula derivation processes, coordinate transformation matrix lookup tables, and calculation examples. Verification is conducted through ADAMS virtual simulation. The model is considered valid when the position and velocity deviations between the calculated model results and the simulation results do not exceed 1%. This model provides core formula support for solving the theoretical relative position and velocity, compensating for the Doppler effect, and correcting motion time difference in dynamic positioning calculations.

[0195] Step S0 requires the simultaneous generation of high-precision 3D digital models for all planetary gears within the gearbox. This model serves as the sole geometric basis for constructing the local geometric coordinate system in crack diagnosis, directly determining the accuracy of crack mechanical mode identification. Modeling can utilize professional mechanical 3D modeling software such as UG, SolidWorks, and CATIA. A parametric 3D model matching the physical part at a 1:1 scale should be created for each planetary gear, with a geometric dimensional error not exceeding 0.01mm. For critical areas prone to crack initiation, such as the tooth surface, tooth root, and gear body surface, the surface curvature, transition fillets, and other structural features should be meticulously reproduced according to the actual machining precision of the equipment. Furthermore, the local coordinate system of the model must be consistent with the planetary gear body coordinate system. Complete overlap ensures a unified coordinate system throughout the entire process. The generated 3D digital model is not merely a geometric model, but an engineering model with precise position mapping and surface vector retrieval capabilities. Inputting any 3D coordinate from the planetary gear body coordinate system into the model allows for precise positioning to the corresponding point on the physical surface of the planetary gear, achieving direct conversion from digital coordinates to physical position. Simultaneously, the model pre-stores the normal and tangential unit vectors for any point on the planetary gear surface. The tangential vector follows the direction of planetary gear meshing motion, and the two vectors satisfy orthogonality and unitity requirements. Inputting position coordinates allows for real-time retrieval of this vector data, providing direct data support for the construction of a local geometric coordinate system. All planetary gear models are named according to the planetary gear identifier, output in common engineering formats such as STEP and IGES, and verified through field measurements using a coordinate measuring machine. A model is considered valid if the deviation between the retrieved normal and tangential vectors and the measured vectors does not exceed 1°.

[0196] In terms of iterative updates, when the physical state of the planetary gearbox changes due to maintenance, modification, or replacement of planetary gears, the entire process of step S0 must be iteratively updated immediately. This involves re-acquiring and verifying the changed structural parameters, correcting the kinematic model, regenerating the three-dimensional digital model of the replaced or repaired parts, and updating the corresponding standardized documents after verification to ensure that the model and parameters always remain consistent with the actual physical state of the equipment.

[0197] Corresponding to the above embodiments, the present invention also proposes a multi-point stress synchronous testing system for planetary gearboxes.

[0198] like Figure 4 As shown, the planetary gearbox multi-point stress synchronous testing system according to an embodiment of the present invention includes a data acquisition module, a dynamic positioning processing module, a polarization analysis processing module, and a diagnostic information generation and output module.

[0199] The data acquisition module is used to synchronously acquire acoustic emission waveform data, planetary carrier speed signal, and planetary carrier real-time phase angle signal. The acoustic emission waveform data is acquired by an acoustic emission sensor array arranged on the planetary gearbox housing. The acoustic emission sensor array contains at least four acoustic emission sensors, and at least one of the acoustic emission sensors is a triaxial sensor. The planetary carrier speed signal and the planetary carrier real-time phase angle signal are acquired by the speed and phase acquisition unit.

[0200] The dynamic positioning processing module is used to perform dynamic positioning calculations when the acoustic emission sensor array captures an acoustic emission event, based on the time difference data of the acoustic emission event arriving at each acoustic emission sensor in the acoustic emission sensor array, the planetary carrier rotation speed signal, and the planetary carrier real-time phase angle signal. The dynamic positioning calculation introduces the kinematic model of the planetary gears relative to the housing, performs motion compensation on the time difference data, and outputs the position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gears in which the event occurred.

[0201] The polarization analysis processing module is used to perform polarization analysis on the longitudinal wave component in the acoustic emission waveform data corresponding to the acoustic emission event, and calculate the principal polarization direction vector of the longitudinal wave component.

[0202] The diagnostic information generation and output module is used to determine the local geometric coordinate system at the corresponding position on the planetary gear based on the position data output by the dynamic positioning processing module; it performs matching analysis between the polarization principal direction vector obtained by the polarization analysis processing module and the local geometric coordinate system, and generates and outputs crack diagnostic information based on the matching results; the crack diagnostic information includes at least the identification of the planetary gear, its position data, and the crack mechanical mode discrimination result.

[0203] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for synchronous multi-point stress testing of a planetary gearbox, characterized in that, The method includes: S1: Synchronously acquire acoustic emission waveform data, planetary carrier rotation speed signal, and planetary carrier real-time phase angle signal; wherein, the acoustic emission waveform data is acquired by an acoustic emission sensor array arranged on the planetary gearbox housing, the acoustic emission sensor array contains at least four acoustic emission sensors, and at least one of the acoustic emission sensors is a triaxial sensor; the planetary carrier rotation speed signal and the planetary carrier real-time phase angle signal are acquired by a rotation speed and phase acquisition unit; S2: When the acoustic emission sensor array captures an acoustic emission event, dynamic positioning calculation is performed based on the time difference data of the acoustic emission event arriving at each acoustic emission sensor in the acoustic emission sensor array, the rotational speed signal of the planetary carrier, and the real-time phase angle signal of the planetary carrier. The dynamic positioning calculation introduces the kinematic model of the planetary gear relative to the housing, performs motion compensation on the time difference data, and outputs the position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gear in which the event occurred. S3: Perform polarization analysis on the longitudinal wave component in the acoustic emission waveform data corresponding to the acoustic emission event, and calculate the principal polarization direction vector of the longitudinal wave component; S4: Based on the position data output in step S2, determine the local geometric coordinate system corresponding to the position on the planetary gear; perform matching analysis between the polarization principal direction vector obtained in step S3 and the local geometric coordinate system, and generate and output crack diagnosis information based on the matching result; the crack diagnosis information includes at least the identifier of the planetary gear, the position data, and the crack mechanical mode discrimination result; The dynamic positioning calculation in step S2 specifically includes: S21: Based on the real-time phase angle signal of the planetary carrier, and according to the kinematic model of the planetary gear, calculate the theoretical relative position and theoretical relative velocity vector of the crack source relative to each acoustic emission sensor in the acoustic emission sensor array at the moment the acoustic emission event occurs, for a hypothetical crack source in the coordinate system of the planetary gear where the crack source is located. S22: Construct a theoretical time difference function with the coordinates of the assumed crack source in the body coordinate system of the planetary gear as variables; the theoretical time difference function is used to calculate the theoretical propagation time of the acoustic emission event from the assumed crack source to each acoustic emission sensor in the acoustic emission sensor array; the calculation of the theoretical propagation time incorporates the effects of the theoretical relative position and the theoretical relative velocity vector; S23: Based on the theoretical time difference function, calculate a set of theoretical propagation times corresponding to the variable, and derive the corresponding theoretical time difference data from the set of theoretical propagation times; iteratively compare the actual time difference data of the acoustic emission event arriving at each acoustic emission sensor with the theoretical time difference data, and with minimizing the error between the two as the optimization objective, solve for the value of the variable that minimizes the error; output the solved value of the variable as the position data. Step S3 specifically includes: S31: Extract the longitudinal wave signal segment that first arrives from the acoustic emission waveform data collected by the triaxial sensor; S32: Construct a covariance matrix for the extracted longitudinal wave signal segment, and perform eigenvalue decomposition on the covariance matrix; S33: The direction of the eigenvector corresponding to the largest eigenvalue is determined as the main polarization direction vector; In step S4, determining the local geometric coordinate system at the corresponding position on the planetary gear specifically involves: Call the pre-stored three-dimensional digital model of the planetary gear; Input the position data output in step S2 into the three-dimensional digital model, and query and obtain the normal vector and tangential vector of the surface position point corresponding to the position data; The local geometric coordinate system is formed by the normal vector and the tangential vector.

2. The method for synchronous multi-point stress testing of a planetary gearbox according to claim 1, characterized in that, In step S4, the matching analysis of the principal polarization direction vector obtained in step S3 with the local geometric coordinate system is specifically performed as follows: Calculate the first angle between the principal polarization direction vector and the normal vector in the local geometric coordinate system; Calculate the second angle between the principal polarization direction vector and the tangential vector in the local geometric coordinate system; Based on the relationship between the first included angle and the second included angle, the crack mechanical mode is determined to be either tension-dominated or shear-dominated.

3. The method for synchronous multi-point stress testing of a planetary gearbox according to claim 1, characterized in that, In step S1, the vibration acceleration signal of the planetary gearbox is also collected synchronously. The method further includes: S5: Based on the location data in the crack diagnosis information output in step S4, extract the sideband energy related to the planetary gear meshing frequency in the vibration acceleration signal within the corresponding time period; S6: Compare the sideband energy with a preset energy threshold. If the sideband energy exceeds the energy threshold, add a load impact intensity warning label to the crack diagnosis information.

4. The method for synchronous multi-point stress testing of a planetary gearbox according to claim 1, characterized in that, The method further includes: S7: Continuously record and store the crack diagnosis information output in step S4 multiple times to form a historical diagnosis sequence; S8: For crack events at the same location in the historical diagnostic sequence, count their frequency of occurrence and calculate the variance of the change in the direction angle of the polarization principal direction vector; S9: If the frequency exceeds a preset frequency threshold and the variance of the direction angle change is less than a preset variance threshold, then generate a crack stability propagation early warning information about the location data.

5. The method for synchronous multi-point stress testing of a planetary gearbox according to claim 1, characterized in that, In step S22, the theoretical time difference function is constructed in the following way: Based on the kinematic model of the planetary gears and the theoretical relative positions, the theoretical propagation distance of the sound wave from the motion source point to each sound emission sensor in the sound emission sensor array is calculated; By combining the components of the theoretical relative velocity vector in the propagation direction, the Doppler effect is used to compensate for the average propagation speed of the sound wave in the material, and the equivalent propagation speed is obtained. Dividing the theoretical propagation distance by the equivalent propagation speed yields the theoretical propagation time to each acoustic emission sensor.

6. A multi-point stress synchronous testing system for a planetary gearbox, characterized in that, The system is used to perform the method as described in any one of claims 1 to 5, the system comprising: The data acquisition module is used to synchronously acquire acoustic emission waveform data, planetary carrier rotation speed signal, and planetary carrier real-time phase angle signal; wherein, the acoustic emission waveform data is acquired by an acoustic emission sensor array arranged on the planetary gearbox housing, the acoustic emission sensor array includes at least four acoustic emission sensors, and at least one of the acoustic emission sensors is a triaxial sensor; the planetary carrier rotation speed signal and the planetary carrier real-time phase angle signal are acquired by a rotation speed and phase acquisition unit; The dynamic positioning processing module is used to perform dynamic positioning calculations when the acoustic emission sensor array captures an acoustic emission event, based on the time difference data of the acoustic emission event arriving at each acoustic emission sensor in the acoustic emission sensor array, the rotational speed signal of the planetary carrier, and the real-time phase angle signal of the planetary carrier. The dynamic positioning calculation introduces the kinematic model of the planetary gears relative to the housing, performs motion compensation on the time difference data, and outputs the position data of the crack source corresponding to the acoustic emission event in the body coordinate system of the planetary gear in which the event occurred. The polarization analysis processing module is used to perform polarization analysis on the longitudinal wave component in the acoustic emission waveform data corresponding to the acoustic emission event, and calculate the principal polarization direction vector of the longitudinal wave component. The diagnostic information generation and output module is used to determine the local geometric coordinate system of the corresponding position on the planetary gear based on the position data output by the dynamic positioning processing module; to perform matching analysis between the polarization principal direction vector obtained by the polarization analysis processing module and the local geometric coordinate system, and to generate and output crack diagnostic information based on the matching result; the crack diagnostic information includes at least the identifier of the planetary gear, the position data, and the crack mechanical mode discrimination result.

Images