Sun gear fault tooth positioning method based on rotor sensing and tidal cycle effect

Abstract

The invention relates to the technical field of fault diagnosis, in particular to a sun gear fault tooth positioning method based on combination of rotor sensing and tidal period effect, which comprises the following steps of: fixing a sensor module on an input shaft of a planetary gear box through a thermocouple adhesive, establishing a reference frame of a fixed shaft system and a rotating coordinate system, collecting radial vibration signals, and determining the radial vibration signals; mathematical modeling analysis is carried out on the collected radial vibration signals, and low-frequency components of the vibration signals are extracted by using a low-pass filter and can be used as absolute phase reference signals; the method comprises the following steps: performing time domain compensation on a meshing event of a sun gear and a single planetary gear by calculating meshing dynamic excitation time offset of the sun gear and simplifying a gearbox model, and ensuring the integrity of the single-tooth meshing times of the sun gear in a complete meshing period, so as to accurately analyze the meshing characteristics of each single tooth of the sun gear and the planetary gear; a sun gear fault working condition is simulated through an experiment, and periodic change of fault tooth meshing is analyzed; deterministic components and random components in the vibration signals are separated through cepstrum analysis; a band-pass filter is used for extracting gear meshing components from deterministic components; a minimum entropy deconvolution filtering technology and a square envelope technology are used for carrying out fault feature enhancement on random components; finally, absolute phase reference provided by gravitational acceleration of a rotor perception sensor is used, and a tidal period effect is combined; and calculating a single-tooth meshing frequency code of the fault tooth lag absolute phase reference point, and determining the position of the fault tooth by using different codes corresponding to different fault positions. The complexity of the monitoring system is effectively reduced, the influence of monitoring equipment on the operation of the planetary gear box is reduced, and the fault positioning accuracy of the sun gear is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a method for locating a sun gear fault tooth based on rotor perception and tidal cycle effects. Background Art

[0002] Planetary gearboxes, as important components in rotating mechanical transmission systems, are widely used in many industrial and mechanical equipment. Their failures not only lead to equipment downtime and huge economic losses, but may also cause serious safety accidents. Therefore, how to efficiently and accurately detect early faults in gearboxes is crucial.

[0003] Traditional vibration signal monitoring methods usually require sensors to be installed on the equipment housing to obtain vibration information of rotating parts. However, due to the complex structure of the planetary gearbox, the transmission path of vibration signals collected by external sensors is complex and changeable, the signal-to-noise ratio is low, and early weak fault signals cannot be accurately obtained.

[0004] In recent years, the development of micro-electromechanical systems (MEMS) technology has provided a new solution for gearbox condition monitoring. MEMS accelerometers have the characteristics of small size, low power consumption and high sensitivity. They can be embedded inside rotating parts to collect signals to achieve the purpose of real-time monitoring of operating conditions. Compared with traditional vibration sensors, MEMS accelerometers have more flexible installation locations and are suitable for detection in small spaces.

[0005] Some existing research works have explored how to use MEMS sensors for condition monitoring of rotating machinery. For example, Al barbar et al. compared the performance of different types of MEMS accelerometers with traditional piezoelectric sensors, and the results showed that MEMS sensors have broad application prospects in rotating machinery fault diagnosis. Aswin et al. verified the vibration measurement capability of MEMS accelerometers under non-steady-state conditions by installing them on wind turbine blades. Koene et al. developed a wireless sensor node consisting of a MEMS accelerometer and a microcontroller with a WLAN module, which is fixed on the bearing seat. Experimental results show that the sensor node can not only detect speed information and rotor harmonic components, but also detect bearing vibration. Jiménez et al. designed an intelligent rotor assembly with integrated MEMS accelerometers, calibrated the axial position of the sensor through mechanical structure, and used the output signal to build a virtual encoder, achieving reliable speed estimation at low speed. In order to solve the problem that MEMS sensors are affected by gravity acceleration, Mones et al. installed two MEMS accelerometers on the input shaft of the planetary gearbox relative to each other, so that the projections of the gravity acceleration of the sensor in the tangential direction of the shaft offset each other, reducing the influence of gravity acceleration on the signal. After studying the mathematical model of the orthogonal output of MEMS accelerometers, Feng et al. proposed a new algorithm that moves the output signal in the orthogonal direction by π / 2 through Hilbert transform, combines it with the output signal in the tangential direction, and compensates the combined signal in the frequency domain to reconstruct the tangential acceleration of the rotor with high precision, thus realizing online detection of leakage faults in reciprocating compressors. With the development of technology, a variety of signal processing methods have been developed to analyze the vibration signals of planetary gearboxes. In order to detect early faults of planetary gear rings, Mark et al. proposed a simple frequency domain method to eliminate the influence of amplitude changes caused by sensors and structural paths. In addition, when rotating components fail, they will generate pulse responses when passing through the rotor. The phase relationship between these pulse responses and gear meshing can help locate faulty planetary gears.

[0006] Peng found that the sensors mounted on the planetary carrier could capture vibration signals from the faulty planetary gears, which exhibited an asymmetric pulse response, with the direction of the asymmetry affected by the position of the faulty gear. Based on this finding, they developed a metric that can accurately locate the faulty planetary gear. Further, Peng conducted a deeper analysis and proposed that the meshing phase relationship is the root cause of the asymmetry in the time domain signal. Using the relationship between the fault pulse and the meshing phase, they proposed an improved metric, called meshing phase difference, to more clearly distinguish the faulty planetary gears.

[0007] Although these studies have shown that MEMS sensors have good performance, existing solutions still have shortcomings, especially in complex planetary gearbox structures, how to accurately locate the sun gear fault remains a challenge.

[0008] Some existing research works have explored how to use MEMS sensors for condition monitoring of rotating machinery. For example, Albarbar et al. compared the performance of different types of MEMS accelerometers with traditional piezoelectric sensors, and the results showed that MEMS sensors have broad application prospects in rotating machinery fault diagnosis. In addition, Aswin et al. verified the vibration measurement capability of MEMS accelerometers under non-steady-state conditions by installing them on wind turbine blades. Although these studies have shown that MEMS sensors have good performance, existing solutions still have shortcomings, especially in complex planetary gearbox structures, how to accurately locate the sun gear fault remains a challenge.

[0009] China Patent Authorization Announcement No.: CN118332318B discloses a method, device, medium and equipment for locating faults of wind turbine gearbox bearings, the method comprising: collecting the actual temperature and original feature data of the wind turbine gearbox bearing to be tested, and preprocessing the original feature data; constructing a wind turbine gearbox bearing temperature prediction model, and training the constructed model; inputting the preprocessed original feature data of the wind turbine gearbox bearing to be tested into the trained wind turbine gearbox bearing temperature prediction model to obtain the predicted temperature of the wind turbine gearbox bearing to be tested; comparing the predicted temperature with the warning temperature to determine whether the wind turbine gearbox bearing to be tested is in a high temperature warning; using a preset interpretation model to perform contribution analysis on the original feature data of the wind turbine gearbox bearing to be tested in a high temperature warning to identify key feature data, and locating faults of the wind turbine gearbox bearing to be tested based on the key feature data. This application can find key features and achieve accurate positioning of bearing faults.

[0010] The present invention uses rotor sensing (ORS) technology to directly install the sensor on the rotating shaft to capture vibration signals and provide an absolute phase reference during the rotation process. On the rotating parts, combined with the tidal cycle effect model, it can accurately locate the faulty sun gear tooth in the planetary gearbox. By analyzing the single-tooth meshing number encoding between the absolute phase reference and the faulty sun gear tooth in the tidal cycle to distinguish different fault locations, the problem of low signal-to-noise ratio of signal acquisition in the prior art is optimized, and a new solution is provided for fault location of large rotating machinery.

[0011] The phase difference between the analysis sensor and the faulty gear is used to generate corresponding codes, thereby effectively distinguishing different fault locations and solving the signal transmission problem and the inaccurate fault location problem in the prior art. Summary of the invention

[0012] To this end, the present invention provides a sun gear fault tooth locating method based on rotor perception and tidal cycle effect, which is used to optimize the problem of low signal-to-noise ratio of signal acquisition in the prior art, and provides a new solution for fault locating of large rotating machinery. To achieve the above purpose, the present invention provides a sun gear fault tooth locating method based on rotor perception and tidal cycle effect, wherein the sensor is fixed on the axial surface of the input shaft of the planetary gearbox by thermocouple adhesive, and is characterized in that, in a single test, it includes:

[0013] Define the fixed axis system and rotating coordinate system, collect radial vibration signals, and use a low-pass filter to obtain the absolute phase reference signal after formula derivation and modeling analysis;

[0014] The dynamic excitation time offset of the meshing between the sun gear and the planetary gear is calculated in combination with the planetary gearbox parameters. The meshing period of the sun gear is compensated in the time domain by simplifying the model as the meshing event between the sun gear and the single planetary gear to ensure the complete meshing period of the sun gear.

[0015] Cepstrum analysis is used to separate the deterministic and random components in the vibration signal, a bandpass filter is used to extract the gear meshing component, and the minimum entropy deconvolution filtering technology and square envelope technology are used to enhance the fault pulse signal in the random component to find the phase relationship between the fault pulse and the gear meshing component.

[0016] The absolute phase reference signal is introduced for phase correction. By analyzing the periodic changes in the meshing position of the faulty tooth during the tidal cycle, the single tooth meshing times encoding of the sun gear faulty tooth lagging the absolute phase reference point is calculated to achieve accurate positioning of the faulty tooth.

[0017] Furthermore, the process of signal acquisition and analysis includes:

[0018] Use MEMS rotor sensing sensors to collect radial vibration and torsional vibration signals, perform mathematical modeling on the collected radial vibration and torsional vibration signals, analyze the projection changes of gravity acceleration in two directions, model and analyze the low-frequency signals of radial vibration and torsional vibration, and use low-pass filters to obtain absolute phase references;

[0019] The deterministic and random components in the vibration signal are separated by cepstrum analysis, the gear meshing components are extracted by bandpass filter, and the fault pulse signal in the random component is enhanced by minimum entropy deconvolution filtering technology and square envelope technology to find the phase relationship between the fault pulse and the gear meshing component.

[0020] The absolute phase reference signal is introduced to correct the phase relationship between the fault pulse and the gear meshing component, and the number of single-tooth meshing times between the faulty tooth of the sun gear and the absolute phase reference point is calculated.

[0021] Furthermore, the determination of tidal cycle effects includes:

[0022] Study the meshing mode between the sun gear faulty tooth and the planetary gear, and analyze the meshing position change between the sun gear faulty tooth and the planetary gear;

[0023] Combined with the parameters of the sun gear and the planetary gears, the number of rotations of the sun gear and the corresponding number of faulty tooth engagements during the tidal cycle are determined.

[0024] Furthermore, the step of locating the faulty tooth includes:

[0025] Time domain compensation of radial vibration signal to ensure the acquisition of the complete sun gear meshing cycle;

[0026] Cepstrum analysis is used to separate the deterministic and random components of the signal. A bandpass filter is used to extract the gear meshing component in the deterministic signal. The minimum entropy deconvolution filtering technique and square envelope analysis are used to enhance the fault characteristic signal in the random signal, and the phase relationship between the fault pulse and the gear meshing component is found.

[0027] Based on the tidal cycle effect and the absolute phase reference signal provided by the rotor sensing sensor, the single tooth meshing number encoding of the absolute phase reference point of the faulty sun gear tooth lagging behind the tidal cycle is determined, the unique identification of the faulty tooth is generated, and the position of the faulty tooth is determined.

[0028] Compared with the prior art, the beneficial effect of the present invention lies in that, by utilizing the rotor sensing (ORS) technology, the sensor is directly installed on the input shaft of the planetary gearbox, which can capture the dynamic vibration signal in the rotating system in real time, effectively solving the problem of signal quality degradation caused by the long signal transmission path of traditional external sensors, thereby improving the accuracy of fault diagnosis.

[0029] Furthermore, the present invention combines the tidal cycle effect to perform periodic analysis on the meshing position of the faulty sun gear tooth and the planetary gear, and uses the position change of the meshing of the faulty sun gear tooth to generate a single tooth meshing number code of the lagging absolute phase reference point, thereby accurately locating the faulty tooth of the sun gear, providing a new solution for fault locating of large rotating machinery.

[0030] Furthermore, the present invention not only improves the sensitivity and reliability of signal acquisition by installing a MEMS accelerometer on the rotating component, but also realizes efficient detection of early fault signals of the sun gear, solving the problem that early fault signals are weak and easily ignored in traditional methods.

[0031] Furthermore, cepstrum analysis is used to separate the deterministic and random components in the vibration signal, a bandpass filter is used to extract the gear meshing component, and the minimum entropy deconvolution filtering technology and square envelope analysis are used to enhance the fault signal in the random component. The phase relationship between the fault pulse signal and the gear meshing is found, and combined with the absolute phase reference provided by the rotor sensing sensor, a comprehensive analysis of the complex sun gear fault tooth meshing state can be achieved, thereby further improving the diagnostic accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Flow chart of the method for locating faulty sun gear teeth based on rotor sensing and tidal cycle effect of the present invention

[0033] Figure 2 This is a schematic diagram of the installation of sensor nodes on a rotor according to an embodiment of the present invention;

[0034] Figure 3 is the spectrum of the analog signal of the embodiment of the present invention;

[0035] Figure 4 This is a schematic diagram of a tidal cycle according to an embodiment of the present invention;

[0036] Figure 5 This is a flow chart of a sun gear fault location algorithm according to an embodiment of the present invention;

[0037] Figure 6 This is a time domain waveform diagram of an ORS vibration signal at 1000 rpm and 50% load according to an embodiment of the present invention;

[0038] Figure 7 The ORS vibration spectrum diagram of the embodiment of the present invention at 1000rpm and 50% load:

[0039] Figure 8 Graph showing various phase relationships of the gear meshing components and the square envelope (SE) of the impulse responses (IRs) within a tidal cycle according to an embodiment of the present invention.

[0040] Fig. 9 This is a cepstrum analysis diagram of the radial vibration signal of an embodiment of the present invention.

[0041] Fig.10 The square envelope (SE) of the gear meshing component and impulse response of the embodiment of the present invention shows various phase relationship diagrams within two adjacent complete sun gear meshing cycles. DETAILED DESCRIPTION

[0042] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0043] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0044] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0045] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0046] See also Figure 1 As shown, it is a flow chart of the sun gear fault tooth positioning method based on rotor perception and tidal cycle effect of the present invention. The sensor module is fixed on the axial surface of the planetary gearbox input shaft by thermocouple adhesive. In a single test, it includes:

[0047] Step S1, collecting radial vibration signals, and performing cepstrum analysis on the radial vibration signals to separate deterministic components and random components in the vibration signals;

[0048] Step S2, using a bandpass filter to extract the gear meshing signal from the deterministic component, and a low-pass filter to extract the low-frequency component of the vibration signal as an absolute phase reference signal;

[0049] Step S3, using minimum entropy deconvolution filtering technology and square envelope technology to enhance the fault characteristics of random components;

[0050] Step S4, performing a multi-cycle test on the rotation of the sun gear to analyze the phase relationship between the enhanced fault pulse and the gear meshing component;

[0051] Step S5, performing phase correction according to the absolute phase reference provided by the rotor sensing sensor, finding the single tooth meshing number code of the sun gear faulty tooth lagging absolute phase reference point within the tidal cycle, and finally determining the position of the faulty tooth;

[0052] The present invention uses rotor sensing (ORS) technology to directly install the sensor on the rotating shaft of the rotating component, and combines the tidal cycle effect to accurately locate the faulty tooth of the sun gear in the planetary gearbox. By analyzing the number of single-tooth meshing intervals between the sensor and the faulty gear in the tidal cycle, different fault locations are distinguished, the problem of low signal-to-noise ratio of signal acquisition in the prior art is optimized, and a new solution is provided for fault location of large rotating machinery.

[0053] Specifically, the steps of signal acquisition and separation of deterministic components and random components in vibration signals include:

[0054] Use MEMS rotor sensing sensors to collect radial vibration and torsional vibration signals, perform mathematical modeling on the collected radial vibration and torsional vibration signals, analyze the projection changes of gravity acceleration in two directions, and model and analyze the low-frequency signals of radial vibration and torsional vibration. Use low-pass filters to obtain absolute phase references; use cepstrum analysis to separate deterministic and random components in vibration signals;

[0055] Specifically, the steps of processing the separated deterministic components in the vibration signal include:

[0056] A bandpass filter is used to extract the gear meshing signal in the deterministic component, and a low-pass filter is used to extract the low-frequency component of the vibration signal as the absolute phase reference signal;

[0057] Specifically, the steps of processing the separated random components in the vibration signal include:

[0058] Fault feature enhancement of random components using minimum entropy deconvolution filtering, pulse bandwidth selection, and square envelope techniques;

[0059] Specifically, the steps of analyzing the phase relationship between the fault pulse and the gear meshing component include:

[0060] Perform multi-cycle tests on the rotation of the sun gear and analyze the phase difference between the enhanced fault pulse and the gear mesh component;

[0061] Specifically, the steps of determining the position of the faulty tooth include:

[0062] Phase correction is performed based on the tidal cycle effect and the absolute phase reference provided by the rotor sensing sensor, and the single tooth meshing number encoding of the absolute phase reference point of the faulty sun gear tooth lagging behind the tidal cycle is determined to generate a unique identifier for the faulty tooth and determine the position of the faulty tooth.

[0063] The purpose of the present invention is to provide a method for locating a faulty sun gear tooth based on a combination of rotor sensing (ORS) technology and tidal cycle effects, so as to solve the technical problems existing in the existing conventional fault diagnosis methods, such as the complex and changeable signal transmission path resulting in a low signal-to-noise ratio of the collected signal, difficulty in fault location, inability to achieve real-time monitoring, and reliance on complex signal processing methods.

[0064] The present invention utilizes rotor sensing (ORS) technology to directly install a MEMS accelerometer on a rotating shaft, which can more accurately capture the dynamic characteristics of a rotating system and improve signal quality and the accuracy of fault diagnosis.

[0065] To achieve the above purpose, the specific technical solution of the method for locating the faulty sun gear tooth based on rotor sensing (ORS) technology and tidal cycle effect of the present invention is as follows:

[0066] The method for locating sun gear fault based on rotor sensing (ORS) technology and tidal cycle effect includes the following steps:

[0067] The sensor module is fixed on the axial surface of the planetary gearbox input shaft by thermocouple adhesive, such as Figure 2 As shown. Figure 2 In (a), two coordinate systems are defined: one is the fixed axis system XYZ, whose direction remains unchanged during the rotation of the rotor; the other is the rotating coordinate system UVW, whose direction changes as the rotor rotates. In the rotating coordinate system, the three-axis directions of the sensor are defined as follows: the U direction represents the torsional vibration of the rotor system, the V direction represents the radial vibration, and the W direction represents the vibration along the axis. Assuming that the initial phase of the ORS sensor relative to the stationary system is θ0, for simplicity, any measured rotation parameter (e.g., angular velocity ω) is assumed to have an initial phase of zero relative to the stationary coordinate system. Therefore, θ0 also represents the phase difference between the sensor and the initial rotor rotation phase.

[0068] See also Figure 2 As shown, it is a schematic diagram of the installation of sensor nodes on the rotor of an embodiment of the present invention. In the figure, the ORS sensor node in (a) is installed on the input shaft of the planetary gearbox in (b), and a MEMS sensor is provided in (b).

[0069] In the gravitational field, the acceleration is 1.0g, but the direction is opposite to the earth's gravity and can be sensed by the accelerometer. When the rotor shaft rotates around the center point o at an angular velocity ω, the torsional and radial vibration signals of the ORS sensor can be expressed as:

[0070] u(t)=x(t)cos(ωt)-y(t)sin(ωt)-g×cos(ωt)

[0071] (1)v(t)=x(t)sin(ωt)+y(t)cos(ωt)-g×sin(ωt)+ω 2 r (2)

[0072] Where x(t) and y(t) represent the vibration in the horizontal and vertical directions respectively, θ0=ωt represents the angular displacement caused by the rotation of the rotor, ω 2 r represents the centrifugal acceleration experienced by the sensor. r is the radius of the shaft and is a constant when the rotor system runs at a constant speed. In dynamic analysis, this constant centrifugal acceleration of the sensor can be ignored. We can regard the collected signals in the x(t) and y(t) directions as periodic and thus represent them through the Fourier series:

[0073]

[0074] Among them, α and β represent the vertical and horizontal fundamental frequencies of the rotor respectively. Ax(n) and φx(n) represent the amplitude and phase of the nth harmonic in α, while Ay(n) and φy(n) similarly describe the amplitude and phase in β. Therefore, the torsional and radial vibration signals of the ORS sensor can be expressed as:

[0075]

[0076] Based on the derived equations (5) and (6), simulation analysis was performed to analyze the torsional and radial vibrations when the sun gear speed was set to fs = 16.6 Hz and there was a local fault in the sun gear. The parameters α and β were designed to include the gear meshing frequency f m and fault characteristic frequency f f In addition, the influence of gravity acceleration during rotation is also considered, and the projection change of gravity acceleration on the rotating coordinate axis is analyzed. The spectrum of the simulation signal is as follows: Figure 3 As shown in Figure 2, the frequency spectra of the torsional and radial vibration signals are shown respectively. It can be observed that the rotation frequency modulated sidebands appear near the gear meshing frequency, and the rotation frequency shows a relatively high amplitude in the low frequency range. Figure 3 (c) and Figure 3(d) shows the projected spectrum of gravity acceleration on the U and V axes, indicating the presence of a unit g component at the rotation frequency. This indicates that the rotation frequency dominates the low-frequency part of the signal. During the rotation of the ORS, the projection components of the sensor's gravity acceleration in the torsional and radial vibration directions change with time in the form of a sine wave. However, the direction of the sensor's gravity acceleration remains unchanged. Therefore, for radial vibration, when the sensor is located at the highest and lowest points of the shaft, it corresponds to the peak or valley of the sine wave, respectively. Torsional vibration exhibits a π / 2 phase difference with respect to radial vibration. The above analysis proves that the sensor's gravity acceleration can provide position information of the sensor during rotation. In addition, since the sensor is mounted on the input shaft, the phase difference between the sensor position and the faulty gear remains unchanged, so that the sensor's position can be used as an absolute phase reference.

[0077] When the sun gear rotates one circle, the number of meshing events is less than the number of teeth of the sun gear. To facilitate the identification of the phase difference between the faulty tooth and the sensor, the system can be simplified to the meshing of the sun gear and a single planetary gear to ensure that the number of meshing events is equal to the number of sun gear teeth. Therefore, compensation is required in the time domain to obtain a complete sun gear meshing cycle. The planetary gearbox used in this experiment includes a sun gear connected to the input shaft, three planetary gears connected to the output shaft, a planet carrier, and an external ring gear. The rotation of the planet carrier is driven by the meshing between the sun gear and the planetary gears for speed increase or speed decrease. The meshing frequency can be calculated by the following formula:

[0078] f m =z s (f s -f c ) (7)

[0079] Where, fs represents the rotational frequency of the sun gear; f c Represents the rotational frequency of the planet carrier; z s represents the number of teeth on the sun gear. The mounting characteristics of the planetary gearbox result in alternating single-tooth and double-tooth meshing between the sun gear and the planet gears. Due to the different mounting angles ψi of the planet gears, the time each planet gear passes through the sensor is different and is related to the rotation speed of the planet carrier. Therefore, different planet gears have a time offset in the direction of the dynamic excitation force and the transfer path function. The dynamic excitation time offset t at the meshing point between the sun gear and the planet gear si It can be expressed as:

[0080]

[0081] Among them, f nsc Indicates the relative speed between the sun gear and the planet carrier; ψ iis the installation angle of the ith planetary gear; N is the number of planetary gears. Assuming the gear ratio is r, when the sun gear rotates by 2π, the planet carrier rotates by 2π / r. In order to obtain the complete meshing period (t p ), the signal needs to be compensated in the time domain as follows:

[0082]

[0083] Tidal cycle effect in planetary gearbox: When there is a fault in the sun gear, the faulty tooth meshes with planetary gears 1, 2 and 3 in sequence, and then meshes with planetary gear 1 again, repeating the cycle. The change in the meshing position of the faulty tooth is very similar to the tide in natural phenomena, so it is called the tidal cycle effect. Therefore, if Figure 4 As shown, after running for a period of time, the faulty tooth of the sun gear meshes with the specific planetary gear again at the same position.

[0084] For a planetary gearbox equipped with N equally spaced planetary gears, the phase difference between adjacent planetary gears is 2π / N. When the faulty sun gear tooth meshes with planetary gear 2, the rotation angles of the sun gear and planetary gear 1 are expressed as and Indicates that:

[0085]

[0086] From one faulty gear meshing to the next, the sun gear must rotate an additional 2π / N angle relative to the planet carrier. Based on equation (11) and the angular velocity difference between the sun gear and the planet gear, the time difference t between the faulty gear meshing with two different planet gears can be determined. s :

[0087]

[0088] Due to the transmission ratio r, the speed relationship between the sun gear and the planetary gear can be expressed as:

[0089]

[0090] where z r Indicates the number of teeth on the outer ring gear. The rotation angle of the sun gear and planetary gear can be expressed as:

[0091]

[0092] After the faulty tooth experiences n meshing events, when the faulty tooth of the sun gear meshes with the specific planetary gear again at the same position, the number of rotations of the sun gear and the planetary gear in one tidal cycle can be expressed as:

[0093]

[0094] It can be seen that the tidal period depends only on the parameters of the sun gear and the planetary gear. s +z r and z s The greatest common divisor (GCD) z g , it means that when the sun gear rotates (z s +z r ) / z s During the period, the planetary gear rotates z s / z g cycle, which just constitutes a complete tidal cycle.

[0095] See also Figure 4 As shown, it is a schematic diagram of the tidal cycle of an embodiment of the present invention. The signal processing method used in this patent includes three parts: gear meshing component extraction, fault feature enhancement and extraction, and determination of the sun gear fault position.

[0096] Gear mesh components and absolute phase reference extraction

[0097] First, the collected original vibration signal is subjected to cepstrum analysis according to formula (18). The cepstrum of signal x is calculated by Fourier transforming x and taking its complex natural logarithm, and then inverse Fourier transforming the obtained sequence. This analysis method helps to extract periodic signals in the original spectrum that are difficult to identify by the naked eye, and aggregates the sideband spectrum lines in the original spectrum into a single spectrum line. Then a bandpass filter is used to extract the gear meshing component from the deterministic signal, and the center frequency is set to the meshing frequency. The output of the ORS sensor is affected by the gravitational acceleration. When the sensor is located at the top or bottom of the rotor shaft, the projection direction of the gravitational acceleration in the direction of radial vibration is opposite, corresponding to the peak and trough of the vibration signal time domain waveform respectively. The deterministic component of the radial vibration signal is low-pass filtered, and the resulting waveform is similar to a sine signal, which effectively indicates the sensor position as an absolute phase reference.

[0098]

[0099] Enhancement and extraction of fault pulses For random components, the application of fault pulse enhancement technology can enhance fault information. Minimum entropy deconvolution (MED) filters are particularly suitable for this purpose and can better characterize transient phenomena. When using MED, a delay must be arbitrarily selected to ensure that the deconvolution filter maintains causality. Power spectral density (PSD) is a measure that describes the energy distribution of a signal in the frequency domain. By analyzing the power spectral density, we can understand the frequency components with higher or lower energy levels in the signal, thereby identifying frequency bands with obvious fault characteristics. Square envelope processing of the signal can effectively emphasize the amplitude effect of the signal, which is very effective for signal analysis in nonlinear systems. The square envelope (SE) of the signal x(n) is obtained by calculating the square of the modulus of the analytical signal formed by x(n) and its Hilbert transform, see formula (19):

[0100] S SE (x[n])=|x[n]| 2 =|x[n]+j·Hilbert(x[n])| 2 (19)

[0101] By comparing the phase relationship between the fault pulse and the gear meshing, the distribution pattern of the fault pulse within the complete sun gear meshing cycle can be preliminarily determined. Subsequently, the phase calibration is performed using the absolute phase reference extracted in the first part to identify the phase difference between the sensor and the fault pulse. However, due to the tidal cycle effect, the meshing position of the faulty tooth presents a periodic change pattern. Based on this phenomenon, the single tooth meshing number encoding between the sensor and the fault pulse is finally obtained. This encoding can be used to identify faulty sun gear teeth at different positions.

[0102] See also Figure 5 As shown, it is a flow chart of the sun gear fault location algorithm according to an embodiment of the present invention, and the process is specifically described by taking the following embodiment as an example:

[0103] Embodiment 1:

[0104] The first step is the experimental setup: The ORS system used in this study is mainly composed of a micro MEMS sensor module and a micro processor module. The sensor uses the ADXL357 accelerometer, and the parameters are shown in Table 1. It has a wide dynamic range, high resolution, and excellent stability. The sampling rate is set to 4000Hz, which can cover multiple harmonics of the meshing frequency of the sun gear and the planetary gear. The processing module uses the nRF52840 processor from Nordic Semiconductor, equipped with a 32-bit Cortex TM-M4 core, 1MB flash memory and 256kB RAM, and integrated low-power Bluetooth to support wireless data transmission. This processor was selected for its small size, high computing power and low power consumption. In addition, the system is equipped with a 3.7V lithium battery as a power source. Please refer to Table 1 for the main parameters of ADXL357:

[0105] Table 1 Main parameters of ADXL357

[0106]

[0107] The host computer software is developed on an Android phone and is able to receive and display acceleration data from the processor module. The received data can be saved to local storage for subsequent offline analysis. In addition, the user can remotely adjust the sampling rate and dynamic range of the ORS sensor as needed through the host interface.

[0108] In order to evaluate the proposed method for locating the faulty tooth of the sun gear, a case study of a planetary gearbox test bench was conducted. The test bench includes a three-phase variable frequency motor, a torque-speed sensor, a planetary gearbox, a magnetic powder brake, a circulating water cooling system, a coupling, and a system control box. The control box is used to control the speed of the AC motor, the load generated by the magnetic powder brake, and the water circulating cooling system. The three-phase variable frequency motor has a rated voltage of 380V, a rated current of 10.7A, a power of 5.5kW, and a maximum speed of 3000r / min. The maximum torque load of the magnetic powder brake is 1.2A. The ORS sensor module is located on the shaft surface connected to the input shaft of the sun gear. The radius of the shaft is about 25 mm. Therefore, at a speed of 1200 rpm, the centrifugal acceleration of the ORS sensor is about 35g, which is within the measurement range of the sensor. In order to simulate the crack fault of the sun gear, a 1.34 mm crack was artificially introduced at the root of the sun gear tooth. In this experimental study, the sun gear crack fault has a phase difference of about 90° relative to the ORS sensor. The speed of the sun gear was set to three different speeds: 800 rpm, 1000 rpm, and 1200 rpm, and the load was set to three different levels: 30% (0.4 A), 50% (0.6 A), and 70% (0.8 A). The sampling frequency of the ORS was set to 4000 Hz to acquire 15 seconds of data under each condition.

[0109] The second step is ORS signal analysis: At a speed of 1000 rpm and 50% load, the time domain vibration signal collected by ORS is as follows: Figure 6As shown, the U, V, and W axes correspond to torsional, radial, and axial vibrations, respectively. Since the ORS sensor rotates with the rotor, the time domain waveforms of torsional and radial vibrations show a periodic variation pattern, which is caused by the variation in the projection of gravity acceleration on these two measurement axes. In addition, the vibrations caused by gear meshing and gear faults are coupled into the signal as high-frequency components. It is worth noting that Figure 6 The amplitude of the axial vibration signal in (c) is smaller than the signals collected from the other two axes.

[0110] See also Figure 7 As shown, it is a time domain waveform diagram of the ORS vibration signal at 1000rpm and 50% load according to an embodiment of the present invention. Figure 7 The spectrum of the ORS signal acquired at 1000 rpm and 50% load on three axes is shown. Figure 7 The amplitude of the axial acceleration spectrum in (c) is smaller than the spectra of the other two axes. This is because there is no obvious axial excitation during the rotation of the planetary gearbox. The rotation frequency of the sun gear shows a relatively high amplitude, which clearly reflects the speed of the input shaft. The meshing frequency of the sun gear and the planetary gear and its harmonics are clearly visible in both the torsional and radial vibration spectra. According to the previous analysis, a low-pass filter can be used to extract the low-frequency components in the radial vibration signal as an absolute phase reference. When the sensor is located at the highest point of the rotor shaft, the collected radial vibration signal is in the same direction as the gravity acceleration sensed by the sensor, corresponding to the peak of the radial vibration time domain signal.

[0111] See also Figure 8 As shown, it is an ORS vibration spectrum diagram of an embodiment of the present invention at 1000rpm and 50% load, wherein,

[0112] Step 3 Fault pulse distribution pattern analysis: The vibration caused by gear meshing, even in the presence of a fault, still mainly shows periodic characteristics. Meshing with a faulty gear can produce a broadband pulse signal, which also shows a periodic pattern under slight random speed and load fluctuations. In order to find the distribution pattern of the fault pulse, the most impactful frequency band of the signal is extracted from the radial vibration signal. The third harmonic of the meshing frequency is selected as the center frequency of the bandpass filter with a bandwidth of 400Hz. Then, by compensating the time offset of the sun gear meshing in the time domain, the complete meshing cycle of the sun gear can be obtained.

[0113] Step 4: Locate the faulty sun gear tooth:

[0114] (1) Extraction of gear meshing components and fault pulses:

[0115] In this process, the deterministic components and random components in the radial vibration are first separated. By performing cepstrum analysis on the radial vibration signal, such as Fig. 9 As shown, it is a cepstral analysis diagram of the radial vibration signal of an embodiment of the present invention, which can identify the periodic components in the signal. The analysis results show that there are multiple obvious periodic pulses, in which specific time intervals (Δt1 and Δt2) represent the meshing frequency between the sun gear and the planetary gear and the rotation frequency of the planet carrier, respectively. By setting a threshold of 0.05, a bandpass filter can be used to extract components exceeding the threshold, which are regarded as deterministic components in this study. After separation, the remaining part of the original vibration signal constitutes a discrete random component. In order to further enhance the fault pulse, the minimum entropy deconvolution filtering (MED) technology is first used. This method relies on the kurtosis difference between the target signal and the interference component, and optimizes the kurtosis of the filtered signal to solve the filter coefficient. Subsequently, after power spectral density (PSD) analysis, it is found that the signal in the 800-1000Hz frequency band is the strongest, so the MED enhanced signal is bandpass filtered. In order to further highlight the fault pulse, the square envelope (SE) of the bandpass filtered signal is calculated and normalized, and the results show that the pulse caused by the fault is significantly enhanced. Then the time domain signal is segmented according to the compensated complete sun gear meshing cycle. Assuming that the three highest peaks in a complete meshing cycle correspond to the meshing of the faulty tooth, the peak search interval is set to 0.02 seconds to avoid misjudging adjacent single-tooth and double-tooth meshing events as two fault pulses. It is worth noting that if the starting point of the complete meshing cycle of the sun gear coincides with the meshing of the faulty tooth, four peaks will appear in the previous cycle.

[0116] (2) Phase calibration using absolute phase reference:

[0117] In order to more intuitively observe the number of single-tooth meshing between the fault pulse and the sensor in a complete sun gear meshing cycle, the x-axis is converted to the sun gear meshing cycle. After obtaining the position information of the sensor during the rotation process, the extracted fault pulse and the gear meshing component are plotted to display two adjacent sun gear meshing cycles, such as Fig.10 As shown, it is the square envelope (SE) of the gear meshing component and pulse response of an embodiment of the present invention, showing various phase relationship diagrams within two adjacent complete sun gear meshing cycles. The interval between two adjacent fault pulses corresponds to approximately seven single tooth meshings. This is because the meshing position between the sun gear and the two adjacent planetary gears in this study differs by seven single teeth. According to the above analysis, the lagging single tooth meshing count between the absolute phase reference and the fault pulse can be calculated. However, it was observed that in the 11th complete sun gear meshing cycle, the number of lagging single tooth meshings was 3, while in the 12th cycle, the value became 1. There is no consistency in the number of single tooth meshings of the fault pulse lagging the absolute phase reference point between the two cycles. This is due to the change in the position of the planetary gear during rotation, which in turn causes the meshing position of the fault tooth to change.

[0118] (3) Calculation of delayed single tooth meshing count code:

[0119] According to formula (17), we can calculate that the sun gear rotates five times in one tidal cycle, corresponding to four complete sun gear meshing cycles, during which the sun gear faulty tooth meshes with different planetary gears a total of 12 times. The planetary gears are evenly distributed, and the meshing position of each planetary gear with the sun gear is seven teeth apart, which means that after seven single-tooth meshings, the sun gear tooth faulty tooth will begin to mesh with the next planetary gear in the rotation direction. Due to the unique structure of the planetary gearbox, three areas of the sun gear teeth mesh with three different planetary gears at the same time. Therefore, the faulty tooth can be identified as one of the three potential fault locations. Assume that we divide the sun gear teeth into three evenly distributed areas, and the faulty tooth numbers range from 1 to 6 in the three areas. The number of single-tooth meshings of the fault pulse lagging the absolute phase reference point calculated in one tidal cycle is as follows: Figure 8 As shown. In two adjacent complete meshing cycles of the sun gear, the difference in the number of lagging single tooth meshing corresponds to the rotation angle of the planet carrier. According to the above analysis, Figure 8 As shown, it can be observed that the single tooth meshing number encoding of the fault pulse lagging the absolute phase reference point within one tidal cycle. It can be deduced that as the fault tooth of the sun gear changes, the encoding will also change accordingly, which can be attributed to the meshing characteristics of the planetary gearbox. Therefore, the encoding corresponding to different sun gear fault teeth is derived and listed in Table 2. Here, the lagging single tooth meshing number represents the number of teeth between the absolute phase reference and the nearest potential fault position in the direction of rotation. This encoding will help identify potential faulty sun gears.

[0120] See also Figure 8 As shown, it is a diagram of various phase relationships between the gear meshing components and the square envelope (SE) of the impulse response (IRs) according to an embodiment of the present invention.

[0121] In this figure, please refer to Table 2, which is the delayed single tooth meshing encoding of different sun gear tooth fault positions:

[0122] Table 2 Delayed single tooth meshing coding of different sun gear tooth fault positions

[0123]

[0124] The early fault signals of the planetary gearbox are not obvious. The present invention utilizes direct measurement from the rotating shaft, which is different from the traditional external sensors that are difficult to capture the tiny early fault signals inside the gearbox, thereby effectively reducing the equipment maintenance costs and risks.

[0125] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for locating faulty sun gear teeth based on rotor sensing and tidal cycle effect, wherein the sensor is fixed on the axial surface of the input shaft of the planetary gearbox by means of a thermocouple adhesive, characterized in that: Included in a single test: Define the fixed axis system and rotating coordinate system, collect radial vibration signals, and use a low-pass filter to obtain the absolute phase reference signal after formula derivation and modeling analysis; The dynamic excitation time offset of the meshing between the sun gear and the planetary gear is calculated in combination with the planetary gearbox parameters. The meshing period of the sun gear is compensated in the time domain by simplifying the model as the meshing event between the sun gear and the single planetary gear to ensure the complete meshing period of the sun gear. By analyzing the influence of tidal cycle effect on the meshing position of the sun gear faulty tooth, the periodic change of the meshing position of the sun gear faulty tooth is revealed; Cepstrum analysis is used to separate the deterministic and random components in the vibration signal, a bandpass filter is used to extract the gear meshing components, and the minimum entropy deconvolution filtering technology and square envelope technology are used to enhance the fault pulse signal in the random component. Finally, the single tooth meshing number encoding of the absolute phase reference point of the sun gear faulty tooth lag is calculated to achieve accurate positioning of the faulty tooth.

2. The method for locating faulty sun gear teeth based on rotor sensing and tidal cycle effect according to claim 1, characterized in that: The process of signal acquisition and analysis includes: Use MEMS rotor sensing sensors to collect radial vibration and torsional vibration signals, perform mathematical modeling on the collected radial vibration and torsional vibration signals, analyze the projection changes of gravity acceleration in two directions, model and analyze the low-frequency signals in radial vibration and torsional vibration, and use low-pass filters to obtain absolute phase references; The radial vibration signal is subjected to cepstrum analysis to separate the deterministic and random components in the vibration signal, the gear meshing component is extracted using bandpass filtering, and the fault pulse signal in the random component is enhanced by minimum entropy deconvolution filtering technology and square envelope technology; By analyzing the periodic changes of the meshing position of the faulty tooth during the tidal cycle, a single tooth meshing number code of the interval between the faulty tooth of the sun gear and the absolute phase reference point is constructed.

3. The method for locating faulty sun gear teeth based on rotor sensing and tidal cycle effect according to claim 1 is characterized in that: The determination of tidal cycle effects includes: Study the meshing mode between the sun gear faulty tooth and the planetary gear, and calculate the meshing position change between the sun gear faulty tooth and the planetary gear; Combined with the parameters of the sun gear and the planetary gears, the number of rotations of the sun gear and the corresponding number of faulty tooth engagements during the tidal cycle are determined.

4. The method for locating faulty sun gear teeth based on rotor sensing and tidal cycle effect according to claim 3 is characterized in that: The steps for locating the faulty tooth include: Time domain compensation of radial vibration signal to ensure the acquisition of the complete sun gear meshing cycle; Cepstrum analysis is used to separate the deterministic and random components of the signal. A bandpass filter is used to extract the gear meshing components from the deterministic components, and the minimum entropy deconvolution filtering technology and square envelope technology are used to enhance the fault feature signal. Based on the tidal cycle effect and the absolute phase reference provided by the rotor sensing sensor, the single tooth meshing number encoding of the absolute phase reference point of the faulty sun gear tooth lagging behind the tidal cycle is determined, the unique identification of the faulty tooth is generated, and the position of the faulty tooth is determined.

5. The method for locating faulty sun gear teeth based on rotor sensing and tidal cycle effect according to claim 4 is characterized in that: The precise identification process of the fault location includes: Perform multi-cycle tests on the rotation of the sun gear and analyze the phase difference between the enhanced fault pulse and the gear mesh component; The phase is corrected according to the absolute phase reference provided by the rotor sensing sensor, and the single tooth meshing times encoding of the absolute phase reference point of the faulty sun gear tooth lag within the tidal cycle is found, and the position of the faulty tooth is finally determined.

Images