A signal detection method for an abrasive particle detection device with a U-shaped channel reflux oil circuit

Abstract

The present invention belongs to the field of oil fluid abrasive monitoring, and particularly relates to a signal detection method for an abrasive detection device with a U-shaped channel reflux oil circuit. The device includes an inductive abrasive detection sensor; the inductive abrasive detection sensor includes a U-shaped tube detection component, an excitation coil, an iron core, two detection coils, two oil pipe locking devices, and two magnetic poles; two detection signals of the oil fluid are collected, the movement rate of the abrasive is preliminarily estimated according to the detection signals, and the cut-off frequency of low-pass filtering is set; the detection signals are filtered according to the cut-off frequency, and the filtered signals are denoised to obtain two new detection signals, and the volume size of the abrasive is estimated according to the new detection signals; the present invention can accurately calculate the real-time movement speed of ferromagnetic wear particles in the oil fluid based on the time delay of the collected abrasive signals during oil fluid abrasive monitoring, and use the multi-dimensional least mean square algorithm to realize signal adaptive denoising, and realize abrasive signal recognition and abrasive size estimation.

Description

Technical Field

[0001] The present invention belongs to the field of oil particle monitoring, and particularly relates to a signal detection method for an abrasive particle detection device using a U-shaped channel return oil circuit. Background Art

[0002] Wear particles in a lubricating oil system are one of the important indicators reflecting the health status of mechanical equipment. By online monitoring the size and concentration signals of metal particles in the oil through an inductive abrasive particle detection sensor, the real-time assessment of the wear condition of mechanical equipment can be achieved. Currently, oil particle detection sensors have been widely used in fields such as the wear detection of ship gearboxes, the working state assessment of aeroengines, and the life prediction of wind power turbines. However, in practical applications, the induced signal generated when the abrasive particles pass through the sensor is very weak, and there are inevitably random noises and vibration interferences, which cause the sensor output abrasive particle signal to be distorted, thus severely restricting the detection performance of the inductive abrasive particle detection sensor. In response to this, some scholars have tried to use signal processing methods combined with abrasive particle signal models and recognition indicators to extract abrasive particle feature information. In the abrasive particle feature extraction algorithms related to signal processing, usually the energy of the abrasive particle signal is mainly concentrated in the low-frequency band, and the sampling signal often needs to select an appropriate cut-off frequency based on the frequency of the abrasive particle signal for low-pass filtering to suppress some high-frequency noises and interferences, and selecting an appropriate cut-off frequency will not damage the morphological features of the abrasive particles. If the selected cut-off frequency is too small, it may lead to the loss of the energy of the abrasive particle signal and cause the distortion of its morphological features; conversely, if the selected cut-off frequency is too large, it may leave high-frequency noises remaining, affecting the accurate identification of the subsequent abrasive particle signal. And the frequency of the abrasive particle signal is closely related to the movement speed of the abrasive particles in the oil. Therefore, the accurate calculation of the movement speed of the abrasive particles is of great significance for the extraction of the abrasive particle signal and the estimation of its size.

[0003] Currently, at home and abroad, based on the characteristic that metal abrasive particles are relatively small in the early stage of the mechanical system failure, it is directly assumed that the movement speed of the abrasive particles is equal to the oil flow rate to estimate the frequency range of the abrasive particle signal. However, due to the multi-field coupling effect of magnetic force, gravity, and oil viscosity resistance, the oil flow rate is not the same as the movement speed of the abrasive particles, which will lead to an estimation deviation of the central frequency of the abrasive particle signal; when the sensor structure and driving current are fixed, the amplitude of the abrasive particle induced voltage signal is mainly jointly determined by the volume of the abrasive particles and the movement speed of the abrasive particles; therefore, a large error in the estimation of the movement speed will significantly affect the accuracy of the calculation result of the size of the abrasive particles, thus being not conducive to evaluating the true wear condition of the equipment. Summary of the Invention

[0004] To solve the problems existing in the above prior art, the present invention proposes a signal detection method for a wear particle detection device with a U-shaped channel return oil circuit. The wear particle detection device with a U-shaped channel return oil circuit includes an inductive wear particle detection sensor; the inductive wear particle detection sensor includes a U-shaped tube detection component, an excitation coil, an iron core, a detection coil A, a detection coil B, two oil pipe locking devices, and two magnetic poles; the two oil pipe locking devices are respectively arranged on the two magnetic poles for fixing the U-shaped tube detection component; an air gap is arranged between the two magnetic poles, and the two magnetic poles are connected in parallel and facing each other through the iron core; the excitation coil is wound around the iron core; the two detection coils are respectively wound in the U-shaped tube detection component to form an inductive wear particle detection sensor; using the inductive wear particle detection sensor to detect signals includes:

[0005] S1: Continuously collect signals of the oil fluid containing metal wear particles by using a monitoring device including a detection sensor, a preamplifier, a signal acquisition card, and a gear oil pump; the collected signals include that when the metal wear particles pass through the detection flow channel A, the detection coil A collects the detection signal A; when the metal wear particles return through the detection flow channel B, the detection coil B collects the detection signal B;

[0006] S2: Estimate the corresponding time delay according to the moments corresponding to the same peak-to-peak value of the induced voltage in the detection signal A and the detection signal B, and combine with the known length of the effective detection area of the sensor to preliminarily estimate the movement speed of the wear particles in the oil fluid and the center frequency of the induced voltage; set the cut-off frequency of the low-pass filter according to the movement speed of the wear particles in the oil fluid and the center frequency of the induced voltage;

[0007] S3: Perform low-pass filtering and harmonic elimination processing on the detection signal A and the detection signal B according to the cut-off frequency of the low-pass filter; use the filtered detection signal A as the input signal of the least mean square filter, and at the same time use the detection signal B as the desired signal of the least mean square filter for adaptive noise reduction to obtain the detection signal C; use the detection signal B as the input signal and the detection signal A as the desired signal to input the least mean square filter to obtain the detection signal D;

[0008] S4: Perform sliding correlation coefficient analysis on the detection signal C and the detection signal D, that is, gradually discard the tail data of the detection signal C and the head data of the detection signal D, calculate the Pearson correlation of the remaining part, and when the correlation is the largest, the sliding length of the detection signal C and the detection signal D corresponds to the exact time delay between the detection signal C and the detection signal D; estimate the movement speed of the wear particles according to the exact time delay, and extract the amplitude numerical characteristics of the detection signal C and the detection signal D to estimate the corresponding volume size of the wear particles.

[0009] Preferably, the U-shaped tube detection component includes a U-shaped tube, a first detection oil pipe, and a second detection oil pipe; the first detection oil pipe and the second detection oil pipe are respectively connected to the two pipe orifices of the U-shaped tube.

[0010] Further, a detection coil A is wound around the middle position of the first detection oil pipe, and a detection coil B is wound around the middle position of the second detection oil pipe. Two groups of detection signals that can be used as reference signals for each other are collected through the two detection coils.

[0011] Preferably, a groove is provided at the lower end of the oil pipe locking device, and a threaded hole is provided on the groove; the groove is aligned with the magnetic pole, and an internal hexagon countersunk head bolt is used to fix it on the magnetic pole through a through hole.

[0012] Further, two through holes are provided in the upper part of the oil pipe locking device. The diameters of the two through holes match the diameters of the first detection oil pipe and the second detection oil pipe of the U-shaped pipe detection assembly, and the distance between the two through holes is equal to the distance between the first detection oil pipe and the second detection oil pipe.

[0013] Preferably, the detection signal A includes an abrasive particle signal and noise interference, the detection signal B only includes noise interference, and this interference is highly correlated with the interference in the detection signal A.

[0014] Preferably, the formula for estimating the corresponding time delay according to the moments corresponding to the same peak-to-peak value of the induced voltage in the detection signal A and the detection signal B is:

[0015]

[0016] wherein, l represents the distance from the midpoint of the first detection coil along the pipeline to the midpoint of the second detection coil, t 1max represents the time corresponding to the maximum amplitude of the abrasive particle signal in the detection signal A, t 2max represents the time corresponding to the maximum amplitude of the abrasive particle signal in the detection signal B, t 1min represents the time corresponding to the minimum amplitude of the abrasive particle signal in the detection signal A, t 2min represents the time corresponding to the minimum amplitude of the abrasive particle signal in the detection signal B.

[0017] Preferably, the formulas for initially estimating the movement speed of abrasive particles in the oil and the center frequency of the induced voltage are:

[0018] v lr = l / τ

[0019] wherein, l represents the distance from the midpoint of the first detection coil along the pipeline to the midpoint of the second detection coil, τ represents the initially estimated time delay, and v lr represents the initially estimated movement speed of abrasive particles in the oil;

[0020] f d = 1 / τ

[0021] wherein, f d represents the initially estimated center frequency of the induced voltage.

[0022] Preferably, the process of accurately estimating the moving speed of abrasive grains includes: sliding the detection signal D relative to the detection signal C and calculating the correlation coefficient after sliding; when the abrasive grain signals in the two detection signals coincide, the correlation coefficient value is the largest, and the time corresponding to the sliding distance of the detection signal D is the time for the abrasive grains to pass through the detection area; the calculation formula is:

[0023] v le = l / t cmax

[0024] wherein, t cmax = 2N w / f s represents the time corresponding to the sliding distance of the detection signal D when the correlation coefficient is the largest, N w represents the sliding lengths of the detection signal C and the detection signal D, and f s represents the sampling frequency.

[0025] Advantages of the present invention:

[0026] 1. The structural design of the U-shaped tube enables the metal abrasive grains to pass through the detection coil A. Due to the similar coil parameters and similar environments, the background noises between the detection signal B and the detection signal A are highly correlated. At the same time, when the metal abrasive grains pass through the detection coil B, the background noises in the detection signal A and the detection signal B are highly correlated. Therefore, inputting the detection signal A and the detection signal B into the least mean filter can achieve signal adaptive noise reduction, which is more conducive to filtering out noise signals similar to the real abrasive grain signals, extracting the characteristics of the abrasive grain signals, and retaining more complete abrasive grain signal characteristics.

[0027] 2. According to the different times when the induction signals appear when the abrasive grains pass through the detection coil A and the detection coil B in the U-shaped tube, the moving speed of the abrasive grains in the oil can be calculated, providing a data basis for subsequent signal processing and analysis. Compared with the traditional method of equating the moving speed of the abrasive grains to the oil flow rate, when using the present invention, the flow rates of the abrasive grains flowing through the A tube and the B tube are basically the same, so the obtained moving speed of the abrasive grains is more accurate, the size of the abrasive grains can be obtained more accurately, and no additional equipment is required to measure the moving speed of the abrasive grains.

[0028] 3. By comparing the processing results of the two sections of signals, this abrasive grain detection method for the U-shaped tube channel reflux oil circuit greatly improves the accuracy of abrasive grain signal recognition compared with the traditional abrasive grain sensor detection method, and reduces the situation of misjudging random noise as abrasive grain signals. Description of the Drawings

[0029] Figure 1 is a cross-sectional view of the inductive sensor detection device according to the embodiment provided by the present invention;

[0030] Figure 2 It is an exploded view of the inductive sensor detection device of the embodiment provided by the present invention;

[0031] Figure 3 It is a flowchart of a wear particle detection device based on a U-shaped channel reflux oil circuit and its signal detection method according to the embodiment provided by the present invention;

[0032] Figure 4 It is a schematic diagram of detection signal A and detection signal B according to the embodiment provided by the present invention;

[0033] Figure 5 It is a schematic diagram of signal adaptive noise cancellation according to the embodiment provided by the present invention;

[0034] Among them, 1. U-shaped tube detection component, 2. Detection coil, 3. Oil pipe locking device, 4. Magnetic pole, 5. Excitation coil, 6. Iron core, 7. U-shaped tube, 8. Flexible washer, 9. Socket head cap screw, 10. Flat head bolt, 1a. First detection oil pipe, 2a. Detection coil A, 1b. Second detection oil pipe, 2b. Detection coil B, 3a. First oil pipe locking device, 3b. Second oil pipe locking device. Specific embodiments

[0035] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0036] A wear particle detection device based on a U-shaped channel reflux oil circuit, as Figure 1 and Figure 2 shown, includes a U-shaped tube detection component 1, a detection coil 2, an oil pipe locking device 3, a magnetic pole 4, an excitation coil 5 and an iron core 6; the outer surface of the middle of the U-shaped tube detection component 1 is wound with a detection coil 2 and is connected to a preamplifier; the oil pipe locking device 3 is fastened to both sides of the magnetic pole 4 by screws, and its central hole is used to fix the U-shaped tube detection component 1; there is an air gap between the magnetic poles 4, which are connected in parallel and opposite by the iron core 6; the excitation coil 5 is wound on the iron core 6 and is connected to a regulated DC power supply.

[0037] Specific implementation of a U-shaped tube detection component, where the U-shaped tube detection component 1 is an integral component, including three parts: a first detection oil pipe (1a), a second detection oil pipe (1b), and a U-shaped tube 7. The outer surfaces of the middle parts of the first detection oil pipe (1a) and the second detection oil pipe (1b) are respectively wrapped with a detection coil A (2a) and a detection coil B (2b), and the two are respectively placed parallel to the magnetic poles and locked by an oil pipe locking device 3.

[0038] In this embodiment, the U-shaped tube detection component 1 is fixed by a flat head bolt 10 on the oil pipe locking device 3.

[0039] In this embodiment, the oil pipe locking device 3 is divided into a first oil pipe locking device 3a and a second oil pipe locking device 3b, and the two are fixed at both ends of the magnetic pole 4 by an internal hexagonal countersunk head bolt 9 and a flexible washer 8; the flexible washer 8 is embedded in the internal through hole of the oil pipe locking device 3, and the internal hexagonal countersunk head bolt 9 squeezes the flexible washer 8 after being threadedly connected to fix the oil pipe locking device 3.

[0040] A signal detection method for an abrasive detection device with a U-shaped channel return oil circuit, as Figure 3 shown, this method includes:

[0041] S1. Use a monitoring device including a detection sensor, a preamplifier, a signal acquisition card, and a gear oil pump to continuously collect signals from the oil fluid containing metal abrasives, and obtain the detection signal A and the detection signal B to be processed.

[0042] S2. Use the directly available time delay of the same abrasive signal to preliminarily estimate the movement speed of the abrasives in the oil fluid, and then set the cut-off frequency of the low-pass filter based on this speed.

[0043] S3. Perform preprocessing on the signals. Perform low-pass filtering and harmonic interference suppression on the detection signal A and the detection signal B with the same parameters, and then input the detection signal A and the detection signal B into the least mean square (LMS) filter as the input signal and the desired signal respectively to obtain the adaptive filtering result detection signal C and the detection signal D.

[0044] S4. Calculate the sliding correlation between the detection signal C and the detection signal D, determine the movement time of the abrasives from the detection coil A to the detection coil B, and accurately estimate the movement speed of the abrasives in combination with the size parameters of the U-shaped tube. Extract the numerical characteristics of the abrasive signal, and finally realize the identification and counting of the abrasive signal and the estimation of the size of the wear particles.

[0045] In this embodiment, the rough calculation formula for the movement speed of the abrasives in step S2 is specifically expressed as:

[0046]

[0047] Among them, l represents the distance from the midpoint of the first detection coil along the pipeline to the midpoint of the second detection coil, and t 1max represents the time corresponding to the maximum amplitude of the abrasive particle signal in the detection signal A, and t 2max represents the time corresponding to the maximum amplitude of the abrasive particle signal in the detection signal B, and t 1min represents the time corresponding to the minimum amplitude of the abrasive particle signal in the detection signal A, and t 2min represents the time corresponding to the minimum amplitude of the abrasive particle signal in the detection signal B.

[0048] In signal preprocessing, the cut-off frequency of the low-pass filter is selected as:

[0049] f c =(s d v l ) / l c

[0050] Among them, l c represents the length of the abrasive particle detection area, s d represents the cut-off frequency extension parameter, and v l represents the speed of the abrasive particle detection area.

[0051] An optional implementation manner is that the cut-off frequency extension parameter can be any number between 2.5 and 3.

[0052] In this embodiment, the detection signal A and the detection signal B in step S3 are preprocessed to obtain the detection signal A sequence p(n) and the detection signal B sequence q(n); when using p(n) and q(n) as the input of the LMS algorithm, the adaptive filtering process can be expressed as (1) noise estimation (2) filtering process (3) adaptive process Among them, p(n)=[p(n),p(n - 1),...,p(n - L + 1)] represents a vector with a length of L in the detection signal B, ω represents the weight vector, represents the estimated value of the noise signal in the detection signal A at the nth moment, represents the filtering result, and μ represents the global update step size.

[0053] In this embodiment, step S4 performs a correlation coefficient analysis by sliding the detection signal D and the detection signal C. When the abrasive particle signals in the two detection signals coincide, the correlation coefficient should reach the highest. The time corresponding to the sliding distance of the detection signal D is the time when the abrasive particle passes through the detection area. The precise calculation formula for the movement speed of the abrasive particle in the oil is specifically expressed as v le =l / t cmax , where l represents the distance from the midpoint of the detection coil A along the pipeline to the midpoint of the detection coil B, and t cmaxIt represents the time corresponding to the sliding distance of the detection signal D when the correlation coefficient is the largest.

[0054] To illustrate the situation of the detection device and signal processing for adaptive noise reduction, as well as the accurate speed of abrasive particles in the oil, in this embodiment, three metal abrasive particles with sizes of 400um, 250um, and 130um are selected for experiments. The experimental parameters are set as follows: sampling frequency fs = 5000, sampling time T = 70s for each group, amplifier magnification is 2000, the excitation coil is set in a constant voltage and constant current mode, with a size I = 0.5A, and the gear pump flow rate is set to 510ml / min. Under the action of the gear pump, the abrasive particles will flow through the first detection oil pipe, U-shaped pipe, and second detection oil pipe in sequence. When the abrasive particles flow through detection coil A and detection coil B, they will change the local magnetic field at the sensor air gap, causing the magnetic flux of the detection coil to change, thereby outputting an induced electromotive force. If the operating time of the detection device is taken as the abscissa, and the collected detection signal and reference signal are taken as the ordinates respectively, a change curve can be made as Figure 4 shown.

[0055] Reference Figure 4 , in the time domain diagram, the two abrasive particles with sizes of 250um and 130um are masked by noise, but the signal characteristic information of the abrasive particle with a size of 400um can be directly obtained. According to the calculation formula of the time delay and movement speed of the abrasive particle signal, the movement speed v of the abrasive particle can be initially estimated lr ≈800mm / s, where l = 80cm. Then, by setting the cut-off frequency f according to the size of the abrasive particle movement speed c = 100Hz, where s d takes 2.5, and l c = 20mm. After signal preprocessing and adaptive noise reduction, the final result is as Figure 5 shown. It can be seen that after signal processing, the signal characteristic information of the two abrasive particles with sizes of 250um and 130um can be well reflected. At this time, the movement speed of the abrasive particles in the oil can be further accurately calculated based on the time delay of the abrasive particle signal.

[0056] The above-mentioned embodiments further elaborate on the purpose, technical solutions, and advantages of the present invention. It should be understood that the above-mentioned embodiments are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made to the present invention within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A signal detection method for a wear particle detection device with a U-shaped channel reflux oil circuit, characterized in that, the wear particle detection device of the U-shaped channel reflux oil circuit includes an inductive wear particle detection sensor; the inductive wear particle detection sensor includes a U-shaped tube detection component, an excitation coil, an iron core, a detection coil A, a detection coil B, two oil pipe locking devices and two magnetic poles; the two oil pipe locking devices are respectively arranged on the two magnetic poles for fixing the U-shaped tube detection component; an air gap is arranged between the two magnetic poles, and the two magnetic poles are connected in parallel and facing each other through the iron core; the excitation coil is wound around the iron core; the two detection coils are respectively wound in the U-shaped tube detection component to form an inductive wear particle detection sensor; The signal detection by using the inductive wear particle detection sensor includes: S1: Continuously collect signals of the oil fluid containing metal wear particles by using a monitoring device including a detection sensor, a preamplifier, a signal acquisition card and a gear oil pump; the collected signals include that when the metal wear particles pass through the detection flow channel A, the detection coil A collects the detection signal A; when the metal wear particles reflux through the detection flow channel B, the detection coil B collects the detection signal B; S2: Estimate the corresponding time delay according to the moments corresponding to the same peak-to-peak value of the induced voltage in the detection signal A and the detection signal B, and combine the known length of the effective detection area of the sensor to preliminarily estimate the movement speed of the wear particles in the oil fluid and the center frequency of the induced voltage; set the cut-off frequency of the low-pass filter according to the movement speed of the wear particles in the oil fluid and the center frequency of the induced voltage; S3: Perform low-pass filtering and harmonic elimination processing on the detection signal A and the detection signal B according to the cut-off frequency of the low-pass filter; use the filtered detection signal A as the input signal of the least mean square filter, and at the same time use the detection signal B as the desired signal of the least mean square filter for adaptive noise reduction to obtain the detection signal C; use the detection signal B as the input signal and the detection signal A as the desired signal to input the least mean square filter to obtain the detection signal D; S4: Perform sliding correlation coefficient analysis on the detection signal C and the detection signal D, that is, gradually discard the tail data of the detection signal C and the head data of the detection signal D, calculate the Pearson correlation of the remaining part, and when the correlation is the largest, the precise time delay between the detection signal C and the detection signal D corresponding to the sliding length of the detection signal C and the detection signal D; estimate the movement speed of the wear particles according to the precise time delay, and extract the amplitude numerical characteristics of the detection signal C and the detection signal D to estimate the volume size of the corresponding wear particles; The process of accurately estimating the movement speed of the wear particles includes: sliding the detection signal D relative to the detection signal C and calculating the correlation coefficient after sliding; when the wear particle signals in the two detection signals coincide, the correlation coefficient value is the largest, and the time corresponding to the sliding distance of the detection signal D is the time for the wear particles to pass through the detection area; the calculation formula is: v le = l / t cmax where, t cmax = 2N w / f s represents the time corresponding to the sliding distance of the detection signal D when the correlation coefficient is the largest, N w represents the sliding lengths of the detection signal C and the detection signal D, and f s represents the sampling frequency; l represents the distance from the midpoint of the first detection coil along the pipeline to the midpoint of the second detection coil.

2. The signal detection method for a wear particle detection device with a U-shaped channel reflux oil circuit according to claim 1, characterized in that, The U-shaped tube detection component includes a U-shaped tube, a first detection oil pipe, and a second detection oil pipe; the first detection oil pipe and the second detection oil pipe are respectively connected to two pipe orifices of the U-shaped tube.

3. The signal detection method of a wear particle detection device for a U-shaped channel return oil circuit according to claim 2, wherein, A detection coil A is wound around the middle position of the first detection oil pipe, and a detection coil B is wound around the middle position of the second detection oil pipe. Two groups of detection signals that can be mutually used as reference signals are collected through the two detection coils.

4. The signal detection method of a wear particle detection device for a U-shaped channel return oil circuit according to claim 1, wherein, A groove is provided at the lower end of the oil pipe locking device, and a threaded hole is provided on the groove; the groove is aligned with the magnetic pole, and an internal hexagon countersunk head bolt is used to fix it on the magnetic pole through a through hole.

5. The signal detection method of a wear particle detection device for a U-shaped channel return oil circuit according to claim 4, wherein, Two through holes are provided in the upper part of the oil pipe locking device. The diameters of the two through holes match the diameters of the first detection oil pipe and the second detection oil pipe of the U-shaped tube detection component, and the distance between the two through holes is equal to the distance between the first detection oil pipe and the second detection oil pipe.

6. The signal detection method of a wear particle detection device for a U-shaped channel return oil circuit according to claim 1, wherein, The detection signal A includes a wear particle signal and noise interference, the detection signal B only includes noise interference, and this interference is highly correlated with the interference in the detection signal A.

7. The signal detection method of a wear particle detection device for a U-shaped channel return oil circuit according to claim 1, wherein, The formula for estimating the corresponding time delay according to the moments corresponding to the same peak-to-peak value of the induced voltage in the detection signal A and the detection signal B is: Among them, t 1max represents the time corresponding to the maximum amplitude of the abrasive particle signal in the detection signal A, and t 2max represents the time corresponding to the maximum amplitude of the abrasive particle signal in the detection signal B, and t 1min represents the time corresponding to the minimum amplitude of the abrasive particle signal in the detection signal A, and t 2min represents the time corresponding to the minimum amplitude of the abrasive particle signal in the detection signal B.

8. The signal detection method of a wear particle detection device according to claim 7, wherein, The formula for preliminarily estimating the movement speed of wear particles in the oil and the center frequency of the induced voltage is: v lr = l / τ Among them, v lr represents the preliminarily estimated movement speed of abrasive particles in the oil fluid; f d = 1 / τ Among them, f d represents the center frequency of the induced voltage estimated preliminarily.

Images