A method for enhancing weak output characteristic signals of an oil metal particle sensor

Abstract

The application discloses a weak output characteristic signal enhancement processing method of an oil metal particle sensor, and the method comprises the following steps: setting a sampling frequency and a sampling point number, and collecting original output signals of the oil metal particle sensor when tiny metal particles pass through the oil metal particle sensor; performing mean value removal processing and amplitude scaling processing on the original output signals to obtain weak characteristic signals; enhancing the weak characteristic signals by using a nonlinear system to obtain initial enhanced output weak characteristic signals; and performing point-by-point smoothing filtering and mean value removal processing on the initial enhanced output weak characteristic signals to obtain final enhanced output weak characteristic signals. The application is applied to the field of weak digital signals, can effectively enhance the weak output characteristic signals of 100 mu m tiny metal particles when the tiny metal particles pass through the sensor, and improve the detection sensitivity of the sensor. Moreover, the application has small calculation amount, good robustness and adaptability, can be standardized, and is convenient for embedded microprocessors to realize signal operation and processing.

Description

Technical Field

[0001] This invention relates to the field of weak digital signal technology, specifically a method for enhancing the weak output characteristic signal of a metal particle sensor when tiny metal particles pass through an oil-based metal particle sensor, which can improve the sensor's detection sensitivity for tiny metal particles. Background Technology

[0002] Oil-based metal particle sensors are widely used in online health monitoring and fault prediction of complex mechanical systems such as engines and wind turbine gearboxes. These sensors are online oil detection sensors, typically installed in the lubrication circuit of the mechanical system. They detect the number and size of metal particles in the flowing lubricating oil to monitor whether mechanical components are experiencing abnormal wear or malfunctions.

[0003] According to experimental statistics, the equivalent diameter of most metal particles generated during normal wear of mechanical parts is less than 100 μm, while the size of metal particles generated during early abnormal wear ranges from 100 μm to 150 μm. Therefore, effective monitoring of metal wear particles with an equivalent diameter of around 100 μm is crucial for detecting early wear faults. However, detecting metal wear particles of around 100 μm is quite difficult. On the one hand, because the particle size is small, the output characteristic signal when the particle passes through the sensor is very weak. On the other hand, the sensor is affected by ambient noise such as vibration and environmental interference, which can drown out the originally weak signal, making it difficult to detect the particle characteristic signal.

[0004] Currently, wavelet analysis and time-domain filtering are commonly used to eliminate noise. However, these methods require adjusting many parameters, have a large computational load, poor real-time performance, and cannot effectively extract the output feature signals of tiny metal particles that are almost submerged in noise, which greatly affects the sensor's sensitivity to detect tiny metal particles. Summary of the Invention

[0005] To address the shortcomings of the prior art, this invention provides a method and system for enhancing the weak output characteristic signal of an oil-based metal particle sensor. This method enhances the output characteristic signal of tiny metal particles from the metal particle sensor, thereby improving the sensor's detection sensitivity for tiny metal particles with an equivalent diameter of approximately 100 μm.

[0006] To achieve the above objectives, the present invention provides a method for enhancing the weak output characteristic signal of an oil metal particle sensor, comprising the following steps:

[0007] Step 1: Set the sampling frequency and the number of sampling points, and collect the original output signal s0(n) when tiny metal particles pass through the oil metal particle sensor based on the sampling frequency and the number of sampling points;

[0008] Step 2: Preprocess the original output signal s0(n) to obtain the weak feature signal s(n). The preprocessing includes mean removal and amplitude scaling.

[0009] Step 3: Use a nonlinear system to enhance the weak feature signal s(n) to obtain the initial enhanced weak feature signal x(n);

[0010] Step 4: Perform point-by-point smoothing filtering and mean removal on the initial enhanced weak feature signal x(n) to obtain the final enhanced weak feature signal s2(n).

[0011] In one embodiment, step 1, setting the sampling frequency and the number of sampling points, specifically involves:

[0012] fs = M / T

[0013] N≥3M

[0014] In the formula, fs is the sampling frequency, M is the number of samples corresponding to the width of a single output signal pulse waveform, used to control the sampling frequency, requiring M≥100, T is the width of the characteristic signal pulse waveform output by the oil metal particle sensor, and N is the number of sampling points.

[0015] In one embodiment, step 2, the preprocessing process specifically includes:

[0016] First, the original output signal s0(n) is subjected to mean removal processing, and then the amplitude of the original output signal s0(n) is normalized to the interval [-1,1] through amplitude scaling processing to obtain the weak feature signal s(n).

[0017] In one embodiment, in step 3, the enhancement of the weak feature signal s(n) using a nonlinear system specifically involves using the weak feature signal s(n) as an excitation input to a nonlinear discrete equation, which is:

[0018] x(n+1)=(h+1)·x(n)+h[s(n)-x 3 [n]n=0,1,…N-1

[0019] In the formula, x(n) is the weak feature signal of the nth initial enhanced output, x(n+1) is the weak feature signal of the (n+1)th initial enhanced output, h is the discrete step size, N is the number of sampling points, and the initial value of the nonlinear discrete equation is set to x(0) = 1 or x(0) = -1, h = 1 / M.

[0020] In one embodiment, step 4 specifically includes:

[0021] The weak feature signal x(n) of the initial enhanced output is smoothed point by point to obtain the weak feature signal s1(n) of the initial enhanced output after filtering.

[0022]

[0023] In the formula, K is the number of smoothing points, where K = M / 10;

[0024] The weak feature signal s1(n) of the initial enhanced output after filtering is subjected to mean removal processing to obtain the weak feature signal s2(n) of the final enhanced output.

[0025] To achieve the above objectives, the present invention also provides a system for enhancing and processing the weak output characteristic signal of an oil metal particle sensor, comprising:

[0026] At least one processing unit;

[0027] And at least one memory and bus connected to the processing unit; wherein the processing unit and the memory communicate with each other through the bus;

[0028] The processing unit is used to call program instructions in the memory to execute some or all of the steps of the above method.

[0029] Compared with the prior art, the present invention has the following beneficial technical effects:

[0030] This invention employs a combination of nonlinear signal processing and smoothing filtering to effectively enhance the weak output characteristic signal when 100μm micro-metal particles pass through a sensor, thereby improving the sensor's detection sensitivity. The method provided by this invention not only has low computational complexity and good real-time performance, but also requires few parameters. The only parameters that need to be considered in the entire processing flow are related to the sampling frequency f. s The relevant parameter M is used as a reference. Other parameters such as the number of sampling points N, the step size h, and the number of smoothing points K can be set based on M. The entire signal is normalized, and the signal processing process has good robustness and adaptability. It can be standardized, which makes it easy for embedded microprocessors to implement signal calculation and processing. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0032] Figure 1This is a flowchart of the method for enhancing the output characteristic signal of tiny metal particles in an embodiment of the present invention;

[0033] Figure 2 This is a waveform diagram of the ideal output characteristic signal when metal particles pass through the sensor in an embodiment of the present invention;

[0034] Figure 3 This is a waveform diagram of the output signal with noise interference when a 100μm ferromagnetic metal particle passes through the sensor in an embodiment of the present invention;

[0035] Figure 4 This is a waveform diagram of the sensor's noisy output signal after demeaning and normalization in an embodiment of the present invention.

[0036] Figure 5 This is a waveform diagram of the output signal after the normalized noisy signal is input into the nonlinear system in an embodiment of the present invention;

[0037] Figure 6 This is a waveform diagram of the output signal of the nonlinear system in an embodiment of the present invention after smoothing and filtering.

[0038] Figure 7 This is a waveform diagram of the output signal after the mean is removed from the smoothed filter in an embodiment of the present invention.

[0039] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0040] 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 a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0041] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible for those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0042] like Figure 1 The above is a method for enhancing the weak output characteristic signal of an oil metal particle sensor disclosed in this embodiment, which mainly includes the following steps 1-4.

[0043] Step 1, set a suitable sampling frequency f sThe original output signal s0(n) of tiny metal particles passing through the oil metal particle sensor is acquired based on the sampling frequency and the number of sampling points N.

[0044] Ideally, when a metal particle passes through an electromagnetic metal particle sensor, its output characteristic signal is a positive and negative pulse waveform, such as... Figure 2 As shown. The width T of the pulse waveform is related to the flow rate of the lubricating oil through the sensor, but not directly related to the size of the metal particles. The sampling frequency fs can be set according to the pulse waveform width T. In this embodiment, the sampling frequency is set to fs = M / T, and the number of sampling points is set to N ≥ 3M, where M is the number of samples corresponding to the width of a single output signal pulse waveform, and is a sampling frequency control parameter. The value of M is determined based on the numerical simulation experiment of the model, requiring M ≥ 100 for better model processing performance. After setting the sampling frequency fs and the number of sampling points N, the original output signal of the small metal particles passing through the oil metal particle sensor can be sampled to obtain the original output signal s0(n).

[0045] Step 2: Preprocess the original output signal s0(n) to obtain the weak feature signal s(n). The preprocessing includes mean removal and amplitude scaling, and the specific implementation process is as follows:

[0046] First, the original output signal s0(n) is subjected to mean removal processing, and then the amplitude of the original output signal s0(n) is normalized to the interval [-1,1] through amplitude scaling processing to obtain the weak feature signal s(n).

[0047] Step 3: Enhance the weak feature signal s(n) using a nonlinear system to obtain the initial enhanced weak feature signal x(n). The specific implementation process is as follows:

[0048] Taking the weak characteristic signal s(n) as the excitation input, the nonlinear discrete equation is:

[0049] x(n+1)=(h+1)·x(n)+h[s(n)-x 3 [n]n=0,1,…N-1

[0050] In the formula, x(n) is the weak feature signal of the nth initial enhancement output, x(n+1) is the weak feature signal of the (n+1)th initial enhancement output, h is the discrete step length, and N is the number of sampling points.

[0051] Solving the nonlinear discrete equation using a numerical algorithm yields the initial enhanced output weak characteristic signal x(n). The initial value of the nonlinear discrete equation is set to x(0) = 1 or x(0) = -1. In this case, the nonlinear discrete equation essentially describes the motion of a particle in a bistable potential well subjected to s(n) at a certain initial monostable point x(0) = 1 or x(0) = -1. Due to the nonlinear effect of the single potential well, the nonlinear discrete equation has a good enhancement effect on the weak pulse signal in the noisy input s(n), thereby enhancing and highlighting the weak characteristic signal corresponding to the tiny metal particles.

[0052] Step 4 involves performing point-by-point smoothing filtering and mean removal on the initial enhanced weak feature signal x(n) to obtain the final enhanced weak feature signal s2(n). The specific implementation process is as follows:

[0053] The weak feature signal x(n) of the initial enhanced output is smoothed point by point to obtain the weak feature signal s1(n) of the initial enhanced output after filtering.

[0054]

[0055] In the formula, K is the number of smoothing points;

[0056] The weak feature signal s1(n) of the initial enhanced output after filtering is subjected to mean removal processing to obtain the weak feature signal s2(n) of the final enhanced output.

[0057] In step 3, the dispersion step length is set to h = 1 / M; in step 4, the number of smoothing points is K = M / 10. That is, both the dispersion step length and the number of smoothing points are set based on parameter M, which is related to the width of the pulse waveform of the characteristic signal output by the oil metal particle sensor, thus ensuring that the signal processing method is adaptable to the output characteristic signal.

[0058] The above method is used to enhance the noisy output signal of a metal particle sensor for a tiny ferromagnetic metal particle with an equivalent diameter of 100 μm. The steps are as follows:

[0059] Step 1: Set an appropriate sampling frequency. Taking an oil flow rate of 2 m / s as an example, according to experimental tests, the ideal pulse waveform width T of the characteristic signal output by metal particles at this flow rate is approximately 10 ms. Let f be the sampling frequency. s=200 / T=20KHz, i.e., M=200, set the number of sampling points N=1000, and sample the original output signal of the oil metal particle sensor to obtain s0(n). The smaller the metal particle size, the smaller the amplitude of the output characteristic signal, the greater the interference noise, and the lower the signal-to-noise ratio of the output signal. In a strong noise environment, it is difficult to find a clear pulse characteristic signal corresponding to the metal particle in the output signal when a tiny metal particle with an equivalent diameter of 100μm passes through the sensor. The waveform of s0(n) is as follows: Figure 3 As shown.

[0060] Step 2: Sample signal preprocessing. The mean of the signal s0(n) acquired in Step 1 is removed and normalized to the interval [-1, 1] to obtain s(n), where the waveform of s(n) is as follows: Figure 4 As shown.

[0061] Step 3: Enhance weak feature signals using a nonlinear system. Using s(n) as the excitation, input the following nonlinear discrete equation, where h = 1 / M = 1 / 200 = 0.005. Set the initial value of the nonlinear discrete equation to x(0) = 1. Solve the equation using a numerical algorithm to obtain the output signal x(n), where the waveform of x(n) is as follows: Figure 5 As shown.

[0062] Step 4: Perform point-by-point smoothing filtering on the signal x(n) obtained in Step 3. During filtering, the number of smoothing points K = M / 10 = 20, and the smoothing is performed according to the following formula:

[0063]

[0064] After smoothing and filtering, the output signal s1(n) is obtained, and its waveform is as follows: Figure 6 As shown. The obtained signal s1(n) is subjected to mean-removal processing to obtain the characteristic pulse signal s2(n) corresponding to the final enhanced output of the tiny metal particles, the waveform of which is shown below. Figure 7 As shown.

[0065] The results above demonstrate that the weak signal enhancement method proposed in this invention can effectively enhance the output pulse characteristic signal of 100μm micro ferromagnetic metal particles passing through the sensor in a strong noise background, thereby improving the detection sensitivity of the metal particle sensor for micro ferromagnetic metal particles and providing effective wear information support for early wear fault detection in mechanical systems.

[0066] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for enhancing the weak output characteristic signal of an oil-metal particle sensor, characterized in that, Includes the following steps: Step 1: Set the sampling frequency and the number of sampling points, and collect the raw output signal of tiny metal particles passing through the oil metal particle sensor based on the sampling frequency and the number of sampling points. s 0( n ); Step 2, process the original output signal s 0( n Preprocessing is performed to obtain weak feature signals. s ( n The preprocessing includes mean removal and amplitude scaling. Step 3: Enhance weak feature signals using nonlinear systems. s ( n This yields the weak characteristic signal of the initial enhanced output. x ( n The method of enhancing weak feature signals using nonlinear systems s ( n Specifically, this involves analyzing weak feature signals. s ( n The nonlinear discrete equation used as the excitation input is: In the formula, x ( n ) is the first n The weak characteristic signal of the initial enhanced output, x ( n +1) is the first n +1 initial enhanced weak feature signal of the output, h For the long walk, N This represents the number of sampling points; The initial values ​​of the nonlinear discrete equations are set as follows: x (0) = 1 or x (0) = -1; h = 1 / M ,in, M The number of samples corresponding to the width of a single output signal pulse waveform, used to control the sampling frequency; Step 4: Analyze the weak feature signals of the initial enhanced output. x ( n Point-by-point smoothing filtering and mean removal are performed to obtain the weak characteristic signal of the final enhanced output. s 2( n Specifically, this includes: Weak characteristic signal of the initial enhanced output x ( n Point-by-point smoothing filtering is performed to obtain the weak characteristic signal of the initial enhanced output after filtering. s 1( n ),for; In the formula, K The number of smoothing points; Weak characteristic signal of the initial enhanced output after filtering s 1( n The mean-removing process is performed to obtain the weak feature signal of the final enhanced output. s 2( n ).

2. The method for enhancing the weak output characteristic signal of an oil-metal particle sensor according to claim 1, characterized in that, In step 1, setting the sampling frequency and the number of sampling points specifically involves: f s = M / T N ≥ 3 M In the formula, f s is the sampling frequency. T The width of the characteristic signal pulse waveform output by the oil metal particle sensor. M The number of samples corresponding to the width of a single output signal pulse waveform, used to control the sampling frequency. N This represents the number of sampling points.

3. The method for enhancing the weak output characteristic signal of an oil-metal particle sensor according to claim 1, characterized in that, M ≥ 100。 4. The method for enhancing the weak output characteristic signal of an oil-metal particle sensor according to claim 1, 2, or 3, characterized in that, In step 2, the preprocessing process specifically includes: First, analyze the original output signal. s 0( n The mean is removed, and then the amplitude is scaled to restore the original output signal. s 0( n The signal amplitude is normalized to the interval [-1, 1] to obtain the weak characteristic signal. s ( n ).

5. The method for enhancing the weak output characteristic signal of an oil-metal particle sensor according to claim 1, 2, or 3, characterized in that, K = M / 10, of which, M This represents the number of samples corresponding to the width of a single output signal pulse waveform, used to control the sampling frequency.

Images