PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation

Abstract

The invention discloses a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation, and relates to the technical field of PCCP broken wire signal identification methods, and the method comprises the steps: S1, obtaining a broken wire signal, and analyzing the broken wire signal to obtain the frequency distribution feature of the broken wire signal, the method comprises the steps of S1, determining parameters of a band-pass filter according to frequency distribution characteristics, and setting the band-pass filter according to the parameters, S2, inputting a detection signal, and performing band-pass filtering processing on the detection signal through the set band-pass filter to obtain a filtered signal; s3, performing operation processing on the filtering signal to obtain a target modulus length; and S4, setting a target module length threshold value, and judging the occurrence moment of the PCCP broken wire signal. According to the method, the PCCP broken wire signal can be accurately identified, the operation efficiency is high, and the feasibility of real-time monitoring is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of PCCP broken wire signal identification methods, and in particular to a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation. Background Art

[0002] Prestressed Concrete Cylinder Pipe (PCCP) has the characteristics of long life, good earthquake resistance, and anti-leakage, and is widely used in water supply and drainage projects. The research on electromagnetic nondestructive testing technology for PCCP broken wires mentioned that the commonly used methods in the industry are the in-pipe electromagnetic detection method, far-field eddy current detection method, and orthogonal electromagnetic detection for PCCP broken wires. However, these methods rely on manual or robot-moving detection coils for detection. When facing long-distance PCCP pipelines, they face problems such as long detection cycles and inconvenient operations. In addition, there are teams that use optical fiber sensors to collect broken wire signals for analysis and then conduct detection. The advantages of distributed optical fiber sensing technology are long-distance detection and high sensitivity, which are particularly suitable for the collection and monitoring of PCCP broken wire signals, but the problem at the same time is how to capture instantaneous broken wire signals from continuous broadband background noise.

[0003] Broken wire signals are impact signals, characterized by instantaneity, non-periodicity, and frequency attenuation from high to low. They are often submerged in background noise and difficult to distinguish. Regarding the monitoring of PCCP broken wire signals, the commonly used methods currently include modal decomposition, Hilbert-Huang transform, machine learning, short-time energy spectrum, etc. Modal decomposition separates noise from PCCP broken wire signals by decomposing noisy signals into a series of intrinsic modal components. Hilbert-Huang transform is to perform Hilbert transform on the modal components of empirical mode decomposition, mapping the time domain signal into a time-frequency-amplitude distribution, which is particularly suitable for the identification of abnormal points of non-stationary signals. Machine learning methods are often combined with denoising methods and feature extraction methods to obtain a large number of samples with PCCP broken wire signal label data, train the model, and then realize the identification of PCCP broken wire signals. Vibration signal feature extraction of hydropower unit based on wavelet packet decomposition and CEEMDAN energy entropy The vibration characteristics of the hydropower unit are extracted by wavelet packet decomposition and adaptive noise complete empirical mode decomposition; Planetary transmission system meshing impact extraction technology and verification The instantaneous frequency of the signal is solved by Hilbert-Huang transform to determine the occurrence time of PCCP broken wire signal; Research on heart rate extraction and ECG signal waveform reconstruction based on cardiac impact signal The method of combining adaptive noise complete empirical mode decomposition with convolutional neural network is used to identify cardiac PCCP broken wire signal.

[0004] These methods have complex algorithm design, high computational overhead or require a lot of data samples for training. They are not efficient and easy to use, and it is difficult to achieve online real-time processing. In the direction of time-frequency analysis, the short-time energy spectrum can observe the energy distribution of the signal frequency in the time and frequency dimensions, which is suitable for observing the change of the signal frequency over time. However, this method is constrained by the Heisenberg uncertainty principle, which makes the time resolution and frequency resolution mutually restricted. If the selected parameters are not appropriate, the observation effect will be very poor. In the direction of instantaneous frequency calculation.

[0005] Patent CN107992448B discloses a method for solving the instantaneous frequency based on direct arc cosine based on absolute value. It uses direct arc cosine based on absolute value to solve the instantaneous frequency, bypassing the phase expansion process. Compared with Hilbert-Huang transform and direct arc cosine method based on phase expansion, the algorithm complexity is greatly reduced, and the speed of solving the instantaneous frequency is significantly improved. However, the difference in the instantaneous frequency characteristics between the PCCP broken wire signal and the noise is not significant, which makes subsequent judgment difficult. Summary of the invention

[0006] In order to solve the above problems, the present invention provides a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation to achieve the purpose of accurate positioning.

[0007] In the present invention, a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation is provided, comprising the following steps:

[0008] S1. Obtain a broken wire signal, analyze the broken wire signal to obtain a frequency distribution characteristic of the broken wire signal, determine parameters of a band-pass filter according to the frequency distribution characteristic, and set the band-pass filter according to the parameters;

[0009] S2, input detection signal, and perform bandpass filtering on the detection signal S(t) through the set bandpass filter to obtain the filtered signal S band (t);

[0010] S3, filter signal S band (t) is processed to obtain the target modulus length C(t);

[0011] S4. Set the target module length threshold K to determine the moment when the PCCP wire break signal occurs.

[0012] Preferably, the specific content of S1 is as follows:

[0013] Acquire the detection signal, calculate the short-time energy spectrum of the detection signal, and determine the frequency band distribution range of the broken wire signal and the background noise;

[0014] Set the bandpass filter passband to 10K-12KHz.

[0015] Preferably, the specific steps of acquiring the detection signal, calculating the short-time energy spectrum of the detection signal, and determining the frequency band distribution range of the broken wire signal and the background noise are as follows:

[0016] S101, setting parameters for short-time energy spectrum calculation, wherein the Hamming window size win is set to 10000, the overlap length noverlap is set to 5000, and the FFT calculation length nfft is set to 10000;

[0017] S102, input a detection signal segment S(t) with a sampling frequency Fs of 50KHz and a sampling time of 10S, calculate the short-time energy spectrum, and obtain the time vector T, frequency vector F and corresponding energy S of the detection signal segment S(t);

[0018] S103, making a corresponding short-time energy spectrum according to the time vector T, frequency vector F and corresponding energy S corresponding to the detection signal segment S(t) calculated in S102;

[0019] S104, determining, according to the short-time energy spectrum, a range where the energy difference between the broken wire signal and the background noise frequency band is the largest as the passband of the bandpass filter.

[0020] Preferably, the expression for calculating the short-time energy spectrum in S102 is as follows:

[0021] (S,F,T)=Spec(S(t),win,noverlap,nfft,Fs)

[0022] Where Spec represents the short-time energy spectrum calculation of the detection signal segment, S(t). After the short-time energy spectrum transformation, the resolution of the time vector T is 0.1S and the resolution of the frequency vector F is 5Hz.

[0023] Preferably, the specific steps of operating and processing the filtered signal to obtain the target modulus length in S3 are:

[0024] S301 filters the signal S band (t) is processed by arc cosine to obtain the instantaneous phase

[0025] S302, instantaneous phase Directly perform the forward difference diff operation to obtain the instantaneous frequency F1(t) that changes with time;

[0026] S303, performing a forward difference diff operation on the instantaneous frequency F1(t) to obtain an instantaneous frequency change rate F2(t);

[0027] S304, square the instantaneous frequency change rate F2(t) to obtain the modulus length, and obtain the primary modulus length A1(t);

[0028] S305 , filter A1(t) using a median filter to obtain a target modulus C(t).

[0029] Preferably, the modulus length threshold K in S4 is set according to the calculated background noise modulus length, and is generally set to be more than twice the background noise modulus length;

[0030] If the target module length C(t) is greater than the module length threshold, the corresponding moment is the moment when the PCCP wire break signal occurs;

[0031] If the target module length C(t) is less than or equal to the module length threshold, the corresponding time is not the time when the PCCP wire break signal occurs.

[0032] In summary, the PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation of the present invention has the following advantages over the traditional PCCP broken wire signal recognition method technology:

[0033] (1) The PCCP broken wire signal feature enhancement and positioning method of the present invention adopts the direct inverse cosine method and the forward difference method to solve the square modulus of the instantaneous frequency change rate of the detection signal. It can realize the feature enhancement and identification of the PCCP broken wire signal without complex time-frequency domain processing. The algorithm is simple and efficient, and is more suitable for online operation.

[0034] (2) Compared with the existing direct arccosine instantaneous frequency solution method based on absolute value, the PCCP broken wire signal feature enhancement and positioning method of the present invention can better suppress background noise and process the PCCP broken wire signal feature more significantly. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The present invention is a flow chart of the PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation.

[0038] Figure 2 It is the detection signal of an embodiment of the PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation of the present invention.

[0039] Figure 3 It is a filtered signal of an embodiment of a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation of the present invention.

[0040] Figure 4 It is the instantaneous frequency calculated by the direct arc cosine instantaneous frequency solution method based on absolute value in an embodiment of the present invention.

[0041] Figure 5 It is the calculation result of the PCCP broken wire signal feature enhancement method based on threshold-instantaneous frequency change rate calculation in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The technical solution of the present invention is further described below by means of the accompanying drawings and embodiments. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and numerical values ​​described in these embodiments do not limit the scope of the present invention.

[0043] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0044] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0045] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0046] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0047] like Figure 1 As shown, the present invention proposes a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation, comprising the following steps:

[0048] S1. Obtain a broken wire signal, analyze the broken wire signal to obtain a frequency distribution characteristic of the broken wire signal, determine parameters of a band-pass filter according to the frequency distribution characteristic, and set the band-pass filter according to the parameters;

[0049] Furthermore, the specific contents of S1 are as follows:

[0050] Acquire the detection signal, calculate the short-time energy spectrum of the detection signal, and determine the frequency band distribution range of the broken wire signal and the background noise;

[0051] Set the bandpass filter passband to 10K-12KHz.

[0052] Furthermore, the specific steps of acquiring the detection signal, calculating the short-time energy spectrum of the detection signal, and determining the frequency band distribution range of the broken wire signal and the background noise are as follows:

[0053] S101, setting parameters for short-time energy spectrum calculation, wherein the Hamming window size win is set to 10000, the overlap length noverlap is set to 5000, and the FFT calculation length nfft is set to 10000;

[0054] S102, input a detection signal segment S(t) with a sampling frequency Fs of 50KHz and a sampling time of 10S, calculate the short-time energy spectrum, and obtain the time vector T, frequency vector F and corresponding energy S of the detection signal segment S(t);

[0055] S103, making a corresponding short-time energy spectrum according to the time vector T, frequency vector F and corresponding energy S corresponding to the detection signal segment S(t) calculated in S102;

[0056] S104, determining, according to the short-time energy spectrum, a range where the energy difference between the broken wire signal and the background noise frequency band is the largest as the passband of the bandpass filter.

[0057] Furthermore, the expression for calculating the short-time energy spectrum in S103 is as follows:

[0058] (S,F,T)=Spec(S(t),win,noverlap,nfft,Fs)

[0059] Where Spec represents the short-time energy spectrum calculation of the detection signal segment, S(t). After the short-time energy spectrum transformation, the resolution of the time vector T is 0.1S and the resolution of the frequency vector F is 5Hz.

[0060] S2, input detection signal, and perform bandpass filtering on the detection signal S(t) through the set bandpass filter to obtain the filtered signal S band (t);

[0061] S3, filter signal S band (t) is processed to obtain the target modulus length C(t);

[0062] Furthermore, the specific steps of operating and processing the filtered signal in S3 to obtain the target modulus length are:

[0063] S301, filtering signal S band (t) is processed by arc cosine to obtain the instantaneous phase

[0064] S302, instantaneous phase Directly perform the forward difference diff operation to obtain the instantaneous frequency F1(t) that changes with time. The expression is:

[0065] S303, performing a forward difference diff operation on the instantaneous frequency F1(t) to obtain the instantaneous frequency change rate F2(t), expressed as: F2(t)=diff(F1(t));

[0066] S304, square the instantaneous frequency change rate F2(t) to obtain the modulus length, and obtain the primary modulus length A1(t), which is expressed as: A1(t)=abs(F2(t) 2 );

[0067] S305 , filter A1(t) using a median filter to obtain a target modulus C(t).

[0068] S4. Set the target module length threshold K to determine the moment when the PCCP wire break signal occurs.

[0069] Furthermore, the modulus length threshold K in S4 is set according to the modulus length of the calculated background noise, and is generally set to be more than twice the modulus length of the background noise;

[0070] If the target module length C(t) is greater than the module length threshold, the corresponding moment is the moment when the PCCP wire break signal occurs;

[0071] If the target module length C(t) is less than or equal to the module length threshold, the corresponding time is not the time when the PCCP wire break signal occurs.

[0072] The specific implementation methods of the present invention are as follows:

[0073] A detection signal with a sampling frequency of 50KHz and a sampling time of 10s was selected. The detection signal contains two PCCP broken wire signals, which occurred at around 1.6 seconds and 7.1 seconds, but were submerged due to the large amplitude of the background noise.

[0074] The detection signal S(t) is bandpass filtered to remove most of the background noise and obtain the filtered signal S band (t). The cut-off frequency of the detection signal bandpass filter is set to 10KHz and 12KHz, and the order is 1. The filtered data is as follows Figure 2 As shown, it can be seen that most of the background noise is removed after filtering, but a small amount of noise still exists, interfering with the PCCP broken wire signal recognition;

[0075] For the filtered signal S band (t) Take the inverse cosine to get the instantaneous phase signal of the filtered signal

[0076] For instantaneous phase signal Perform a forward differential diff operation to obtain the instantaneous phase signal The instantaneous frequency F1(t);

[0077] Perform a forward difference diff operation on the instantaneous frequency F1(t) to obtain the rate of change of the instantaneous frequency F2(t);

[0078] Square the instantaneous frequency change rate F2(t), and then take its modulus length to obtain the modulus length A1(t) of the square of the instantaneous frequency change rate, i.e., the primary modulus length;

[0079] Use median filtering to filter the modulus length A1(t) of the square of the instantaneous frequency change rate. The median filter order is set to 10, and the modulus length C(t) of the square of the instantaneous frequency change rate after filtering is obtained, which is the target modulus length. Figure 3 As shown in the figure, the noise in the C(t) signal is suppressed, and the characteristics of the PCCP broken wire signal are enhanced. It can be observed that the PCCP broken wire signal occurs at around 1.6 seconds and 7.1 seconds.

[0080] According to the noise amplitude level in C(t), the PCCP broken wire signal threshold is set to 20, and the moment when the amplitude is higher than the threshold is the moment when the PCCP broken wire signal occurs.

[0081] Excluding data loading and plotting code, this method takes 0.031736 seconds.

[0082] For comparison, a direct arc cosine instantaneous frequency solution method based on absolute value is used to process the same filtered signal S as in the first embodiment. band (t), the method first filters the signal S band (t) Take the inverse cosine to get the instantaneous phase signal of the filtered signal Then the instantaneous phase signal Perform a forward differential diff operation to obtain the instantaneous phase signal The instantaneous frequency F1(t) is as follows. Figure 4 As shown, excluding data loading and plotting code, the method takes 0.026303 seconds.

[0083] contrast Figure 3 and Figure 4 It can be seen that the signal C(t) calculated in the embodiment of the present invention has significantly better characteristics of the PCCP broken wire signal than the signal F1(t) in the comparative example, and has significantly better noise suppression effect. In order to measure the enhancement level of the PCCP broken wire signal characteristics obtained by the two methods, the minimum amplitude value S of the PCCP broken wire signal is used. min The maximum noise amplitude N max The result is expressed in dB, and the expression is as follows:

[0084] τ=S min / N max

[0085] In the above formula, the larger the τ, the more significant the enhancement of the PCCP broken wire signal characteristics. The calculated τ in F1(t) is 8.0825dB, and the τ in C(t) is 30.6226dB. By comparison, it can be found that the PCCP broken wire signal characteristics of C(t) are significantly better than those of F1(t).

[0086] The above comparison shows that, compared with the direct arccosine instantaneous frequency solution method based on absolute value, the PCCP broken wire signal feature enhancement effect of the former is more than 4 times that of the latter under the premise of similar calculation speed, which reflects the advantages of the present invention.

[0087] It should be noted that the calculation process and results of these two methods are calculated using MATLAB R2022b software in the same computer environment. The specific computer environment is as follows:

[0088] CPU: Intel(R)Core(TM)i7-8750HCPU@2.20GHz 2.21GHz

[0089] RAM: 16GB

[0090] ROM: 128GB SSD+1TB HDD

[0091] GPU: NVIDIA GTX1060

[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation, characterized in that: The following steps are involved: S1. Obtain a broken wire signal, analyze the broken wire signal to obtain a frequency distribution characteristic of the broken wire signal, determine parameters of a band-pass filter according to the frequency distribution characteristic, and set the band-pass filter according to the parameters; S2, input detection signal, and perform bandpass filtering on the detection signal S(t) through the set bandpass filter to obtain the filtered signal S band (t); S3, filter signal S band (t) is processed to obtain the target modulus length C(t); S4. Set the target module length threshold K to determine the moment when the PCCP wire break signal occurs.

2. According to claim 1, a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation is characterized in that: The specific contents of S1 are as follows: Acquire the detection signal, calculate the short-time energy spectrum of the detection signal, and determine the frequency band distribution range of the broken wire signal and the background noise; Set the bandpass filter passband to 10K-12KHz.

3. According to claim 1, a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation is characterized in that: The specific steps of obtaining the detection signal, calculating the short-time energy spectrum of the detection signal, and determining the frequency band distribution range of the broken wire signal and the background noise are as follows: S101, setting parameters for short-time energy spectrum calculation; S102, input a detection signal segment S(t) with a sampling frequency Fs of 50KHz and a sampling time of 10S, calculate the short-time energy spectrum, and obtain the time vector T, frequency vector F and corresponding energy S of the detection signal segment S(t); S103, making a corresponding short-time energy spectrum according to the time vector T, frequency vector F and corresponding energy S corresponding to the detection signal segment S(t) calculated in S102; S104, determining, according to the short-time energy spectrum, a range where the energy difference between the broken wire signal and the background noise frequency band is the largest as the passband of the bandpass filter.

4. According to claim 3, a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation is characterized in that: The expression for calculating the short-time energy spectrum in S103 is as follows: (S,F,T)=Spec(S(t),win,noverlap,nfft,Fs); Where Spec represents the short-time energy spectrum calculation of the detection signal segment, S(t). After the short-time energy spectrum transformation, the resolution of the time vector T is 0.1S and the resolution of the frequency vector F is 5Hz.

5. According to claim 1, a PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation is characterized in that: The specific steps of operating and processing the filtered signal in S3 to obtain the target modulus length are: S301 filters the signal S band (t) is processed by arc cosine to obtain the instantaneous phase S302, instantaneous phase Directly perform the forward difference diff operation to obtain the instantaneous frequency F1(t) that changes with time; S303, performing a forward difference diff operation on the instantaneous frequency F1(t) to obtain an instantaneous frequency change rate F2(t); S304, square the instantaneous frequency change rate F2(t) to obtain the modulus length, and obtain the primary modulus length A1(t); S305 , filter A1(t) using a median filter to obtain a target modulus C(t).

6. The PCCP broken wire signal feature enhancement and positioning method based on threshold-instantaneous frequency change rate calculation according to claim 5 is characterized in that: In S4, the modulus length threshold K is set according to the calculated modulus length of the background noise and is set to be more than twice the modulus length of the background noise; If the target module length C(t) is greater than the module length threshold, the corresponding moment is the moment when the PCCP wire break signal occurs; If the target module length C(t) is less than or equal to the module length threshold, the corresponding time is not the time when the PCCP wire break signal occurs.

Images