A cable fault detection method based on limited information degrees of freedom

Abstract

The present invention provides a cable fault detection method based on limited information degrees of freedom. First, a detection pulse signal is input into a cable to be measured, and the corresponding echo signal is received; then, a sampling kernel is used to perform modulation processing on the echo signal to obtain a modulation signal; next, the limited information degrees of freedom of the echo signal are calculated, and the sampling interval is calculated through the limited information degrees of freedom. Subsequently, the modulation signal is sampled at equal intervals to obtain discrete sparse data; by performing parameter estimation on the discrete sparse data, the amplitude parameter and delay parameter of the cable reflection signal are obtained; finally, according to the amplitude and delay parameters, the position of the cable fault point is calculated, and the fault type is judged. The cable fault detection method of the present invention reconstructs the original signal with a relatively low sampling rate and extremely few sampling data, breaking through the technical bottleneck that the prior art can only improve the fault location accuracy by increasing the sampling rate.

Description

Technical Field

[0001] The present invention belongs to the technical field of cable fault detection, and particularly relates to a cable fault detection method based on limited information degrees of freedom. Background Art

[0002] Cables are indispensable energy or information transmission channels on complex equipment such as airplanes and ships. They are distributed in various parts of the equipment, with a large variety and quantity, and play a crucial role in the normal operation of the equipment. Once a cable fails, it is very likely to cause the interruption of signal or power transmission, and even the collapse of the operating system, resulting in serious accidents and huge losses of life and property. Therefore, timely and accurate detection of cable faults, as well as the location and identification of faults, are of great significance for the timely discovery of cable faults and the troubleshooting and repair of faults, and are the key to ensuring the safety and normal operation of the equipment.

[0003] In the aspect of cable fault detection technology, in the early days, people generally used the bridge method to detect cable faults, and later the traveling wave method was gradually applied. According to different injected signals, the traveling wave method is divided into Time Domain Reflection (TDR), Frequency Domain Reflection (FDR), and Spread Spectrum Time Domain Reflection (SSTDR), etc. TDR is the most widely used and classical method. Its basic principle is that when a pulse signal encounters an impedance mismatch point during cable transmission, reflection will occur. According to the waveform characteristics and time difference between the incident wave and the reflected wave, the fault type and location of the cable can be judged. FDR is developed on the basis of TDR. Different from the pulse signal used in TDR detection, its incident signal is a swept-frequency signal. When calculating the position of the fault point, the measured data of the reflected signal needs to be converted into time-domain information through inverse Fourier transform, which is more cumbersome compared with TDR. The detection signal of SSTDR is a pseudo-random code modulated by cosine, and relevant algorithms need to be combined to locate cable faults. The high-speed pseudo-random code generator required by this method has a high cost, and the measurement curve will be affected by the inherent periodicity of the pseudo-random code and large periodic sidelobes will appear, interfering with fault location.

[0004] In summary, the TDR method is the most mature and commonly used method at present, with a simple principle and convenient operation. The key to the detection and positioning accuracy of TDR lies in the accurate measurement of the time interval between the incident wave and the reflected wave, and the accuracy of this time interval measurement depends on the sampling rate. The higher the sampling rate, the higher the time positioning accuracy. In order to obtain higher positioning accuracy, the sampling rate has been increased to the GHz level, resulting in an increase in hardware cost and restricting the popularization and application of cable fault detection technology. Summary of the Invention

[0005] In view of the deficiencies in the prior art, the present invention provides a cable fault detection method based on limited information degrees of freedom to solve the problem of the extremely high sampling rate required in the traditional TDR method.

[0006] The present invention achieves the above technical objectives through the following technical means.

[0007] A cable fault detection method based on limited information degrees of freedom includes the following steps:

[0008] Step 1: Input the detection pulse signal x(t) into the cable to be tested and receive the corresponding echo signal y(t);

[0009] Step 2: Use a sampling kernel to modulate the echo signal y(t) to obtain a modulated signal e(t);

[0010] Step 3: Calculate the limited information degrees of freedom of the echo signal y(t), calculate the sampling interval T through the limited information degrees of freedom, and then perform equally spaced sampling on the modulated signal e(t) to obtain discrete sparse data where N is the number of sampling points;

[0011] Step 4: Perform parameter estimation on the discrete sparse data to obtain the amplitude parameter and the time delay parameter of the cable reflection signal, where L is the number of echo signal pulses;

[0012] Step 5: Calculate the position of the cable fault point according to the amplitude and time delay parameters and judge the fault type.

[0013] Further, in step 1, the generated pulse signal x(t) is first subjected to power amplification processing and then input into the cable to be tested.

[0014] Further, the detection pulse signal x(t) is a voltage pulse signal with a pulse width in the nanosecond range.

[0015] Further, in step 2, the sampling kernel adopts a SoS function structure.

[0016] Further, the sampling kernel is designed by the transfer function approximation method, and the sampling kernel is:

[0017]

[0018] where τ is the sampling interval duration, is a continuous integer set determined by the information degrees of freedom of y(t);

[0019] Subsequently, the echo signal y(t) is input into the sampling kernel to obtain the modulation signal

[0020] Further, in step 3: within the sampling interval duration τ, if the number of detected echo signal pulses is L, then the information degree of freedom is 2L, and the sparse sampling interval

[0021] Further, the number of sampling points N≥2L + 1.

[0022] Further, in step 4, the nulling filter method in the spectral estimation method is used to perform parameter estimation.

[0023] Further, for the echo signal y(t):

[0024]

[0025] where β l is the amplitude, b l is the time delay, α is the Gaussian pulse bandwidth factor, and Z is the set of integers;

[0026] Performing a transformation on the echo signal y(t) gives:

[0027]

[0028] where Y[k] is the Fourier series coefficient of y(t), is the Fourier transform of the basis function of y(t);

[0029] Construct a nulling filter with filter coefficients The Z-transform of the coefficients is:

[0030]

[0031] Let the zeros of Ψ(z) be Let Ψ0 = 1, then Ψ(z) can be factorized as:

[0032]

[0033] Convolve the nulling filter coefficients Ψ k with C[k], and we get:

[0034]

[0035] By performing a discrete Fourier transform on to obtain N Fourier series coefficients, and thus solving for thereby obtaining the time delay parameter of the echo signal y(t)

[0036] Through the following Vandermonde matrix:

[0037]

[0038] The amplitude parameters are solved

[0039] Furthermore, the specific content of step 5 is as follows:

[0040] Through the time delay parameter Calculate the fault point location:

[0041]

[0042] where l is the length of the fault point from the detection end, b i is the time delay of the peak of the incident wave of the cable, b r is the time delay of the peak of the first reflection wave of the cable, V d is the wave velocity of the pulse signal in the cable;

[0043] Based on the incident wave amplitude β i and the first reflection wave amplitude β r Judge the fault type, where when β i β r > 0, it is an open circuit fault, and when β i β r < 0, it is a short circuit fault.

[0044] The beneficial effects of the present invention are:

[0045] The present invention provides a cable fault detection method based on limited information degrees of freedom. This method is based on the sparse sampling principle of limited information rate, specifically a method of sampling according to the number of useful information parameters in the signal, that is, the information degree of freedom. This method has completely deviated from the conventional Nyquist sampling theory framework based on non-loss of frequency information and has been applied and promoted in the radar field. Correspondingly, the cable fault detection method based on limited information degrees of freedom of the present invention can reconstruct the original signal with a lower sampling rate and extremely few sampling data. Therefore, compared with the existing conventional cable fault detection technologies, it breaks through the technical bottleneck of only being able to improve the fault location accuracy by increasing the sampling rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is the flow chart of the cable fault detection method of the present invention;

[0047] Figure 2 is the waveform diagram of the pulse signal in the test of the present invention;

[0048] Figure 3 is the waveform diagram of the fault reflection signal in the test of the present invention;

[0049] Figure 4 This is the waveform diagram of the modulation signal in the test of the present invention;

[0050] Figure 5 This is the amplitude and time delay parameter estimation diagram of the fault reflection signal in the test of the present invention. Detailed implementation manners

[0051] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as a limitation to the present invention.

[0052] I. Technical solution

[0053] As Figure 1 shown, the cable fault detection method based on limited information degrees of freedom includes the following steps:

[0054] S1. Generate a detection pulse signal x(t).

[0055] Specifically, a detection signal excitation module is used to generate the detection pulse signal x(t). This pulse signal is a voltage pulse signal, and the pulse width of the signal x(t) is in the nanosecond level. The narrower the pulse width, the higher the fault location accuracy.

[0056] S2. Amplify the power of the detection pulse signal x(t). Through power amplification, the fault detection range of a long-distance cable is improved. Then, the amplified signal x(t) is input into the cable to be measured, and thus the echo signal y(t) reflected from the cable to be measured is obtained accordingly.

[0057] S3. Use a "sampling kernel" to modulate the echo signal y(t) to obtain a modulation signal e(t).

[0058] In this step, modulation is performed through the sampling kernel, so as to perform frequency screening on y(t), so as to obtain the Fourier coefficients required for reconstructing y(t) from sparse sampling data. The sampling kernel adopts the SoS function structure, that is, its transfer function s(t) is the SoS function. The frequency selection characteristic of the sampling kernel is obtained from the amplitude-frequency characteristic of the transfer function. The better the frequency selection characteristic, the higher the accuracy of reconstructing the original signal from sparse sampling data. In actual use, the sampling kernel can be designed by the transfer function approximation method, and then y(t) is input into the sampling kernel to obtain the signal modulated by the sampling kernel.

[0059] S4. Calculate the finite information degrees of freedom of the echo signal y(t), and calculate the sampling interval T through the finite information degrees of freedom; perform equally-spaced sampling on the modulation signal e(t) to obtain discrete sparse data where N is the number of sampling points;

[0060] Specifically in this step: Let τ be the time length of the sampling interval. The number of echo signal pulses detected within the time length τ is L. Since the number of pulses in the cable fault reflection signal is a finite value, and in addition, each reflected echo signal can be characterized by amplitude and time delay, the information degrees of freedom for detecting the reflected echo signal are finite values. At this time, the calculated sparse sampling interval is Perform sparse sampling on the signal at equal intervals to obtain discrete sparse data The smaller the sampling interval, the higher the sampling frequency. The number of sampling points N should satisfy N ≥ 2L + 1.

[0061] S5. Perform parameter estimation on the collected discrete sparse data to obtain the amplitude parameter of the cable reflection signal and the time delay parameter where L is the number of echo signal pulses detected;

[0062] Specifically, the nulling filter method in the spectral estimation method can be used to perform parameter estimation from to obtain the amplitude and time delay parameters of the echo signal y(t)

[0063] S6. Calculate the position of the cable fault point according to the estimated parameters obtained in the previous step and judge the fault type, so as to realize the detection and location functions of the cable fault.

[0064] Specifically:

[0065] ① For the location of the fault point, it can be calculated through the time delay parameter The corresponding calculation formula is:

[0066]

[0067] where l is the length of the fault point from the detection end, b i is the time delay of the peak of the cable incident wave, b r is the time delay of the peak of the first cable reflection wave, and V d is the wave velocity of the pulse signal in the cable.

[0068] ② For the judgment of the fault type, it is determined by the amplitude β i of the incident wave and the amplitude β r of the first reflection wave. Among them, when β i β r > 0, it is an open circuit fault; when βi β r When it is less than 0, it is a short - circuit fault.

[0069] II. Testing

[0070] S1. Use a signal generator to generate a detection pulse signal x(t). In this embodiment, the pulse width of the pulse signal x(t) is 50 ns, the amplitude is 1 V, and the waveform of the pulse signal x(t) is as Figure 2 shown.

[0071] S2. Amplify the generated pulse signal x(t) through a power amplifier. In this embodiment, the voltage amplitude is increased from 1 V to 5 V. Then connect the amplified pulse signal to one end of the cable under test (this end is the detection end). In this embodiment, an open - circuit fault is set at a distance of 93.8 m from the detection end. Due to the presence of the fault point, a fault reflection signal (echo signal) y(t) will be received at the detection end, and its waveform diagram is as Figure 3 shown.

[0072] S3. Modulate y(t) using a sampling kernel. In this embodiment, the selected sampling kernel is:

[0073]

[0074] where τ is the time length of the sampling interval. In this embodiment, τ = 2 μs, is a set of continuous integers determined by the information degree of freedom of y(t). In this embodiment, is the set of integers {-10, -9,..., 10}, and j is the imaginary unit;

[0075] Input y(t) into the above - mentioned sampling kernel to obtain the signal e(t) modulated by the sampling kernel, and its waveform is as Figure 4 shown.

[0076] S4. Detect the number of echo signal pulses within the time duration τ as L. In this embodiment, the number of pulses L = 2, and the information degree of freedom of the echo signal is 2L = 4. Thus, calculate the sparse sampling interval Finally, in this embodiment, the sampling interval is taken as 0.1 μs, and the number of sampling points N = 21, obtaining discrete sparse data

[0077] S5. Use the annihilating filter method in the spectral estimation method to perform parameter estimation on the collected discrete sparse data In this embodiment, the echo signal y(t) can be expressed as:

[0078]

[0079] where β l is the amplitude, bl where τ is the time delay, α is the Gaussian pulse bandwidth factor, and Z is the set of integers.

[0080] Expanding y(t) in the form of a Fourier series gives:

[0081]

[0082] where Y[k] are the Fourier series coefficients of y(t).

[0083] According to the Poisson summation formula, y(t) can be expressed as:

[0084]

[0085] By comparing the Fourier series form and the Poisson summation form of y(t), we can obtain:

[0086]

[0087] where, is the Fourier transform of the basis function of y(t), and at we have: we have:

[0088]

[0089] Construct a nulling filter with coefficients Its Z-transform is:

[0090]

[0091] Let the zeros of Ψ(z) be i.e.:

[0092]

[0093] Let Ψ0 = 1, then Ψ(z) can be factorized as:

[0094]

[0095] When all the time delay parameters are not equal to each other, the zeros of the nulling filter can uniquely represent the peak time delay parameters of the detected reflected signal y(t). Convolving the nulling filter coefficients Ψ k with C[k] gives:

[0096]

[0097] Expanding gives:

[0098] Ψ1C[k - 1]+Ψ2C[k - 2]+…+Ψ L C[k - L] = -C[k]

[0099] Written in matrix form as:

[0100]

[0101] The equation requires at least 2L consecutive Fourier series coefficients for the system of equations to have a unique solution. In this embodiment, discrete sparse data has been obtained For Performing a discrete Fourier transform can obtain 21 Fourier series coefficients, so can be solved to obtain Thereby obtaining the time delay parameter of the echo signal y(t)

[0102]

[0103] Solving to obtain the amplitude parameter Detecting the amplitude and time delay parameters of the reflected signal y(t) The estimation results are as Figure 5 shown.

[0104] S6. According to the time delay parameter estimated by the nulling filter method, perform fault location. In this embodiment, the estimated time delay of the peak of the incident wave of the cable b i = 752.0 ns, the time delay of the peak of the first reflection wave b r = 1696.1 ns. By Calculating the position at the fault point. In this embodiment, V d = 197.8 m / us, the calculated position at the fault point l = 93.37 m, the actual position of the fault point is 93.8 m, and the fault location error is 0.43 m.

[0105] According to the estimated amplitude of the incident wave β i and the amplitude of the first reflection wave β r Perform a preliminary judgment on the fault type. In this embodiment, the estimated amplitude of the incident wave of the cable β i = 5.0 V, the amplitude of the first reflection wave β r = 3.2 V. Therefore, there is β i β r > 0, and it can be judged that the fault point type is an open circuit fault, and the fault type recognition result is accurate.

[0106] In this embodiment, the sampling rate is only 10 MHz. A comparative test is carried out with the conventional TDR detection method. In the conventional method, the sampling rate is 200 MHz. The sampling rate in the present invention is significantly decreased, and the corresponding cable fault detection comparison results are shown in Table 1 below.

[0107] Table 1: Comparison of cable fault detection results

[0108]

[0109] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention.

[0110] The present invention is not limited to the above embodiments, and any obvious improvements, substitutions or deformations that those skilled in the art can make without departing from the essence of the present invention fall within the protection scope of the present invention.

Claims

1. A cable fault detection method based on limited information degrees of freedom, characterized in that, Including the following steps: Step 1: Input the detection pulse signal x(t) into the cable under test and receive the corresponding echo signal y(t); Step 2: Use the sampling kernel to modulate the echo signal y(t) to obtain the modulation signal e(t); Step 3, calculate the finite information degrees of freedom of the echo signal y(t), calculate the sampling interval T through the finite information degrees of freedom, and then perform equally spaced sampling on the modulation signal e(t) to obtain discrete sparse data where N is the number of sampling points; Step 4, for the discrete sparse data perform parameter estimation to obtain the amplitude parameter of the cable reflection signal and the time delay parameter where L is the number of echo signal pulses; the process of parameter estimation is as follows: For the echo signal y(t): where β l is the amplitude, b l is the time delay, α is the Gaussian pulse bandwidth factor, and Z is the set of integers; The transformation of the echo signal y(t) is as follows: where \(Y[k]\) are the Fourier series coefficients of \(y(t)\), which is the Fourier transform of the basis function of \(y(t)\) ; Construct a nulling filter with filter coefficients as The Z-transform of the coefficients is: Let the zeros of Ψ(z) be Let Ψ0 = 1, then Ψ(z) can be factorized as: Convolve the nulling filter coefficient Ψ k with C[k], resulting in: By performing a discrete Fourier transform to obtain N Fourier series coefficients, and thereby solving for the time delay parameter of the echo signal y(t) can be obtained After that, through the following Vandermonde matrix: Solve for the amplitude parameter Step 5, based on the amplitude and time delay parameters Calculate the location of the cable fault point and determine the fault type.

2. The cable fault detection method based on limited information degrees of freedom according to claim 1, wherein: In Step 1, the generated pulse signal x(t) is first subjected to power amplification processing and then input into the cable under test.

3. The cable fault detection method based on limited information freedom according to claim 2, characterized in that: The detection pulse signal x(t) is a voltage pulse signal with a pulse width at the nanosecond level.

4. The cable fault detection method based on limited information freedom according to claim 1, characterized in that: In Step 2, the sampling kernel adopts a SoS function structure.

5. The cable fault detection method based on limited information freedom according to claim 4, characterized in that: The sampling kernel is designed by the transfer function approximation method, and the sampling kernel is: where τ is the duration of the sampling interval, is a set of consecutive integers determined by the information freedom degree of y(t), and j is the imaginary unit; The echo signal y(t) is then input into the sampling kernel to obtain the modulation signal 6. The cable fault detection method based on limited information degrees of freedom according to claim 1, wherein: In step 3: for the number of echo signal pulses detected within the sampling interval duration τ being L, the information degrees of freedom are 2L, and the sparse sampling interval 7. The cable fault detection method based on limited information degrees of freedom according to claim 6, characterized in that: The number of sampling points N≥2L + 1.

8. The cable fault detection method based on limited information degrees of freedom according to claim 1, characterized in that: In the said step 4, the nulling filter method in the spectral estimation method is used to perform parameter estimation.

9. The cable fault detection method based on limited information degrees of freedom according to claim 1, characterized in that: Step 5 is specifically: By means of the time delay parameter Calculate the location of the fault point: where l is the length of the fault point from the detection end, b i is the time delay of the peak of the incident wave of the cable, b r is the time delay of the peak of the first reflection wave of the cable, V d is the wave velocity of the pulse signal in the cable; By the amplitude β of the incident wave i and the amplitude β of the primary reflected wave r to determine the fault type, where when β i β r > 0, it is an open circuit fault, and when β i β r < 0, it is a short circuit fault.

Images