Distribution cable insulation aging monitoring method based on traveling wave principle

Abstract

The invention discloses a traveling wave principle-based distribution cable insulation aging monitoring method, which is a novel fault point branch circuit judgment method, and is characterized in that after a fault branch circuit is judged by using the method, the time difference of a fault initial traveling wave arriving at an end point and a T node of the branch circuit is further calculated, and the time difference of the fault initial traveling wave arriving at the T node is calculated; and comparing the time difference with the time difference when the fault traveling wave is propagated from each connection point of the branch to the line end measuring point and the T node, determining an overhead line section or a cable line section where the fault is located, and finally realizing accurate calculation of the fault distance by using a double-end traveling wave method. The problem of power distribution cable defect operation and maintenance in the prior art can be fully solved, the service life cycle of the power distribution cable can be prolonged, and the intelligent degree and the operation and maintenance cost of intelligent inspection of the power distribution cable are further reduced. A neural network wavelet and Teager energy operator combined digital signal processing method is adopted to realize accurate positioning of traveling wave arrival time and accurately realize fault distance measurement.

Description

Technical Field

[0001] This application belongs to the field of power distribution cable fault monitoring technology, specifically relating to a method for monitoring the insulation aging of power distribution cables based on the traveling wave principle. Background Technology

[0002] With the acceleration of urbanization and the continuous expansion of urban areas in my country, the demand for power grids is growing rapidly, while the deployment of power distribution lines is gradually increasing. Power distribution cables are buried underground, some in ductwork, making faults difficult to detect and operating in harsh environments. Furthermore, the manufacturing quality of power distribution cables is currently inconsistent, and long lines cannot be fully inspected before commissioning. Especially in new power systems that are susceptible to short-term overload operation, insulation is prone to aging and corrosion, leaving potential safety hazards for power distribution network operation. Current technical methods for condition monitoring of medium and low voltage cables mainly rely on offline high-voltage testing, which is difficult and inefficient, failing to meet the requirements for lean outage planning management of cables within stations and improving the overall level of lean management of medium and low voltage cables. Furthermore, with the rapid growth of urban power grid load and the increasing scarcity of land resources, the dense laying of power transmission and distribution cables, their shared trench laying, and the intrusion of communication optical cables are common occurrences. Fire prevention and external damage protection measures for power distribution cable lines and channels are inadequate, neutral point system upgrades have not been implemented, and the operation of power distribution cable lines with single-phase faults leads to the expansion of accidents and severe risks of cross-sectional loss in cable channels, resulting in frequent major power outages. Once a cable fire or explosion occurs, it will cause significant economic losses to the city and casualties. Because cables are buried underground or in trenches, locating the fault point requires substantial manpower and financial investment.

[0003] Because power distribution networks are complex systems, various situations may occur during their operation, such as short circuits and line breaks. When these situations occur, power outages for maintenance or equipment replacement are necessary. If the fault point can be quickly located using technology, the downtime can be significantly shortened and power supply reliability improved. Therefore, research on fault location methods for power distribution network cables has significant theoretical and practical importance. Due to the use of a three-core unified design in power distribution cables, accurate phase line fault monitoring is difficult. Currently, conventional methods include impedance analysis, phasor analysis, traveling wave analysis, and partial discharge analysis. However, due to the numerous branch links, complex structure, and severe signal attenuation in power distribution networks, accurate fault location is challenging. Among current methods, the traveling wave fault location method is relatively accurate, but due to waveform dispersion and attenuation, this invention proposes a fault detection method based on time-domain and frequency-domain fusion characteristics to accurately calibrate the wavefront arrival time and determine the traveling wave velocity, thereby improving the reliability and accuracy of traveling wave fault location. Summary of the Invention

[0004] The purpose of this application is to overcome the problem of inaccurate fault location in power distribution cables.

[0005] To achieve the above objectives, this application proposes a method for monitoring the insulation aging of distribution cables based on the traveling wave principle. This method effectively solves the problems of defective operation and maintenance in existing distribution cables, while also improving the service life of distribution cables and further reducing the intelligence level and maintenance costs of intelligent inspection of distribution cables. This invention solves the problem of complex branch link structures in distribution networks through a novel ranging algorithm, enabling precise single-line positioning even in multi-layer cable lines and branch links. Furthermore, while the traditional single-end method uses fault traveling waves and reflected waves from the fault point for ranging, it requires identification of reflected waves from the fault point. Although the double-end method uses the line length and the arrival times of the initial traveling waves at both ends of the line for ranging, eliminating the need for reflected wave identification, the ranging effect is unsatisfactory due to the asynchronous clocks at both ends.

[0006] Compared with the prior art, the advantages of the present invention are:

[0007] Based on the analysis of the internal and external modulus of the cable, this invention uses a digital signal processing method combining neural network wavelet and Teager energy operator to achieve accurate positioning of the arrival time of traveling waves, and proposes a ranging method based on the difference in the internal and external modulus of the cable, which can accurately realize fault location. Attached Figure Description

[0008] Figure 1 This is a flowchart of a method for monitoring the aging of power distribution cable insulation based on the traveling wave principle.

[0009] Figure 2 This is a fault traveling wave path diagram for a three-terminal hybrid power distribution line.

[0010] Figure 3 Flowchart for multimodal differential fault feature extraction;

[0011] Figure 4 This is a schematic diagram of the multi-hole algorithm.

[0012] Figure 5 This is a flowchart for multimodal differential fault ranging. Detailed Implementation

[0013] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0014] The specific design steps and process are as follows: Figure 1 As shown in the description below:

[0015] (1) Fault point branch route determination method:

[0016] The structure of power distribution cable networks is complex. Firstly, the boundary between overhead and cable lines is significant, especially in urban areas where special geographical locations such as highways or river crossings necessitate converting cables to overhead installations to address route constraints. Secondly, based on the distribution area structure and load division, the main distribution lines are divided into different branches to supply power to different load areas. These two factors contribute to the complexity of the power distribution path structure. This situation places high demands on the accurate location of line faults. Traveling wave monitoring devices installed at both ends of the main line need to accurately measure the location of the faulty line, requiring precise location methods. Conventional double-ended traveling wave measurement methods are only suitable for relatively simple linear links. For complex links, new judgment criteria need to be added to the original double-ended measurement method. This invention proposes a hybrid multi-segment link fault location strategy that can determine the characteristics of overhead and cable branch lines. The specific architecture is as follows... Figure 2 As shown, Figure 2 This is a typical three-terminal, three-section overhead-cable hybrid transmission line. B1-B2 is the cable section, A1-B1, B1-B2, B2-T, B2-A2, and T-A3 are the overhead sections, and F is the fault point, labeled l1, l2, l3, l4, l5, l6, l7, l8, l9, l10, l11, l12, l13, l10, l11, l12, l13, l14, l15, l16, l17, l18, l18, l19, l10, l11, l12, l13 ...0, l11, l12, l13, l14, l15, l16, l17, l18, l18, l19, l10, l11, l12, 4， l5 represents the section length. Assume the traveling wave velocity is V1 in the overhead line and V in the cable. 2， The time it takes for the fault traveling wave to travel through each segment is:

[0017]

[0018] Based on the relationship between the time difference of the initial traveling wave arriving at each measuring point and the time difference of the traveling wave propagating from node T to each measuring point, the fault branch judgment conditions of a three-terminal hybrid line can be summarized as follows:

[0019] 1) When a fault occurs in branch A1-T, the time difference between the arrival time of the fault traveling wave at end A1 and the arrival time at end A2 is expressed as: It can be represented as:

[0020]

[0021] Where t FT Let F be the time taken for the initial traveling wave to travel from the fault point to the branch node. Since point F is on the A1-T branch, the following characteristic relationship exists:

[0022]

[0023] 2) When the fault points occur in branches A2-T and A3-T respectively, the characteristic relationships of each time difference can be analyzed:

[0024]

[0025] 3) When the fault occurs at node T, the time differences have the following characteristic relationship:

[0026]

[0027] (2) Calling the multimodal difference-based fault location method: After determining the fault branch based on the method in step (1), the next step is to accurately locate the fault point. Generally, the time difference between the zero-mode wave velocity and the linear mode wave velocity of the traveling wave is used to calculate the fault distance and measure the initial arrival time of the traveling wave. However, the large dispersion problem of the traveling wave is ignored, which will affect the speed of the traveling wave. The traveling wave does not use a fixed traveling wave velocity when there is a defect or fault, and there will be a slight difference. This difference will affect the distance measurement. Therefore, this invention proposes a fault location method based on multimodal traveling wave features. The feature extraction method process is as follows: Figure 3 As shown:

[0028] Fault traveling waves are generally represented by the characteristic impedance Z. c The propagation constant γ represents the characteristics of the traveling wave. The wave velocity of traveling waves of all moduli increases with increasing frequency, and remains stable at high frequencies. Comparing the velocities of traveling waves of different moduli, the wave velocities of moduli 4-6 are the same at any frequency, and their wave velocity curves with frequency overlap. However, moduli 1-3 show significant differences, and their wave velocities are all lower than those of moduli 4-6. As the frequency increases, the attenuation coefficients of traveling waves of moduli 1-6 also continuously increase. Comparing the attenuation coefficients of traveling waves of different moduli, the attenuation rates of moduli 4-6 are the slowest and the same. Compared to the three moduli mentioned above, the attenuation rates of moduli 2 and 3 are slightly faster, but the difference is not significant. The attenuation rate of moduli 1 is the largest, and it increases rapidly with increasing frequency, reaching hundreds of times the attenuation coefficients of the other moduli at high frequencies. Therefore, moduli 2-6 with smaller attenuation can be used for ranging. This invention selects the most typical moduli 3 and 6 to achieve accurate ranging.

[0029] Traveling wave ranging is calculated based on the arrival time of the traveling wave at the measuring point, so the identification and calibration of the arrival time are crucial. Traditional identification methods include Fourier transform and fast Fourier transform, but they are only sensitive to time parameters and the influence of the frequency domain is uncontrollable. Therefore, this invention chooses wavelet transform as the basic principle and time-frequency characteristic structure, which can detect the intensity of signal singularity and the distribution and specific location of distortion points.

[0030] This invention employs a fast algorithm using binary wavelets and selects a multi-aperture algorithm for implementation. The principle is as follows: Figure 4 Convolving the original signal f(t) with a high-pass filter yields a signal with a scale of 2. l The high-frequency signal G is convolved with a low-pass filter to obtain a signal with a scale of 2. lThe low-frequency signal H is then used as input, and the filter is further convolved with the current layer's filter by zero-placing at intervals. The decomposition algorithm is expressed as follows:

[0031]

[0032] S 2j f(n) is the wave approximation of the signal, W 2j f(n) represents the wavelet transform of the signal. The fault signal is decomposed into different scales, with the power frequency component corresponding to a larger scale and the high-frequency component corresponding to a smaller scale. Therefore, the quantization error and high-frequency noise are relatively large at the first scale, so the result at the second scale is chosen as the basis for ranging.

[0033] To highlight the abrupt amplitude changes of non-stationary signals, the Teager energy operator is introduced. The energy operator for a continuous signal x(t) = Acos(wt + θ) is expressed as:

[0034] ψ (x(t)) =[x(t)] 2 -x(t)x(t)* (8)

[0035] Substituting x(t) = Acos(wt + θ) into equation (8), we can derive:

[0036] ψ (x(t)) =ψ[Acos(wt+θ)]=A 2 w 2 (9)

[0037] The Teager energy operator can assess the energy of an instantaneous signal from both its amplitude and frequency. Based on this characteristic, the traveling wave signal is first decomposed to separate individual components, and then the Teager energy operator is applied to each component. The time corresponding to the first peak of the processed signal is the time when the initial traveling wave surge of the fault reaches the detection point.

[0038] (3) Cable ranging algorithm based on modulus difference: Signals with moduli 1 to 3 propagate outside the cable's metallic sheath and are called external moduli, while signals with moduli 4 to 6 propagate inside the cable's metallic sheath and are called internal moduli. Since the parameters of the internal and external moduli are different, their transmission speeds are also different. The parameters of the external modulus are much larger than those of the internal modulus, therefore the wave speed of the internal modulus is greater than that of the external modulus. Among the external moduli, modulus 1 has a relatively large attenuation and is not suitable as a ranging modulus. However, moduli 2 and 3 have approximately the same speed, and moduli 4 to 6 also have approximately the same wave speed. Therefore, moduli 3 and 6 are selected for ranging. These two moduli have a good correspondence and a speed difference.

[0039] First, measure the fault on the main line. Assume the line has only one main line with no branches, and its fault component network is a straight line. Assume t0 is the time when the fault occurs, and t... M3 t is the time when the traveling wave modulus 3 reaches the measurement point M. N3 t represents the time when the traveling wave modulus 3 reaches the measurement point N. M6 t is the time when the traveling wave modulus 6 reaches the measurement point M. N6 The time when the traveling wave modulus 6 reaches the measurement point N. V3 is the wave velocity of modulus 3, V6 is the wave velocity of modulus 6, and d M d is the distance between the fault point and the measuring point M. N Let N be the distance between the fault point and the measuring point N, and L be the total length of the line. Then:

[0040]

[0041] If the total length of the line is known to be L, then the double-ended traveling wave ranging formula without wave velocity can be calculated according to formula (10).

[0042] Next, measure the location of the branch link fault. Assume a fault occurs on a branch DA in the main line MN, and the distance from the branch point D is x1. Similarly, the time it takes for the traveling wave modulus 6 to arrive at points M and N on both sides will be earlier than that of the traveling wave modulus 3. Based on the time difference in arrival at point MN, perform modulus differential velocity distance measurement on the line at point MN. From the principle of traveling wave modulus velocity difference, we can obtain:

[0043]

[0044] Assume t0 is the time when the fault occurs, t M3 t is the time when the traveling wave modulus 3 reaches the measurement point M. N3 t represents the time when the traveling wave modulus 3 reaches the measurement point N. M6 t is the time when the traveling wave modulus 6 reaches the measurement point M. N6 The time when the traveling wave modulus 6 reaches the measurement point N. V3 is the wave velocity of modulus 3, V6 is the wave velocity of modulus 3, and d M d is the distance between the fault point and the measuring point M. N This represents the distance between the fault point and the measuring point N.

[0045] Since the fault occurs on a branch line, the fault distance and the total line length are related as follows:

[0046] d M +d N =L+2x1 (12)

[0047] The distance from the fault point to the branch point can then be calculated:

[0048]

[0049] At the same time, the distance between the faulty branch point and the measuring point M can be determined as follows:

[0050] D1 = d M -x1 (14)

[0051] The calculated D1 should be the same as l1. If the fault occurs on the main line of the distributed power grid, the two-end ranging method is valid, but in this case, it is necessary to determine whether the fault point is on the main line. M +d N If the equation equals L, the fault occurs on the main line; if the equation does not hold, the fault occurs on a branch line. Figure 5 The diagram shown is a flowchart of multimodal differential fault location.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for monitoring the aging of power distribution cable insulation based on the traveling wave principle, characterized in that, A novel method for determining faulty branch lines is proposed. After identifying the faulty branch, the method further calculates the time difference between the initial traveling wave of the fault reaching the end point of the branch and node T, and compares it with the time difference between the traveling wave of the fault propagating from each connection point of the branch to the end measurement point and node T. This determines whether the overhead line segment or cable line segment where the fault is located. Finally, the double-ended traveling wave method is used to accurately calculate the fault distance.

2. The method for monitoring the aging of power distribution cable insulation based on the traveling wave principle according to claim 1, characterized in that, A digital signal processing method combining neural network wavelet and Teager energy operator is proposed. First, the traveling wave signal is decomposed into binary wavelet. To avoid noise interference and quantization error, the second frequency band is used as the ranging frequency band. Then, the Teager energy operator is used to process it, which enhances the abrupt change characteristics and accurately locates the arrival time of the traveling wave.

3. The method for monitoring the aging of power distribution cable insulation based on the traveling wave principle according to claim 1, characterized in that, A dual-end traveling wave ranging algorithm based on modulus time difference is proposed. It is not affected by line parameters and time synchronization, does not require wave velocity calculation, and only needs to know the accurate line length to accurately locate the distance between the fault point and the measurement point, and there is no ranging dead zone. For distribution network cable lines with branches, a single-end ranging algorithm is adopted to accurately identify the branch where the fault occurs and give the fault distance.

Images