High altitude area cable fault early warning and fault positioning method

Abstract

The invention discloses a high altitude area cable fault early warning and fault positioning method comprising the following steps: monitoring terminals are installed at two ends of a cable line, respectively collect traveling wave signals generated by a cable having a partial discharge source / fault point, carry out on-site data processing on the traveling wave signals, and carry out feature extraction by applying wavelet analysis; and the partial discharge pulse signals are uploaded to a background master station through a wireless or optical fiber communication channel, the background master station draws a partial discharge map according to long-period statistical information such as the amplitude, the number of times and the corresponding power frequency phase of the partial discharge pulse signals, and judgment and fault early warning are carried out on the partial discharge type and the development trend of the cable according to the change condition of the map. According to the invention, the wavelet analysis technology is used to extract the main characteristic quantity of the partial discharge signal, the main characteristic quantity is sent to the background master station through a wireless or optical fiber communication channel, and the arrival time of the wave head of the voltage traveling wave is determined and the fault distance of the power grid is calculated through the wave head identification method of the voltage traveling wave, thereby realizing the fault location of the voltage traveling wave of the power grid. Therefore, the position of the fault point is accurately calculated.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault location, and in particular to a method for early warning and locating cable faults in high altitude areas. Background Art

[0002] The cable ratio of distribution network lines in my country is constantly increasing. During operation, factors such as stress, heat, and improper operation during construction and installation will cause local defects (such as water trees, electrical trees, etc.) in the cable body, intermediate joints and other accessories, leading to discharge. Partial discharge of cables increases dielectric loss and reduces insulation level, while traditional dielectric loss angle and leakage current detection often find it difficult to detect local defects in cables. As partial discharge continues, it will eventually cause insulation breakdown and lead to permanent failure of the cable line. For high-altitude areas, the dielectric constant is relatively low, and the cable insulation problem is more prominent. The resulting operation and maintenance and power outage problems seriously affect the power supply reliability of the distribution network.

[0003] At present, the domestic cable insulation status detection generally adopts the regular inspection method, such as the widely used oscillation wave detection method. However, the regular inspection requires power outage, and the cable status will also change during the regular inspection cycle. In addition, with the increase of cable lines, the regular inspection method also increases the workload of the maintenance unit and has a certain impact on production scheduling. From the perspective of technological development trends, with the increase in the number and length of cable lines, traditional line inspection or offline positioning methods are no longer applicable. Summary of the invention

[0004] The purpose of the present invention is to provide a cable fault early warning and fault location method in high altitude areas to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for early warning and locating cable faults in high altitude areas, comprising:

[0007] The monitoring terminals are installed at both ends of the cable line to collect the traveling wave signals generated by the partial discharge source / fault point of the cable, process the traveling wave signals on-site, extract features using wavelet analysis, and send them to the background master station through wireless or optical fiber communication channels. The background master station draws a partial discharge map based on the long-period statistical information such as the amplitude, number, and corresponding power frequency phase of the partial discharge pulse signal, and judges the type and development trend of the cable partial discharge and issues fault warnings based on changes in the map.

[0008] Furthermore, the traveling wave signal is processed on-site and feature extraction is performed using wavelet analysis, which specifically includes denoising the voltage traveling wave signal using a wavelet analysis algorithm to obtain a smooth voltage traveling wave signal, identifying the voltage traveling wave head, and calculating the cable fault distance based on the two arrival times of the voltage traveling wave head to achieve fault warning.

[0009] Furthermore, the voltage traveling wave signal denoising includes:

[0010] Wavelet decomposition of voltage traveling wave signals, including wavelet coefficients;

[0011] Determining a wavelet coefficient threshold using the wavelet coefficients, including a wavelet coefficient set;

[0012] The denoised voltage traveling wave signal is reconstructed using the wavelet coefficient set reconstruction algorithm.

[0013] Furthermore, the expression of the voltage traveling wave signal decomposed by wavelet is as follows:

[0014]

[0015] Among them, x x (a) represents the initial approximation coefficient; E(a) is the voltage traveling wave signal; x y (a) is the voltage traveling wave signal at scale 2 y The approximation coefficient under the decomposition filter coefficient h h , g h ;z y (a) is the voltage traveling wave signal at scale 2 y The wavelet coefficients below.

[0016] Furthermore, the expression of the wavelet coefficient threshold is as follows:

[0017]

[0018] The wavelet coefficients are processed with hard threshold, and the expression is as follows:

[0019]

[0020] in, is the wavelet coefficient threshold; M() is the median removal function; κ is a constant; γ is the length of the yth layer wavelet coefficient after the voltage traveling wave signal is decomposed; ι(z y (a)) is the set of retained wavelet coefficients.

[0021] Furthermore, the expression of the reconstructed denoised voltage traveling wave signal is as follows:

[0022]

[0023] in, is the reconstruction result of the voltage traveling wave signal, are the reconstruction results of the wavelet coefficients of the y-th layer and the y-1-th layer, respectively, and the reconstruction filter coefficient σ h , h Determined by the wavelet basis function.

[0024] Furthermore, according to the two arrival times of the voltage wave head, the cable fault distance is calculated specifically including:

[0025] Get the best hyperparameters. The hyperparameter evaluation indicator is the root mean square error, which is expressed as follows:

[0026]

[0027] Among them, A spu is the root mean square error of hyperparameter training, C is the length of the training set, υ ψa is the predicted value under the ψ group of hyperparameters, Z a is the original value;

[0028] In A spu When the calculation result is minimized, it is the optimal hyperparameter, and the arrival time of the voltage traveling wave head is obtained. The expression is as follows:

[0029]

[0030] Where T is the arrival time of the voltage traveling wave head; ν is the stationarity of the residual sequence, ν is the standard deviation value of stability.

[0031] Furthermore, according to A spu The two arrival times of the voltage traveling wave head are calculated, which are T1 and T2 respectively, and the grid fault distance is calculated. The expression is as follows:

[0032]

[0033] Wherein, W is the distance between the fault point and the traveling wave measurement point, S is the total length of the transmission line; V represents the transmission speed of the voltage traveling wave signal.

[0034] Furthermore, the monitoring terminal includes:

[0035] Cable partial discharge sensor, used to measure the traveling wave signal generated by the cable partial discharge source / fault point;

[0036] The integrated board for synchronization, acquisition and processing of FPGA, ARM and DSP is used to complete the acquisition of traveling wave signals and perform local data processing and storage of traveling wave signals;

[0037] The communication unit is used for the integrated board to upload the processed data to the background master station using wireless / optical fiber communication technology.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] In the present invention, the terminal device collects the traveling wave signal (sampling rate is greater than 50MHz) generated by the partial discharge source / fault point of the cable, and the terminals rely on pulse transmission and reception to achieve self-synchronization (synchronization error is less than 10ns). The lower computer terminal uses wavelet analysis technology to extract the main feature quantity of the partial discharge signal and sends it to the background main station through a wireless or optical fiber communication channel. The background draws a partial discharge spectrum based on the amplitude, number, corresponding power frequency phase and other long-period statistical information of the partial discharge pulse signal, and judges the type and development trend of the cable partial discharge and provides fault warning according to the changes in the spectrum.

[0040] The present invention determines the arrival time of the voltage traveling wave head through a voltage traveling wave head identification method, calculates the power grid fault distance, realizes power grid voltage traveling wave fault positioning, and thus accurately calculates the position of the fault point. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a system architecture diagram of the present invention.

[0042] Figure 2 This is the cable fault early warning and positioning process of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] See also Figure 1 to Figure 2 , the present invention provides a technical solution:

[0045] The distribution network cable partial discharge early warning and fault location system consists of signal sensors, monitoring terminal devices installed at both ends of the cable line, and a data analysis platform. The cable monitoring terminal device realizes data collection and local data processing, and uses wavelet analysis technology to extract features. The data analysis platform realizes long-term data analysis, pattern recognition, and positioning calculation functions. Among them, the application of wavelet analysis technology for feature extraction is the key to locating the partial discharge source / fault point.

[0046] The basic working mode of the cable partial discharge warning and fault location system is that the terminal device collects the traveling wave signal (sampling rate is greater than 50MHz) generated by the partial discharge source / fault point of the cable, and the terminals rely on pulse transmission and reception to achieve self-synchronization (synchronization error is less than 10ns). The lower computer terminal uses edge computing technology to extract the main characteristic quantity of the partial discharge signal and sends it to the background main station through wireless or optical fiber communication channels. The background draws a partial discharge map based on the amplitude, number, corresponding power frequency phase and other long-period statistical information of the partial discharge pulse signal. According to the changes in the map, the type and development trend of the cable partial discharge are judged and fault warning is issued. The warning information includes: two-dimensional / three-dimensional map, discharge type, partial discharge source location, and the fault location information includes: fault point location, fault type, fault waveform, and absolute fault time.

[0047] The positioning is completed by calculating the time difference between the initial traveling wave of the fault and the two ends of the cable, combined with parameters such as cable length and propagation speed. Since the two-end method relies on the initial traveling wave, it does not need to identify the reflected wave and has high reliability.

[0048] The cable partial discharge signal is weak, and the effective high-frequency component exceeds 20MHz. It is technically difficult to measure the partial discharge signal in the grounding wire and extract the partial discharge characteristics. It is necessary to develop a broadband partial discharge signal sensor, apply advanced digital conversion and FPGA technology, design a partial discharge signal acquisition system with a sampling speed of 100MHz, realize high-speed signal acquisition and storage, and apply edge computing technology on the terminal side based on the principle of wavelet transform to extract the characteristics of the partial discharge source / fault point.

[0049] Generally, the cable length is short, the cable partial discharge signal frequency is high, and the traditional GPS / BD timing is used for synchronous fault location between terminals, which has a large error and is difficult to meet the requirements of field use. This project uses pulse intra-synchronization technology, where terminals send and receive pulse signals to each other, and uses FPGA technology to achieve precise synchronization between terminals. The synchronization error is less than 10ns, ensuring that the cable partial discharge source / fault point location error is less than 1%.

[0050] The on-site data processing of the traveling wave signal and the application of wavelet analysis for feature extraction specifically include the use of wavelet analysis algorithm to denoise the voltage traveling wave signal to obtain a smooth voltage traveling wave signal, identify the voltage traveling wave head, calculate the cable fault distance based on the two arrival times of the voltage traveling wave head, and realize fault warning.

[0051] Voltage traveling wave signal denoising includes:

[0052] Wavelet decomposition of voltage traveling wave signals, including wavelet coefficients;

[0053] Determining a wavelet coefficient threshold using the wavelet coefficients, including a wavelet coefficient set;

[0054] The denoised voltage traveling wave signal is reconstructed using the wavelet coefficient set reconstruction algorithm.

[0055] The expression of wavelet decomposition voltage traveling wave signal is as follows:

[0056]

[0057] Among them, x x (a) represents the initial approximation coefficient; E(a) is the voltage traveling wave signal; x y (a) is the voltage traveling wave signal at scale 2 y The approximation coefficient under the decomposition filter coefficient h h , g h ; z y (a) is the voltage traveling wave signal at scale 2 y The wavelet coefficients below.

[0058] The grid fault point transmits traveling wave signals (voltage and current) to both ends of the transmission line. Based on the voltage traveling wave signals, the grid fault is accurately located. Specifically:

[0059] A capacitor is installed between the power grid and the transformer, which can convert the current traveling wave signal into a voltage traveling wave signal using the following formula:

[0060]

[0061] Where E is the grid voltage traveling wave signal, c represents the rated value of the installed capacitor, I is the grid current traveling wave signal, T x is the time for collecting voltage traveling wave signal.

[0062] The voltage traveling wave signal is denoised by wavelet analysis algorithm, which solves the problem that the voltage traveling wave signal is based on the acquired voltage traveling wave signal and there is a lot of noise in the voltage traveling wave signal, which reduces the accuracy of fault location.

[0063] The expression of wavelet coefficient threshold is as follows:

[0064]

[0065] The wavelet coefficients are processed with hard threshold, and the expression is as follows:

[0066]

[0067] in, is the wavelet coefficient threshold; M() is the median removal function; κ is a constant; γ is the length of the yth layer wavelet coefficient after the voltage traveling wave signal is decomposed; ι(z y (a)) is the set of retained wavelet coefficients.

[0068] The expression of reconstructed denoised voltage traveling wave signal is as follows:

[0069]

[0070] in, is the reconstruction result of the voltage traveling wave signal, are the reconstruction results of the wavelet coefficients of the y-th layer and the y-1-th layer, respectively, and the reconstruction filter coefficient σ h , h Determined by the wavelet basis function.

[0071] The application of wavelet analysis algorithm can effectively remove the noise of voltage traveling wave signal and obtain a smooth voltage traveling wave signal, providing a supporting basis for the subsequent voltage traveling wave head identification.

[0072] According to the two arrival times of the voltage wave head, the cable fault distance is calculated including:

[0073] Get the best hyperparameters. The hyperparameter evaluation indicator is the root mean square error, which is expressed as follows:

[0074]

[0075] Among them, A spu is the root mean square error of hyperparameter training, C is the length of the training set, υ ψa is the predicted value under the ψ group of hyperparameters, Z a is the original value;

[0076] In A spu When the calculation result is minimized, it is the optimal hyperparameter, and the arrival time of the voltage traveling wave head is obtained. The expression is as follows:

[0077]

[0078] Where T is the arrival time of the voltage traveling wave head; ν is the stationarity of the residual sequence, ν is the standard deviation value of stability.

[0079] The above process completes the identification of the voltage traveling wave head and determines the calculation method of its arrival time, providing support for subsequent research.

[0080] The voltage traveling wave signal has reflection characteristics. According to A spu The two arrival times of the voltage traveling wave head are calculated, which are T1 and T2 respectively. Based on this, the grid fault distance is calculated. The expression is as follows:

[0081]

[0082] Wherein, W is the distance between the fault point and the traveling wave measurement point, S is the total length of the transmission line; V represents the transmission speed of the voltage traveling wave signal.

[0083] According to the formula The calculation results can determine the location information of the power grid fault, thereby realizing the power grid voltage traveling wave fault location and providing accurate fault location information for the stable operation of the power grid.

[0084] Traditional cable partial discharge pattern recognition is based on the statistical feature analysis of two-dimensional and three-dimensional maps, extracting the fractal dimension and other feature quantities of the map, and then using neural networks or clustering algorithms for pattern recognition. In applications, it exposes the problem of insufficient anti-interference ability, especially weak anti-pulse interference ability. In response to the above problems, the project research analyzes the correlation of pulse signals based on cross wavelet transform to eliminate the influence of white noise and short-time pulse interference. In response to periodic pulse interference, the project research constructs a strong classifier based on multiple weak classifiers, that is, the Adaboosting ensemble learning algorithm is used to complete pattern recognition based on multiple feature quantities such as fractal dimension and graphic features. The key issues to be solved are: setting of weak classifier coefficients and the influence of abnormal samples on pattern recognition.

[0085] The distributed cable monitoring terminal is the hardware platform studied in this project, including the following contents:

[0086] 1) Development of distributed terminals based on SOC architecture processors

[0087] Develop an integrated board for synchronization, acquisition, and processing based on FPGA+ARM+DSP, complete high-speed acquisition and storage, acquisition speed of 100MHz, and synchronization error of less than 10ns; use edge computing technology to extract the characteristics of partial discharge sources / fault points based on the principle of wavelet transform; use wireless / optical fiber communication technology to upload data to the main station. The key issues to be solved are: high-speed signal acquisition and data storage, self-synchronization technology, and feature extraction.

[0088] 2) Research and development of data processing master station

[0089] Adopting object-oriented development mode, develop data processing master station software, draw partial discharge map according to partial discharge monitoring data sent by each terminal, predict partial discharge development trend, identify discharge type, realize advanced functions such as partial discharge source / fault location, insulation state analysis, and online detection of device status; the system is based on B / S architecture, and completes the development of visual and dynamic human-machine interface. Focus on solving partial discharge source, fault location, insulation state analysis, etc.

[0090] Specifically, the monitoring terminal includes:

[0091] Cable partial discharge sensor, used to measure the traveling wave signal generated by the cable partial discharge source / fault point;

[0092] The integrated board for synchronization, acquisition and processing of FPGA, ARM and DSP is used to complete the acquisition of traveling wave signals and perform local data processing and storage of traveling wave signals;

[0093] The communication unit is used for the integrated board to upload the processed data to the background master station using wireless / optical fiber communication technology.

[0094] At present, in the field of cable partial discharge detection in distribution networks, in principle, cable partial discharge sensors can be divided into inductive and capacitive types. Capacitive sensors are based on the principle of electrostatic coupling. They are generally composed of patch electrodes or ring-type capacitive sensors attached to the semiconductor layer. They have advantages in sensitivity and are easy to shield external interference, but they require contact measurement. Therefore, their application is subject to certain restrictions. They have been used for 400kV and 500kV cable line monitoring in Germany, Japan and other countries. Inductive sensors generally use Rogowski coils, which can achieve non-contact measurement and fast dynamic response based on the principle of electromagnetic coupling. However, electromagnetic coupling is relatively susceptible to interference and is not easy to achieve good shielding. The core material of traditional Rogowski coils is generally an open structure with a bandpass characteristic. The signal bandwidth is 0 to 30MHz, and it is difficult to achieve full-band coverage. This project develops a closed Rogowski coil installed in a metal chassis to solve the bandwidth and interference problems.

[0095] The present invention, the undescribed parts are prior art.

[0096] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for early warning and locating cable faults in high altitude areas, characterized in that: include: The monitoring terminals are installed at both ends of the cable line to collect the traveling wave signals generated by the partial discharge source / fault point of the cable, process the traveling wave signals on-site, extract features using wavelet analysis, and send them to the background master station through wireless or optical fiber communication channels. The background master station draws a partial discharge map based on the long-period statistical information such as the amplitude, number, and corresponding power frequency phase of the partial discharge pulse signal, and judges the type and development trend of the cable partial discharge and issues fault warnings based on changes in the map.

2. A method for early warning and locating cable faults in high altitude areas as claimed in claim 1, characterized in that: The on-site data processing of the traveling wave signal and the application of wavelet analysis for feature extraction specifically include the use of wavelet analysis algorithm to denoise the voltage traveling wave signal to obtain a smooth voltage traveling wave signal, identify the voltage traveling wave head, calculate the cable fault distance based on the two arrival times of the voltage traveling wave head, and realize fault warning.

3. A method for early warning and locating cable faults in high altitude areas as claimed in claim 2, characterized in that: Voltage traveling wave signal denoising includes: Wavelet decomposition of voltage traveling wave signals, including wavelet coefficients; Determining a wavelet coefficient threshold using the wavelet coefficients, including a wavelet coefficient set; The denoised voltage traveling wave signal is reconstructed using the wavelet coefficient set reconstruction algorithm.

4. A method for early warning and locating cable faults in high altitude areas as claimed in claim 3, characterized in that: The expression of wavelet decomposition voltage traveling wave signal is as follows: Among them, x x (a) represents the initial approximation coefficient; E(a) is the voltage traveling wave signal; x y (a) is the voltage traveling wave signal at scale 2 y The approximation coefficient under the decomposition filter coefficient h h , g h ;z y (a) is the voltage traveling wave signal at scale 2 y The wavelet coefficients below.

5. A method for early warning and locating cable faults in high altitude areas as claimed in claim 4, characterized in that: The expression of wavelet coefficient threshold is as follows: The wavelet coefficients are processed with hard threshold, and the expression is as follows: Among them, θ is the wavelet coefficient threshold; M() is the median removal function; κ is a constant; γ is the length of the yth layer wavelet coefficient after the voltage traveling wave signal is decomposed; ι(z y (a)) is the set of retained wavelet coefficients.

6. A method for early warning and locating cable faults in high altitude areas as claimed in claim 5, characterized in that: The expression of reconstructed denoised voltage traveling wave signal is as follows: in, is the reconstruction result of the voltage traveling wave signal, are the reconstruction results of the wavelet coefficients of the y-th layer and the y-1-th layer, respectively, and the reconstruction filter coefficient σ h , h Determined by the wavelet basis function.

7. A method for early warning and locating cable faults in high altitude areas as claimed in claim 6, characterized in that: According to the two arrival times of the voltage wave head, the cable fault distance is calculated including: Get the best hyperparameters. The hyperparameter evaluation indicator is the root mean square error, which is expressed as follows: Among them, A spu is the root mean square error of hyperparameter training, C is the length of the training set, υ ψa is the predicted value under the ψ group of hyperparameters, Z a is the original value; In A spu When the calculation result is minimized, it is the optimal hyperparameter, and the arrival time of the voltage traveling wave head is obtained. The expression is as follows: Where T is the arrival time of the voltage traveling wave head; ν is the stationarity of the residual sequence, ν is the standard deviation value of stability.

8. A method for early warning and locating cable faults in high altitude areas as claimed in claim 6, characterized in that: According to A spu The two arrival times of the voltage traveling wave head are calculated, which are T1 and T2 respectively, and the grid fault distance is calculated. The expression is as follows: Wherein, W is the distance between the fault point and the traveling wave measurement point, S is the total length of the transmission line; V represents the transmission speed of the voltage traveling wave signal.

9. A method for early warning and locating cable faults in high altitude areas as claimed in claim 1, characterized in that: The monitoring terminals include: Cable partial discharge sensor, used to measure the traveling wave signal generated by the cable partial discharge source / fault point; The integrated board for synchronization, acquisition and processing of FPGA, ARM and DSP is used to complete the acquisition of traveling wave signals and perform on-site data processing and storage of traveling wave signals; the communication unit is used for the integrated board to upload processed data to the background main station using wireless / optical fiber communication technology.

Images