High-voltage cable sheath circulating current monitoring device and fault positioning method

Abstract

The invention discloses a high-voltage cable sheath circulating current monitoring device and a fault positioning method, and particularly relates to the technical field of cable fault monitoring. Comprising a distributed optical fiber sensor spirally wound on the surface of a cable, quantum magnetometer array modules arranged at equal phase difference of 90 degrees, a climbing robot used for positioning and clamping, and an edge calculation module integrated with an improved WP-CNN processing chip. The fault positioning method is realized by acquiring a magnetic field gradient tensor, constructing a fault feature vector, performing wavelet packet decomposition, performing convolutional neural network operation, calculating a fault point distance, and finally moving and clamping and fixing a fault point by a climbing robot. The cable state can be accurately monitored, fault features can be efficiently extracted, fault positioning can be accurately carried out, fault processing and continuous monitoring are facilitated, and safe and stable operation of the high-voltage cable is effectively guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault monitoring. More specifically, the present invention relates to a high-voltage cable sheath circulating current monitoring device and a fault location method. Background Art

[0002] In the modern power transmission system, high-voltage cables, as key infrastructure, undertake the important task of efficiently and stably delivering a large amount of electric energy to various regions. With the acceleration of the urbanization process and the continuous expansion of the industrial scale, the social demand for electricity has shown an explosive growth. This has made the application of high-voltage cables more and more extensive, and its coverage extends from the core areas of cities to remote rural areas, and from prosperous commercial areas to various industrial parks.

[0003] However, high-voltage cables face many challenges during operation, and faults occur from time to time. Due to being in a working environment of high voltage and large current for a long time, the insulating materials inside the cable will gradually age, resulting in a decline in insulation performance, and then causing faults such as short circuits and leakage. At the same time, external environmental factors, such as chemical corrosion in the soil, seepage erosion of groundwater, extrusion and collision of mechanical external forces, and extreme climate conditions (such as lightning strikes, heavy rains, high temperatures, etc.), may all cause serious damage to the structure and performance of the cable, affecting its normal operation.

[0004] Current cable fault monitoring technologies have obvious limitations in dealing with these complex problems. On the one hand, existing monitoring devices can often only monitor some parameters of the cable. For example, they can only detect temperature or current, and cannot comprehensively and integrally reflect the actual operating state of the cable. This makes it difficult to detect some potential fault hazards in time, resulting in the fact that faults cannot be effectively controlled in the initial stage of development, and ultimately leading to serious power outages. On the other hand, existing fault location methods have low accuracy. When facing a long-distance and multi-branch high-voltage cable network, it is difficult to quickly and accurately determine the specific location of the fault point. This will not only increase the difficulty and time cost of fault troubleshooting, but also prolong the power outage time, bringing huge losses to the social economy.

[0005] In addition, traditional cable monitoring systems also have problems such as complex installation, difficult maintenance, and high costs. Many monitoring devices need to carry out large-scale wiring and transformation on the cable line, which not only increases the difficulty and risk of construction, but also causes certain interference to the normal operation of the cable. Moreover, these devices need to be maintained and calibrated regularly during operation, and the maintenance cost is relatively high. For some remote areas or areas with harsh environments, it is even more difficult to carry out maintenance work.

[0006] Therefore, a high-voltage cable sheath circulating current monitoring device and a fault location method are proposed. Summary of the Invention

[0007] In order to overcome the above-mentioned defects of the prior art, the present invention provides a high-voltage cable sheath circulating current monitoring device and a fault location method to solve the problems raised in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions: A high-voltage cable sheath circulating current monitoring device, characterized in that it includes:

[0009] A distributed optical fiber sensor helically wound on the surface of the cable, adopting a double-helix orthogonal arrangement, and the pitch is 1.2-1.5 times the diameter of the cable;

[0010] A quantum magnetometer array module arranged at equal phase differences of 90° in the four quadrants of the cable circumference;

[0011] A mobile positioning module sleeved on the cable for moving to the fault point to clamp both ends of the fault point;

[0012] An edge computing module integrated with an improved WP-CNN processing chip, with an operation speed ≥ 2 TFLOPS.

[0013] Preferably, the mobile positioning module is a climbing robot, and self-locking clamping mechanisms are arranged on both sides of the climbing robot.

[0014] Preferably, the self-locking clamping mechanism includes variable-curvature grippers driven by SMA and a contact force feedback unit with piezoelectric ceramics built-in.

[0015] Preferably, the climbing robot adopts a vibration adsorption - electromagnetic composite drive method, with a stepping accuracy of 0.1 mm, a maximum climbing angle of 90°, and a load capacity ≥ 20 kg.

[0016] Preferably, the distributed optical fiber sensor adopts space division multiplexing technology to achieve dual-mode measurement of BOTDR and Φ-OTDR in a single optical fiber, with a spatial resolution of 5 cm and a sampling frequency of 1 kHz.

[0017] A fault location method based on the above-mentioned high-voltage cable sheath circulating current monitoring device includes the following steps:

[0018] Step 1: Obtain the magnetic field gradient tensor through the quantum magnetometer array module. Using the quantum magnetometer array, measure the rate of change of the magnetic field components in three spatial directions (x, y, z), and combine these rates of change into a magnetic field gradient tensor, whose form is Among them, are respectively the rates of change of the magnetic field components B x , B y , B z in the x, y, z directions;

[0019] Step 2: Based on the obtained fault feature vector of the magnetic field gradient tensor component, comprehensively consider the differential value I of the sheath circulating current d , temperature gradient , the previously obtained magnetic field gradient tensor , and the vibration signal power spectrum S(f) to form a fault feature vector F, expressed as

[0020] Step 3: Decompose through wavelet packet according to the information provided by the fault feature vector. The wavelet packet decomposition formula is:

[0021]

[0022] where h m is the wavelet packet decomposition filter of the m-th layer, and x n is the signal of the n-th channel;

[0023] Step 4: Input the fault feature vector of the component into the convolutional neural network. After multiple-layer operations, realize the fusion and extraction of fault features. The specific formula is y = σ(W3·ReLU(W2·ReLU(W1·F + b1)+b2)+b3), where W i is the weight matrix obtained through training, and σ is the Sigmoid activation function;

[0024] Step 5: Then calculate the distance between the fault point and the climbing robot according to the electromagnetic propagation speed, signal transmission time difference, cable attenuation correction factor, and reflection coefficient. The specific formula is

[0025]

[0026] where v p is the electromagnetic propagation speed, Δt is the signal transmission time difference, K is the cable attenuation correction factor, and Γ is the reflection coefficient;

[0027] Step 6: The climbing robot moves to the fault point according to the distance from the fault point, clamps and fixes both ends of the fault point to realize fault point positioning and manifestation.

[0028] Preferably, the correction formula for the propagation speed in Step 5 is:

[0029]

[0030] where c is the speed of light, ε r is the equivalent dielectric constant, and Z0 and Y0 are the impedance and admittance per unit length.

[0031] Preferably, in Step 1, the quantum positioning algorithm is adopted. Through the eigenvalue decomposition of the measured magnetic field gradient tensor, its decomposition formula is:

[0032]

[0033] Then calculate the position of the fault point. The formula for calculating the position of the fault point is:

[0034]

[0035] where is the eigenvalue benchmark in the normal state. By comparing the difference between the eigenvalue and the normal state, the position closest to the fault state is found, thereby determining the position of the fault point.

[0036] Preferably, the mathematical relationship between the clamping force and displacement of the climbing robot in step six is:

[0037] F(x) = F0·e μ(θ+βx)

[0038] where μ is the friction coefficient, θ is the initial contact angle, α is the curvature adjustment factor, and x is the displacement.

[0039] The technical effects and advantages of the present invention:

[0040] 1. By adopting the double - helix orthogonal arrangement and space - division multiplexing technology for distributed optical fiber sensors, dual - mode measurements of BOTDR and Φ - OTDR are realized. The spatial resolution reaches 5 cm and the sampling frequency is 1 kHz, which can accurately obtain cable temperature and strain information, providing a reliable basis for early fault warning.

[0041] 2. By obtaining the magnetic field gradient tensor through the quantum magnetometer array, combining the differential value of the sheath circulating current, temperature gradient, and vibration signal power spectrum to construct a fault feature vector, and then through wavelet packet decomposition and convolutional neural network operations, the fault features can be effectively fused and extracted, improving the accuracy of fault judgment.

[0042] 3. By calculating the fault point distance using parameters such as the electromagnetic propagation speed and signal transmission time difference, and correcting the propagation speed, combined with the quantum positioning algorithm, the position of the fault point can be accurately determined with high positioning accuracy.

[0043] 4. By using the vibration adsorption - electromagnetic composite drive for the climbing robot, with a stepping accuracy of 0.1 mm, a maximum climbing angle of 90°, and a load - bearing capacity of ≥20 kg, and the self - locking clamping mechanism with variable - curvature jaws and contact - force feedback unit cooperating, it can stably clamp the fault point. At the same time, it can carry equipment for preliminary detection or marking, facilitating subsequent processing by maintenance personnel. Moreover, during and after the fault handling process, each sensor continuously collects data, repeating the fault judgment process, which can monitor the cable operation status in real - time, timely discover new hidden dangers, and ensure the long - term stable operation of the cable. Brief Description of the Drawings

[0044] Figure 1Schematic diagram of the fault location method of the present invention. Detailed implementation manners

[0045] Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0046] A high-voltage cable sheath circulating current monitoring device provided by the present invention includes:

[0047] A distributed optical fiber sensor spirally wound on the surface of the cable, adopting a double-helix orthogonal arrangement mode, and the pitch is 1.2-1.5 times the diameter of the cable;

[0048] A quantum magnetometer array module arranged at four quadrant positions on the circumference of the cable with an equal phase difference of 90°;

[0049] A mobile positioning module sleeved on the cable for moving to the fault point and clamping both ends of the fault point;

[0050] An edge computing module integrated with an improved WP-CNN processing chip, and the operation speed ≥ 2 TFLOPS.

[0051] The mobile positioning module is a climbing robot, and self-locking clamping mechanisms are arranged on both sides of the climbing robot.

[0052] The self-locking clamping mechanism includes variable-curvature grippers driven by SMA and a contact force feedback unit with piezoelectric ceramics built therein.

[0053] The climbing robot adopts a vibration adsorption-electromagnetic composite drive mode, with a stepping accuracy of 0.1 mm, a maximum climbing angle of 90°, and a load capacity ≥ 20 kg.

[0054] The distributed optical fiber sensor adopts space division multiplexing technology to realize BOTDR and Φ-OTDR dual-mode measurement in a single optical fiber, with a spatial resolution of 5 cm and a sampling frequency of 1 kHz.

[0055] As Figure 1 shown, the present invention also provides a high-voltage cable sheath circulating current monitoring fault location method, including the following steps:

[0056] Step 1: Obtain the magnetic field gradient tensor through the quantum magnetometer array module. Using the quantum magnetometer array, measure the change rates of the magnetic field components in three spatial directions (x, y, z), and combine these change rates into a magnetic field gradient tensor, whose form is where are respectively the change rates of the magnetic field components B x , B y , B z in the x, y, and z directions;

[0057] Step 2: Based on the obtained fault feature vector of the magnetic field gradient tensor component, comprehensively combine the differential value I of the sheath circulating current d , the temperature gradient the previously obtained magnetic field gradient tensor and the vibration signal power spectrum S(f) to form a fault feature vector F, expressed as

[0058] Step 3: Decompose through wavelet packet according to the information provided by the fault feature vector. The wavelet packet decomposition formula is:

[0059]

[0060] where h m is the wavelet packet decomposition filter of the m-th layer, and x n is the signal of the n-th channel;

[0061] Step 4: Input the fault feature vector of the component into the convolutional neural network. Through multi-layer operations, realize the fusion and extraction of fault features. The specific formula is y = σ(W3·ReLU(W2·ReLU(W1·F + b1) + b2) + b3), where W i is the weight matrix obtained through training, and σ is the Sigmoid activation function;

[0062] Step 5: Then calculate the distance between the fault point and the climbing robot according to the electromagnetic propagation speed, signal transmission time difference, cable attenuation correction factor, and reflection coefficient. The specific formula is

[0063]

[0064] where v p is the electromagnetic propagation speed, Δt is the signal transmission time difference, K is the cable attenuation correction factor, and Γ is the reflection coefficient;

[0065] Step 6: The climbing robot moves to the fault point according to the distance from the fault point, clamps and fixes both ends of the fault point to realize fault point positioning and manifestation.

[0066] The correction formula for the propagation speed in Step 5 is:

[0067]

[0068] where c is the speed of light, ε r is the equivalent dielectric constant, and Z0 and Y0 are the impedance and admittance per unit length.

[0069] In Step 1, the quantum positioning algorithm is adopted. Through the eigenvalue decomposition of the measured magnetic field gradient tensor, its decomposition formula is:

[0070]

[0071] Then calculate the location of the fault point. The calculation formula for the fault point location is:

[0072]

[0073] Where, is the eigenvalue benchmark in the normal state. By comparing the difference between the eigenvalue and the normal state, the location closest to the fault state is found, thereby determining the fault point location.

[0074] The mathematical relationship between the clamping force and displacement of the climbing robot in step six is:

[0075] F(x) = F0·e μ(θ+αx)

[0076] Where, μ is the friction coefficient, θ is the initial contact angle, α is the curvature adjustment factor, and x is the displacement.

[0077] During specific implementation, a distributed fiber optic sensor with a double - helix orthogonal arrangement is helically wound on the surface of the high - voltage cable, ensuring that the pitch is 1.2 - 1.5 times the cable diameter. Using space - division multiplexing technology, a single fiber realizes dual - mode measurements of BOTDR and Φ - OTDR, ensuring a spatial resolution of 5 cm and a sampling frequency of 1 kHz to accurately obtain cable temperature and strain information;

[0078] Then, the quantum magnetometer array module is installed at four - quadrant positions on the cable circumference with an equal phase difference of 90°, enabling it to accurately measure the change rate of the magnetic field components in three spatial directions (x, y, z), providing data support for obtaining the magnetic field gradient tensor in the subsequent process;

[0079] And the climbing robot is sleeved on the cable, and the self - locking clamping mechanisms on both sides are installed and debugged in place. The variable - curvature jaws driven by SMA can adjust the jaw curvature according to the actual situation of the cable, enhancing the clamping stability. The contact - force feedback unit with built - in piezoelectric ceramics can monitor and feedback the clamping force in real time to ensure reliable clamping.

[0080] Finally, the integrated improved WP - CNN processing chip and the edge - computing module with an operation speed ≥2 TFLOPS are installed and initialized, providing computing power guarantee for subsequent data processing and analysis.

[0081] Specifically, using the quantum magnetometer array module, continuously measure the change rate of the magnetic field components in the x, y, and z directions, and calculate the magnetic field gradient tensor according to the formula For example, collect data once every certain time interval (such as 1 second) to obtain the magnetic field gradient tensor value at the current moment.

[0082] Synchronous sheath circulating current differential value I d , temperature gradient and vibration signal power spectrum S(f), combined with the obtained magnetic field gradient tensor According to the formula Construct a fault feature vector. The acquisition frequency is consistent with the acquisition frequency of the magnetic field gradient tensor to ensure the real-time and accuracy of the data.

[0083] Then, for each channel signal in the fault feature vector, according to the formula WPD m,n (k) = ∑ l h m (l - 2k)x n (l) to perform wavelet packet decomposition, where, select appropriate wavelet packet decomposition layer number m and filter h according to the signal characteristics m , and process each channel signal x n to extract signal features at different frequency scales.

[0084] Input the data after wavelet packet decomposition into a convolutional neural network, perform multi-layer operations according to the formula y = σ(W3·ReLU(W2·ReLU(W1·F + b1) + b2) + b3), and pre-train the convolutional neural network with a large number of sample data in fault and normal states in advance to obtain an appropriate weight matrix W i , and realize the fusion and extraction of fault features through the Sigmoid activation function σ, and output the fault judgment result.

[0085] Then, according to the formula Calculate the distance between the fault point and the climbing robot, where, the electromagnetic propagation speed v p According to the formula perform correction calculation, c is the speed of light, ε r is the equivalent dielectric constant, Z0 and Y0 are the impedance and admittance per unit length, the signal transmission time difference Δt, the cable attenuation correction factor K and the reflection coefficient are obtained by measuring with relevant sensors or based on cable parameters and empirical data.

[0086] The climbing robot adopts a vibration adsorption - electromagnetic composite drive method, and moves towards the fault point with a stepping accuracy of 0.1 mm according to the calculated fault distance. After reaching the fault point, clamp and fix both ends of the fault point through a self-locking clamping mechanism to realize the positioning and manifestation of the fault point. The relationship between the clamping force and displacement follows the formula F(x) = F0·e μ(θ+αx), by adjusting the jaw displacement x and using the SMA drive and contact force feedback unit, ensure an appropriate clamping force to stably fix the fault point. Thus, when the fault point has a fracture tendency, the clamping force can be increased according to the change in the friction force with the cable, avoiding the direct breakage and separation of the cable.

[0087] After the climbing robot fixes the fault point, it can carry relevant equipment to conduct preliminary detection or marking of the fault point, providing convenience for subsequent maintenance personnel to handle the fault. For example, install a camera to capture images of the fault point, or equip it with simple detection tools to preliminarily judge the fault type. At the same time, a positioning system can also be installed in the climbing robot so that maintenance personnel can directly reach the fault point for maintenance.

[0088] During or after the fault handling process, distributed optical fiber sensors, quantum magnetometer array modules, etc. continuously collect data, repeat the above data processing and fault judgment processes, and monitor the cable operation status in real time to timely discover new potential fault hazards.

[0089] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A high-voltage cable sheath circulating current monitoring device, characterized in that, Including: A distributed optical fiber sensor helically wound around the surface of the cable, adopting a double-helix orthogonal arrangement mode, and the pitch is 1.2 - 1.5 times the diameter of the cable; A quantum magnetometer array module arranged at the four quadrant positions of the cable circumference with an equal phase difference of 90°; A mobile positioning module sleeved on the cable and used to move to the fault point to clamp both ends of the fault point; An edge computing module integrated with an improved WP-CNN processing chip, and the operation speed ≥ 2 TFLOPS.

2. The high-voltage cable sheath circulating current monitoring device according to claim 1, characterized in that, The mobile positioning module is a climbing robot, and self-locking clamping mechanisms are arranged on both sides of the climbing robot.

3. The high-voltage cable sheath circulating current monitoring device according to claim 2, characterized in that, The self-locking clamping mechanism includes variable-curvature grippers driven by SMA and a contact force feedback unit with piezoelectric ceramics built-in.

4. The high-voltage cable sheath circulating current monitoring device according to claim 3, characterized in that, The climbing robot adopts a vibration adsorption - electromagnetic composite drive mode, with a stepping accuracy of 0.1 mm, a maximum climbing angle of 90°, and a load capacity ≥ 20 kg.

5. The high-voltage cable sheath circulating current monitoring device according to claim 1, characterized in that The distributed optical fiber sensor adopts space division multiplexing technology to achieve BOTDR and Φ-OTDR dual-mode measurement in a single optical fiber, with a spatial resolution of 5 cm and a sampling frequency of 1 kHz.

6. A fault location method for a high-voltage cable sheath circulating current monitoring device according to any one of claims 1-5, characterized in that, Including the following steps: Step 1: Obtain the magnetic field gradient tensor through the quantum magnetometer array module. Using the quantum magnetometer array, measure the rate of change of the magnetic field components in the three spatial directions (x, y, z), and combine these rates of change into the magnetic field gradient tensor, which has the form where are the rates of change of the magnetic field components B x , B y , B z in the x, y, and z directions, respectively; Step 2: According to the obtained fault feature vector of the magnetic field gradient tensor component, comprehensively combine the differential value I of the sheath circulating current d , the temperature gradient , the magnetic field gradient tensor obtained previously , and the vibration signal power spectrum S(f) to form a fault feature vector F, expressed as Step Three: Decompose through wavelet packet according to the information provided by the fault feature vector, and the wavelet packet decomposition formula is: where h m is the wavelet packet decomposition filter of the m-th layer, and x n is the signal of the n-th channel; Step 4: Input the fault feature vector of the component into the convolutional neural network. Through multiple-layer operations, the fusion and extraction of fault features are realized. The specific formula is y = σ(W3·ReLU(W2·ReLU(W1·F + b1)+b2)+b3), where W i is the weight matrix obtained through training, and σ is the Sigmoid activation function; Step Five: Then calculate the distance between the fault point and the climbing robot according to the electromagnetic propagation speed, signal transmission time difference, cable attenuation correction factor, and reflection coefficient. The specific formula is where v p is the electromagnetic propagation speed, Δt is the signal transmission time difference, K is the cable attenuation correction factor, and is the reflection coefficient; Step Six: The climbing robot moves to the fault point according to the distance from the fault point, clamps and fixes both ends of the fault point, and realizes fault point positioning and manifestation.

7. A high-voltage cable sheath circulating current monitoring fault location method according to claim 6, characterized in that, The correction formula for the propagation speed in Step Five is: where c is the speed of light, ε r is the equivalent permittivity, and Z0 and Y0 are the impedance and admittance per unit length.

8. A high-voltage cable sheath circulating current monitoring fault location method according to claim 6, characterized in that In Step One, a quantum positioning algorithm is adopted. Through the eigenvalue decomposition of the measured magnetic field gradient tensor, the decomposition formula is: Then calculate the position of the fault point, and the fault point position calculation formula is: Among them, is the normal state eigenvalue benchmark. By comparing the eigenvalue with that in the normal state, the position closest to the fault state is found, thereby determining the fault point position.

9. A method for monitoring and fault locating of the sheath circulating current of a high-voltage cable according to claim 6, characterized in that The mathematical relationship between the clamping force and displacement of the climbing robot in Step Six is: F(x) = F0·e μ(θ+αx) Where, μ is the friction coefficient, θ is the initial contact angle, α is the curvature adjustment factor, and x is the displacement amount.

Images