A non-contact current traveling wave fault location method for multi-terminal overhead distribution lines

Abstract

The application discloses a kind of multi-terminal overhead distribution line non-contact current traveling wave fault location method, belong to power system relay protection technical field;The application is arranged tunnel magneto-resistive sensor array at each terminal, non-contact acquisition fault traveling wave magnetic field signal;According to Biot-Savart law, the vector superposition method is used to calculate the resultant magnetic flux density and project to each coordinate axis, introduce spatial projection transformation and attitude error compensation mechanism;Dynamic structure element morphological filtering algorithm is used to accurately identify the arrival time of traveling wave front under strong noise;Based on fault influence degree matrix and fault direction encoding matrix, branch fault, bus node fault and node line fault are distinguished by hierarchical discrimination logic;Finally, combined with double-ended traveling wave method, fault accurate positioning is completed;Through testing, the fault section identification accuracy of the application is high, the positioning error is small, no power outage installation is needed, the anti-interference ability is strong, suitable for complex multi-branch distribution network, and the engineering practicability is strong.

Description

Technical Field

[0001] This invention relates to the field of power distribution network technology, specifically to a fault location method for overhead power distribution lines based on non-contact magnetic field sensing and dynamic morphological analysis, which is particularly suitable for complex power distribution networks with multi-terminal power supply and multi-level branch structure. Background Technology

[0002] As the "last mile" connecting users to the power system, the reliability of power distribution lines directly affects the quality of power supply. With the advancement of new power systems, the structure of modern power distribution networks is becoming increasingly complex, with multi-terminal power supply, closed-loop operation, and multi-level branch wiring becoming more and more common. Once a permanent fault occurs in a line, if the fault point cannot be quickly and accurately located, it will lead to large-scale and prolonged power outages, causing serious economic losses.

[0003] Currently, existing power distribution line fault location technologies mainly suffer from the following types of defects: 1. Contact Traveling Wave Detection Technology: The traveling wave method itself has the advantages of high positioning accuracy and is not affected by fault type. However, this type of solution must connect capacitor dividers, Rogowski coils, and other acquisition equipment in series or parallel in the high-voltage circuit. Power must be cut off during the construction of such devices, making on-site operations difficult. The equipment procurement and subsequent operation and maintenance costs are high, making it impossible to implement on a large scale in distribution networks with a large number of nodes and complex topologies.

[0004] 2. Traditional non-contact detection solutions: The industry has attempted to use Hall effect sensors and fluxgate magnetometers for non-contact magnetic field measurement. However, these devices generally suffer from low detection sensitivity and narrow effective operating bandwidth, making it difficult to capture the weak high-frequency traveling wave signals generated by faults. Furthermore, existing fault diagnosis logic designs for multi-terminal power grids are simple, lacking a systematic hierarchical identification system. This makes it impossible to distinguish whether a fault is located on a main line, branch line, or bus node, resulting in ambiguous location results that fail to meet maintenance requirements.

[0005] 3. Impedance-based fault location method: This method relies on line impedance to calculate the fault distance, and the calculation results are highly dependent on the accuracy of line parameters. Outdoor lines are easily affected by factors such as terrain, mutual inductance of adjacent lines, and cable joints. In high-impedance grounding fault scenarios, the location error will increase significantly, and in severe cases, it will fail directly.

[0006] In summary, developing an intelligent fault location method that is easy to deploy, has strong anti-interference capabilities, and can adapt to complex multi-terminal topologies is a technical challenge that urgently needs to be solved in the current power system. Summary of the Invention

[0007] The main objective of this invention is to provide a non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines, aiming to solve existing technical problems.

[0008] To achieve the above objectives, the present invention provides a non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines, comprising the following steps: Step S1: Deploy tunnel magnetoresistive sensors at each terminal node, with the sensitive axis of the tunnel magnetoresistive sensor perpendicular to the line direction, to non-contactly collect the spatial alternating magnetic flux density signal generated by the initial fault current traveling wave, and synchronize the clock of each terminal. Step S2: Establish a three-dimensional spatial coordinate system with the sensor's sensing center as the origin. According to the Biot-Savart law, the three-phase current traveling waves are regarded as independent magnetic field sources. The magnetic flux density vector generated by each phase current at the sensor position is calculated using the vector superposition method. After superposition, the composite magnetic flux density vector is obtained. Then, this vector is projected onto the X-axis and Y-axis directions of the sensor coordinate system to obtain the components of the composite magnetic flux density. The attitude error of the composite magnetic flux density is compensated using a rotation calibration matrix to obtain the corrected magnetic flux density components. Step S3: The corrected magnetic flux density signal is processed using a dynamic structural element morphological filtering algorithm. The length of the structural element of the dynamic structural element filter is adaptively adjusted according to the local variance characteristics of the signal. The traveling wave envelope features are extracted using morphological dilation and erosion operations. By calculating the cross-correlation function between the dilation signal and the erosion signal, the incident wave and the reflected wave are identified according to the time delay direction corresponding to the cross-correlation peak. The abrupt change point of the incident traveling wave is identified as the wavefront arrival time. Step S4: Using the line section between each terminal and the adjacent bus node as the basic unit, pre-calibrate the reference magnetic flux density value of each terminal when a unit amplitude traveling wave signal is injected; after a system fault occurs, obtain the initial traveling wave magnetic flux density amplitude measured by each terminal, and construct a fault influence matrix using the ratio of the fault amplitude of each terminal to the corresponding reference value as elements. A ; Step S5: Calculate the fault impact matrix. A A sign decision function is applied to each element in the matrix. This sign decision function sets a tolerance range with the value 1 as a reference point, mapping each element to three encoding states: "+1", "0", and "-1", thus forming a fault location encoding matrix. C ; Step S6: Based on the fault location coding matrix C The proportion of "-1" and "+1" in the off-diagonal elements is counted by row. By detecting the complementary and symmetrical relationship between row code and column code, fault segment identification is performed according to the hierarchical logic, which includes branch fault identification, bus node fault identification and inter-node fault identification. Step S7: After determining the fault section, select two terminals located at both ends of the fault section or close to the fault point as ranging ends. Extract the arrival time of the initial traveling wave detected by each of the two ranging ends. Combine the line length between the two ranging ends and the traveling wave propagation speed, and use the double-end traveling wave ranging formula to calculate the location of the fault point.

[0009] Furthermore, in step S1, the tunnel magnetoresistive sensor is installed below the crossarm or on the side wall of the main tower material, with an installation height of not less than 0.5m and not more than 2.0m from the plane where the three-phase conductors are located; the system uses a sampling rate of not less than 4MHz for real-time data acquisition, and realizes clock synchronization of the acquisition process of each terminal through a timing module. The timing module uses a Beidou satellite navigation system or GPS global positioning system receiving unit, and each terminal realizes second pulse synchronization by receiving satellite timing signals, with a synchronization accuracy better than 1μs.

[0010] Furthermore, in step S2, according to the Biot-Savart law, the magnetic flux density vector generated by the three-phase current traveling wave at the sensor location can be expressed as:

[0011] in, I k These are the instantaneous values ​​of the current in each phase. d k This represents the straight-line spatial distance from the sensor to each phase conductor. e dk Let be the unit direction vector of each phase conductor of the sensor; project this vector onto the X and Y axes of the sensor coordinate system respectively to obtain the component expressions of the composite magnetic flux density: ; ; in, m 0 is the permeability of free space. , , These are the instantaneous values ​​of the traveling waves of the three-phase currents A, B, and C, respectively. , , These represent the straight-line spatial distances from the sensor to the three phase conductors A, B, and C, respectively. α , β These are the angles between the horizontal projection and the vertical direction of the connection between the A-phase and C-phase conductors and the sensor; The spatial projection transformation and attitude error compensation are achieved through the following rotation matrix: ; in, iA pitch angle deviation is installed on the sensor, which is obtained by real-time measurement using a tilt sensor or by manual calibration during on-site installation.

[0012] Furthermore, in step S3, the dynamic structuring element morphological filtering algorithm employs a one-dimensional flat adaptive structuring element. i ( m The structuring element's magnitude is set to 0 to reduce computational complexity. The structuring element length... M Based on magnetic flux density signal B ( n The local variance characteristics of ) are adaptively determined, and the determination method is as follows:

[0013] in, s B and m B These represent the standard deviation and mean of the magnetic flux density signal within the current analysis window, respectively. l This is a preset scaling factor used to map the relative fluctuation amplitude of the signal to a suitable structural element length dimension. M min and M max These are the preset upper and lower limits for the length of the structural elements. This design allows the filter to automatically increase the filtering scale to smooth noise when the signal fluctuates wildly, and automatically decrease the filtering scale to preserve detailed features when the signal is calm.

[0014] Furthermore, in step S3, the expansion operation... Corrosion Operation The expressions are as follows: ; ; To distinguish between incident and reflected waves, the cross-correlation function between the expansion and corrosion signals is calculated:

[0015] Take the cross-correlation function Latency corresponding to the first peak t p ;like t p >0 indicates that the expansion signal leads the corrosion signal in phase, and is determined to be an incident traveling wave; if t p <0 indicates that the expansion signal lags behind the corrosion signal in phase, and is determined to be a reflected traveling wave; the abrupt change point of the incident traveling wave is identified as the wavefront arrival time.

[0016] Further, in step S4, the reference magnetic flux density The specific calibration method is as follows: at the bus node J Towards the first i A unit amplitude pulsed traveling wave signal is injected into the line section where each terminal is located. The tunnel magnetoresistive sensor at that terminal measures and records the corresponding synthetic magnetic flux density amplitude, which is then calibrated as follows: The rise time of the pulse current is no greater than 1 μs, and the pulse width is no greater than 2 μs. The fault impact matrix A for n × n 3D matrix n For the total number of terminals, its number is... i Line 1 j Column elements r ij Defined as: ; in, Let be the amplitude of the initial traveling wave flux density measured at the i-th terminal. For the first j The reference magnetic flux density value of each terminal in the corresponding line segment, with diagonal elements marked as invalid elements.

[0017] Furthermore, in step S5, the sign decision function The expression is: ; in, d The preset error tolerance threshold is used to overcome numerical fluctuations caused by field measurement noise and sensor inconsistencies. d The value range is from 0.01 to 0.05; the fault location coding matrix C elements in ,Right now Diagonal elements are marked as invalid.

[0018] Furthermore, in step S6, the hierarchical logic specifically includes: (1) Branch fault identification: encoding the fault location matrix C For each row k in the array, calculate the percentage of off-diagonal elements that have the value "-1". The percentage of values ​​with a sum of "+1" If a terminal exists k satisfy If the value is 1, and the percentage of off-diagonal elements with a value of "+1" in the corresponding column of this terminal is also 1, then the fault is determined to occur in this terminal. k On the connected branch lines; (2) Bus node fault identification: If the fault location coding matrix C If all off-diagonal elements are 0, then the fault is located at the bus node. (3) Fault identification between nodes: When the above-mentioned branch fault characteristics and bus node fault characteristics are not met, the fault is determined to be located on the line segment between adjacent bus nodes. The specific location method is: traverse all combinations of adjacent rows and columns in the fault location coding matrix, and locate the smallest closed area that simultaneously satisfies the characteristics of "row direction value is -1 and column direction value is +1", which is the specific inter-node segment where the fault is located.

[0019] Further, in step S7, the formula for the two-end traveling wave ranging is: ; in, D This represents the distance between the fault point and one of the ranging terminals. L The length of the line between the two ranging ends. v For the speed of travel wave propagation, , These represent the arrival times of the initial traveling waves detected by the two ranging ends; The traveling wave propagation speed v The method for obtaining this is as follows: Calculations are performed using the known length parameters of the line segment and the historical traveling wave transmission delay data recorded by the terminals at both ends of the segment during normal system operation. v = L / t prop ,in t prop This represents the measured time delay of the traveling wave propagating between the two ends of this section.

[0020] Furthermore, in step S7, the following consistency verification method is used to determine whether the initial traveling waves detected by the two ranging ends are the same wavefront: calculate the arrival time difference Δ between the two traveling waves detected at both ends. T = | T L - T R |, if Δ T Satisfy Δ T ≤ L / v If the polarity of the traveling wave amplitude detected at both ends is the same, then it is determined to be the same wavefront; otherwise, wavefront identification is performed again. If multiple terminals exist within the faulty section, the terminals closest to the fault point at both ends of the faulty section should be selected as ranging terminals, or the two terminals with the highest magnetic flux density signal-to-noise ratio should be selected. The signal-to-noise ratio is calculated by using the root mean square value of the signal from the 10 sampling points before the fault occurred as the noise power estimate.P noise The signal power estimate is taken as the peak value of the signal at the five sampling points after the arrival of the traveling wave following the fault. P signal Signal-to-noise ratio (SNR) = 10 × log 10 ( P signal / P noise ).

[0021] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0022] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0023] The beneficial effects of this invention are reflected in: 1. Non-contact installation, no power outage required: The tunnel magnetoresistive sensor achieves non-contact magnetic field induction, eliminating the need for direct electrical connection with high-voltage lines, significantly reducing construction difficulty and safety risks; 2. Strong anti-interference capability and high positioning accuracy: Through dynamic structural element morphological filtering and phase discrimination mechanism, background noise is effectively suppressed, wavefront is accurately identified, and positioning error is less than 100 meters; 3. Adaptable to complex topologies and clear hierarchical identification: Utilizing dual-matrix criteria and hierarchical logic, it uniformly realizes the identification of fault sections in branches, buses, and nodes, breaking through the limitations of the traditional double-end traveling wave method on topology structures; 4. Wide range of applicable operating conditions: It has good adaptability to various fault types, transition resistances and initial phase angles of faults; 5. Low computational complexity and strong real-time performance: The algorithm mainly uses basic arithmetic operations, which meets the timeliness requirements of rapid self-healing of distribution networks. Attached Figure Description

[0024] Figure 1 This is an overall flowchart of the fault location method of the present invention; Figure 2 This is a schematic diagram of the magnetic flux density measurement of a three-phase overhead line according to the present invention; Figure 3 This is a schematic diagram of the three-terminal power distribution network of the present invention; Figure 4 This is a schematic diagram of the multi-terminal power distribution network of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Please see Figure 1 This invention provides a non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines, comprising the following steps: Step S1: Deploy tunnel magnetoresistive sensors at each terminal node, with the sensitive axis of the tunnel magnetoresistive sensor perpendicular to the line direction, to non-contactly collect the spatial alternating magnetic flux density signal generated by the initial fault current traveling wave, and synchronize the clock of each terminal. Specifically, tunnel magnetoresistive sensors are deployed at each terminal node. The sensors are fixed to the tower or crossarm with their sensitive axis perpendicular to the line direction. They utilize the tunnel magnetoresistive effect to non-contactly sense the spatial alternating magnetic flux density generated by the initial fault current traveling wave. The system uses a sampling rate of at least 4MHz for real-time data acquisition and synchronizes the clock of each terminal's acquisition process through a timing module. This sampling parameter satisfies the Nyquist sampling theorem and can completely cover the main spectrum of the fault traveling wave in the distribution line. The timing module uses a BeiDou satellite navigation system or GPS global positioning system receiver unit. Each terminal achieves pulse-per-second (PPS) synchronization by receiving satellite timing signals, with a synchronization accuracy better than 1μs. When a fault traveling wave front is detected, each terminal records the absolute timestamp of the wave front arrival based on its local PPS. The tunnel magnetoresistive sensors are installed below the crossarm or on the side wall of the main tower material, at a height of at least 0.5m and no more than 2.0m above the plane of the three-phase conductors to avoid the sensors entering the near-field saturation region of the conductors while ensuring sufficient magnetic flux density signal amplitude. For nearby conductors such as lightning protection wires and OPGW optical cables in the line, the power frequency magnetic field frequency generated by them is much lower than the frequency band of the fault traveling wave signal (the main spectral components of the traveling wave signal are above 10kHz). It can be suppressed by morphological filtering and high-pass digital filtering in step S3, without the need for special shielding treatment at the sensor installation level.

[0027] Step S2: Establish a three-dimensional spatial coordinate system with the sensor's sensing center as the origin. According to the Biot-Savart law, the three-phase current traveling waves are regarded as independent magnetic field sources. The magnetic flux density vector generated by each phase current at the sensor position is calculated using the vector superposition method. After superposition, the composite magnetic flux density vector is obtained. Then, this vector is projected onto the X-axis and Y-axis directions of the sensor coordinate system to obtain the components of the composite magnetic flux density. The attitude error of the composite magnetic flux density is compensated using a rotation calibration matrix to obtain the corrected magnetic flux density components. Specifically, a three-dimensional spatial coordinate system is established with the sensor's sensing center as the origin. Based on the coordinate positions of each phase conductor in space, the magnetic flux density vector components generated by each phase current at that origin are calculated. The magnetic flux density vector generated by the three-phase current traveling wave at the sensor location can be expressed as:

[0028] in, I k These are the instantaneous values ​​of the current in each phase. d k This represents the straight-line spatial distance from the sensor to each phase conductor. e dk Let be the unit direction vector of each phase conductor of the sensor; project this vector onto the X and Y axes of the sensor coordinate system respectively to obtain the component expressions of the composite magnetic flux density: ; ; For three-dimensional space Z Axial components B z Because the overhead conductors are laid out along the Z-axis, the magnetic field generated by the traveling waves of the three-phase current is... Z The directions cancel each other out, and the theoretical value is always 0; at the same time, the sensitive direction of the sensor used in this scheme is focused on... X , Y Planar, therefore it will not be considered in subsequent calculations and signal processing. B z Quantity.

[0029] in, m 0 is the permeability of free space. I a , I b , I c These are the instantaneous values ​​of the traveling waves of the three-phase currents A, B, and C, respectively. d a , d b , d cThese represent the straight-line spatial distances from the sensor to the three phase conductors A, B, and C, respectively. α , β These are the angles between the horizontal projection and the vertical direction of the connection between phase A and phase C conductors and the sensor, respectively. The above formula is derived based on a three-phase conductor horizontal arrangement scenario. If the line adopts other arrangement forms such as triangular or vertical, it is only necessary to re-substitute the conductor spatial distance and angle parameters. The solution has universal adaptability.

[0030] To address potential attitude deviations during on-site installation, a rotation calibration matrix is ​​introduced to perform projection transformation and error compensation on the synthesized magnetic flux density. The corrected magnetic flux density components... The calculation is as follows: ; in, i A pitch angle deviation is measured for the sensor, which is obtained in real time by an tilt sensor or through manual calibration during on-site installation. This correction eliminates the influence of the sensor's non-ideal installation posture on the measurement results. It should be noted that the sensor is installed with its sensitive axis perpendicular to the line direction; the main posture deviation originates from pitch angle deviation caused by tower tilt or mounting bracket deformation. Therefore, only the pitch angle needs to be corrected. i Calibration is required. Roll angle deviation can be eliminated through horizontal calibration during installation. Yaw angle deviation does not affect the measurement of the magnetic field component perpendicular to the track direction, so it is not considered. As an alternative, if the sensor is a three-axis tunnel magnetoresistive chip, a three-dimensional rotation matrix can also be used. B x , B y , B z Three components are used for full attitude correction.

[0031] Step S3: The corrected magnetic flux density signal is processed using a dynamic structural element morphological filtering algorithm. The length of the structural element of the dynamic structural element filter is adaptively adjusted according to the local variance characteristics of the signal. The traveling wave envelope features are extracted using morphological dilation and erosion operations. By calculating the cross-correlation function between the dilation signal and the erosion signal, the incident wave and the reflected wave are identified according to the time delay direction corresponding to the cross-correlation peak. The abrupt change point of the incident traveling wave is identified as the wavefront arrival time. Specifically, a dynamic structural element morphological filtering algorithm is used to process the corrected magnetic flux density signal. B ( n This algorithm extracts and identifies the initial traveling wave front and timing of the fault. It employs a one-dimensional flat adaptive structuring element. i ( m The amplitude of the structuring element is uniformly set to 0, and the calculation is performed solely based on the signal shape to reduce computational complexity. The length of the structuring element is...M Based on magnetic flux density signal B ( n The local variance characteristics of ) are adaptively determined, and the determination method is as follows:

[0032] in, s B and m B These represent the standard deviation and mean of the magnetic flux density signal within the current analysis window, respectively. l This is a preset scaling factor used to map the relative fluctuation amplitude of the signal to a suitable structural element length dimension. M min and M max These are the preset upper and lower limits for the length of the structural elements. This design allows the filter to automatically increase the filtering scale to smooth noise when the signal fluctuates wildly, and automatically decrease the filtering scale to preserve detailed features when the signal is calm.

[0033] Based on the fundamental dilation and erosion operators of mathematical morphology, morphological filtering and dilation operations are performed on the magnetic flux density signal. B d ( n ) and corrosion operation B e ( n The expressions for ) are as follows: ; ; To distinguish between incident and reflected waves, the cross-correlation function between the expansion and corrosion signals is calculated:

[0034] Take the cross-correlation function Latency corresponding to the first peak t p ;like t p >0 indicates that the expansion signal leads the corrosion signal in phase, and is determined to be an incident traveling wave; if t p <0 indicates that the expansion signal lags behind the corrosion signal in phase, and is determined to be a reflected traveling wave; the abrupt change point of the incident traveling wave is identified as the wavefront arrival time.

[0035] Step S4: Using the line section between each terminal and the adjacent bus node as the basic unit, pre-calibrate the reference magnetic flux density value of each terminal when a unit amplitude traveling wave signal is injected; after a system fault occurs, obtain the initial traveling wave magnetic flux density amplitude measured by each terminal, and construct a fault influence matrix using the ratio of the fault amplitude of each terminal to the corresponding reference value as elements. A ; Specifically, taking the line segment between each terminal and the adjacent bus node as the basic unit, the reference magnetic flux density value when a unit amplitude traveling wave signal is injected into the beginning of each terminal in that segment is pre-calibrated and denoted as . subscript i Indicates the terminal number. J This indicates the corresponding bus node number. When a terminal connects to multiple line segments simultaneously (i.e., the terminal is a T-junction or a bus node), the terminal has multiple corresponding bus node numbers. J 1, J 2, ..., J m Correspondingly, there are multiple reference magnetic flux density values. , ,..., When constructing the fault flux ratio matrix, the elements in the row and column corresponding to the terminal that involve the terminal are taken as the first bus node traversed by the shortest path between the terminal and the corresponding terminal. J The value of is taken. For other unselected reference magnetic flux density values ​​involved in this terminal, they are used as auxiliary judgments in subsequent inter-node fault determination, but are not used as the main basis for matrix construction.

[0036] The reference magnetic flux density The specific calibration method is as follows: before the system is put into operation or during maintenance, for the first... i Each terminal and its corresponding bus node J At the bus node J A pulse current with a unit amplitude (e.g., 1A) and traveling wave characteristics is injected into the line section where the terminal is located. The rise time of the pulse current is no greater than 1μs to simulate the frequency characteristics of an actual fault traveling wave. The amplitude of the synthesized magnetic flux density signal measured at this time is measured and recorded by the tunnel magnetoresistive sensor at the terminal, and this amplitude is calibrated as... The waveform of the pulsed current is a unipolar square wave or a bipolar exponentially decaying wave, with a pulse width not exceeding 2μs, to ensure that its spectrum covers the main frequency components (10kHz~1MHz) of the fault traveling wave. The injection operation is performed at the beginning of the line using a portable pulsed current generator during system outage maintenance, or alternatively, calibration is completed through a field current-carrying test before the new line is put into operation. Alternatively, the calibration process can also be calculated using a high-precision electromagnetic transient simulation model.

[0037] When a system failure occurs, the amplitude of the initial traveling wave flux density measured by each terminal is acquired. B if For those with n A network of individual terminals, building n × n Dimensional Fault Influence Matrix A , its first i Line 1 j Column elements r ij Defined as: ; Among them, diagonal elements r ii Marked as invalid element.

[0038] by Figure 3 The following explanation uses a three-terminal power distribution network as an example. The network includes main terminals M1 and M2, branch terminal Z1, bus node J1, and line sections M1-J1, J1-M2, and J1-Z1. Assume the fault occurs at a point on the branch line J1-Z1 belonging to terminal Z1, and the amplitudes of the initial fault current traveling wave flux density measured at each terminal are as follows: , , The corresponding reference magnetic flux density is calibrated as follows: , , The final matrix form is: ; Where, matrix A The first row corresponds to terminal M1, the second row to terminal M2, and the third row to terminal Z1; the first column corresponds to terminal M1, the second column to terminal M2, and the third column to terminal Z1. The diagonal elements are marked with " / ", indicating that the ratio of a terminal to itself is meaningless.

[0039] Step S5: Calculate the fault impact matrix. A A sign decision function is applied to each element in the matrix. This sign decision function sets a tolerance range with the value 1 as a reference point, mapping each element to three encoding states: "+1", "0", and "-1", thus forming a fault location encoding matrix. C ; Specifically, define the sign decision function. as follows: ; in, d The preset error tolerance threshold is used to overcome numerical fluctuations caused by field measurement noise and sensor inconsistencies. d The value ranges from 0.01 to 0.05; the fault impact matrix... A Each element in Substituting the symbols into the symbol decision function, we obtain the fault location coding matrix. C Its elements ,Right now Diagonal elements are marked as invalid.

[0040] To facilitate a clear understanding of the correspondence between the value patterns of the matrix elements and the discrimination logic, a typical numerical case satisfying the ratio relationship defined in step S4 is presented here. (Continuing with step S4...) Figure 3 Taking the three-terminal network shown as an example, in this case, assuming the fault occurs on the branch line to which Z1 belongs (i.e., the J1-Z1 section), the fault impact matrix is... A The values ​​of each element and their relative sizes are as follows: ; Taking δ=0.02, a sign decision function is applied to each element to obtain the corresponding fault location coding matrix. C : .

[0041] Step S6: Based on the fault location coding matrix C Count the non-diagonal elements by row. The ratio of "1" to "+1" is determined by detecting the complementary and symmetrical relationship between row and column codes, and fault segment identification is performed according to hierarchical logic. The hierarchical logic includes branch fault identification, bus node fault identification, and inter-node fault identification. Specifically, based on the aforementioned fault location coding matrix C Fault section identification is performed according to the following hierarchical logic: 1. Branch fault identification: encoding the fault location matrix C Each line in k Calculate the percentage of off-diagonal elements that have the value "-1". The percentage of values ​​with a sum of "+1" If a terminal exists k satisfy If the value is 1, and the percentage of off-diagonal elements with a value of "+1" in the corresponding column of this terminal is also 1, then the fault is determined to occur in this terminal. k On the connected branch lines; using the matrix constructed in step S5 C For example, observe that the off-diagonal elements in the third row (corresponding to terminal Z1) are all "-1" and the off-diagonal elements in the third column (corresponding to terminal Z1) are all "+1". This satisfies the above discrimination conditions, so it is determined that the fault occurs on the branch line J1-Z1 connected to terminal Z1. 2. Busbar node fault detection: If the fault location coding matrix C If all off-diagonal elements are 0, then the fault is located at the bus node. 3. Fault Identification Between Nodes: When the characteristics of branch faults and bus node faults mentioned above are not met, the fault is determined to be located on the line segment between adjacent bus nodes. The specific location method is as follows: traverse all combinations of adjacent rows and columns in the fault location coding matrix, and locate the smallest closed region that simultaneously satisfies the characteristic of "row direction value is -1 and column direction value is +1", which is the specific inter-node segment where the fault is located.

[0042] Step S7: After determining the fault section, select two terminals located at both ends of the fault section or close to the fault point as ranging ends. Extract the arrival time of the initial traveling wave detected by each of the two ranging ends. Combine the line length between the two ranging ends and the traveling wave propagation speed, and use the double-end traveling wave ranging formula to calculate the location of the fault point.

[0043] Specifically, after determining the faulty section, two terminals located at either end of the faulty section or close to the fault point are selected as ranging terminals, and the arrival times of the initial traveling waves detected by each of the two ranging terminals are extracted. T L , T R Combined with the line length between the two ranging ends L and the speed of travel wave propagation v The distance between the fault point and one of the ranging ends is calculated using the two-end traveling wave ranging formula. D : ； The traveling wave propagation speed v The method for obtaining this is as follows: Calculations are performed using the known length parameters of the line segment and the historical traveling wave transmission delay data recorded by the terminals at both ends of the segment during normal system operation. v = L / t prop ,in t prop This represents the measured time delay of the traveling wave propagating between the two ends of this section; alternatively, theoretically calculated values ​​of the line parameters can also be used. ,in L and C These are the inductance and capacitance parameters per unit length of the line, respectively. Traveling wave propagation speed. v Before positioning, the data is pre-calculated and stored in the computing units of each terminal. To ensure that the initial traveling waves detected by the two ranging terminals originate from the same fault point and have the same wavefront, the following consistency verification method is used: calculate the arrival time difference Δ between the traveling waves detected at both ends. T= | T L - T R |, if Δ T Satisfy Δ T ≤ L / v (That is, the arrival time difference between the two ends does not exceed the full propagation time of the traveling wave on the line segment), and the polarity of the traveling wave amplitude detected at both ends is the same, then it is determined to be the same wavefront; otherwise, wavefront identification is performed again. Preferably, if there are multiple terminals in the fault segment, the terminals closest to the fault point at both ends of the fault segment are selected as ranging terminals; if there are multiple terminals in the fault segment, the two terminals with the highest magnetic flux density signal-to-noise ratio are selected. The signal-to-noise ratio is calculated as follows: the root mean square value of the signal at the 10 sampling points before the fault occurs is used as the noise power estimate. P noise The signal power estimate is taken as the peak value of the signal at the five sampling points after the arrival of the traveling wave following the fault. P signal Signal-to-noise ratio (SNR) = 10 × log 10 ( P signal / P noise ).

[0044] The following are specific examples: Simulation Model Construction: A 10kV, 50Hz multi-terminal overhead distribution network model was built based on the electromagnetic transient simulation platform PSCAD / EMTDC. This scheme can also be adapted to overhead lines with higher voltage levels. The model includes multiple power supply terminals (main terminals), multiple branch terminals, and multiple sets of intermediate bus nodes, covering branches and trunk lines of different lengths. Frequency-dependent (phase) models are used for each line section to accurately simulate the propagation characteristics of traveling waves. Tunnel magnetoresistive sensor models are set at each terminal node, with the sensors arranged in a predetermined orientation and the sampling frequency set to 4MHz. The simulation conditions cover various fault types (single-phase ground fault LG, phase-to-phase short circuit LL, two-phase ground fault LLG, three-phase short circuit LLLG), transition resistance (0~200Ω), fault initial phase angle (0°~120°), and noise level (35dB signal-to-noise ratio).

[0045] Non-contact magnetic flux density measurement and correction: A fault point is set up within the simulated circuit, and the traveling wave current excited by the fault generates an alternating magnetic field in space. Tunnel magnetoresistive sensors at each terminal collect the magnetic field signal. Following the method described in step S2, the resultant magnetic flux density vector generated by the three-phase currents at the sensor is first calculated in vector superposition form, and then projected onto each axis of the sensor coordinate system to obtain... B x and B yThe components are then attitude-compensated by a rotation matrix. The spatial relationship between each sensor and the three-phase conductors is as follows: Figure 2 As shown.

[0046] Dynamic Structural Element Morphological Filtering and Wavefront Extraction: The magnetic flux density signals acquired and corrected by each terminal are input into the dynamic structural element morphological filtering module. Following the method described in step S3, the algorithm adaptively determines the length of the structural element based on the signal's local variance characteristics and performs dilation and erosion operations on the signal. By calculating the cross-correlation function between the dilated and eroded signals, the initial incident traveling wavefront is identified based on the peak time delay direction, and its absolute arrival time at each terminal is recorded.

[0047] Reference flux density calibration and matrix construction: In the simulation model, the reference flux density is calibrated for each terminal according to the method described in step S4. Specifically, a unit pulse traveling wave signal with an amplitude of 1A is injected at the beginning of the line segment connected to each terminal (near the bus node), and the amplitude of the synthesized flux density measured at that terminal is recorded as the reference value. subscript i Indicates the terminal number. J This indicates the corresponding bus node number.

[0048] by Figure 3 The following explanation uses a three-terminal power distribution network as an example. The network includes main terminals M1 and M2, branch terminal Z1, bus node J1, and line sections M1-J1, J1-M2, and J1-Z1. The pre-calibrated reference magnetic flux densities of each terminal are denoted as follows: , , .

[0049] Assuming the fault occurs at a point on branch line J1-Z1 to which terminal Z1 belongs, the magnetic flux density amplitudes generated by the initial fault current traveling wave measured by the sensors at each terminal are as follows: , , Construct the fault impact matrix according to step S4. A (Example values): ; Apply a sign decision function to each element of the matrix (take...) d =0.02), thus obtaining the fault location coding matrix. C : ; Observation Matrix C The off-diagonal elements in the 3rd row (corresponding to terminal Z1) are all -1, and the off-diagonal elements in the 3rd column are all +1. =1 and the corresponding column "+1" accounts for 1. According to the judgment logic of step S6 (1), the fault is determined to occur on the branch line J1-Z1 connected to Z1.

[0050] for Figure 4 The complex network shown includes more terminals (main terminals M1-M2, branch terminals Z1-Z6, and bus nodes J1-J6), and the above matrix is ​​extended to... n × n dimension( n (Total number of terminals), by applying the hierarchical discrimination logic described in step S6 row by row and column by column, the faulty segment can be identified. The matrix expansion rule is: the first... i Line 1 j Column elements represent terminals i The amplitude of the fault magnetic flux density and the terminal j The ratio of the reference magnetic flux density value; the physical meaning and symbol marking rules of each element are completely consistent with those of the three-terminal network case.

[0051] Precise fault location: After determining the faulty section, select the terminals at both ends of the section as ranging ends, extract the arrival times of the traveling wave obtained in step S3, and calculate the fault location using the double-ended traveling wave formula described in step S7. Traveling wave propagation speed. v The data is obtained in advance through actual measurement or theoretical calculation on this line section and stored in the computing unit of each terminal for use during positioning.

[0052] Simulation Results and Data Analysis: Tests were conducted under different fault distances, fault types, impedances, initial phase angles, and noise conditions. The results are as follows: Table 1. Fault distance 10km (J1-J2 section)

[0053] Table 2 Fault distance 22km (J3 node)

[0054] Table 3. Fault distance 35km (J5-Z5 section)

[0055] Table 4. Fault distance 60km (J6-Z6 section)

[0056] Note: All simulation results in Tables 1-4 were measured under a signal-to-noise ratio of 35dB.

[0057] Test results show that the fault section identification accuracy is 100%; the positioning error is less than 100 meters under all operating conditions; the fault impedance, initial phase angle, and noise have little impact on accuracy and strong stability; the method is applicable to complex multi-terminal and multi-branch power distribution networks and has engineering promotion value.

[0058] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0059] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0060] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to perform any of the methods described in the above embodiments.

[0061] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0062] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0063] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0064] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0065] 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 foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines, characterized in that, Includes the following steps: Step S1: Deploy tunnel magnetoresistive sensors at each terminal node, with the sensitive axis of the tunnel magnetoresistive sensor perpendicular to the line direction, to non-contactly collect the spatial alternating magnetic flux density signal generated by the initial fault current traveling wave, and synchronize the clock of each terminal. Step S2: Establish a three-dimensional spatial coordinate system with the sensor's sensitive center as the origin. Based on the Biot-Savart law, use the vector superposition method to calculate the projection components of the combined magnetic flux density generated by the three-phase current traveling wave at the sensor position on each coordinate axis. Perform spatial projection transformation and attitude error compensation on the obtained components to obtain the corrected magnetic flux density components. Step S3: The corrected magnetic flux density signal is processed using a dynamic structural element morphological filtering algorithm. The length of the structural element of the filter is adaptively determined according to the local time-varying characteristics of the signal. The traveling wave envelope features are extracted using morphological dilation and erosion operations. The incident wave and the reflected wave are identified based on the cross-correlation function characteristics of the dilation signal and the erosion signal at the abrupt change point. The abrupt change point of the incident traveling wave is identified as the wavefront arrival time. Step S4: Using the line section between each terminal and the adjacent bus node as the basic unit, pre-calibrate the reference magnetic flux density value of each terminal when a unit amplitude traveling wave signal is injected; after a system fault occurs, obtain the initial traveling wave magnetic flux density amplitude measured by each terminal, and construct a fault influence matrix by comparing the fault amplitude of each terminal with the corresponding reference value. A ; Step S5: Calculate the fault impact matrix. A A sign decision function is applied to each element in the matrix, which maps each element to three states: "-1", "0", and "+1", forming a fault location coding matrix C. Step S6: Based on the fault location coding matrix C For each row, its encoding sequence is calculated. By detecting the complementary symmetry between row encoding and column encoding, fault segment discrimination is performed according to hierarchical logic. The hierarchical logic includes branch fault discrimination, bus node fault discrimination and inter-node fault discrimination. Step S7: After determining the fault section, select two terminals located at both ends of the fault section or close to the fault point as ranging ends. Extract the arrival time of the initial traveling wave detected by each of the two ranging ends. Combine the line length between the two ranging ends and the traveling wave propagation speed, and use the double-end traveling wave ranging formula to calculate the location of the fault point.

2. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S1, the tunnel magnetoresistive sensor is installed below the crossarm or on the side wall of the main tower material, with an installation height of not less than 0.5m and not more than 2.0m from the plane where the three-phase conductors are located. The system uses a sampling rate of not less than 4MHz for real-time data acquisition and realizes clock synchronization of the acquisition process of each terminal through a timing module. The timing module uses a Beidou satellite navigation system or GPS global positioning system receiving unit. Each terminal realizes second pulse synchronization by receiving satellite timing signals, with a synchronization accuracy better than 1μs.

3. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S2, according to the Biot-Savart law, the magnetic flux density vector generated by the three-phase current traveling wave at the sensor location is expressed as: in, I k These are the instantaneous values ​​of the current in each phase. d k This represents the straight-line spatial distance from the sensor to each phase conductor. e dk Let be the unit direction vector of each phase conductor of the sensor; project this vector onto the X and Y axes of the sensor coordinate system respectively to obtain the component expressions of the composite magnetic flux density: ； ； in, μ 0 is the permeability of free space. , , These are the instantaneous values ​​of the traveling waves of the three-phase currents A, B, and C, respectively. , , α and β are the straight-line spatial distances from the sensor to the three-phase conductors A, B, and C, respectively, and α and β are the angles between the horizontal projection and the vertical direction of the line connecting the A-phase and C-phase conductors to the sensor, respectively. The spatial projection transformation and attitude error compensation are achieved through the following rotation transformation: ； in, θ A pitch angle deviation is installed on the sensor, which is obtained by real-time measurement using a tilt sensor or by manual calibration during on-site installation.

4. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S3, the dynamic structuring element morphological filtering algorithm employs a one-dimensional flat adaptive structuring element. i ( m The structuring element amplitude is set to 0, and the structuring element length is... M Based on magnetic flux density signal B ( n The local variance characteristics of ) are adaptively determined, and the determination method is as follows: in, σ B and μ B These represent the standard deviation and mean of the magnetic flux density signal within the current analysis window, respectively. λ This is a preset proportional coefficient. M min and M max These are the preset upper and lower limits for the length of the structuring element.

5. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S3, the expansion operation is performed. Corrosion Operation The expressions are as follows: ； ； To distinguish between incident and reflected waves, the cross-correlation function between the expansion and corrosion signals is calculated: Take the cross-correlation function Latency corresponding to the first peak τ p ;like τ p >0 indicates that the expansion signal leads the corrosion signal in phase, and is determined to be an incident traveling wave; if τ p <0 indicates that the expansion signal lags behind the corrosion signal in phase, and is determined to be a reflected traveling wave; the abrupt change point of the incident traveling wave is identified as the wavefront arrival time.

6. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S4, the reference magnetic flux density The specific calibration method is as follows: at the bus node J Towards the first i A unit amplitude pulse current with traveling wave characteristics is injected into the line section where each terminal is located. The amplitude of the synthesized magnetic flux density signal measured at this time is recorded by the tunnel magnetoresistive sensor at the terminal, and this amplitude is calibrated as... The rise time of the pulse current is no greater than 1 μs, and the pulse width is no greater than 2 μs. The fault impact matrix A for n × n 3D matrix n For the total number of terminals, its number is... i Line 1 j Column elements r ij Defined as: ； in, For the first i The amplitude of the initial traveling wave flux density measured at each terminal during the fault. For the first j The reference magnetic flux density value of each terminal in the corresponding line segment, with diagonal elements marked as invalid elements.

7. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S5, the symbol decision function Defined as: in, δ The preset error tolerance threshold, the error tolerance threshold δ The value range is from 0.01 to 0.05; The fault location coding matrix C elements in ,Right now Diagonal elements are marked as invalid.

8. The non-contact current traveling wave fault location method for multi-terminal overhead distribution lines according to claim 1, characterized in that, In step S6, the hierarchical logic specifically includes: (1) Branch fault identification: encoding the fault location matrix C Each line in k Calculate the percentage of off-diagonal elements that have the value "-1". The percentage of values ​​with a sum of "+1" If a terminal exists k satisfy If the value is 1, and the percentage of off-diagonal elements with a value of "+1" in the corresponding column of this terminal is also 1, then the fault is determined to occur in this terminal. k On the connected branch lines; (2) Bus node fault identification: If the fault location coding matrix C If all off-diagonal elements are 0, then the fault is located at the bus node. (3) Fault identification between nodes: When the above-mentioned branch fault characteristics and bus node fault characteristics are not met, the fault is determined to be located on the line section between adjacent bus nodes. The specific location method is to traverse all combinations of adjacent rows and columns in the fault location coding matrix and locate the smallest closed area that simultaneously satisfies the characteristics of "row direction value is -1 and column direction value is +1".

9. The non-contact current traveling wave fault location method for multi-terminal overhead power distribution lines according to claim 1, characterized in that, In step S7, the double-ended traveling wave ranging formula is: ； in, D This represents the distance between the fault point and one of the ranging terminals. L The distance between the two ranging ends is denoted as , and v is the traveling wave propagation speed. , These represent the arrival times of the initial traveling waves detected by the two ranging ends; The traveling wave propagation speed v The method for obtaining this is as follows: Calculations are performed using the known length parameters of the line segment and the historical traveling wave transmission delay data recorded by the terminals at both ends of the segment during normal system operation. v = L / t prop ,in t prop This represents the measured time delay of the traveling wave propagating between the two ends of this section.

10. The non-contact current traveling wave fault location method for multi-terminal overhead distribution lines according to claim 1, characterized in that, In step S7, the following consistency check method is used to determine whether the initial traveling waves detected by the two ranging ends are the same wavefront: calculate the arrival time difference Δ of the traveling waves detected at both ends. T = | T L - T R |, if Δ T Satisfy Δ T ≤ L / v If the polarity of the traveling wave amplitude detected at both ends is the same, then it is determined to be the same wavefront; otherwise, wavefront identification is performed again. If multiple terminals exist within the faulty section, the terminals closest to the fault point at both ends of the faulty section should be selected as ranging terminals, or the two terminals with the highest magnetic flux density signal-to-noise ratio should be selected. The signal-to-noise ratio is calculated by using the root mean square value of the signal from the 10 sampling points before the fault occurred as the noise power estimate. P noise The signal power estimate is taken as the peak value of the signal at the five sampling points after the arrival of the traveling wave following the fault. P signal Signal-to-noise ratio (SNR) = 10 × log 10 ( P signal / P noise ).

Images