A mine power supply fault diagnosis positioning method and system

Abstract

This application relates to the field of mine power supply technology and discloses a method and system for fault diagnosis and location in mine power supply. The method includes: collecting the open / closed state sequence of each circuit breaker and identifying topology change events through jumper identification; incrementally updating the equivalent impedance of affected nodes based on the topology change events to obtain impedance increments; substituting the impedance increments into the setting formula and iteratively verifying the coordination to obtain a network-wide setting value vector, which is then archived to obtain a topology time archive; when a fault occurs, retrieving the topology time archive to construct a node admittance matrix, calculating the theoretical value of the short-circuit current, and outputting the fault location result after normalized matching error calculation. This application improves the accuracy of relay protection coordination and the precision of fault section location in power supply systems under dynamic topology evolution scenarios in mines.

Description

Technical Field

[0001] This application relates to the field of mine power supply technology, and in particular to a method and system for diagnosing and locating mine power supply faults. Background Technology

[0002] The underground power supply system in the mine adopts a multi-level radial network structure, with power supplied sequentially from the surface substation to the substations in each mining area via main shaft cables. Substations at each level are interconnected through high-voltage switchgear and feeder lines. The existing mine power supply system's relay protection devices are set based on the three-stage current protection principle. The instantaneous overcurrent setting and overcurrent delay setting values ​​for each level of protection device are manually issued by professionals after offline calculation based on the system's rated short-circuit current level. The operating power supply for each level of integrated protection device is AC, drawn from the secondary side of the voltage transformer. Monitoring and dispatching of the entire power supply system rely on data exchange between the ground-based control center and each substation via a fixed communication network.

[0003] However, the aforementioned existing technologies have the following shortcomings in actual mining applications: As underground mining faces extend deeper with production, new substations, extended feeder lines, or switched tie lines are required every few weeks to months, leading to frequent changes in the power supply network topology. After each topology change, it takes 3 to 14 days for professionals to recalculate and distribute the protection settings to each integrated protection device. During this setting lag, the system operates in a mismatched state between the old settings and the new topology, and the coordination between the instantaneous overcurrent setting values ​​of each level of protection device and the actual short-circuit current level has changed. Misalignment can easily trigger cascading trips, causing large-scale power outages. Furthermore, the underground AC operating power supply becomes unstable due to a sudden drop in bus voltage when a short-circuit fault occurs, resulting in the relay protection device being in an unstable power supply state precisely at the fault moment when reliable operation is most needed, significantly increasing the risk of failure to operate. Existing fault location methods use the current system topology as the basis for calculations. When the topology has changed but the settings have not been updated, the topology used for calculations is inconsistent with the actual topology of the system at the time of the fault, leading to deviations in the theoretical value of the short-circuit current calculation and a significant increase in fault location errors. Summary of the Invention

[0004] This application provides a method and system for diagnosing and locating power supply faults in mines. It solves the problems of delayed protection setting, high risk of cascading tripping, and large location errors caused by inconsistencies between the calculated and actual topologies of underground power supply systems in mines due to frequent topology changes. It improves the accuracy of relay protection coordination and the precision of fault section location in dynamic topology evolution scenarios in mines.

[0005] Firstly, this application provides a method for diagnosing and locating power supply faults in mines, the method comprising: Step S1: Collect the open / closed state sequence of each circuit breaker in the integrated protection device's input circuit, and perform jump identification processing on the state sequence based on the majority voting rule to obtain the topology change event carrying the node number and timestamp; Step S2: Triggered by the topology change event, the equivalent impedance of each node in the set of affected nodes with a hop distance of no more than two hops is incrementally updated according to the feeder parameters of the hop node to obtain the impedance increment of each affected node. Step S3: Substitute the impedance increment into the instantaneous overcurrent setting formula, and obtain the network-wide setting value vector through the cooperation and verification iteration between adjacent nodes. Then, archive the setting value vector together with the current network-wide switch state matrix and node impedance matrix to obtain a topology time file with timestamps. Step S4: When a fault occurs, the topology time archive is retrieved using the fault timestamp as the query key. The network switch state matrix and node impedance matrix closest to the fault time are extracted. The node admittance matrix is ​​constructed and the theoretical value of the short-circuit current of each candidate section is calculated. The theoretical value is normalized and matched with the measured fault current of each node to calculate the error. The feeder section with the smallest error is output as the fault location result.

[0006] Secondly, this application provides a mine power supply fault diagnosis and location system, the mine power supply fault diagnosis and location system comprising: The jump module is used to collect the open and closed state sequence of each circuit breaker in the integrated protection device's input circuit, and perform jump identification processing on the state sequence based on the majority voting rule to obtain the topology change event carrying the node number and timestamp. The update module is used to incrementally update the equivalent impedance of each node in the set of affected nodes whose hop distance does not exceed two hops, based on the feeder parameters of the hop node and triggered by the topology change event, so as to obtain the impedance increment of each affected node. The archiving module is used to substitute the impedance increment into the instantaneous overcurrent setting formula, and obtain the network setting value vector through the cooperation and verification iteration between adjacent nodes. The setting value vector is then archived together with the current network switch state matrix and node impedance matrix to obtain a topology time archive with timestamps. The calculation module is used to retrieve the topology time archive by using the fault timestamp as the query key when a fault occurs, extract the network-wide switch state matrix and node impedance matrix that are closest to the fault time, construct the node admittance matrix and calculate the theoretical value of the short-circuit current of each candidate section, perform normalization matching error calculation between the theoretical value and the measured fault current of each node, and output the feeder section with the smallest error as the fault location result.

[0007] Thirdly, a mine power supply fault diagnosis and location device is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the mine power supply fault diagnosis and location device to execute the aforementioned mine power supply fault diagnosis and location method.

[0008] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed on a computer, cause the computer to perform the aforementioned method for diagnosing and locating power supply faults in mines.

[0009] The technical solution provided in this application introduces a switch-state sequence jump identification mechanism based on majority voting rules into the integrated protection device's input circuit. This directly maps each switch-state operation of the underground mine circuit breaker to a topology change event carrying a node number and a timestamp. Utilizing the physical constraint that topology changes in the mine power supply system must be achieved through switching operations, zero-delay perception of topology changes is achieved without adding any additional sensors, fundamentally eliminating the lag window between manual resetting operations and topology changes in existing technologies. Triggered by this topology change event, the equivalent impedance of each node within the affected node set (within a jump distance of no more than two jumps) is incrementally updated based on the jump node feeder parameters. This narrows the data exchange range for resetting recalculation to impedance increments rather than the entire network impedance data, and is combined with verification of the instantaneous overcurrent setting values ​​between adjacent nodes. The iterative protocol enables each integrated protection device to independently calculate its setting value and perform cross-sectional verification locally. Even when a single node loses communication, the remaining nodes can still independently complete the setting, breaking through the strong dependence of existing centralized setting systems on the complete communication network. The final effective network-wide setting value vector, the current network-wide switch state matrix, and the node impedance matrix are jointly archived using timestamps as indices to form a topology time archive. This establishes a precise temporal correspondence between setting values ​​and topology states, allowing the retrieval of historical snapshots that are completely consistent with the actual topology at the time of the fault using the fault timestamp as the query key when a fault occurs. Under this deterministic topology, the node admittance matrix is ​​constructed and the theoretical value of the short-circuit current of each candidate segment is calculated. Then, the normalized matching error is calculated by comparing it with the measured fault current of each node, completely eliminating the interference of the inconsistency between the calculated topology and the actual topology on the fault location accuracy in existing technologies.

[0010] The aforementioned technical features work together to address the specific application scenario of dynamic topology evolution in mines. The introduction of the majority voting jump identification algorithm makes topology sensing robust against contact jitter. The impedance increment switching protocol compresses the communication overhead of distributed computing to the minimum necessary amount. The normalized matching error algorithm uses the sum of the theoretical values ​​of the two end nodes as the normalization benchmark to eliminate the influence of the difference in current dimensions between nodes with different power supply capacities on error comparison. The historical snapshot retrieval mechanism of the topology time archive improves the positioning accuracy of the algorithm from relying on the current topology assumption to relying on historical deterministic topology facts. Each algorithm feature makes an irreplaceable technical contribution to the specific scenario of frequent topology changes in mine power supply systems, rather than a simple transplantation and application of general power system protection algorithms. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic diagram of one embodiment of the mine power supply fault diagnosis and location method in this application; Figure 2 This is a schematic diagram of the power supply network node impedance matrix in an embodiment of this application; Figure 3 This is a schematic diagram of the simulated waveforms of the three-phase current before and after a three-phase short-circuit fault in the underground power supply system of the mine, as described in this application. Detailed Implementation

[0013] This application provides a method and system for diagnosing and locating power supply faults in mines. The terms "first," "second," "third," "fourth," etc. (if present)," in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0014] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1One embodiment of the mine power supply fault diagnosis and location method in this application includes: Step S1: Collect the open / closed state sequence of each circuit breaker in the integrated protection device's input circuit, and perform jump identification processing on the state sequence based on the majority voting rule to obtain the topology change event carrying the node number and timestamp; Specifically, a topology change event refers to a structured record generated after the integrated protection device detects a change in the circuit breaker's open / closed state in its input circuit. Its components include node number, UTC millisecond timestamp, operation type, and the MD5 hash value of the adjacent node's switch state before the operation. The MD5 hash value is used to verify the integrity of the topology state before the operation when retrieving the topology time file in step S4. The majority voting rule adopts a criterion that at least two out of three consecutive sampling points support the same change direction. This criterion corresponds to a 60ms debounce window with a 20ms sampling period, which can eliminate misjudgments caused by circuit breaker contact jitter without affecting the recognition timeliness of normal opening and closing operations.

[0015] Step S2: Triggered by a topology change event, the equivalent impedance of each node in the set of affected nodes with a hop distance of no more than two hops is incrementally updated based on the feeder parameters of the hop node to obtain the impedance increment of each affected node. Specifically, the affected node set refers to the set of all nodes whose hop distance from the topology-changing node is no more than two hops. The boundary of two hops is set based on the following: due to the attenuation of line impedance, the short-circuit current level of nodes exceeding two hops is affected by the current topology change by no more than 3% of the rated short-circuit current, which is negligible within the engineering setting accuracy range. This boundary reduces the calculation range while ensuring that the setting accuracy is not affected. The impedance increment refers to the difference between the new and old equivalent impedance values ​​of each affected node, stored in the form of a binary tuple of the difference in resistive component and the difference in inductive reactance component, and is the direct input for the setting value calculation in step S3.

[0016] Step S3: Substitute the impedance increment into the instantaneous overcurrent setting formula, and obtain the network setting value vector through the cooperation and verification iteration between adjacent nodes. Then, archive the setting value vector together with the current network switch state matrix and node impedance matrix to obtain the topology time file with timestamp. Specifically, the network-wide setting value vector refers to the vector formed by arranging the instantaneous overcurrent setting value, overcurrent setting value, and overcurrent delay setting value of all nodes in the network in order of node number. The reliability coefficient is set to 1.3, based on the provisions of the Coal Mine Safety Regulations regarding the reliability coefficient of instantaneous overcurrent protection of underground high-voltage protection devices. The upper limit of the coordination verification iteration is set to 5 times, which is an engineering value for controlling the setting response time under the premise of ensuring the convergence of coordination relationships. The topology time archive refers to a complete archived record indexed by timestamps, containing the network-wide setting value vector, the network-wide switch state matrix, the node impedance matrix, and SHA-256 hash values. The SHA-256 hash values ​​are used to verify that the archived data has not been tampered with or damaged during step S4 retrieval.

[0017] Step S4: When a fault occurs, the topology time archive is retrieved using the fault timestamp as the query key. The network switch state matrix and node impedance matrix closest to the fault time are extracted. The node admittance matrix is ​​constructed and the theoretical value of the short-circuit current of each candidate section is calculated. The theoretical value is normalized and matched with the measured fault current of each node to calculate the error. The feeder section with the smallest error is output as the fault location result. Specifically, the topology time file closest to the fault time refers to the record with the largest timestamp among all records whose timestamp is not greater than the fault timestamp. This retrieval rule ensures that the extracted topology state is the actual operating topology of the system at the time of the fault, not the topology state after the fault. The normalized matching error is calculated by dividing the sum of the absolute differences between the measured fault current and the theoretical short-circuit current at both ends of each candidate feeder section by the sum of the theoretical short-circuit current values ​​at both ends. The graded judgment standard of 0.15 corresponding to unique location output and 0.30 corresponding to candidate section output is set based on the comprehensive engineering statistics of current transformer measurement error and line parameter error in the 10kV power supply system of the mine.

[0018] In one specific embodiment, step S1 includes: With a sampling period of 20ms, the opening and closing status of each circuit breaker in the integrated protection device input circuit is collected. The closing status is recorded as 1 and the opening status is recorded as 0. 600 consecutive sampling points are stored in the state sequence buffer area corresponding to each circuit breaker to obtain the opening and closing status sequence of each circuit breaker. Based on the majority voting rule, the transition identification process for three consecutive sampling points in the split-combination state sequence is performed using the following formula: ; in, For the first The state value of each sampling point = +1 corresponds to the closing trip. = -1 corresponds to tripping the circuit breaker. = 0 indicates no transition. The above formula is executed one by one for each sampling point in the separation and combination state sequence to obtain the transition determination result of each sampling point. In the transition determination result The node number corresponding to the sampling point, the UTC millisecond timestamp of the transition time, the operation type, and the MD5 hash value of the switch status of each adjacent node before the operation are written into the topology change event record table to obtain the topology change event carrying the node number and timestamp. The topology change events are broadcast to all integrated protection devices across the network via UDP multicast, resulting in the topology change events to be processed in the local topology change queue of each integrated protection device.

[0019] Specifically, the sampling period is set to 20ms, corresponding to one power frequency cycle of a 50Hz power grid. This sampling frequency can completely capture the state changes during a single opening and closing operation of the circuit breaker, while avoiding waste of storage resources due to excessively high sampling frequencies. The state sequence buffer stores 600 sampling points corresponding to a 12-second historical state window. This window length covers the maximum delay time from receiving the protection action command to the circuit breaker completing the opening and closing operation of the underground mine protection device, ensuring that any normal operation falls within the buffer window. The majority voting rule adopts the judgment criterion that at least two out of three consecutive sampling points support the same transition direction, corresponding to a 60ms debouncing window. This window length is set based on engineering statistics showing that the duration of contact jitter in underground mine circuit breakers does not exceed 40ms, which can eliminate misjudgments caused by contact jitter without affecting the identification of normal operations. In the transition judgment result... The sampling point is the moment when the state changes effectively, corresponding to a complete circuit breaker opening or closing operation.

[0020] In the topology change event log table, the node number adopts a two-level combination format of substation number and feeder number to uniquely identify each electrical node in the underground power supply network of the mine; the UTC millisecond timestamp is provided by the GPS timing module with millisecond accuracy, and is used in step S4 to establish a precise time sequence correspondence between the fault time and the historical topology status when retrieving the topology time file using the fault timestamp as the query key; the MD5 hash value of the switch status of each adjacent node before the operation is calculated by combining the switch status of all the current adjacent nodes of this node, and is used to verify the integrity of the topology status data before the operation when retrieving the topology time file in step S4, to prevent data corruption during the storage process from causing distortion of the topology status based on the positioning calculation; the topology change event is broadcast in UDP multicast mode. UDP multicast does not require the establishment of point-to-point connection and can reach all integrated protection devices in the entire network with a single transmission. In the industrial Ethernet environment of the mine, it can ensure that all integrated protection devices in the entire network complete the reception within 100ms, meeting the real-time requirements of topology change notification.

[0021] In one specific embodiment, step S2 includes: Based on the node number in the topology change event, and according to the power supply network adjacency table pre-stored by each integrated protection device, all nodes whose hop distance to the hop node corresponding to the node number does not exceed two hops are screened to obtain the set of affected nodes.

[0022] Using the old equivalent impedance values ​​of each node in the affected node set and the positive sequence impedance of the newly added feeder at the switching node as input, the unit resistance and unit inductive reactance of the newly added feeder are read from the line parameter database. The resistance and inductive reactance components of the positive sequence impedance of the newly added feeder are obtained by multiplying them by the line length. Thevenin parallel update processing is performed on the equivalent resistance values ​​of each node in the affected node set and the positive sequence impedance of the newly added feeder to obtain the new equivalent impedance values ​​of each affected node. The impedance increment of each affected node is obtained by calculating the difference between the resistance and inductive reactance components of the new equivalent impedance and the resistance and inductive reactance components of the corresponding old equivalent impedance. Each impedance increment, along with its corresponding node number and CRC32 checksum, is encapsulated into an impedance increment message and sent via unicast to the corresponding node in the affected node set.

[0023] Specifically, the power supply network adjacency table is pre-stored in the non-volatile memory of each integrated protection device in a sparse matrix format. It records the feeder impedance values ​​of the current node and all one-hop adjacent nodes and two-hop adjacent nodes. The impedance values ​​are stored in the form of a binary pair of resistance and inductive reactance components, in ohms, with a precision of 4 decimal places. The boundary between the number of hops and the two-hop is based on the fact that the change in short-circuit current of nodes outside this range affected by the current topology change does not exceed 3% of the rated short-circuit current. This is negligible within the engineering setting accuracy range of the 10kV power supply system in the mine. This boundary ensures that the setting accuracy is not affected while reducing the calculation range of the set of affected nodes. Thevenin parallel update processing refers to calculating the original equivalent impedance of each affected node and the newly added positive sequence impedance of the feeder according to the parallel impedance formula, that is, the product of the two impedances divided by the sum of the two impedances. The result is the new equivalent impedance value. The resistance component of the newly added positive sequence impedance of the feeder is obtained by multiplying the unit resistance read from the line parameter database by the line length, and the inductive reactance component is obtained by multiplying the unit inductive reactance by the line length.

[0024] The impedance increment is represented by the difference between the new equivalent impedance value and the old equivalent impedance value. Specifically, it is a binary tuple consisting of the difference in resistive components and the difference in inductive reactance components. This binary tuple, together with the corresponding node number and CRC32 checksum, is encapsulated into an impedance increment message. The CRC32 checksum is calculated from all bytes of the message and is used by the receiving node to verify that the message has not been corrupted during transmission before performing the setting calculation in step S3. The impedance increment message is sent using unicast instead of multicast because the impedance increment received by each affected node differs due to the different node locations. Multicast cannot achieve differentiated distribution based on node location, while unicast ensures that each affected node only receives the impedance increment corresponding to its own node, avoiding data confusion between nodes that could lead to errors in the setting value calculation in step S3.

[0025] Figure 2 This is a schematic diagram of the power supply network node impedance matrix in an embodiment of this application. The rows and columns in the diagram correspond to the node numbers in the power supply network, the diagonal elements represent the self-equivalent impedance magnitude of each node, the off-diagonal elements represent the feeder impedance magnitude between two nodes, the color depth corresponds to the magnitude of the impedance magnitude, the unit of the value is ohms, and the zero value cell indicates that there is no direct feeder connection between two nodes. This matrix is ​​incrementally updated and included in the topology time archive after each topology change event is triggered. The target archive is retrieved using the fault timestamp as the query key and extracted to construct the node admittance matrix.

[0026] In one specific embodiment, step S3 includes: Add each impedance increment to the old equivalent impedance value of the corresponding node to obtain the updated equivalent impedance value of each affected node. Calculate the candidate instantaneous overcurrent setting value for each affected node using the following formula: ; in, The reliability factor is set to 1.3. The system's rated voltage. and These are the resistive and inductive components of the impedance increment, respectively. The above formula is applied to each node in the set of affected nodes one by one to obtain the candidate values ​​for the instantaneous overcurrent setting of each affected node. The instantaneous overcurrent setting candidate value for the i-th affected node; The resistance component of the old equivalent impedance of the i-th affected node, in Ω; The inductive reactance component of the old equivalent impedance value of the i-th affected node; Using each instantaneous overcurrent setting candidate value as input, a setting coordination request is sent to the upper-level node of each affected node. If the upper-level node fails the verification, it replies with a suggested correction amount. Each affected node corrects the instantaneous overcurrent setting candidate value according to the suggested correction amount and re-requests verification. The iteration limit is 5 times. The instantaneous overcurrent setting value, overcurrent setting value, and overcurrent delay setting value that finally pass the verification are combined to obtain the setting value vector of the entire network. The entire network setting value vector is distributed to the current effective register area of ​​each integrated protection device in a two-stage manner of pre-writing and atomic switching, so as to obtain the effective confirmation of the entire network setting value vector; Using the effective time of the network-wide setting value vector as the timestamp, the network-wide setting value vector, the current network-wide switch state matrix, the node impedance matrix, and the SHA-256 hash values ​​of each field are written into the time series database to obtain a topology time file with timestamps.

[0027] Specifically, in the instantaneous overcurrent setting formula, the reliability coefficient of 1.3 is taken according to the "Coal Mine Safety Regulations," which stipulates that the reliability coefficient of instantaneous overcurrent protection for underground high-voltage integrated protection devices should not be less than 1.3; the system rated voltage... The rated line voltage of the corresponding 10kV power supply system in the mine is 10000V; the resistive component of the old equivalent impedance value. With resistance component The resistance component of the impedance increment is read from the pre-stored adjacency table of each node. With resistance component The impedance increment message from step S2, after substituting the four into the formula, yields the result. This refers to the instantaneous overcurrent protection operating current value that the i-th affected node should be set after this topology change. The verification rule for the setting coordination request is that the candidate value of the instantaneous overcurrent protection setting of the lower-level node must be less than the quotient of the candidate value of the instantaneous overcurrent protection setting of the upper-level node divided by the reliability coefficient 1.3. If the verification fails, the upper-level node will multiply the candidate value of its own instantaneous overcurrent protection setting by the reliability coefficient 1.3 and the difference between it and the candidate value of the lower-level node as the suggested correction amount. The lower-level node will add the suggested correction amount to the candidate value of the instantaneous overcurrent protection setting and re-initiate the verification request. The iteration limit of 5 times is set according to the engineering statistics that the maximum number of levels in the multi-level power distribution system in the mine does not exceed 5 levels. The overcurrent setting value is calculated by multiplying the maximum load current of this feeder by the self-starting coefficient 2.5, then by the reliability coefficient 1.25, and then dividing by the return coefficient 0.85. The overcurrent delay setting value is determined by decreasing step by step with a time difference of 0.5 seconds from the upper-level protection. The combination of the three constitutes the setting value vector of the entire network.

[0028] In the pre-writing phase, after the entire network setting value vector is written into the temporary storage register area of ​​each integrated protection device, each node will transmit the received setting value data back along the original path. The sender confirms the pre-writing success only when the byte sequence of the original data and the transmitted data are completely consistent. In the atomic switching phase, the ground control center simultaneously sends a switching execution command to all nodes that have been confirmed to have successfully pre-written the data. Each node copies the setting value from the temporary storage register area to the currently effective register area within the next 20ms sampling period. The two-stage design ensures that the entire network protection coordination relationship takes effect synchronously within the same sampling period, eliminating coordination errors during the period when some nodes have switched while others have not. The SHA-256 hash value in the topology time file is calculated by concatenating the byte content of all fields of the entire network setting value vector, the entire network switch status matrix, the node impedance matrix, and the timestamp. When retrieving the target file in step S4, the SHA-256 hash value of the extracted data is recalculated and compared with the stored value in the file. The file data is confirmed to be complete only when the two are consistent, which is used to eliminate the impact of data corruption during storage on the fault location calculation results.

[0029] In one specific embodiment, the topology time archive is retrieved using the fault timestamp as the query key, and the network-wide switch state matrix and node impedance matrix closest to the fault time are extracted, including: Using the UTC millisecond timestamp of the fault occurrence time as the query key, all records that satisfy the condition that the file timestamp is not greater than the fault timestamp are retrieved from the time-series database of the topology time archive. The record with the largest file timestamp is taken as the target file. The SHA-256 hash value of the target file is subjected to integrity verification. After the verification is passed, the whole network switch state matrix and node impedance matrix are extracted from the target file to obtain the topology snapshot at the fault time. Using the network-wide switch state matrix from the topology snapshot at the time of the fault as input, the admittance values ​​of each feeder are obtained by taking the reciprocal of the feeder impedance in the closed state in the node impedance matrix, and the admittance values ​​of feeders in the open state are set to zero. The node admittance matrix is ​​obtained by filling each node in the network row by row according to the rule that the diagonal elements are the sum of the admittance values ​​of all operating feeders connected to the corresponding node, and the off-diagonal elements are the negative of the admittance values ​​of feeders between two nodes.

[0030] Specifically, the topology snapshot at the time of the fault refers to the combination of the network-wide switch state matrix and the node impedance matrix extracted from the target archive, corresponding to the actual operating state of the system at the time of the fault. The network-wide switch state matrix is ​​an N×1 binary vector, where N is the total number of circuit breakers in the network. A value of 1 indicates the corresponding circuit breaker is in the closed-operation state, and 0 indicates it is in the open-operation state. The retrieval rule for the target archive adopts a strategy where the archive timestamp is no greater than the fault timestamp, and the maximum value is taken. Physically, this means taking the system state after the last topology change before the fault occurred, which is the state corresponding to the time of the fault. The actual system topology is completely consistent, eliminating the interference of switch status changes caused by protection actions after the fault occurred on the topology restoration; the execution method of SHA-256 hash value integrity verification is to recalculate the SHA-256 hash value of all fields of the whole network setting value vector, the whole network switch status matrix, the node impedance matrix and the timestamp of the extracted target file, and compare the calculation result with the SHA-256 hash value stored in the target file byte by byte. Only when all are consistent is the file data confirmed to be complete. If there is any difference in any byte, the file data is judged to be corrupted and a verification failure alarm is reported to the central control center.

[0031] The node admittance matrix is ​​an N×N complex matrix, where N is the total number of nodes in the network. The admittance value of each feeder is calculated by taking the reciprocal of its corresponding feeder impedance; that is, the admittance value equals the reciprocal of the complex factor of the feeder impedance. Specifically, the real part of the admittance is obtained by dividing the resistive component by the square of the impedance magnitude, and the imaginary part is obtained by taking the negative of the inductive reactance component and dividing it by the square of the impedance magnitude. The physical basis for setting the admittance value of a feeder in the open state to zero is that the open circuit breaker disconnects the corresponding feeder from the electrical circuit. This feeder does not participate in current conduction, its equivalent impedance is infinite, and the corresponding admittance value is zero. The diagonal elements of the point admittance matrix are the sum of the admittance values ​​of all feeders connected to that node and in operation, reflecting the total admittance of that node to ground; the off-diagonal elements are the negatives of the feeder admittance values ​​connecting two nodes, reflecting the mutual admittance relationship between the two nodes. After filling each node in the network row by row according to the above rules, the resulting node admittance matrix completely describes the electrical connection relationship of the system at the time of the fault. By solving the matrix equation with the system rated voltage as the boundary condition, the normal operating voltage amplitude of each node can be obtained, and then the theoretical value of the short-circuit current of each candidate section can be calculated.

[0032] In one specific embodiment, constructing the nodal admittance matrix and calculating the theoretical short-circuit current value for each candidate segment includes: Using the node admittance matrix and the system rated voltage as input, the voltage distribution of each node in the entire network under normal operating conditions is solved to obtain the normal operating voltage amplitude of each node. Using the normal operating voltage amplitude of each node and the node impedance matrix in the topology snapshot at the time of the fault as input, a three-phase metallic short circuit is assumed to occur at each node in the entire network. The short-circuit current is calculated by dividing the normal operating voltage amplitude of each node by the equivalent impedance magnitude of the corresponding node, and the theoretical value of the short-circuit current at the end node of each candidate section is obtained.

[0033] Specifically, the calculation method for the normal operating voltage amplitude of each node using the node admittance matrix and the system rated voltage is as follows: the product of the node admittance matrix and the voltage vector of each node is equal to the injected current vector of each node. The system rated voltage of 10000V is used as the known boundary condition of the power supply node. The terminal load node is treated as an unloaded state, i.e., the injected current is zero. The linear equation system above is solved node by node using the Gauss-Seidel iteration method. In each iteration, the voltage of each node in the previous round is substituted into the right side of the equation system to update the voltage of each node. When the change in the voltage amplitude of each node between two adjacent iterations does not exceed 0.001V, convergence is determined. The obtained voltage amplitude of each node is the voltage amplitude of the node under normal operating conditions. The assumption of three-phase metallic short circuit is that the three phases at the fault point are directly short-circuited and the transition resistance is zero. This assumption corresponds to the maximum short-circuit current under the most severe fault condition. The theoretical value of the short-circuit current calculated in this way is the maximum fault current that the protection device at the node should theoretically detect under the current topology condition.

[0034] The equivalent impedance magnitude of each node is calculated by taking the square root of the sum of the squares of the resistive and inductive components of the equivalent impedance of the corresponding node in the topology snapshot at the fault time. The theoretical value of the short-circuit current of each node is calculated by dividing the normal operating voltage amplitude of the node by the equivalent impedance magnitude of the corresponding node. The candidate section end nodes refer to the nodes that constitute the electrical connection points at both ends of the feeder section, that is, two nodes that are directly adjacent in the power supply network adjacency table and have a connection between them and the operating feeder. Each candidate feeder section corresponds to two end nodes. The theoretical values ​​of the short-circuit current at the two end nodes together constitute the theoretical value combination of the short-circuit current of the candidate section. In the normalized matching error calculation, they are compared with the measured fault current of the corresponding node. By judging the joint error of the two end nodes rather than the single error of a single end node, the interference of single node measurement deviation on the fault section location result is eliminated.

[0035] In one specific embodiment, the normalized matching error is calculated by comparing the theoretical value with the measured fault current at each node, and the feeder section with the smallest error is output as the fault location result, including: Collect the measured fault current reported by each integrated protection device in the entire network at the time of the fault occurrence. Subtract the measured fault current and the theoretical short-circuit current of each candidate feeder section at both ends of the entire network and take the absolute value. Divide the sum of the absolute values ​​of the two ends of the nodes by the sum of the theoretical short-circuit current of the two ends of the nodes. Perform the above calculation on each candidate feeder section in the entire network to obtain the normalized matching error of each candidate feeder section. The normalized matching errors of each candidate feeder segment are sorted in ascending order. When the minimum normalized matching error is less than 0.15, the corresponding candidate feeder segment is taken as the unique fault location result. When the minimum normalized matching error is not less than 0.15 and less than 0.30, the top three candidate feeder segments with the smallest normalized matching errors are taken as the fault location results. When the minimum normalized matching error is not less than 0.30, the zero-sequence current direction auxiliary criterion is triggered to perform a secondary judgment on each candidate feeder segment to obtain the fault location result.

[0036] Specifically, the numerator of the normalized matching error is the sum of the absolute values ​​of the differences between the measured fault current and the theoretical short-circuit current at the two ends of the candidate feeder section, and the denominator is the sum of the theoretical short-circuit current at the two ends. The purpose of using the sum of the theoretical values ​​at the two ends as the normalization benchmark is to eliminate the influence of the difference in current dimensions between nodes with different voltage levels and power supply capacities on the error comparison, so that the error values ​​of each candidate feeder section in the entire network are comparable. The error threshold of 0.15 corresponds to the upper limit of engineering statistics after the comprehensive superposition of the rated measurement error of the current transformer of 0.5 level and the line parameter error in the 10kV power supply system of the mine. When the error is lower than 0.15, it indicates that the degree of agreement between the measured fault current and the theoretical value is sufficient to uniquely determine the fault section. The threshold of 0.30 corresponds to the maximum allowable deviation between the measured value and the theoretical value of the three-phase metallic short circuit caused by the attenuation of the fault current amplitude in the high-resistance grounding fault scenario. When it exceeds 0.30, it indicates that relying solely on current amplitude matching is not enough for reliable location, and directional information must be introduced to assist in the judgment.

[0037] The measured fault current is calculated from the effective value of the three-phase current synchronously sampled by each integrated protection device at the moment of protection operation, and the sampling time corresponds to the fault timestamp; the execution logic of the zero-sequence current direction auxiliary criterion is as follows: extract the upstream node set and downstream node set of the nodes at both ends of the candidate feeder section from the topology snapshot at the time of fault, and determine the direction of the angle between the measured zero-sequence current phasor of each node and the zero-sequence voltage phasor of the bus. Angles between 90° and 270° are determined to be positive, i.e., pointing to the fault point. Angles between 0° and 90° and between 270° and 360° are considered negative. The direction between ° is determined to be opposite, i.e., away from the fault point. All nodes in the upstream node set should be in the positive direction, and all nodes in the downstream node set should be in the opposite direction. When the actual coincidence rate is not less than 80%, the candidate section is determined to pass the direction verification. The 80% coincidence rate threshold is set based on engineering statistics in the high resistance grounding fault scenario in mines, where no more than 20% of the nodes fail to determine the direction due to excessive transition resistance causing the zero-sequence current amplitude to be lower than the minimum accurate operating current of the transformer. The candidate feeder section with the highest coincidence rate is taken as the final fault location result output.

[0038] Figure 3 This is a schematic diagram of the simulated three-phase current waveforms before and after a three-phase short-circuit fault in the underground power supply system of the mine in this application embodiment. The three sub-graphs in the figure correspond to the current waveforms of phase A, phase B, and phase C, respectively. The vertical axis represents the current amplitude in amperes, and the horizontal axis represents the time in milliseconds. The vertical dashed line marks the time of the fault occurrence, t=40ms. Before the fault, the three-phase current amplitude is about 100A. After the fault occurs, the current amplitude of each phase rises rapidly to over 850A and is accompanied by a decrease in the DC component. At this moment, each integrated protection device collects the effective value of the three-phase current as the measured fault current input for the normalization matching error calculation in step S4.

[0039] The above describes the mine power supply fault diagnosis and location method in the embodiments of this application. The following describes the mine power supply fault diagnosis and location system in the embodiments of this application. One embodiment of the mine power supply fault diagnosis and location system in the embodiments of this application includes: The jump module is used to collect the open and closed state sequence of each circuit breaker in the integrated protection device's input circuit, and perform jump identification processing on the state sequence based on the majority voting rule to obtain the topology change event carrying the node number and timestamp. The update module is used to incrementally update the equivalent impedance of each node in the set of affected nodes whose hop distance does not exceed two hops, based on the feeder parameters of the hop node and triggered by the topology change event, so as to obtain the impedance increment of each affected node. The archiving module is used to substitute the impedance increment into the instantaneous overcurrent setting formula, and obtain the network setting value vector through the cooperation and verification iteration between adjacent nodes. The setting value vector is then archived together with the current network switch state matrix and node impedance matrix to obtain a topology time archive with timestamps. The calculation module is used to retrieve the topology time archive by using the fault timestamp as the query key when a fault occurs, extract the network-wide switch state matrix and node impedance matrix that are closest to the fault time, construct the node admittance matrix and calculate the theoretical value of the short-circuit current of each candidate section, perform normalization matching error calculation between the theoretical value and the measured fault current of each node, and output the feeder section with the smallest error as the fault location result.

[0040] This invention also provides a mine power supply fault diagnosis and location device, which can be a server. The device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor, designed as a computer, provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores the data corresponding to this embodiment. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0041] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the mine power supply fault diagnosis and location method.

[0042] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0043] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a mine power supply fault diagnosis and location device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0044] 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 mine power supply fault diagnosis positioning method, characterized in that, The method includes: Step S1: Collect the open / closed state sequence of each circuit breaker in the integrated protection device's input circuit, and perform jump identification processing on the state sequence based on the majority voting rule to obtain the topology change event carrying the node number and timestamp; Step S2: Triggered by the topology change event, the equivalent impedance of each node in the set of affected nodes with a hop distance of no more than two hops is incrementally updated according to the feeder parameters of the hop node to obtain the impedance increment of each affected node. Step S3: Substitute the impedance increment into the instantaneous overcurrent setting formula, and obtain the network-wide setting value vector through the cooperation and verification iteration between adjacent nodes. Then, archive the setting value vector together with the current network-wide switch state matrix and node impedance matrix to obtain a topology time file with timestamps. Step S4: When a fault occurs, the topology time archive is retrieved using the fault timestamp as the query key. The network switch state matrix and node impedance matrix closest to the fault time are extracted. The node admittance matrix is ​​constructed and the theoretical value of the short-circuit current of each candidate section is calculated. The theoretical value is normalized and matched with the measured fault current of each node to calculate the error. The feeder section with the smallest error is output as the fault location result.

2. The mine power supply fault diagnosis and positioning method according to claim 1, characterized in that, Step S1 includes: With a sampling period of 20ms, the opening and closing status of each circuit breaker in the integrated protection device input circuit is collected. The closing status is recorded as 1 and the opening status is recorded as 0. 600 consecutive sampling points are stored in the state sequence buffer area corresponding to each circuit breaker to obtain the opening and closing status sequence of each circuit breaker. Based on the majority voting rule, the transition identification process for three consecutive sampling points in the separation and combination state sequence is performed using the following formula: ； wherein, is the state value of the i-th sampling point, = +1 corresponds to a closing jump, = -1 corresponds to an opening jump, = 0 indicates no jump, the above formula is executed for each sampling point in the on-off state sequence to obtain the jump determination result of each sampling point;​ In the transition determination result The node number corresponding to the sampling point, the UTC millisecond timestamp of the transition time, the operation type, and the MD5 hash value of the switch status of each adjacent node before the operation are written into the topology change event record table to obtain the topology change event carrying the node number and timestamp. The topology change events are broadcast to all integrated protection devices across the network via UDP multicast, resulting in the topology change events to be processed in the local topology change queue of each integrated protection device.

3. The method for diagnosing and locating power supply faults in mines according to claim 1, characterized in that, Step S2 includes: Based on the node number in the topology change event, and according to the power supply network adjacency table pre-stored by each integrated protection device, all nodes whose hop distance to the hop node corresponding to the node number does not exceed two hops are screened to obtain the set of affected nodes. Using the old equivalent impedance values ​​of each node in the affected node set and the positive sequence impedance of the newly added feeder at the switching node as input, the unit resistance and unit inductive reactance of the newly added feeder are read from the line parameter database, and multiplied by the line length to obtain the resistance component and inductive reactance component of the positive sequence impedance of the newly added feeder. Thevenin parallel update processing is performed on the old equivalent impedance values ​​of each node in the affected node set and the positive sequence impedance of the newly added feeder to obtain the new equivalent impedance values ​​of each affected node. The impedance increment of each affected node is obtained by calculating the difference between the resistance and inductive reactance components of the new equivalent impedance values ​​and the resistance and inductive reactance components of the corresponding old equivalent impedance values. Each impedance increment, along with its corresponding node number and CRC32 checksum, is encapsulated into an impedance increment message and sent via unicast to the corresponding node in the affected node set.

4. The method for diagnosing and locating power supply faults in mines according to claim 1, characterized in that, Step S3 includes: The updated equivalent impedance value of each affected node is obtained by adding the impedance increment of each node to the old equivalent impedance value of the corresponding node. The candidate value of the instantaneous overcurrent setting of each affected node is then calculated using the following formula: ； in, The reliability factor is set to 1.

3. The system's rated voltage. and The resistive and inductive components of the impedance increment are respectively represented by the above formula. The above formula is applied to each node in the set of affected nodes one by one to obtain the candidate value of the instantaneous overcurrent setting for each affected node. The instantaneous overcurrent setting candidate value for the i-th affected node; The resistance component of the old equivalent impedance of the i-th affected node, in Ω; The inductive reactance component of the old equivalent impedance value of the i-th affected node; Using the current instantaneous trip setting candidate values ​​as input, a setting coordination request is sent to the upper-level node of each affected node. If the upper-level node fails the verification, it replies with a suggested correction amount. Each affected node corrects the current instantaneous trip setting candidate value according to the suggested correction amount and re-requests verification. The iteration limit is 5 times. The current instantaneous trip setting value, overcurrent setting value, and overcurrent delay setting value that finally pass the verification are combined to obtain the setting value vector of the entire network. The entire network setting value vector is sent to the current effective register area of ​​each integrated protection device in a two-stage manner of pre-writing and atomic switching to obtain the effective confirmation of the entire network setting value vector; Using the effective time of the network-wide setting value vector as a timestamp, the network-wide setting value vector, the current network-wide switching state matrix, the node impedance matrix, and the SHA-256 hash values ​​of each field are written into the time series database to obtain the topology time file with timestamp.

5. The method for diagnosing and locating power supply faults in mines according to claim 1, characterized in that, In step S4, the topology time archive is retrieved using the fault timestamp as the query key, and the network-wide switch state matrix and node impedance matrix closest to the fault time are extracted, including: Using the UTC millisecond timestamp of the fault occurrence time as the query key, all records that satisfy the condition that the archive timestamp is not greater than the fault timestamp are retrieved from the time-series database of the topology time archive. The record with the largest archive timestamp is taken as the target archive. The SHA-256 hash value of the target archive is subjected to integrity verification. After the verification is passed, the whole network switch state matrix and node impedance matrix are extracted from the target archive to obtain the topology snapshot at the fault time. Using the network-wide switch state matrix in the topology snapshot at the time of the fault as input, the admittance value of each feeder is obtained by taking the reciprocal of the feeder impedance in the closed state in the node impedance matrix, and the admittance value of the feeder in the open state is set to zero. The node admittance matrix is ​​obtained by filling each node in the network row by row according to the rule that the diagonal element is the sum of the admittance values ​​of all operating feeders connected to the corresponding node, and the off-diagonal element is the negative of the admittance value of the feeder between two nodes.

6. The method for diagnosing and locating power supply faults in mines according to claim 5, characterized in that, In step S4, the nodal admittance matrix is ​​constructed and the theoretical short-circuit current value of each candidate segment is calculated, including: Using the node admittance matrix and the system rated voltage as input, the voltage distribution of each node in the entire network under normal operating conditions is solved to obtain the normal operating voltage amplitude of each node. Using the normal operating voltage amplitude of each node and the node impedance matrix in the topology snapshot at the fault time as input, a three-phase metallic short circuit is assumed to occur at each node in the entire network. The short-circuit current is calculated by dividing the normal operating voltage amplitude of each node by the equivalent impedance magnitude of the corresponding node, and the theoretical value of the short-circuit current at the end node of each candidate section is obtained.

7. The method for diagnosing and locating power supply faults in mines according to claim 6, characterized in that, In step S4, the theoretical value and the measured fault current at each node are normalized and matched to calculate the error. The feeder section with the smallest error is output as the fault location result, including: Collect the measured fault current reported by each integrated protection device in the entire network at the time of the fault occurrence. Subtract the measured fault current at both ends of each candidate feeder section from the theoretical value of the short-circuit current and take the absolute value. Divide the sum of the absolute values ​​of the two ends by the sum of the theoretical values ​​of the short-circuit current at both ends. Perform the above calculation on each candidate feeder section in the entire network to obtain the normalized matching error of each candidate feeder section. The normalized matching errors of each candidate feeder segment are arranged in ascending order. When the minimum normalized matching error is less than 0.15, the corresponding candidate feeder segment is taken as the unique fault location result. When the minimum normalized matching error is not less than 0.15 and less than 0.30, the top three candidate feeder segments with the smallest normalized matching errors are taken as the fault location results. When the minimum normalized matching error is not less than 0.30, the zero-sequence current direction auxiliary criterion is triggered to perform a secondary judgment on each candidate feeder segment to obtain the fault location result.

8. A fault diagnosis and location system for power supply in mines, characterized in that, For implementing the mine power supply fault diagnosis and location method as described in any one of claims 1-7, the mine power supply fault diagnosis and location system comprises: The jump module is used to collect the open and closed state sequence of each circuit breaker in the integrated protection device's input circuit, and perform jump identification processing on the state sequence based on the majority voting rule to obtain the topology change event carrying the node number and timestamp. The update module is used to incrementally update the equivalent impedance of each node in the set of affected nodes whose hop distance does not exceed two hops, based on the feeder parameters of the hop node and triggered by the topology change event, so as to obtain the impedance increment of each affected node. The archiving module is used to substitute the impedance increment into the instantaneous overcurrent setting formula, and obtain the network setting value vector through the cooperation and verification iteration between adjacent nodes. The setting value vector is then archived together with the current network switch state matrix and node impedance matrix to obtain a topology time archive with timestamps. The calculation module is used to retrieve the topology time archive by using the fault timestamp as the query key when a fault occurs, extract the network-wide switch state matrix and node impedance matrix that are closest to the fault time, construct the node admittance matrix and calculate the theoretical value of the short-circuit current of each candidate section, perform normalization matching error calculation between the theoretical value and the measured fault current of each node, and output the feeder section with the smallest error as the fault location result.

9. A fault diagnosis and location device for power supply in mines, characterized in that, The method includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the mine power supply fault diagnosis and location method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the mine power supply fault diagnosis and location method as described in any one of claims 1 to 7.

Images