Intelligent Fault Sensing and Decision Support System for Distribution Networks
By dividing the topology into time periods and calculating electrical disturbances in the distribution network, cross-topology power supply paths are identified, solving the problem of misjudgment caused by distributed power sources and terminal clock drift in distribution network fault handling, and realizing more reliable fault location and isolated power transfer.
Patent Information
- Application Number
- CN202611196260.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-07
- Publication Date
- 2026-11-10
AI Technical Summary
Existing fault handling methods for distribution networks can lead to errors in fault section judgment when faced with factors such as differences in fault current characteristics of distributed power sources and terminal clock drift. This can even result in the generation of incorrect power transfer and closing commands, causing the fault point to become energized again and the protection system to operate repeatedly.
The topology time period generation module divides the front topology and the back topology. The electrical disturbance module calculates the current disturbance energy and voltage drop energy. The reachability coefficient calculation module and the pseudo-continuity coefficient calculation module identify the cross-topology power supply path. The fault location module determines the unique fault branch. The isolation and transfer module performs safety confirmation. Finally, the control commands for isolation and transfer are generated.
It reduces basic data errors caused by cross-topology electrical quantity mixing, identifies false power supply paths, reduces the risk of secondary faults caused by accidental closing, and improves the reliability of fault handling.
Smart Images

Figure CN123066478A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of smart grid technology, specifically to a distribution network fault status intelligent perception and auxiliary decision-making system. Background Technology
[0002] The distribution network is the final link in the power system that supplies power to users. Rapid fault location, isolation, and restoration of power to non-faulty sections are the core functions of distribution automation systems. Existing distribution automation systems typically collect line current amplitudes, node voltage amplitudes, switch position status, and protection action signals through feeder terminals and substation terminals. Based on the fusion of multi-source data within a unified time window, they determine the faulty section and generate switch opening and closing control commands to automate fault handling.
[0003] As the penetration rate of distributed generation in power distribution networks increases, the fault current of inverter-type distributed generation is constrained by the current limiting of the converter, with an amplitude close to the rated current and a relatively long duration, which is significantly different from the high-current short-circuit characteristics of traditional synchronous power sources. At the same time, the distribution terminals are affected by factors such as clock drift, communication delay, and interruption of time synchronization links, resulting in inherent errors in the timestamps of data transmitted by different terminals. Measurement data of the same fault event may overlap or be misaligned in time.
[0004] Existing fault handling methods typically integrate multi-terminal data using a fixed-length time window, assuming that data within the window belongs to the same topology. They fail to consider the impact of topology abrupt changes caused by switch opening, nor do they retain or process the time errors in the measurement data. When the switch opening occurs within the fault time window, the electrical quantities before and after the opening may be pieced together into a non-existent power supply path, leading to errors in fault section identification and even generating incorrect power transfer and closing commands, causing problems such as the fault point to re-energize and protection systems to repeatedly operate. Summary of the Invention
[0005] This application provides an intelligent perception and auxiliary decision-making system for distribution network fault conditions, which solves the technical problems mentioned in the background.
[0006] This application provides an intelligent fault state perception and auxiliary decision-making system for distribution networks, including: The topology period generation module is used to divide the front topology period and the back topology period based on the base diagram, switch status, protection action flag, timestamp, and maximum error boundary of timestamp, and to generate front topology, back topology, measurement interval and electrical quantity baseline data. The electrical disturbance module is used to process the current amplitude and voltage amplitude according to the overlap length between the measurement interval and the previous and subsequent topological time periods, and to obtain the interval overlap coefficient, current disturbance energy, voltage drop energy and branch fault evidence coefficient. The reachability coefficient calculation module is used to construct a set of paths from power nodes to candidate branches in the front and back topologies, and determine the single-topology power supply reachability coefficient based on the fault evidence coefficient of the branch within the path. The pseudo-continuity coefficient calculation module is used to merge the closed branches of the previous topology and the subsequent topology to form a union graph. It uses the branch fault evidence coefficient to determine the hybrid reachability coefficient and the time cross-boundary coincidence coefficient, and generates the cross-topology power supply pseudo-continuity coefficient. The fault location module is used to determine the confidence coefficient of the faulty branch by combining the branch fault evidence coefficient, the single-topology power supply reachability coefficient and the cross-topology power supply pseudo-continuity coefficient, and select the unique faulty branch. The isolation transfer module is used to determine the set of disconnecting switches and healthy power-off load blocks based on the subsequent topology and the only faulty branch, calculate the current envelope and safety margin factor of candidate tie switches, and select the only tie switch. The control message output module is used to receive the position feedback of the disconnecting switch and the amplitude of the isolation boundary current, confirm the isolation conditions, and generate a unique fault branch identifier, a list of disconnecting switch opening commands, and a unique tie switch closing command.
[0007] The beneficial effects of this application are as follows: Addressing the fault handling needs in distribution network topology abrupt changes, this application retains the time error information of measurement data, reducing fundamental data errors caused by the mixing of electrical quantities across topologies. Through dual-time period electrical quantity calculation and path evidence analysis, it can identify false power supply paths formed by cross-topology splicing, reducing misjudgment of fault sections during topology transient processes. Employing a conservative current envelope and dual isolation confirmation mechanism, it can verify the line's operational safety conditions before power transfer, reducing the risk of secondary faults caused by erroneous closing. No additional hardware is required; it is compatible with existing distribution automation terminal data collection and distribution network operation scenarios containing distributed power sources, improving the reliability of the fault handling process. Attached Figure Description
[0008] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0009] Figure 1 The calculation flowchart of the intelligent perception and auxiliary decision-making system for distribution network fault status provided in the embodiments of this application is shown below. Figure 2 This is a reliability analysis diagram of candidate branch fault location for cross-topological pseudo-continuity suppression provided in an embodiment of this application. Figure 3A comprehensive evaluation diagram of the current carrying capacity and safety margin of the candidate interconnecting switch provided in the embodiments of this application. Detailed Implementation
[0010] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.
[0011] It should be noted that, in the description of this application, 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. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.
[0012] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0013] like Figure 1 As shown, the intelligent fault situation perception and auxiliary decision-making system for distribution networks includes: The topology period generation module is used to divide the front topology period and the back topology period based on the base diagram, switch status, protection action flag, timestamp, and maximum error boundary of timestamp, and to generate front topology, back topology, measurement interval and electrical quantity baseline data. The electrical disturbance module is used to process the current amplitude and voltage amplitude according to the overlap length between the measurement interval and the previous and subsequent topological time periods, and to obtain the interval overlap coefficient, current disturbance energy, voltage drop energy and branch fault evidence coefficient. The reachability coefficient calculation module is used to construct a set of paths from power nodes to candidate branches in the front and back topologies, and determine the single-topology power supply reachability coefficient based on the fault evidence coefficient of the branch within the path. The pseudo-continuity coefficient calculation module is used to merge the closed branches of the previous topology and the subsequent topology to form a union graph. It uses the branch fault evidence coefficient to determine the hybrid reachability coefficient and the time cross-boundary coincidence coefficient, and generates the cross-topology power supply pseudo-continuity coefficient. The fault location module is used to determine the confidence coefficient of the faulty branch by combining the branch fault evidence coefficient, the single-topology power supply reachability coefficient and the cross-topology power supply pseudo-continuity coefficient, and select the unique faulty branch. The isolation transfer module is used to determine the set of disconnecting switches and healthy power-off load blocks based on the subsequent topology and the only faulty branch, calculate the current envelope and safety margin factor of candidate tie switches, and select the only tie switch. The control message output module is used to receive the position feedback of the disconnecting switch and the amplitude of the isolation boundary current, confirm the isolation conditions, and generate a unique fault branch identifier, a list of disconnecting switch opening commands, and a unique tie switch closing command.
[0014] In one alternative implementation, the topology time-period generation module is configured to perform the following operations: Step S101: Verify the correspondence between the switch equipment identifier, power node identifier and the branch and node in the basic diagram. Based on the protection action flag, determine the first protection action switch state transition time as the topology boundary time, and classify the protection switch actions that occur consecutively within the anti-jitter time into the same topology transition event. Read the branch rated current amplitude, node rated voltage amplitude, branch allowable continuous current, line equivalent impedance amplitude, network rated voltage amplitude, allowable minimum voltage amplitude, switch remote control attribute, switch normally open and normally closed attribute, single-sided fault analysis window length, stability window length, state holding time and no-current threshold. The specific basic diagram is as follows: ; in, Represents the basic diagram. Represents a set of nodes. Represents the set of branches; The stabilization time and the length of the single-sided fault analysis window satisfy the following: ; in, Indicates the image stabilization time. Indicates the length of the single-sided fault analysis window; The basic diagram refers to the network topology model generated after extracting the connection relationships between nodes and branches from the equipment ledger database of the power distribution automation system and verifying the topology connectivity. The rated current amplitude of the branch refers to the rated current value of the branch that is read from the factory nameplate parameters of the current transformer and feeder terminal, and then entered into the system after on-site ratio verification. The rated voltage amplitude of a node refers to the node rated voltage value that is determined according to the design value of the voltage level of the distribution network and entered into the system after the voltage transformer ratio is verified on site. The allowable continuous current of a branch refers to the maximum current value that a branch can carry over a long period of time, which is obtained after correction based on conductor type, cable cross-section, and ambient temperature, extracted from the line design manual or equipment ledger. The equivalent impedance magnitude of a line refers to the positive sequence impedance value of a branch calculated based on the impedance parameters per unit length of the line and the actual length, and converted to the system rated voltage reference. The rated voltage amplitude of the network refers to the nominal voltage value of the system extracted from the distribution network planning and design documents, which is determined according to the voltage level of the distribution network. The minimum allowable voltage amplitude is preferably 0.7 to 0.9 times the rated voltage. The value is a preset parameter that combines the national standard for power supply voltage deviation and the threshold for low voltage protection of the distribution network. The remote control attribute of the switch refers to the attribute parameters extracted from the capability ledger of the distribution automation terminal equipment, which indicate whether the switch supports remote issuance of opening and closing commands by the distribution master station; Normally open and normally closed switch attributes are attribute parameters extracted from the normal operation mode file of the distribution network, which identify whether the switch is in the open or closed position under normal operation. The anti-shake time refers to a preset value of 20 milliseconds to 200 milliseconds, which is set by combining the action time of the switching mechanism, the jitter duration of the auxiliary contact, and the anti-shake requirements of the protection action signal. The length of the single-sided fault analysis window is a preset parameter set by combining the transient duration of the distribution network fault, the protection action exit time, and the effective duration of the fault recording, with the preferred value being between 100 milliseconds and 1000 milliseconds. The stable window length refers to a preset parameter set by taking into account the distribution network load fluctuation cycle and the current and voltage stability time requirements during normal operation, with an optimal value of 5 to 30 minutes. The state holding time is a preset parameter with a preferred value of 1 to 5 seconds, which is set by combining the stabilization time after the switch auxiliary contact changes position and the delay time of the terminal position signal transmission. The no-current threshold is a preset parameter with a preferred value of 0.5 amperes to 2 amperes, which is set in combination with the minimum measurement accuracy of the distribution terminal and the range of the line no-load charging current. The topology boundary moment refers to the timestamp value of the signal recorded when the first protective action switch changes from the closed to the open position. Step S102: Read the switch states that have continuously reached the anti-jitter time before and after the topology boundary moment to form the front topology and the back topology; the front topology is composed of closed branches that have continuously reached the anti-jitter time before the topology boundary moment, and the back topology is composed of closed branches that have continuously reached the anti-jitter time after the topology boundary moment; switches with repeated state changes or invalid quality bits are treated as open states. The pre-topology refers to the network topology structure consisting of a set of closed branches that have maintained no switching state changes for more than the debounce time before the topology boundary time. Post-topology refers to the network topology structure consisting of a set of closed branches that have maintained no switching state changes for more than the debounce time after the topology boundary time. Step S103: Establish the pre-topology time period and the post-topology time period according to the topology boundary time and the length of the single-sided fault analysis window; divide the load blocks according to the power failure connection relationship in the post-topology; calculate the branch baseline current amplitude, node baseline voltage amplitude, load block boundary baseline current and branch original load baseline current within the stability window before protection action to form electrical quantity baseline data; The preceding topological time period and the following topological time period are respectively: ; in, Indicates the previous topological period. Indicates the post-topological period. Indicates the topological boundary time. Indicates the length of the single-sided fault analysis window; The stable window length satisfies: ; in, Indicates the stable window length. Indicates the length of the single-sided fault analysis window; There are no protection activations or switching changes within the stable window; the branch baseline current amplitude is the median current amplitude of the corresponding branch within the stable window, the node baseline voltage amplitude is the median voltage amplitude of the corresponding node within the stable window, the load block boundary baseline current is the median boundary current amplitude of the corresponding load block within the stable window, and the branch original load baseline current is the median current amplitude of the corresponding branch when it bears the original healthy load within the stable window. The pre-topology period refers to the time interval formed by extending the length of the single-sided fault analysis window forward from the topology boundary time. The post-topology period refers to the time interval formed by extending the length of the single-sided fault analysis window backward from the topology boundary time. The branch baseline current amplitude refers to the current reference value obtained by taking the median of all sampling points of the branch current amplitude within the stable window before the fault occurs. The node baseline voltage magnitude refers to the reference voltage value obtained by taking the median of all sampling points of the node voltage magnitude within the stability window before the fault occurs. The load block boundary baseline current refers to the reference current value obtained by taking the median of the current amplitude sampling points at the boundary between the load block and the main grid within the stability window before the fault occurs. The original load baseline current of a branch refers to the reference value of the current obtained by taking the median of the current amplitude sampling points when the branch is carrying the original healthy load within the stable window before the fault occurs. Step S104: Determine the possible measurement range based on the measurement packet timestamp and the maximum error boundary of the timestamp, and read the sampling duration of the measurement packet; current measurement packets and voltage measurement packets with out-of-range or invalid quality bits are not included in the calculation of current disturbance energy and voltage drop energy, respectively, and the corresponding possible measurement ranges are retained; The formula for calculating the possible measurement interval is as follows: ; in, Indicates measurement package The possible measurement range, Indicates measurement package timestamp, Indicates measurement package The maximum error boundary of the timestamp; When the timestamp of the same terminal goes backward, the maximum error boundary of the timestamp is expanded to cover the backward interval.
[0015] The measurement possible interval refers to the closed interval formed by taking the timestamp of the measurement package as the center and extending it forward and backward to the maximum error boundary of each extended timestamp; The maximum error boundary of timestamps refers to the collected parameters obtained by reading the clock precision parameters and time synchronization error parameters configured in the terminal. The error boundary of the terminal using satellite time synchronization is 1 millisecond, and the error boundary of the terminal using network time synchronization is 10 milliseconds. The sampling duration refers to the collection parameters obtained by reading the terminal sampling point interval configuration and message packaging cycle configuration. When sampling 32 points per cycle and packaging a message every 20 milliseconds, the sampling duration is 20 milliseconds. Furthermore, an invalid quality bit means that the quality identifier carried in the message indicates that the message cannot be used for the corresponding data calculation. For switch status messages, an invalid quality bit includes terminal communication interruption, device self-test abnormality, switch status acquisition abnormality, or auxiliary contact disconnection; for current measurement packets and voltage measurement packets, an invalid quality bit includes terminal sampling abnormality, communication verification error, or device lockout. Different types of messages are judged according to their corresponding quality identifiers. Within the anti-jitter time window, if the open / closed position signal of the same switch alternates twice or more, it is judged as repeated state changes; for switches with repeated state changes or invalid quality bits, they are uniformly treated as open state when generating the corresponding topology and are not included in the closed branch set. For example, if a switch sends open, closed, and open status signals successively within 10 to 30 milliseconds before the topology boundary time, and the anti-jitter time is 50 milliseconds, the switch is judged as having repeated state changes and is treated as open in the previous topology.
[0016] Furthermore, in the post-topology, closed branches are traversed starting from all power nodes, and reachable nodes are marked as energized nodes, with the remaining nodes forming a set of de-energized nodes. Within the set of de-energized nodes, the largest connected subgraph is identified based on branch connectivity, with each largest connected subgraph representing a de-energized load block. Once the unique faulty branch is determined, connected subgraphs containing the unique faulty branch are excluded from the de-energized load blocks to determine healthy de-energized load blocks.
[0017] Furthermore, the measurement packets sent from the same terminal are traversed in chronological order. When the timestamp of a later measurement packet is earlier than that of a previous measurement packet, a timestamp backtracking is determined to have occurred. The difference between the timestamps of the previous and later measurement packets is determined as the backtracking interval length. For the measurement packets affected by this timestamp backtracking, their maximum timestamp error boundary is expanded to cover the backtracking interval, ensuring that the possible measurement interval includes the range of time backtracking.
[0018] Furthermore, starting from the topology boundary moment, we trace back and continuously search for time periods without protection start signals, without switch status changes, and without current surges exceeding 20% of the rated current. When the length of this time period reaches the length of the stable window, we extract this time period as the stable window. When calculating the median, the number of sampling points is no less than 30. If the number of sampling points is insufficient, we continue to expand the window forward until the number requirement is met.
[0019] Understandably, the topology boundary is defined by the moment the first protective action switch transition occurs, and the preceding and following topologies are generated based on the anti-jitter time. The preceding and following topologies correspond to the network states before and after the topology boundary moment, respectively. For closed branches that do not undergo state changes during the topology transition, they can simultaneously belong to the sets of closed branches in both the preceding and following topologies. For each measurement packet sent by a terminal, its topology time period is not determined based on the timestamp, but rather the error range of the timestamp is retained to generate a corresponding possible measurement interval. This preserves the uncertainty of the measurement packet's time assignment, avoiding misclassification of measurement data due to terminal clock errors. The above embodiment completes the division of topology time periods at the data access level, retains the time error information of the measurement data, and avoids mixing electrical quantities belonging to different topology states into the same analysis set. This provides accurate topology and time references for subsequent power supply path determination, reducing basic data errors caused by topology changes.
[0020] In one alternative implementation, the electrical disturbance module is configured to perform the following operations: Step S201: Calculate the overlap length between the possible measurement interval and the previous topological time period and the subsequent topological time period, and determine the ratio of the overlap length to the length of the possible measurement interval as the interval overlap coefficient. The formula for calculating the interval overlap coefficient is as follows: ; in, Indicates measurement package During topological time The coefficient of coincidence within the interval, Indicates the previous topological period. Indicates the post-topological period. Indicates measurement package The possible measurement range, Indicates topological time period , Indicates the length of the time interval; When the measurement interval does not intersect with the corresponding topological time period, the interval overlap coefficient is 0. The interval overlap coefficient is a proportional value obtained by dividing the intersection length of a possible interval and its corresponding topological time period by the total length of the possible intervals. Step S202: Based on the product of the square of the difference between the cumulative current amplitude of the interval overlap coefficient and the branch baseline current amplitude and the sampling duration, and normalized by the product of the square of the branch rated current amplitude and the weighted sampling duration, the current disturbance energy is obtained. The formula for calculating the energy of the current disturbance is as follows: ; in, Indicates a branch During topological time Energy of internal current disturbance. Indicates a branch The effective current measurement package set, Indicates measurement package During topological time The coefficient of coincidence within the interval, Indicates measurement package The current amplitude, Indicates a branch The amplitude of the branch baseline current, Indicates a branch The rated current amplitude of the branch circuit, Indicates measurement package The sampling duration; When the weighted sampling duration is 0, the current disturbance energy is 0; The effective current measurement packet set refers to the set of measurement packets remaining after removing all current measurement packets corresponding to the branch, such as those with out-of-range or invalid quality bits. Current disturbance energy refers to the value obtained by multiplying the square of the difference between the effective current measurement packet and the baseline current by the sampling duration, with the interval overlap coefficient as the weight, and then dividing by the product of the square of the rated current and the weighted total duration. Step S203: Based on the interval overlap coefficient, accumulate the product of the positive difference square of the node baseline voltage amplitude and the voltage amplitude and the sampling duration, and normalize it using the product of the square of the node rated voltage amplitude and the weighted sampling duration to obtain the voltage drop energy. The formula for calculating the voltage drop energy is as follows: ; in, Represents a node During topological time Energy from voltage drop within the device. Represents a node Effective voltage measurement package set, Indicates measurement package During topological time The coefficient of coincidence within the interval, Represents a node The node baseline voltage amplitude, Indicates measurement package voltage amplitude, Represents a node The rated voltage amplitude of the node, Indicates measurement package Sampling duration, Indicates taking The larger of 0; When the weighted sampling duration is 0, the voltage drop energy is 0; The effective voltage measurement packet set refers to the set of measurement packets remaining after removing out-of-range and invalid quality bits from all voltage measurement packets corresponding to a node. Voltage drop energy refers to the value obtained by weighting the effective voltage measurement package with the interval overlap coefficient, accumulating the product of the square of the positive difference between the baseline voltage and the measured voltage and the sampling duration, and then dividing it by the product of the square of the rated voltage and the weighted total duration. Step S204: Select the voltage drop energy with the larger value at both ends of the branch, calculate the square root of the product of the selected voltage drop energy and the current disturbance energy, and obtain the branch fault evidence coefficient. The formula for calculating the branch fault evidence coefficient is as follows: ; in, Indicates a branch During topological time The branch fault evidence coefficient within the branch is Indicates a branch During topological time Energy of internal current disturbance. Indicates a branch first endpoint During topological time Energy from voltage drop within the device. Indicates a branch The second endpoint During topological time The energy of voltage drop within.
[0021] The branch fault evidence coefficient is the square root of the product of the energy of the current disturbance in the branch and the energy of the voltage drop at the two ends of the branch. Furthermore, whether the current measurement packet and voltage measurement packet exceed the range limit is determined according to the upper and lower limits of the range configured for the corresponding measurement channel. In this embodiment, the upper limit of the current measurement channel is set to 1.2 times the rated current amplitude of the corresponding branch, and the upper limit of the voltage measurement channel is set to 1.2 times the rated voltage amplitude of the corresponding node; when the measured value exceeds the corresponding upper limit or falls below the lower limit, it is determined to be an out-of-range measurement. After removing invalid packets, if the weighted total duration is 0, the current disturbance energy of the corresponding branch is set to 0, and the voltage drop energy of the corresponding node is set to 0. For example, if the rated current amplitude of a branch is 600 amperes, the upper limit of the range of the corresponding current measurement channel is set to 1.2 times it, i.e., 720 amperes; when the current amplitude sent by a current measurement packet is 800 amperes, the measured value exceeds the upper limit of the range, so the measurement packet is removed from the set of valid current measurement packets.
[0022] Furthermore, the difference between the baseline voltage and the measured voltage is calculated, and only positive differences are included in the accumulation. If the difference is negative or zero, the contribution of the measurement packet is zero. The basis for only accumulating voltage drops is that the voltage rise during the fault process belongs to the recovery process after the fault is cleared and does not reflect the electrical disturbance characteristics of the fault point. Including it in the calculation will offset the evidence of voltage drop caused by the fault.
[0023] Understandably, by using the overlap ratio between the possible measurement interval and the topological time period as the time weight, time-weighted energy values are calculated for current disturbances and voltage drops respectively. Then, the current energy and voltage energy are coupled to generate fault evidence coefficients for a single branch, without forcibly assigning measurement data with time errors to a single moment or a single topological time period. The above embodiment preserves the time uncertainty of cross-topological measurement packets, avoids the loss of fault information caused by forced time period assignment, verifies single-branch fault characteristics through the coupling of current and voltage quantities, reduces misjudgments of branch faults caused by fluctuations in a single electrical quantity, and improves the reliability of single-branch fault evidence.
[0024] In one alternative implementation, the reachability coefficient calculation module is configured to perform the following operations: Step S301: Determine the branches in the connected domain where the first protected device is located and the adjacent power failure connected domain as candidate branches; The first protective action device is the protective action switch corresponding to the topology boundary moment; Candidate branches refer to the set of all branches within the connectivity of the first activated protective switch, and all branches within the adjacent power failure connectivity. Step S302: For each candidate branch, construct a set of simple paths from the power node to the candidate branch and passing through the candidate branch in the front topology and back topology respectively; each simple path starts from any power node and terminates when it passes through the corresponding candidate branch. The set of simple paths refers to the set of all paths in a single topology that start from any power node, pass through candidate branches, and do not repeatedly pass through the same node or branch. Step S303: For each simple path, select the branch fault evidence coefficient with the smallest value within the path, and then select the path result with the largest value from the set of simple paths to obtain the front topology power supply reachability coefficient and the back topology power supply reachability coefficient respectively, and refer to the two as the single topology power supply reachability coefficient. The formula for calculating the single-topology power accessibility coefficient is as follows: ; in, Indicate candidate branch During topological time The single-topology power supply reachability factor within the system, hour This represents the topological energy accessibility factor. hour This represents the topology's power accessibility factor. Indicate candidate branch During topological time A set of simple paths within, Indicates a simple path. Represents a simple path The branch road in the middle, Indicates a branch During topological time The coefficient of evidence for branch faults within the circuit; When the set of simple paths is empty, the reachability coefficient of a single topology is 0.
[0025] The single-topology power reachability coefficient is a value obtained by taking the minimum value of the fault evidence coefficients of all branches for each path in the simple path set, and then taking the maximum value among all the minimum values of all paths. Furthermore, in the subsequent topology, a de-energized connected region separated from the first activated protective switch by only a tripping switch is considered an adjacent de-energized connected region. The boundary of the connected region is defined by the tripping switch and does not include de-energized areas indirectly connected by other closed switches. For example, after a 10 kV feeder outgoing switch trips, the busbar side is a energized connected region, and the line side is a de-energized connected region. This de-energized connected region is separated from the energized region on the busbar side only by the outgoing switch and is considered an adjacent de-energized connected region, with all its internal branches included in the candidate branch range.
[0026] Furthermore, a depth-first search is used to traverse closed branches starting from the power node. When the search path passes through a candidate branch, it stops extending and the path is recorded. If a node that has already been visited is encountered during the search, the current branch is terminated to avoid forming a loop. The path must end with a candidate branch and does not include other branches extending after the candidate branch.
[0027] Understandably, in a single real topology, a wide-path computation method is used to calculate the power supply reachability coefficient. For each power supply path, the value of the branch with the weakest evidence is taken as the path evidence, and then the maximum evidence value among all paths is taken as the reachability coefficient of that candidate branch. A weighted summation method is not used to calculate the path evidence. The above embodiment uses the branch with the weakest evidence on the power supply path as the bottleneck of physical connectivity, avoiding a single high-amplitude measurement point dominating the fault judgment result. This ensures that the evidence of power supply paths within the same topology conforms to the physical laws of electrical connectivity, improving the rationality of power supply reachability judgment.
[0028] In one alternative implementation, the pseudo-continuity coefficient calculation module is configured to perform the following operations: Step S401: Merge the closed branches of the pre-topology and post-topology to form a union graph, and take the branch with the larger value in the pre-topology time period and post-topology time period as the merged branch fault coefficient. The formula for calculating the union graph is as follows: ; in, Represents a union graph. Represents a set of nodes. Denotes the set of closed branches in the pre-topology. Represent the set of closed branches in the post-topology; The formula for calculating the fault coefficient of the merged branch is as follows: ; in, Indicates a branch The fault coefficient of the merged branch, Indicates a branch The branch fault evidence coefficient in the previous topology period, Indicates a branch Branch fault evidence coefficients in the later topology period; A union graph is a virtual network topology structure formed by taking the union of the set of closed branches of the previous topology and the set of closed branches of the subsequent topology. The combined branch fault coefficient refers to the value obtained by taking the larger value between the branch fault evidence coefficient of the previous topology period and the branch fault evidence coefficient of the subsequent topology period for each branch. Step S402: Construct a set of union paths from power nodes to candidate branches and passing through candidate branches in the union graph, and sequentially select the minimum fault coefficient of the merged branch within the path and the maximum value between paths to obtain the mixed reachability coefficient. The formula for calculating the hybrid accessibility coefficient is as follows: ; in, Indicate candidate branch The mixing reachability coefficient, Indicate candidate branch The set of union paths in a union graph. Represents the union path, Representing the union path The branch road in the middle, Indicates a branch The fault coefficient of the merged branch; The hybrid reachability coefficient refers to the reachability coefficient value from the power source to the candidate branch calculated in the union graph using the same wide-path calculation method as the single-topology power supply reachability coefficient; Step S403: Determine the optimal hybrid path according to the number of branches, total impedance amplitude, and device identifier dictionary order; combine the current measurement package of the optimal hybrid path with the terminal voltage measurement package to form a hybrid path measurement package set; when there are multiple union paths with hybrid reachability coefficients, select the union path with the fewest number of branches, the smallest total impedance amplitude, and the smallest device identifier dictionary order as the optimal hybrid path in sequence. The hybrid optimal path refers to the path that achieves the hybrid reachability coefficient in the union graph, and is a single path uniquely determined according to a preset priority. The hybrid path measurement package set refers to the set consisting of current measurement packages corresponding to all branches on the hybrid optimal path, and voltage measurement packages corresponding to the nodes at both ends of all branches; Step S404: Calculate the time cross-boundary overlap coefficient based on the overlap length between each possible measurement interval in the mixed path measurement package set and the previous and subsequent topological time periods. The formula for calculating the time crossover coincidence coefficient is as follows: ; in, Indicate candidate branch The time-crossing overlap coefficient Indicate candidate branch A collection of mixed path measurement packages, This indicates the number of measurement packets in the mixed path measurement packet set. This represents the measurement packets in the mixed path measurement packet set. Indicates measurement package The possible measurement range, Indicates the previous topological period. Indicates the post-topological period. Indicates the length of the time interval; When the mixed path measurement package set is empty, the time cross-boundary overlap coefficient is 0. The time-boundary overlap coefficient refers to the value obtained by calculating the degree to which the possible measurement intervals of each measurement packet in the mixed path measurement packet set simultaneously cover the preceding and following topological time periods, and taking the average value of all measurement packets; Step S405: Calculate the positive difference between the hybrid accessibility coefficient and the larger of the front topology power accessibility coefficient and the back topology power accessibility coefficient; multiply the ratio of the positive difference to the hybrid accessibility coefficient by the coefficient obtained by dividing the sum of 1 and the time cross-boundary coincidence coefficient by 2 to obtain the cross-topology power pseudo-continuity coefficient. The formula for calculating the pseudo-continuity coefficient of cross-topology power supply is as follows: ; in, Indicate candidate branch The pseudocontinuity coefficient of cross-topological energy supply, Indicate candidate branch The mixing reachability coefficient, Indicate candidate branch The topological energy accessibility factor, Indicate candidate branch The power availability factor of the back topology. Indicate candidate branch The time-crossing overlap coefficient Indicates taking The larger of 0; When the hybrid reachability coefficient is 0, the pseudo-continuity coefficient of cross-topology power supply is 0.
[0029] The cross-topology power supply pseudo-continuity coefficient is a value calculated by combining the positive difference between the hybrid reachability coefficient and the single-topology maximum reachability coefficient with the time cross-boundary coincidence coefficient. Furthermore, when multiple paths achieve a mixed reachability coefficient, the first priority is to select the path with the fewest branches. If the number of branches is the same, the second priority is to select the path with the smallest sum of equivalent impedance values of all branches. If both the number of branches and the total impedance are the same, the third priority is to sort the branches lexicographically by their device identifiers and select the path with the highest sorting order. The total impedance is the sum of the equivalent impedance values of all branches on the path, and the device identifiers are sorted lexicographically by their character codes in ascending order. For example, if two paths both achieve a mixed reachability coefficient of 0.8, the first path contains 5 branches with a total impedance of 0.4 ohms, and the second path contains 4 branches with a total impedance of 0.5 ohms. The second path with fewer branches is preferred as the optimal mixed path.
[0030] Furthermore, when the difference between the hybrid reachability coefficient and the maximum reachability coefficient of a single topology is positive, the ratio of this difference to the hybrid reachability coefficient is taken as the topology pseudo-increment; the time cross-boundary coincidence coefficient is added by 1 and divided by 2 as the time weight, and the topology pseudo-increment is multiplied by the time weight to obtain the pseudo-continuity coefficient; when the hybrid reachability coefficient is 0, or the hybrid reachability coefficient is not greater than the maximum reachability coefficient of a single topology, the pseudo-continuity coefficient is set to 0.
[0031] Understandably, constructing a union graph of the pre- and post-topologies simulates the data fusion effect of a traditional unified time window. It calculates the maximum reachability evidence when topological mutual exclusion is ignored, compares it with the reachability evidence under the real topology to obtain spurious increments, and then combines this with the time boundary degree of the measurement packets to generate a cross-topology power supply pseudo-continuity coefficient. The above embodiment constructs a comparison result of erroneous data fusion using a virtual union graph, which can quantitatively identify spurious power supply paths formed solely by cross-topology data splicing. This provides a quantifiable basis for suppressing spurious evidence during subsequent fault location, improving fault identification capabilities in topology change scenarios.
[0032] In one alternative implementation, the fault location module is configured to perform the following operations: Step S501: Calculate the square root of the product of the branch fault evidence coefficient and the previous topology power accessibility coefficient of the candidate branch in the previous topology period, and then multiply it by 1 and the difference between the cross-topology power pseudo-continuity coefficient to obtain the confidence coefficient of the faulty branch. The formula for calculating the confidence coefficient of the faulty branch is as follows: ; in, Indicate candidate branch The confidence coefficient of the faulty branch, Indicate candidate branch The branch fault evidence coefficient in the previous topology period, Indicate candidate branch The topological energy accessibility factor, Indicate candidate branch The pseudo-continuity coefficient of cross-topological energy supply; The confidence coefficient of a faulty branch is the value obtained by taking the square root of the product of the branch fault evidence coefficient of the candidate branch in the previous topology and the power reachability coefficient of the previous topology, and then multiplying it by 1 and subtracting the difference between the pseudo-continuity coefficient of cross-topology power supply. Step S502: Select the candidate branch with the largest confidence coefficient of the fault branch as the only fault branch; when the confidence coefficients of the fault branches are the same, first select the candidate branch with the smallest cumulative impedance amplitude to the first protected device, and then select the candidate branch with the smallest dictionary order of the device identifier. The cumulative impedance magnitude refers to the sum of the equivalent impedance magnitudes of the branches along the power supply path from the power node to the candidate branch in the previous topology, starting from the power supply end of the candidate branch on the power supply path and ending at the location of the first activated protection device. When there are multiple power supply paths, the path with the fewest branches is selected first. When the number of branches is the same, the path with the smallest cumulative impedance magnitude is selected. The unique faulty branch is the branch with the highest confidence coefficient among all candidate branches. In step S503, when the confidence coefficient of the faulty branch of all candidate branches is 0, the first branch of the downstream protection zone of the first protection device is determined as the only faulty branch, and a transfer prohibition sign is generated.
[0033] The transfer prohibition flag is a status flag used to prevent the automatic closing of the tie switch. When the confidence coefficient of the faulty branch for all candidate branches is 0, or when there is no candidate tie switch, the transfer prohibition flag is set to an active state; otherwise, the transfer prohibition flag remains inactive. Furthermore, when calculating the cumulative impedance magnitude, the power supply path from the power source to the candidate branch in the previous topology is followed. From the power source end of the candidate branch to the installation location of the first-operating protection device, the equivalent impedance magnitude of the line is accumulated branch by branch, without accumulating impedance in the reverse power supply direction. For example, if the power source end of a candidate branch is 0.1 ohms away from the branch switch, the branch switch is 0.2 ohms away from the outgoing protection switch, and the outgoing protection is the first-operating device, then the cumulative impedance magnitude of this candidate branch is 0.3 ohms.
[0034] Furthermore, when the confidence coefficient of the fault branch for all candidate branches is 0, the protection range of the first-acting protection device is taken as the downstream protection zone, and the first branch closest to the protection device within the protection zone is taken as the first branch segment, which is then identified as the fault branch. The downstream protection zone is the set of all branches covered by the protection range set by the protection device, and the first branch segment is the first branch segment connected to the outlet of the protection installation location.
[0035] Understandably, the fault evidence of the candidate branch itself is coupled with the power availability evidence within the preceding topology. Then, false evidence is suppressed using a cross-topology pseudo-continuity coefficient to obtain the confidence coefficient of the faulty branch. The branch with the highest confidence coefficient is selected as the unique faulty branch, avoiding the simultaneous output of multiple faulty branches. This embodiment, by suppressing false fault candidates formed through cross-topology splicing using a pseudo-continuity coefficient, can identify the true faulty branch during the topological transient process of switch tripping, avoiding misjudging the fault point as being on the non-faulty side of the tripping switch and improving the accuracy of fault section location.
[0036] In one alternative implementation, the isolation transfer module is configured to perform the following operations: Step S601: Construct a path from the power node to any endpoint of the unique fault branch in the back topology; select the first closed switch with remote tripping capability along each path from the endpoint of the unique fault branch towards the power node; when there are multiple sets of isolating switches that can cut off all paths from the power node to the endpoint of the unique fault branch, first select the set of isolating switches with the fewest number of switches, then select the set of isolating switches with the smallest sum of cumulative impedance amplitudes from each switch to the unique fault branch; if the path still exists after deleting the set of isolating switches, select the remotely controllable switch closest to the unique fault branch along the uncut path towards the power side and add it to the set of isolating switches. A set of disconnecting switches refers to a set of remotely controllable switches in a back topology that can disconnect the power supply path from all power nodes to any endpoint of a faulty branch. Step S602: Identify healthy power-depleted load blocks from the post-topology after deleting the set of disconnecting switches; the healthy power-depleted load block is the largest connected subgraph that does not contain a unique fault branch and is in a power-depleted state; the normally open remote controllable switch that connects the energized connected domain to the healthy power-depleted load block, does not form a path from the power node to the unique fault branch after closing, and makes the load block to be restored have only one power supply path is determined as a candidate tie switch, and the healthy power-depleted load block that restores power supply after the candidate tie switch is closed is determined as the load block to be restored; A healthy power failure load block refers to the largest connected subgraph in the subsequent topology that does not contain faulty branches and has no power connection. A candidate tie switch is a normally open switch that is in the open state, has remote control capability, and meets the conditions of energized connection of the domain, isolation of fault branches, and single power supply path of the load block to be restored. Whether the candidate tie switch meets the thermal capacity constraint and voltage constraint after closing is determined based on the safety margin factor obtained by subsequent calculation. A load block to be restored refers to a healthy power-loss load block that can obtain power from the backup power source after the candidate tie switch is closed. Step S603: For the power supply path after the candidate tie switch is closed, add the original load baseline current of the branch to the load block boundary baseline current of the downstream load block to be restored to obtain the current envelope. The formula for calculating the current envelope is as follows: ; in, Indicates candidate contact switch After closing the branch The current envelope, Indicates a branch The original load baseline current of the branch is taken as the branch within the stable window before the protection operation. The median current amplitude when the original healthy load is borne. Indicates candidate contact switch After closure, it is located in the branch. The downstream set of load blocks to be restored Indicates the load block to be restored The load block boundary baseline current; Current envelope refers to the conservative upper limit of the predicted current in each branch of the power supply path after the tie switch is closed. Step S604: Calculate the thermal capacity margin and voltage drop margin of the candidate tie switch power supply path, and determine the smaller value between the thermal capacity margin and the voltage drop margin as the safety margin factor. The formula for calculating the safety margin factor is as follows: ; in, Indicates candidate contact switch Safety margin factor, Indicates candidate contact switch The power supply path after closure Indicates a branch in the power supply path. Indicates a branch The branch allows continuous current. Indicates candidate contact switch After closing the branch The current envelope, Indicates the rated voltage amplitude of the network. Indicates the minimum allowable voltage amplitude. Indicates a branch The equivalent impedance amplitude of the line is calculated based on a voltage reference that is the same as the rated voltage amplitude of the network; the first item represents the thermal capacity margin, and the second item represents the voltage drop margin. The safety margin factor is a value obtained by taking the minimum value of the thermal capacity margin and voltage drop margin on the power supply path; Step S605: Select the candidate tie switch with the largest safety margin coefficient as the sole tie switch; when the safety margin coefficients are the same, select the candidate tie switch with the largest number of load blocks to be restored, the smallest number of additional switch actions, and the smallest equipment identifier dictionary order in sequence; when there is no candidate tie switch, set the sole tie switch field to an empty identifier and generate a transfer prohibition flag.
[0037] The unique tie switch refers to the switch with the largest safety margin factor among all candidate tie switches. Furthermore, in the post-topology, for each path from a power node to any endpoint of the uniquely faulty branch, the first closed switch with remote tripping capability on that path is selected in the direction from the endpoint of the uniquely faulty branch to the power node. The selected switches for each path are deduplicated to form an initial set of isolating switches. After deleting the switches in the initial set of isolating switches, a connectivity check is performed again. If there are still uninterrupted power supply paths, the nearest remotely controllable closed switch to the uniquely faulty branch is selected along that uninterrupted path towards the power node and added to the isolating switch set, until there are no more paths from the power node to any endpoint of the uniquely faulty branch. When multiple switch sets exist that can all achieve isolation, the set with the fewest switches is selected first; when the number of switches is the same, the set with the smallest sum of cumulative impedance values from each switch to the uniquely faulty branch is selected. For example, if the faulty branch has two power supplies, the first scheme requires disconnecting two switches with a total distance corresponding to an impedance of 0.6 ohms, and the second scheme requires disconnecting two switches with a total distance corresponding to an impedance of 0.4 ohms. The second scheme with the smaller total impedance is preferentially selected as the isolating switch set.
[0038] Furthermore, the regions in the topology that are connected to the power source are identified as energized connected regions, and the regions that are not connected to the power source and do not contain faulty branches are identified as healthy power-depleted load blocks. All normally open remote controllable switches are traversed. If a switch is connected to an energized connected region and a healthy power-depleted load block at both ends respectively, and does not form a path from the power source to the faulty branch after being closed, and each load block to be restored only receives power from a single power source through this switch, then it is determined to be a candidate tie switch. The verification order is as follows: verification of the connectivity attributes at both ends, verification of the connectivity of the faulty branch, and verification of the single power supply path.
[0039] Furthermore, the product of the line equivalent impedance magnitude and the current envelope is the conservative voltage drop value on the line. The total voltage drop is obtained by summing the voltage drops on all power supply paths. The conservative voltage drop design is to take the direct sum of all load currents, without considering the phase cancellation of load currents, to obtain the maximum possible value of voltage drop. The thermal capacity margin is the ratio of the branch allowable continuous current minus the current envelope to the allowable continuous current. The voltage drop margin is the ratio of the difference between the rated voltage and the minimum allowable voltage minus the total voltage drop to the difference between the rated voltage and the minimum allowable voltage. The smaller of the two margins is taken as the safety margin factor.
[0040] Understandably, a minimum set of disconnecting switches is generated based on the post-fault topology. A conservative approach of directly adding load currents is used to calculate the branch current envelope, without setting artificial weights. The minimum margin between thermal capacity and voltage constraints is taken as the safety margin, and the tie switch with the largest safety margin is selected as the sole transfer switch. This embodiment uses a conservative current envelope to calculate the branch current after transfer, and simultaneously verifies thermal stability and voltage quality through an unweighted minimum operator. This avoids the situation where insufficient margin in a single indicator is masked by high margins in other indicators, ensuring that all line operating parameters meet safety requirements after the tie switch is closed.
[0041] In one alternative implementation, the control message output module is configured to perform the following operations: Step S701: Generate an event identifier and execution batch number based on the topology transition event, and send a tripping command to the switches in the disconnector set according to the execution batch; the tripping command includes an event identifier, device identifier, action type, execution batch number and integrity verification field; An event identifier is a unique numerical code generated in chronological order when the system detects a topology transition fault event. The execution batch number refers to the batch code generated in the order of the switch commands issued under the same fault event; Step S702: Receive the position feedback of the disconnecting switch and the amplitude of the isolation boundary current; when the disconnecting switch remains in the open state for the duration of the state holding time and the amplitude of the isolation boundary current is less than the no-current threshold, confirm the isolation condition; The no-current threshold satisfies: ; in, Indicates the threshold of no current flow. Represents a set of disconnecting switches. This indicates the branch containing the switch in the disconnector switch set. Indicates a branch The branch allows continuous current; The isolation condition is met when all switches in the disconnecting switch set remain in the open state for the duration of the state holding time, and the amplitude of each isolation boundary current is less than the no-current threshold. Step S703: When there is a unique tie switch, the isolation condition is met, the safety margin coefficient of the unique tie switch is greater than 0 and no transfer prohibition flag is generated, the closing enable bit of the unique tie switch closing command is set to 1; otherwise, the closing enable bit is set to 0. The closing enable bit is a binary flag bit in the control message that indicates whether the closing command of the tie switch is allowed to be executed; Step S704: Write the unique fault branch into the unique fault branch identifier, generate a list of disconnector opening commands in ascending order of equipment identifiers, write the unique tie switch and closing enable bit into the unique tie switch closing command, and form a control message. Control messages are executable messages generated according to the power distribution terminal communication protocol format, which include fault section identifiers, switch opening commands, switch closing commands, and verification fields. Step S705: Perform length verification, device existence verification, switch current status verification, and integrity verification on the control message, and send the verified control message to the distribution automation terminal.
[0042] Furthermore, the state holding time begins from the moment the distribution master station receives the last disconnect signal from the disconnector; the isolation boundary current is the measured current value of the disconnector on the fault side, using the current sampling value sent by the terminal in real time, with a sampling interval of no more than 20 milliseconds; isolation is determined to be complete when all disconnectors remain in the disconnected position for the state holding time, and the current amplitude of all isolation boundaries is lower than the no-current threshold. For example, if the state holding time is 2 seconds, the no-current threshold is 1 ampere, and the master station receives the disconnect signal of the last disconnector at 10:00:00, and all switches remain in the disconnected position until 10:00:02, and the current on the fault side of all switches is less than 1 ampere, the fault branch is determined to be completely isolated.
[0043] Furthermore, length verification checks whether the total message length meets the fixed length requirements stipulated in the communication protocol; device existence verification checks whether all switch identifiers in the message exist in the system device register; switch current status verification is performed based on the command to be executed and its execution time. Before sending the disconnector switch opening command, the corresponding disconnector switch is checked to be in the closed position; before sending the tie switch closing command, the corresponding tie switch is checked to be in the open position; for disconnectors that have already been opened and whose isolation conditions have been confirmed, their open position status is taken as the isolation completion result and is no longer verified according to the closed position condition before the opening command was sent; integrity verification refers to calculating the integrity verification value according to the verification rules stipulated in the communication protocol adopted by the distribution automation terminal, based on the corresponding fields in the control message, and comparing the calculation result with the integrity verification field carried in the control message. If they match, the integrity verification passes; if they do not match, the integrity verification fails, and the corresponding control command is not sent to the distribution automation terminal. If any verification fails, the message is discarded and a new control message is generated, and no verification failure message is sent to the terminal.
[0044] Furthermore, the verification sequence is as follows: isolation completion verification, safety margin greater than 0 verification, and transfer prohibition flag invalid verification. When all conditions are met, the closing enable position is 1, allowing the execution of the unique tie switch closing command. When any condition is not met, the closing enable position is 0, the control message still retains the unique tie switch field and the closing enable bit, but the corresponding closing command cannot be executed. When there is no unique tie switch, the unique tie switch field is set to an empty flag, and the closing enable bit is set to 0.
[0045] Understandably, the system employs a dual-condition approach—the duration of the switch's open position and the current-free determination at the isolation boundary—to confirm the complete disconnection of the faulty branch. The isolation status, power transfer safety margin, and power transfer prohibition flag are used as interlocking conditions to control the closing enable bit. Control messages are generated in a fixed sequence: first, the isolation switch is opened, then the connection switch is closed. This embodiment, through dual isolation confirmation and a closing interlocking mechanism, avoids mistakenly sending closing commands to faulty branches that are not fully disconnected, prevents the fault point from becoming energized again during power transfer, and reduces the risk of secondary faults, equipment damage, and the expansion of the power outage area.
[0046] It should be noted that this system is deployed on the distribution automation master station server, accessing real-time data collected from feeder terminals and substation terminals through the existing distribution communication network. It interfaces with the equipment ledger, topology model, and protection configuration data of the distribution automation system, without altering the existing terminal sampling configuration and communication protocol during operation. The system monitors switch status and protection action signals in real time. Upon detecting a fault trip event, it automatically initiates the fault handling process, ultimately outputting three types of results that can be executed by the terminals: a unique fault branch identifier, a list of disconnector switch opening commands, and a tie switch closing command with enabled status.
[0047] Taking a typical 10kV overhead feeder scenario as an example, this feeder connects to a distributed photovoltaic system with a total capacity of 3MW. The outgoing switch is equipped with overcurrent protection, a sectionalizing switch is installed in the middle section of the line, and the end is interconnected with adjacent feeders through a tie switch. When a two-phase short-circuit fault occurs in the middle section of the feeder, the outgoing switch protection trips. After the trip, the downstream distributed photovoltaic system continues to provide a current-limiting current of 1.2 times the rated current. Due to network time delay, the current measurement timestamps of some terminals have a 10ms error, spanning the switch tripping time. After the system starts, it first generates dual topologies before and after the trip, calculates electrical disturbance evidence for each time period, identifies false power supply paths spliced across topologies, determines that the fault point is located in the branch between the middle section sectionalizing switch and the outgoing switch, generates a tripping command to disconnect the middle section sectionalizing switch, verifies the current and voltage margin transferred by the tie switch after the isolation is confirmed, and closes the tie switch to restore the power supply to the non-faulty loads downstream of the sectionalizing switch. The entire process requires no manual intervention, and control commands can be issued to the corresponding terminals for execution.
[0048] like Figure 2As shown in the figure, this diagram reflects the power supply reachability, pseudo-continuity, and fault location results of 30 candidate branches during topology state switching. The upper part shows that the mixed reachability coefficients of branches 07, 12, 23, and 27 are significantly higher than the maximum reachability coefficient of a single topology, and they also have high cross-topology power supply pseudo-continuity coefficients. This indicates that the strong reachability evidence for these branches mainly comes from cross-time period combinations of data before and after the circuit breaker trip, rather than a continuous and complete power supply path under any real topology. The lower part further demonstrates the impact of pseudo-continuity suppression on fault confidence. The final confidence of the aforementioned false candidate branches is significantly reduced, while branch 17, due to its strong fault evidence and the previous topology power supply reachability evidence, and its low pseudo-continuity, was ultimately identified as the only fault branch with the highest confidence of 0.9023. This result shows that jointly analyzing current, voltage, and network topology, and identifying false paths caused by topology state inconsistencies, helps reduce misjudgments of fault sections during switching operations; further details are omitted here.
[0049] like Figure 3 As shown in the figure, this diagram reflects the transfer carrying capacity and operational safety level of 30 candidate tie switches after the faulty branch is isolated. The left side compares the allowable continuous current and predicted current envelope on the power supply path of each candidate scheme. All predicted values are lower than the corresponding allowable values, indicating that these schemes do not directly exceed the line's continuous current carrying capacity under the current simulated load conditions, but the remaining capacity varies significantly among different schemes. The right side simultaneously provides the thermal capacity margin, voltage drop margin, and comprehensive safety margin. The comprehensive result is the smaller of the two constraints; therefore, any value approaching the limit will lower the safety level of the candidate scheme. Tie switch 14 has a predicted current of 330 amps, a thermal capacity margin of 0.450, a voltage drop margin of 0.550, and a comprehensive safety margin of 0.450, the highest among all options, thus it was selected as the sole tie switch. Although tie switches 21, 29, and others still meet the basic current conditions, their comprehensive margins are lower, making them more susceptible to load increases or voltage drops during operation. Distribution network restoration studies typically require tie switch reconfiguration to simultaneously meet branch current and node voltage limits; this will not be elaborated upon here.
[0050] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and all such modifications and variations fall within the scope defined by the appended claims.
Claims
1. A power distribution network fault situation intelligent perception and auxiliary decision-making system, characterized in that, include: The topology period generation module divides the previous topology period into the next topology period based on the base diagram, switch status, protection action flag, timestamp, and maximum error boundary of timestamp, and generates the previous topology, next topology, measurement interval, and electrical quantity baseline data. The electrical disturbance module processes the current amplitude and voltage amplitude according to the overlap length between the measurement interval and the previous and subsequent topological time periods to obtain the interval overlap coefficient, current disturbance energy, voltage drop energy and branch fault evidence coefficient. The reachability coefficient calculation module constructs a set of paths from power nodes to candidate branches in the front and back topologies, and determines the single-topology power supply reachability coefficient based on the fault evidence coefficient of the branch within the path. The pseudo-continuity coefficient calculation module merges the closed branches of the pre-topology and post-topology to form a union graph, uses the branch fault evidence coefficient to determine the hybrid reachability coefficient and the time cross-boundary coincidence coefficient, and generates the cross-topology power supply pseudo-continuity coefficient. The fault location module combines the branch fault evidence coefficient, the single-topology power supply reachability coefficient, and the cross-topology power supply pseudo-continuity coefficient to determine the confidence coefficient of the faulty branch and select the unique faulty branch. The isolation transfer module determines the set of disconnecting switches and healthy power-off load blocks based on the subsequent topology and the only faulty branch, calculates the current envelope and safety margin factor of candidate tie switches, and selects the only tie switch. The control message output module receives feedback on the position of the disconnecting switch and the amplitude of the isolation boundary current, confirms the isolation conditions, and generates a unique fault branch identifier, a list of disconnecting switch opening commands, and a unique tie switch closing command.
2. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The topology time period generation module is configured to execute: Verify the correspondence between the switch equipment identification, power node identification and the branches and nodes in the basic diagram. Based on the protection action flag, determine the first protection action switch state transition time as the topology boundary time, and classify the protection switch actions that occur consecutively within the anti-jitter time into the same topology transition event. Read the rated current amplitude of the branch, the rated voltage amplitude of the node, the allowable continuous current of the branch, the equivalent impedance amplitude of the line, the rated voltage amplitude of the network, the minimum allowable voltage amplitude, the remote control attributes of the switch, the normally open and normally closed attributes of the switch, the length of the single-sided fault analysis window, the length of the stability window, the state holding time, and the no-current threshold. Read the switching states that continuously reach the debouncing time before and after the topology boundary moment to form the front topology and the back topology; Establish the pre-topology time period and post-topology time period according to the topology boundary time and the length of the single-sided fault analysis window. Divide the load blocks according to the power failure connectivity in the post-topology. Calculate the branch baseline current amplitude, node baseline voltage amplitude, load block boundary baseline current and branch original load baseline current within the stability window before protection operation to form electrical quantity baseline data. The measurement interval is determined based on the measurement packet timestamp and the maximum error boundary of the timestamp, and the sampling duration of the measurement packet is read.
3. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The electrical disturbance module is configured to perform: Calculate the overlap length between the measurement interval and the preceding and following topological time periods, and determine the ratio of the overlap length to the measurement interval length as the interval overlap coefficient; The current disturbance energy is obtained by multiplying the square of the difference between the cumulative current amplitude of the interval overlap coefficient and the baseline current amplitude of the branch and the sampling duration, and normalizing it by using the square of the rated current amplitude of the branch and the weighted sampling duration. The voltage drop energy is obtained by multiplying the cumulative node baseline voltage amplitude by the square of the positive difference between the voltage amplitude and the sampling duration based on the interval overlap coefficient, and then normalizing the product of the square of the node rated voltage amplitude and the weighted sampling duration. Select the voltage drop energy with the larger value at both ends of the branch, calculate the square root of the product of the selected voltage drop energy and the current disturbance energy, and obtain the branch fault evidence coefficient.
4. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The reachability coefficient calculation module is configured to perform: The branches in the connected domain where the first protective device is located and the adjacent power failure connected domain are identified as candidate branches; For each candidate branch, a set of simple paths from the power node to the candidate branch and passing through the candidate branch is constructed in both the front and back topologies. For each simple path, the smallest branch fault evidence coefficient is selected within the path. Then, the path with the largest value is selected from the set of simple paths to obtain the front topology power accessibility coefficient and the back topology power accessibility coefficient, which are collectively referred to as the single topology power accessibility coefficient.
5. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The pseudo-continuity coefficient calculation module is configured to perform: The closed branches of the pre-merge topology and the post-merge topology form a union graph. The fault evidence coefficient of the branch with the larger value in the pre-merge topology time period and the post-merge topology time period is taken as the fault coefficient of the merged branch. In the union graph, construct a set of union paths from power nodes to candidate branches and passing through candidate branches. Then, select the minimum fault coefficient of the merged branch within the path and the maximum value between paths to obtain the mixed reachability coefficient. The optimal hybrid path is determined based on the number of branches, total impedance amplitude, and lexicographical order of equipment identification. The current measurement package of the optimal hybrid path is combined with the terminal voltage measurement package to form a hybrid path measurement package set. The time crossover coincidence coefficient is calculated based on the overlap length between each measurement interval in the mixed path measurement package set and the preceding and following topological time periods. Calculate the positive difference between the hybrid accessibility coefficient and the larger of the forward topology power accessibility coefficient and the backward topology power accessibility coefficient; multiply the ratio of the positive difference to the hybrid accessibility coefficient by a coefficient obtained by dividing the sum of 1 and the time cross-boundary coincidence coefficient by 2 to obtain the cross-topology power pseudo-continuity coefficient.
6. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The fault location module is configured to execute: Calculate the square root of the product of the branch fault evidence coefficient and the previous topology power accessibility coefficient of the candidate branch in the previous topology period, and then multiply it by 1 and the difference between the cross-topology power pseudo-continuity coefficient to obtain the confidence coefficient of the faulty branch. The candidate branch with the highest confidence coefficient of the faulty branch is selected as the unique faulty branch. When the confidence coefficients of the fault branches are the same, the unique fault branch is determined by the cumulative impedance amplitude from the candidate branch to the first protected device and the dictionary order of the device identification. When the confidence coefficient of the faulty branch of all candidate branches is 0, the first branch of the downstream protection zone of the first protective device is determined as the only faulty branch, and a transfer prohibition sign is generated.
7. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The isolation and transfer module is configured to execute: In the post-topology, construct the path from the power node to any endpoint of the unique fault branch; along each path, select the first closed switch with remote tripping capability from the endpoint of the unique fault branch towards the power node; determine the disconnector switch set according to the number of switches and the sum of the cumulative impedance amplitudes of each switch to the unique fault branch, so that there is no path from the power node to any endpoint of the unique fault branch in the post-topology after deleting the disconnector switch set. Identify healthy power-loss load blocks from the post-topology after deleting the set of disconnecting switches; identify normally open remote controllable switches that connect energized domains and healthy power-loss load blocks, do not form a path from a power node to the only fault branch after closing, and ensure that the load block to be restored has only one power supply path as candidate tie switches; and identify healthy power-loss load blocks that are restored to power supply after the candidate tie switches are closed as load blocks to be restored. For the power supply path after the candidate tie switch is closed, the original load baseline current of the branch is added to the load block boundary baseline current of the downstream load block to be restored to obtain the current envelope; Calculate the thermal capacity margin and voltage drop margin of the candidate tie switch power supply path, and determine the smaller value of the thermal capacity margin and voltage drop margin as the safety margin factor; The candidate tie switch with the largest safety margin coefficient is selected as the sole tie switch. When the safety margin coefficients are the same, the unique tie switch is determined according to the number of load blocks to be restored, the number of additional switch actions, and the dictionary order of equipment identification; if there is no candidate tie switch, the unique tie switch field is set to an empty identifier and a transfer prohibition flag is generated.
8. The intelligent sensing and auxiliary decision-making system for distribution network fault conditions according to claim 1, characterized in that, The control message output module is configured to execute: Based on the topology transition event, an event identifier and execution batch number are generated, and a tripping command is sent to the switches in the disconnector set according to the execution batch; the tripping command includes an event identifier, device identifier, action type, execution batch number, and integrity verification field; Receive the position feedback of the disconnecting switch and the amplitude of the isolation boundary current; confirm the isolation condition when the disconnecting switch remains in the open state for the duration of the state holding time and the amplitude of the isolation boundary current is less than the no-current threshold. When a unique interconnecting switch exists, the isolation condition is met, the safety margin coefficient of the unique interconnecting switch is greater than 0, and no transfer prohibition flag is generated, the closing enable bit of the unique interconnecting switch closing command is set to 1; otherwise, the closing enable bit is set to 0. Write the unique fault branch into the unique fault branch identifier, generate a list of disconnector opening commands in ascending order of equipment identifiers, write the unique tie switch and closing enable bit into the unique tie switch closing command, and form a control message. The control message is checked for length, device existence, switch current status and integrity. Control messages that pass the checks are sent to the distribution automation terminal.
