Rapid fault location and isolation method for power distribution network based on transient characteristic information

Abstract

The invention discloses a rapid fault positioning and isolating method for a power distribution network based on transient characteristic information, and the method comprises the steps: carrying out the preliminary positioning through the current information of a selected key switch, and judging the rough section range of a fault; if the preliminary positioning shows that the fault is in two certain sections, calculating the minimum unit of the fault by combining the transient voltage and current information of the switch flowing through the large fault current and the line model; if the calculated fault distance indicates that the fault only exists in a certain section, fault positioning is completed; if the fault possibly exists in the two sections, the position of the fault is confirmed by further combining the voltage data of the other key switches and the voltage data of the transformer area, and finally the fault is positioned in the single section; after the fault positioning is completed, the section switches and the interconnection switches around the fault section are quickly switched off, and the fault section is isolated. The method improves the fault recovery speed, improves the fault positioning precision, reduces the influence degree of the service life of the switch, and reduces the equipment maintenance cost and the fault risk.

Description

Technical Field

[0001] The present invention relates to a data modeling and prediction method based on artificial intelligence technology, and particularly to an intelligent data modeling method applicable to power prediction of a wind farm. Technical Background

[0002] In modern power systems, the reliability and power supply quality of the distribution network directly affect the daily life of the general public and the normal operation of industrial production. Fault location, as a key link in distribution network management, plays a crucial role. Quickly and accurately locating the fault point can not only significantly shorten the power outage time, improve power supply reliability, but also effectively reduce the economic losses and social impacts caused by power outages. In addition, accurate fault location can optimize the allocation of emergency repair resources, reduce operation and maintenance costs, and at the same time provide data support for the intelligent operation and maintenance, upgrading and transformation of the distribution network. With the access of distributed energy and the development of smart grids, the complexity of the distribution network is increasing continuously, and the efficiency and intelligence level of fault location technology have also become an important guarantee for improving the overall performance of the power grid. Therefore, researching and applying advanced fault location technology has extremely important practical significance for improving the operation efficiency, reliability and intelligence level of the distribution network.

[0003] However, the existing manual fault location methods have the disadvantages of long time consumption, slow recovery speed, limited location accuracy, etc. For example, fault troubleshooting relies on manual line patrol and on-site operations, requiring multiple trial power transmissions and switch operations, resulting in more time consumption and damage to the switch life. At the same time, manual judgment is easily affected by subjective factors, unable to monitor and accurately locate the fault point in real time, especially with poor adaptability in complex fault situations. In addition, this method has high requirements for switch equipment configuration and personnel professionalism, and is prone to expanding the power outage scope during the fault handling process, affecting power supply reliability and user satisfaction, and it is difficult to meet the requirements for rapid power restoration and high power supply reliability under the background of the new power system.

[0004] The current solution technologies for manual fault location are listed as follows:

[0005] 1. A protection control method and intelligent complete set of equipment for self-healing of 10kV distribution network cable lines disclosed in Chinese Patent Publication No. CN105162256B, which relies on a complex communication network and a variety of intelligent devices, has high requirements for communication stability and equipment reliability. Once the communication is interrupted or the equipment fails, it may affect the self-healing function of the fault. At the same time, the system construction cost is relatively high, and the adaptability to changes in the distribution network topology structure is limited, and it is difficult to quickly respond to grid frame adjustments.

[0006] 2. A fault location method applicable to the closed-loop operation of a distribution network disclosed in Chinese Patent Publication No. CN105548801 B has relatively high requirements for the configuration and communication of switch terminals. If a switch terminal or communication fails, it may affect the accuracy of fault location. At the same time, its location accuracy is limited by the switch distribution, making it difficult to accurately locate the fault point, and it is mainly applicable to distribution networks in closed-loop operation, with a limited scope of application.

[0007] 3. A distribution network fault location method, device and system disclosed in Chinese Patent Publication No. CN111323677 B require complex information exchange and synchronization adjustment between master and slave intelligent distribution terminals, have high requirements for communication synchronization, and have a high system complexity and cost. In addition, its fault location depends on the accuracy of the network topology. If the topology information is incorrect, it may lead to location failure or misjudgment.

[0008] 4. A distribution network ground fault section selection method and system disclosed in Chinese Patent Publication No. CN113156267 B have high requirements for data collection and processing. The extraction of characteristic quantities is easily affected by noise interference, which affects the accuracy of fault section selection. At the same time, it is only applicable to single-phase ground faults, and the stepped cooperation requirements of circuit breakers are strict. If the setting is unreasonable, it may expand the power outage range.

[0009] 5. A local type distribution network single-phase ground fault section location method disclosed in Chinese Patent Publication No. CN114089109 B relies on the GOOSE communication network to transmit transient direction parameters, and communication failures may affect the location results. In addition, it is only applicable to single-phase ground faults, with a limited scope of application, and has insufficient adaptability to changes in the distribution network topology. Parameters need to be reconfigured during topology adjustment.

[0010] 6. A distribution network fault detection method based on a multi-scale graph convolutional siamese network disclosed in Chinese Patent Publication No. CN115659258 B is based on a multi-scale graph convolutional siamese network, has high requirements for data quality and quantity, and has a complex training process and high computational cost. The model has limited adaptability to changes in the distribution network topology, and the fault location accuracy may be affected by the model performance and the complexity of fault characteristics, making it difficult to achieve high-precision location.

[0011] Therefore, although there are innovations in the field of distribution network fault location and self-healing in the existing technologies, there are still some common deficiencies. For example, most patents rely on complex communication networks and high-precision data acquisition devices, which have extremely high requirements for communication stability, equipment reliability, and data quality. Once the communication is interrupted, the equipment fails, or the data is interfered, it will affect the reliability of the fault location and self-healing functions. At the same time, most of these systems have limited adaptability to the changes in the topological structure of the distribution network, are difficult to quickly respond to the grid framework adjustment, and have high construction costs. The positioning accuracy of some existing technologies is limited by the complexity of the switch distribution and fault characteristics, and it is difficult to achieve high-precision fault location, especially in the case of complex faults or atypical faults. In addition, some methods are only applicable to specific types of faults, such as single-phase grounding faults, with limited application scope and cannot meet the diverse fault scenario requirements of the distribution network. Summary of the Invention

[0012] The purpose of the present invention is to provide a method for rapid fault location and isolation of a distribution network based on transient characteristic information, so as to improve the fault recovery speed, improve the fault location accuracy, and reduce the impact on the switch life to reduce the equipment maintenance cost and fault risk.

[0013] A method for rapid fault location and isolation of a distribution network based on transient characteristic information provided by the present invention includes four steps: selection of key switches, preliminary location, calculation of fault distance, and confirmation of fault location; specifically as follows:

[0014] First, perform preliminary location through the current information of the selected key switches to judge the approximate section range of the fault; divide the distribution network line into several minimum units at intervals of several meters. If the preliminary location shows that the fault is in a certain two sections, then combine the transient voltage and current information of the switch through which the large fault current flows and the line model to calculate the minimum unit where the fault is located; if the calculated fault distance indicates that the fault only exists in a certain section, then the fault location is completed; if the fault may exist in two sections, then further combine the voltage data of the remaining key switches and the substation area to confirm the fault location, and finally locate the fault in a single section; after the fault location is completed, quickly disconnect the sectional switches and tie switches around the fault section to isolate the fault area.

[0015] In the above method for rapid fault location and isolation of a distribution network based on transient characteristic information, the selection of the key switches includes the following steps:

[0016] (1) Determine the starting switch, starting from the switch at the head of the line as the starting point of the selection process;

[0017] (2) Obtain the outermost switches of the section, and obtain the outermost switches of two sections downstream from the starting switch, which include tie switches and sectional switches;

[0018] (3) Remove the tie switch. Among the outermost switches obtained, remove the tie switch, and the remaining switches are the critical switches;

[0019] (4) Repeat the operation. Starting from the newly determined critical switches, repeat steps (2) and (3) to gradually traverse the entire power grid architecture until all sections are covered;

[0020] (5) Determine the final critical switches. After the traversal is completed, all the determined critical switches are used as the core measurement points for fault location, and voltage and current measurement devices are installed on the required switches.

[0021] In the above-mentioned method for rapid fault location and isolation of a distribution network based on transient characteristic information, the following principles are followed in the selection process of the critical switches:

[0022] The number of outermost switches for every two sections shall not be less than two;

[0023] If there is only one section due to line end limitations, the outermost switch of this section can be accepted as the outermost switch;

[0024] If the above principles cannot be satisfied, the demarcation point of two sections can be selected as one of the outermost switches.

[0025] In the above-mentioned method for rapid fault location and isolation of a distribution network based on transient characteristic information, the preliminary location includes the following steps:

[0026] Step of statistically analyzing normal current data: Statistically analyze the current amplitudes of each critical switch during the normal operation of the distribution network, and calculate its average value μ and standard deviation σ;

[0027] Step of setting a threshold: Set the threshold as μ + kσ to distinguish normal current fluctuations from fault currents;

[0028] Step of detecting fault current: When a fault occurs, the critical switch will detect a significant change in current. The current amplitude of the switch upstream of the fault point will increase significantly, while the current amplitude of the switch downstream of the fault point will rapidly decrease or approach the normal level;

[0029] Step of judging the fault section: If the current amplitude of a certain critical switch is significantly higher than that of other critical switches, and the current amplitude of the critical switch downstream of it decreases significantly or approaches the normal level, it is preliminarily judged that the fault is located in the section corresponding to this critical switch;

[0030] Step of recording the preliminary location result: Record the range of the fault section obtained by the preliminary location to provide a basis for further calculation of the fault distance.

[0031] In the above-mentioned method for rapid fault location and isolation of a distribution network based on transient characteristic information, the fault distance calculation includes the following steps:

[0032] Load modeling step: Mathematically model the load through substation area information and equivalent the load to a constant impedance.

[0033] Line modeling step: Based on line parameters and types, conduct a detailed model of the line, divide the line into several minimum units, and provide a basis for subsequent fault location.

[0034] Data acquisition and storage step: When a fault occurs, quickly acquire and store the voltage and current data in the key switches.

[0035] Iterative calculation of fault distance step: Take the minimum unit as the fault point, based on different fault types, calculate the current of the power supply side (i.e., the key switch at the most upstream) in the fault section according to the composite sequence network, compare the measured current data with the calculated current data, calculate the three-phase comprehensive mean square error, traverse all minimum units, record the two units with the smallest three-phase comprehensive mean square error, and determine whether these two units are adjacent.

[0036] Determination of fault distance range step: If the two units with the smallest three-phase comprehensive mean square error are adjacent, the fault is located between these two units. If not, there may be a second fault location that needs further confirmation.

[0037] In the above-mentioned fast fault location and isolation method for a distribution network based on transient characteristic information, the confirmation of the fault location includes the following steps:

[0038] Analysis of voltage data step: For the two possible fault locations initially located, calculate the voltage data of the remaining key switches and the substation area.

[0039] Calculation of voltage error step: Compare the theoretically calculated voltage amplitude with the actually measured voltage amplitude and calculate the voltage comprehensive mean square error.

[0040] Exclusion of invalid solutions step: If the voltage comprehensive mean square error at a certain fault location is significantly greater than that at other locations, exclude this location as an invalid solution.

[0041] Confirmation of the final fault location step: Confirm the fault location with the smallest and reasonable voltage comprehensive mean square error as the true fault point and record the fault location.

[0042] Isolation of the fault area step: According to the confirmed fault location, quickly disconnect the sectional switches and tie switches around the fault section to isolate the fault area and prevent the expansion of the fault.

[0043] In the above-mentioned fast fault location and isolation method for a distribution network based on transient characteristic information, the step of iterative calculation of the fault distance includes:

[0044] When the fault type is a single-phase grounding fault, according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a single-phase grounding fault are calculated. The positive, negative, and zero-sequence components of phase B and phase C can be obtained by rotating 120° counterclockwise or clockwise, and then the positive, negative, and zero-sequence components are superimposed to obtain the full current components. The calculated current is iteratively compared with the measured voltage and current at the protection to calculate the root mean square error of each phase current. After calculating each phase current, the root mean square error of the three-phase synthesis is calculated.

[0045] In the above-mentioned method for rapid fault location and isolation of a distribution network based on transient characteristic information, the steps of iteratively calculating the fault distance include:

[0046] When the fault type is a three-phase short circuit in the composite sequence network, according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a three-phase short circuit fault can be calculated. Then, the full current components, the root mean square error of each phase current, and the root mean square error of the three-phase synthesis are calculated.

[0047] In the above-mentioned method for rapid fault location and isolation of a distribution network based on transient characteristic information, the steps of iteratively calculating the fault distance include:

[0048] When the fault type is a two-phase short circuit in the composite sequence network, according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a two-phase short circuit fault can be calculated. Next, the full current components, the root mean square error of each phase current, and the root mean square error of the three-phase synthesis are calculated.

[0049] In the above-mentioned method for rapid fault location and isolation of a distribution network based on transient characteristic information, the steps of iteratively calculating the fault distance include:

[0050] When the fault type is a two-phase grounding in the composite sequence network, according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a two-phase grounding fault can be calculated. Next, the full current components, the root mean square error of each phase current, and the root mean square error of the three-phase synthesis are calculated.

[0051] The beneficial effects of the present invention are:

[0052] 1. Improve the fault recovery speed: Existing strategies require manual operation for fault troubleshooting and power supply restoration when facing switches without automatic switching functions. Especially in complex fault situations, it takes a long time and cannot meet the high requirements of urban distribution networks for power supply reliability.

[0053] 2. Improve the fault location accuracy: Existing strategies can only judge the general fault section in complex fault situations and cannot accurately determine the fault location, resulting in low fault troubleshooting efficiency and increasing the time and difficulty of power supply restoration.

[0054] 3. Reduce the impact on switch life: Frequent test transmission and switch operations will seriously affect the switch life, increase equipment maintenance costs and failure risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flow chart of the method for rapid fault location and isolation of distribution network based on transient characteristic information of the present invention;

[0056] Figure 2 This is one of the application scenarios of the present invention, a topological diagram of a 10kV distribution network. DETAILED DESCRIPTION

[0057] The present invention will be further described below in conjunction with the accompanying drawings.

[0058] See also Figure 1 The present invention proposes a method for rapid fault location and isolation of a distribution network based on transient characteristic information. First, preliminary positioning is performed through the current information of key switches to determine the approximate section range of the fault. If the preliminary positioning shows that the fault may be in two sections, the transient voltage and current information of the switch through which the large fault current flows and the line model are combined to calculate the minimum unit where the fault is located. If the calculated fault distance indicates that the fault only exists in one section, the fault location is completed; if the fault may exist in two sections, the voltage data of the remaining key switches and the substation are further combined to confirm the fault location, and finally the fault is located in a single section. After the fault location is completed, the section switches and connecting switches around the fault section are quickly disconnected to isolate the fault area.

[0059] As can be seen from the above, the present invention mainly includes four key links: selection of key switches, preliminary positioning, fault distance calculation and fault location confirmation. The following is a detailed description of these four links:

[0060] 1. Selection of key switches

[0061] The selection of key switches is the basis of the rapid power restoration and self-healing strategy, and the purpose is to achieve accurate positioning of the fault section through a small amount of hardware modification and measurement devices.

[0062] (1) Determine the start switch

[0063] Start with the switch at the head of the line, usually the substation outgoing line switch, as the starting point of the selection process.

[0064] (2) Get the outermost switch of the segment

[0065] From the starting switch, obtain the outermost switches of the two sections downstream, which may include tie switches and section switches.

[0066] (3) Eliminate the contact switch

[0067] Among the obtained outermost switches, remove the tie switches (switches used to connect different lines), and the remaining switches are the key switches.

[0068] (4) Repeat the operation

[0069] Taking the newly determined key switches as the starting point, repeat steps 2 and 3 to gradually traverse the entire power grid structure until all sections are covered. The following principles should be followed during the selection process:

[0070] The number of outermost switches for every two sections shall not be less than two.

[0071] If there is only one section due to line end limitations, the outermost switch of this section can be accepted as the outermost switch.

[0072] If the above principles cannot be satisfied, the demarcation point between two sections can be selected as one of the outermost switches.

[0073] (5) Determine the final key switches

[0074] After the traversal is completed, all the determined key switches are used as the core measurement points for fault location, and voltage and current measurement devices are installed on the required switches.

[0075] 2. Preliminary location

[0076] Preliminary location is to quickly judge the approximate section range of the fault through the current information of the key switches.

[0077] (1) Statistic of normal current data

[0078] Statistic the current amplitudes of each key switch during the normal operation of the distribution network, and calculate their average value μ and standard deviation σ.

[0079] (2) Set the threshold

[0080] Set the threshold as μ + kσ, where k is usually taken as 2 or 3, which is used to distinguish normal current fluctuations and fault currents.

[0081] (3) Detection of fault current

[0082] When a fault occurs, the key switches will detect a significant change in current. The current amplitude of the switches upstream of the fault point will increase significantly, while the current amplitude of the switches downstream of the fault point will rapidly decrease or approach the normal level.

[0083] (4) Judgment of the fault section

[0084] If the current amplitude of a certain key switch is significantly higher than that of other key switches, and the current amplitude of the key switches downstream of it decreases significantly or approaches the normal level, it is preliminarily judged that the fault is located in the section corresponding to this key switch.

[0085] (5) Record the preliminary positioning result

[0086] Record the range of the fault section obtained from the preliminary positioning, providing a basis for further fault distance calculation.

[0087] 3. Fault distance calculation

[0088] The purpose of fault distance calculation is to further determine the specific location of the fault within the section.

[0089] (1) Load modeling

[0090] Perform a mathematical model of the load through the substation area information, and equivalent the load to a constant impedance.

[0091] (2) Line modeling

[0092] Based on the line parameters and types, such as impedance parameters and overhead lines or cables, conduct a detailed model of the line. Divide the line into several minimum units, with each unit being 50 meters, providing a basis for subsequent fault location.

[0093] (3) Data acquisition and storage

[0094] When a fault occurs, quickly acquire and store the voltage and current data in the key switches. These data are important bases for fault distance calculation.

[0095] (4) Iterative calculation of fault distance

[0096] Divide the distribution network line into several minimum units at intervals of 10 meters to set fault points. Based on different fault types such as single-phase grounding, three-phase short circuit, and two-phase short circuit, calculate the current of the key switch on the power supply side (i.e., the most upstream) within the fault section based on their respective composite sequence networks.

[0097] In the case of a single-phase grounding fault, taking phase A grounding as an example, from the composite sequence network in the prior art, it can be known that according to the composite sequence network, the positive, negative, and zero sequence currents of phase A at the most upstream key switch during a single-phase grounding fault can be calculated as shown in the following formula.

[0098]

[0099]

[0100] Where: I A + 、I A - 、I A 0 are respectively the positive sequence current, negative sequence current, and zero sequence current of phase A at the key switch on the power supply side within the fault section. E is the equivalent voltage source of the superior power grid. Z S+ , Z S - , Z S 0 are the equivalent positive-sequence impedance, negative-sequence impedance, and zero-sequence impedance of the upstream power grid, respectively. Z1 + , Z1 - , Z1 0 are the equivalent positive-sequence impedance, negative-sequence impedance, and zero-sequence impedance between the fault point and the key switch on the power supply side within the fault section, respectively. Z2 + , Z2 - , Z2 0 are the equivalent positive-sequence impedance, negative-sequence impedance, and zero-sequence impedance from the fault point to the end of the line, respectively.

[0101] The positive, negative, and zero-sequence components of phase B and phase C can be obtained by rotating counterclockwise or clockwise by 120°, and then the positive, negative, and zero-sequence components are superimposed to obtain the total current components, as shown in the following formula.

[0102] I A = I A + + I A - + I A 0 (4)

[0103] I B = I A + ∠-120° + I A - ∠120° + I A 0 (5)

[0104] I C = I A + ∠120° + I A - ∠-120° + I A 0 (6)

[0105] Where: I A , I B , I C are the phase A current, phase B current, and phase C current at the key switch on the power supply side within the fault section, respectively.

[0106] The current calculated according to the above formula is iteratively compared with the measured voltage and current at the protection to obtain the root mean square error (RMSE) of each phase current.

[0107] The RMSE is defined as shown in the following formula.

[0108]

[0109] Where: x i and y i are the waveforms at both ends for calculating the RMSE, and n is the number of data in x i and y i .

[0110] After calculating the RMSE of each phase current, calculate the comprehensive RMSE of the three phases Z , as shown in the following formula.

[0111]

[0112] Where: RMSE Z is the comprehensive root mean square error of the three phases, RMSE A , RMSE B , RMSE C are the root mean square errors of phase A, phase B, and phase C respectively.

[0113] During a three-phase short-circuit fault, from the composite sequence network in the prior art, it can be known that according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a three-phase short-circuit fault can be calculated, as shown in the following formula.

[0114]

[0115] I A - = 0 (10)

[0116] I A 0 = 0 (11)

[0117] Next, calculate the total current component, the root mean square error of each phase current, and the comprehensive RMSE of the three phases Z , and the calculation method is the same as that for a single-phase short-circuit fault.

[0118] During a two-phase short-circuit fault, from the composite sequence network in the prior art, it can be known that according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a two-phase short-circuit fault can be calculated, as shown in the following formula.

[0119]

[0120] Next, calculate the total current component, the root mean square error of each phase current, and the comprehensive RMSE of the three phases Z , and the calculation method is the same as that for a single-phase short-circuit fault.

[0121] During a two-phase grounding fault, from the composite sequence network in the prior art, it can be known that according to the composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a two-phase grounding fault can be calculated, as shown in the following formula.

[0122]

[0123] Next, calculate the full component of the current, the root mean square error of each phase current, and the RMSE of the three-phase synthesis. Z The calculation method is the same as that for single-phase short-circuit faults.

[0124] Compare the measured current data with the calculated current data to calculate the three-phase synthesis mean square error. Traverse all the smallest units, record the two units with the smallest three-phase synthesis mean square error, and determine whether these two units are adjacent.

[0125] (5) Determine the fault distance range

[0126] If the two units with the smallest three-phase synthesis mean square error are adjacent, the fault is located between these two units. If they are not adjacent, there may be a second fault location that needs to be further confirmed.

[0127] 4. Fault location confirmation

[0128] The fault location is confirmed by further analyzing the voltage data, excluding the invalid solutions caused by line branches, and finally determining the fault location.

[0129] (1) Analyze the voltage data

[0130] For the two possible fault locations initially located, calculate the voltage data of the remaining key switches and the substation area.

[0131] (2) Calculate the voltage error

[0132] Compare the theoretically calculated voltage amplitude with the actually measured voltage amplitude to calculate the comprehensive mean square error of the voltage.

[0133] (3) Exclude invalid solutions

[0134] If the comprehensive mean square error of the voltage at a certain fault location is significantly greater than that at other locations, then exclude this location as an invalid solution.

[0135] (4) Confirm the final fault location

[0136] Confirm the fault location with the smallest and reasonable comprehensive mean square error of the voltage as the true fault point, and record the fault location.

[0137] (5) Isolate the fault area

[0138] According to the confirmed fault location, quickly disconnect the sectional switches and tie switches around the fault section to isolate the fault area and prevent the fault from expanding.

[0139] See Figure 2, the shown 10 kV distribution network was built in PSCAD (Power System Computer Aided Design) and Matlab (Matrix Laboratory) for simulation analysis and verification. In the figure, Protections 6, 8, 10, and 11 are key switches.

[0140] 1. Fault at F3

[0141] When a fault occurs at f3, which is 1,090.245 meters downstream of Protection 6 and within Section 2, the protection usually trips in about two cycles. Assuming Protection 6 trips in 40 ms, the error is calculated based on the measured current data 40 ms after the fault. Assume the fault calculates the current in different minimum units and compares the RMSE with the measured voltage and current. Z , number of iterations, calculation time, minimum RMSE Z And the range of the obtained fault distance is shown in the following table. The calculation environment is CPU: AMD Ryzen 5 PRO 4650G, 6 cores and 12 threads; GPU: AMD Radeon EX 6650XT, 8GB video memory; Memory: 24GB DDR4 - 2400MHz; Storage: 500GB M.2 SSD; Operating System: Windows 10 Professional Edition.

[0142] Table 1 Calculation results of the proposed strategy

[0143]

[0144] It can be seen that the fault is within Section 2. After determining that the fault is within Section 2, Protections 8 and 7 are quickly disconnected to isolate the fault area and prevent the fault from expanding.

[0145] 2. Fault at F4

[0146] When a fault occurs at f4 (1,802.95 meters downstream of Protection 7 and within Section 3), the protection usually trips in about two cycles. Assuming Protection 6 trips in 40 ms. Assume the fault calculates the current in different minimum units and compares the RMSE with the measured voltage and current. Z , number of iterations, minimum RMSE Z And the obtained fault range is shown in the following table.

[0147] Table 2 Calculation results of the proposed strategy

[0148]

[0149]

[0150] It can be seen that the fault is within Section 3. After determining that the fault is within Section 3, Protection 7 is quickly disconnected to isolate the fault area and prevent the fault from expanding.

[0151] 3. Fault at F5

[0152] When a fault occurs at F5 (160 meters downstream of Protection 8, within Section 4), the protection usually trips in about two cycles. Assume that Protection 6 trips in 40 ms. Assume that the fault calculates the current at different minimum units and compares the RMSE of the measured voltage and current Z , number of iterations, minimum RMSE Z And the obtained fault ranges are shown in the following table

[0153] Calculation results of the strategy proposed in Table 3

[0154]

[0155] It can be seen that the fault is within Section 4 or Section 5, and the fault location needs to be further confirmed. Assume the first and second fault locations, and the voltage amplitudes of Protection 10 and Protection 11 calculated, as well as the RMSE UZ of the actual measured values, are shown in the following table

[0156] Results of further fault location confirmation in Table 4

[0157]

[0158] It can be seen that the voltage amplitude calculated at the second fault location is significantly greater than the actual measured value at the first fault location, so the second fault location can be excluded as an invalid solution. After determining that the fault is within Section 4, quickly disconnect Protections 8, 9, and 10 to isolate the fault area and prevent the fault from expanding

[0159] 4. Fault at F6

[0160] When a fault occurs at F6 (150 meters downstream of Protection 9, within Section 5), the protection usually trips in about two cycles. Assume that Protection 6 trips in 40 ms. Assume that the fault calculates the current at different minimum units and compares the RMSE of the measured voltage and current Z , number of iterations, minimum RMSE Z And the obtained fault ranges are shown in the following table

[0161] Calculation results of the strategy proposed in Table 5

[0162]

[0163] It can be seen that although there are two fault locations, they are both within Section 5, so there is no need to further confirm the fault location. After determining that the fault is within Section 5, quickly disconnect Protections 9 and 11 to isolate the fault area and prevent the fault from expanding

[0164] 5. Fault at F7

[0165] When a fault occurs at f7 (40 meters downstream of switch 4 in the direction of branch 4 of protection 11, located within section 6), the protection usually trips in about two cycles. Assume that protection 6 trips in 40 ms. Assume that the fault calculates the current in different minimum units and compares the RMSE with the measured voltage and current. Z , number of iterations, minimum RMSE Z And the obtained fault ranges are shown in the following table.

[0166] Calculation results of the proposed strategy in Table 6

[0167]

[0168] It can be seen that although there are two fault locations, both are within section 6. Therefore, there is no need to further confirm the fault location. After determining that the fault is in section 6, quickly disconnect protection 11 to isolate the fault area and prevent the expansion of the fault.

[0169] From the faults at F3, F4, F5, F6, and F7 above, it can be seen that the present invention can achieve accurate fault location only relying on the transient data of key switches within an extremely short time (the calculation time is all 3.3 ms and below).

[0170] In summary, compared with the prior art, first, the present invention reduces the dependence on communication and equipment. By selecting key switches and analyzing transient characteristic information, it reduces the dependence on complex communication networks and high-precision equipment. Only a measurement device needs to be installed on a small number of key switches to achieve fault location, reducing the system construction cost and operation and maintenance difficulty; second, the present invention achieves high-precision fault location. By combining transient voltage and current information and line models, through iterative calculation and multi-dimensional data analysis, it can accurately calculate the fault distance and locate the fault within a single section. At the same time, by comparing voltage amplitudes and analyzing mean square errors, invalid solutions are excluded to ensure the high precision and reliability of the location; third, the present invention has fast response and efficient recovery. It can complete the preliminary location and accurate calculation of the fault section within a short time, significantly shortening the fault location time. After fault location, through reasonable fault isolation and power supply restoration logic, it can quickly restore the power supply of the non-fault area and reduce the power outage time; fourth, the present invention has wide applicability. It is applicable to various fault types, such as single-phase grounding, three-phase short circuit, two-phase short circuit, etc. And through optimized algorithms and models, it can handle complex faults and non-typical fault scenarios. At the same time, the present invention has strong adaptability to changes in the operation mode and topological structure of the distribution network and does not require frequent parameter adjustment.

[0171] The present invention has been described in detail with reference to the embodiments accompanied by drawings. Those of ordinary skill in the art can make various variations of the present invention according to the above description. Therefore, certain details in the embodiments should not constitute a limitation to the present invention, and the protection scope of the present invention will be defined by the scope defined in the appended claims.

Claims

1. A rapid fault location and isolation method for a distribution network based on transient characteristic information, which includes four steps: selection of key switches, preliminary location, fault distance calculation, and fault location confirmation; specifically as follows: First, conduct a preliminary location through the current information of the selected key switches to determine the approximate section range of the fault; divide the distribution network line into several minimum units at intervals of several meters. If the preliminary location shows that the fault is within a certain two sections, then combine the transient voltage and current information of the switches through which the large fault current flows and the line model to calculate the minimum unit where the fault is located. If the calculated fault distance indicates that the fault exists only in a certain section, the fault location is completed; If the fault may exist in two sections, the voltage data of the remaining key switches and the substation are further combined to confirm the fault location, and finally the fault is located in a single section; After the fault location is completed, the sectional switches and tie switches around the fault section are quickly disconnected to isolate the fault area.

2. A rapid fault location and isolation method for a distribution network based on transient characteristic information according to claim 1, wherein the selection of the key switches includes the following steps: (1) Determine the starting switch, starting from the switch at the head of the line as the starting point of the selection process; (2) Obtain the outermost switches of the sections, obtain the outermost switches of two sections downstream from the starting switch, which include tie switches and sectional switches; (3) Eliminate the tie switches. Among the obtained outermost switches, eliminate the tie switches, and the remaining switches are the key switches; (4) Repeat the operation. Taking the newly determined key switch as the starting point, repeat steps (2) and (3) to gradually traverse the entire power grid structure until all sections are covered; (5) Determine the final key switches. After the traversal is completed, all the determined key switches are used as the core measurement points for fault location, and voltage and current measurement devices are installed on the required switches.

3. A rapid fault location and isolation method for a distribution network based on transient characteristic information according to claim 2, wherein the following principles are followed during the selection of the key switches: The number of outermost switches for every two sections shall not be less than two; If there is only one section due to the limitation of the line end, the outermost switch of this section can be accepted as the outermost switch; If the above principles cannot be met, the demarcation point of two sections can be selected as one of the outermost switches.

4. A rapid fault location and isolation method for a distribution network based on transient characteristic information according to claim 1, wherein the preliminary location includes the following steps: Step of statistically analyzing normal current data: Statistically analyze the current amplitudes of each key switch during the normal operation of the distribution network, and calculate their average value μ and standard deviation σ; Step of setting a threshold: Set the threshold as μ + kσ to distinguish normal current fluctuations from fault currents; Step of detecting fault current: When a fault occurs, the key switch will detect a significant change in current. The current amplitude of the switch upstream of the fault point will increase significantly, while the current amplitude of the switch downstream of the fault point will rapidly decrease or approach the normal level; Step of judging the fault section: If the current amplitude of a certain key switch is significantly higher than that of other key switches, and the current amplitude of the key switch downstream of it decreases significantly or approaches the normal level, it is preliminarily judged that the fault is located in the section corresponding to this key switch; Step of recording the preliminary location result: Record the range of the fault section obtained from the preliminary location to provide a basis for further fault distance calculation.

5. A method for rapid fault location and isolation of a distribution network based on transient characteristic information according to claim 1, wherein the fault distance calculation includes the following steps: Load modeling step, mathematically modeling the load through substation area information and equivalent the load to a constant impedance; Line modeling step, based on line parameters and types, detailedly model the line, divide the line into several minimum units, providing a basis for subsequent fault location; Data acquisition and storage step, when a fault occurs, quickly acquire and store the voltage and current data in the key switches; Iterative calculation of fault distance step, taking the minimum unit as the fault point, according to different fault types, calculate the current of the power supply side (i.e., the most upstream key switch) in the fault section based on the composite sequence network, compare the measured current data with the calculated current data, calculate the three-phase comprehensive mean square error, traverse all minimum units, record the two units with the smallest three-phase comprehensive mean square error, and determine whether these two units are adjacent; Determine the fault distance range step, if the two units with the smallest three-phase comprehensive mean square error are adjacent, the fault is located between these two units. If not adjacent, there may be a second fault location that needs further confirmation.

6. A method for rapid fault location and isolation of a distribution network based on transient characteristic information according to claim 1, wherein the fault location confirmation includes the following steps: Analyze voltage data step, for the two possible fault locations preliminarily located, calculate the voltage data of the remaining key switches and substations; Calculate voltage error step, compare the theoretically calculated voltage amplitude with the actual measured voltage amplitude, and calculate the comprehensive mean square error of voltage; Exclude invalid solution step, if the comprehensive mean square error of voltage at a certain fault location is significantly greater than that at other locations, exclude this location as an invalid solution; Confirm the final fault location step, confirm the fault location with the smallest and reasonable comprehensive mean square error of voltage as the true fault point, and record the fault location; Isolate the fault area step, according to the confirmed fault location, quickly disconnect the sectional switches and tie switches around the fault section to isolate the fault area and prevent the fault from expanding.

7. A method for rapid fault location and isolation of a distribution network based on transient characteristic information according to claim 5, wherein the step of iterative calculation of fault distance includes: When the fault type is single-phase grounding fault, according to the composite sequence network, calculate the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during single-phase grounding fault. The positive, negative, and zero-sequence components of phase B and phase C can be obtained by rotating 120° counterclockwise or clockwise, and then the positive, negative, and zero-sequence components are superimposed to obtain the full current component. Compare the calculated current with the measured voltage and current at the protection to iterate and calculate the root mean square error of each phase current. After calculating each phase current, calculate the three-phase comprehensive root mean square error.

8. A method for rapid fault location and isolation of a distribution network based on transient characteristic information according to claim 5, wherein the step of iterative calculation of fault distance includes: When the fault type is a three-phase short circuit composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a three-phase short circuit fault can be calculated according to the composite sequence network. Then, the total current component, the root mean square error of each phase current, and the root mean square error of the three-phase synthesis are calculated.

9. A method for rapid fault location and isolation of a distribution network based on transient characteristic information according to claim 5, wherein the step of iteratively calculating the fault distance includes: When the fault type is a two-phase short circuit composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a two-phase short circuit fault can be calculated according to the composite sequence network. Next, the total current component, the root mean square error of each phase current, and the root mean square error of the three-phase synthesis are calculated.

10. A method for rapid fault location and isolation of a distribution network based on transient characteristic information according to claim 5, wherein the step of iteratively calculating the fault distance includes: When the fault type is a two-phase grounding composite sequence network, the positive, negative, and zero-sequence currents of phase A at the most upstream key switch during a two-phase grounding fault can be calculated according to the composite sequence network. Next, the total current component, the root mean square error of each phase current, and the root mean square error of the three-phase synthesis are calculated.

Images