A primary and secondary fusion ring network box protection setting value adjustment method

Abstract

The application provides a primary and secondary fusion ring network box protection setting value adjustment method, which comprises the following steps: obtaining zero sequence current amplitude data from a zero sequence acquisition unit, obtaining sheath potential data from a sheath potential acquisition unit, obtaining bus voltage data from a bus voltage acquisition module, and collecting the data into an initial input set; dynamically adjusting a grounding protection discrimination criterion according to an adjustment direction, taking the adjustment direction and the bus voltage data as inputs, adopting a threshold comparison mechanism to form an updated grounding protection discrimination criterion, and using the updated grounding protection discrimination criterion to locate a real grounding branch; and matching the updated grounding protection discrimination criterion with the zero sequence current amplitude data, generating a protection setting value adjustment instruction after the protection setting value adjustment condition is established, updating the ring network box protection setting value, and completing the identification and setting value adjustment of the real grounding branch.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for adjusting the protection settings of a primary and secondary integrated ring network box. Background Technology

[0002] In the field of power system operation and maintenance, ring main units (RMS) are core node equipment in urban distribution networks, responsible for power branch distribution, load balancing, and fault protection. Their stable operation directly determines the continuity and security of power supply. Especially in densely populated urban environments, RMS connects multiple cable lines, forming a complex ring network structure. If a ground fault occurs and the protection settings are not properly adjusted, it may not only fail to quickly isolate the fault point but also trigger a chain reaction, leading to widespread power outages and severely impacting residents' daily lives, hospital power supply, and industrial production continuity. This risk is particularly prominent during peak load periods or in severe weather, as urban distribution networks often face multiple uncertainties such as cable aging and external damage. Currently, the judgment of ground faults in RMS and the adjustment of protection settings generally rely on simple current amplitude detection methods. While this method is simple to operate, it reveals significant shortcomings in complex power grid environments, especially when dealing with multi-branch cable networks. Simply relying on current magnitude to determine the fault location often leads to inaccuracies. This is because the capacitive effect of cable branches and differences in sheath grounding conditions can interfere with the current distribution pattern, causing significant current backflow even in distant branches, thus misleading fault location. For example, in actual operation, due to differences in cable sheath grounding conditions and the influence of branch capacitance, the distribution of zero-sequence current is subject to multiple interferences, exhibiting non-intuitive characteristics. Zero-sequence current should be an important basis for judging the location of ground faults; theoretically, the larger the current, the closer the fault point is to the detection location. This interference not only misleads the identification of fault branches but also amplifies the contradictions in protection setting values. Traditional settings are designed based on ideal zero-sequence current distribution. Once the actual scenario deviates, such as inconsistent sheath grounding or network asymmetry, the protection action will be delayed or malfunction. If a near-end fault is not isolated, the fault current will continue to flow, heating the cable insulation; if a far-end misjudgment occurs, irrelevant branches will trip, increasing the load transfer pressure and potentially triggering cascading faults. More seriously, urban ring main units are often embedded underground or adjacent to buildings, with limited maintenance windows. Manually adjusting settings is time-consuming and cannot adapt to dynamic changes in real time, such as the load difference between daytime peak and nighttime off-peak periods, or grounding interference introduced by temporary construction. Therefore, accurately identifying the true source of zero-sequence current has become a bottleneck for dynamically adjusting protection settings, and it is urgent to overcome the location problem under multiple interferences to ensure the safety and stability of complex power grids. Summary of the Invention

[0003] This invention provides a method for adjusting the protection settings of a primary and secondary integrated ring main unit, the method comprising: Zero-sequence current amplitude data is obtained from the zero-sequence acquisition unit, sheath potential data is obtained from the sheath potential acquisition unit, and bus voltage data is obtained from the bus voltage acquisition module, and then collected into an initial input set. The phase difference between the zero-sequence current amplitude data and the sheath potential data in the initial input set is calculated using the Fourier transform algorithm. The phase difference is compared with the bus voltage data, and the branch state that exceeds the preset limit is determined to be the active return current state of the far-end branch. The cable branch capacitance is identified by the active return current pattern of the far-end branch, and the cable branch capacitance is decomposed by wavelet transform algorithm. The equivalent components obtained by decomposition are substituted into the zero-sequence current amplitude data for adjustment to form updated zero-sequence current amplitude data. The updated zero-sequence current amplitude data is combined with the sheath potential data to extract the sheath grounding distribution. The nodes in the sheath grounding distribution that match the zero-sequence current amplitude data are identified as the actual grounding branch locations, forming the adjustment direction of the grounding fault discrimination conditions. Based on the aforementioned adjustment direction, the grounding protection discrimination criteria are dynamically adjusted. Using the adjustment direction and bus voltage data as input, a threshold comparison mechanism is used to form an updated grounding protection discrimination criterion to locate the actual grounding branch. The updated grounding protection discrimination criteria are matched with the zero-sequence current amplitude data. After the protection setting adjustment conditions are met, a protection setting adjustment command is generated to update the ring network box protection setting, thus completing the identification and setting adjustment of the actual grounding branch.

[0004] Furthermore, the acquisition of zero-sequence current amplitude data from the zero-sequence acquisition unit, sheath potential data from the sheath potential acquisition unit, and bus voltage data from the bus voltage acquisition module, combined into an initial input set, includes: The zero-sequence current amplitude data of each branch outgoing line side of the ring network box are obtained from the zero-sequence acquisition unit and synchronously read at a uniform sampling interval to obtain a zero-sequence current amplitude data sequence with branch number identifier. The sheath potential data of the cable sheath grounding point is obtained from the sheath potential acquisition unit, and the three-phase voltage data of the incoming bus of the ring network box is obtained from the bus voltage acquisition module. The channels are associated according to the branch number and the sampling timestamp, and the data are collected with the zero-sequence current amplitude data sequence to form the initial input set.

[0005] Furthermore, the step of using the Fourier transform algorithm to calculate the phase difference between the zero-sequence current amplitude data and the sheath potential data in the initial input set, comparing it with the bus voltage data, and determining the branch state where the phase difference exceeds a preset limit as a remote branch with active return current includes: Zero-sequence current amplitude data and sheath potential data are retrieved from the initial input set according to the branch number. The Fourier transform algorithm is applied to the two co-source time-series waveforms to extract the power frequency components. The corresponding power frequency phase angles are obtained and subtracted to obtain the phase difference sequence of each branch. Using the power frequency phase angle extracted by Fourier transform of the zero-sequence voltage component in the bus voltage data as a common reference benchmark, the phase difference sequence is aligned with the reference benchmark according to the branch number to obtain a normalized phase difference sequence with the zero-sequence voltage phase of the bus as the zero point. If a branch in the normalized phase difference sequence exceeds the upper and lower limits of the phase difference exceeding the limit determination window, it is determined that the branch is in the active return state of the far-end branch.

[0006] Furthermore, the process of identifying cable branch capacitance through the active return current pattern of the remote branch, decomposing the cable branch capacitance using a wavelet transform algorithm, and substituting the decomposed equivalent components into the zero-sequence current amplitude data for adjustment to form updated zero-sequence current amplitude data includes: Zero-sequence current amplitude data of the corresponding branch is retrieved from the set of branches in the active return current state of the far-end branch. Wavelet transform algorithm is used to decompose the data at multiple scales to obtain the approximate component of the low-frequency band and the detailed component of the high-frequency band near the power frequency. The energy ratio of the detailed component is used as the basis for identifying the activity level of the cable branch capacitance. For the detailed components, the power frequency phase angle corresponding to the sheath potential data of the corresponding branch is taken as the reference direction according to the branch number. The detailed components are projected along a direction that leads the reference direction by 90 degrees to obtain the capacitive equivalent component, and projected along a direction in the same direction as the reference direction to obtain the resistive equivalent component. The return current composite amplitude is obtained by summing the amplitudes of the capacitive equivalent component and the resistive equivalent component according to the branch number. The corresponding return current composite amplitude is then subtracted from the zero-sequence current amplitude data to obtain the remaining amplitude as the updated zero-sequence current amplitude data.

[0007] Furthermore, the step of combining the updated zero-sequence current amplitude data with the sheath potential data to extract the sheath grounding distribution, and identifying the nodes in the sheath grounding distribution that match the zero-sequence current amplitude data as the actual grounding branch locations, thus forming the adjustment direction for the grounding fault discrimination conditions, includes: The remaining amplitude is retrieved from the updated zero-sequence current amplitude data according to the branch number, and a joint amplitude potential record is established with the sheath potential data of the corresponding branch according to the sampling timestamp. According to the amplitude potential joint record, the potential amplitude of the sheath is spatially arranged according to the identification of the potential acquisition point along the sheath, and the high and low distribution pattern of the potential of the sheath along the line is used as the sheath grounding distribution. The node location is compared with the remaining amplitude. If the potential amplitude of the protective layer at the node location rises above a preset potential rise threshold and the remaining amplitude of the corresponding branch exceeds a preset amplitude threshold, then the node is determined to be the location of the actual grounding branch.

[0008] Furthermore, based on the adjustment direction, the grounding protection discrimination criterion is dynamically adjusted. Using the adjustment direction and bus voltage data as input, a threshold comparison mechanism is employed to form an updated grounding protection discrimination criterion, locating the actual grounding branch, including: The branch number and node position identifier are retrieved from the adjustment direction. The zero-sequence voltage amplitude is retrieved from the bus voltage data and compared with the preset zero-sequence voltage start threshold. If the zero-sequence voltage amplitude exceeds the zero-sequence voltage start threshold, the corresponding branch is marked as a branch to be adjusted. The zero-sequence current action threshold reference value corresponding to the branch to be adjusted is retrieved from the pre-registered setting value table in the ring network box protection configuration. The amplitude is reduced according to the pre-registered reduction level table, using the node position identifier as the basis for selecting the reduction level, and the adjusted zero-sequence current action threshold is obtained. The updated grounding protection discrimination criterion is formed by combining the adjusted zero-sequence current operating threshold with the bus zero-sequence voltage over-limit indicator.

[0009] Furthermore, the process of matching the updated grounding protection discrimination criteria with the zero-sequence current amplitude data, generating a protection setting adjustment command after the protection setting adjustment conditions are met, updating the ring main unit protection settings, and completing the identification and setting adjustment of the actual grounding branch includes: From the updated grounding protection discrimination criteria, retrieve the AND gate logic criterion entry according to the branch number, substitute the corresponding branch zero-sequence current amplitude data into the AND gate logic criterion entry, and obtain the branch location mark bit through two input logic gate operations; The branch positioning mark is accumulated as true according to the number of consecutive sampling frames. If the number of consecutive accumulated frames exceeds the preset setting adjustment duration frame threshold, the adjusted zero-sequence current action threshold is taken from the setting value table as the target setting value and encapsulated with the branch number to form the protection setting value adjustment instruction. The protection setting adjustment command is sent to the corresponding protection device along the internal data link of the ring network box, and the original zero-sequence current action threshold in the setting register is replaced with the target setting value.

[0010] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a method for adjusting the protection settings of a primary and secondary integrated ring main unit (RMU). It proposes a complete solution to the operational scenario of ground fault location and dynamic adjustment of protection settings in RMUs. This problem involves the acquisition and analysis of zero-sequence current, sheath potential, and bus voltage data. The core challenge is accurately identifying the actual grounding branch and dynamically adjusting the protection settings. This invention uses Fourier transform to calculate the phase difference and combines it with bus voltage comparison to identify the active return current pattern of the distant branch; it uses wavelet transform to decompose the cable branch capacitance and adjust the zero-sequence current data; furthermore, it extracts the grounding distribution using sheath potential to confirm the location of the actual grounding branch; finally, it dynamically adjusts the protection discrimination criteria based on a threshold comparison mechanism to generate setting adjustment instructions. This invention achieves accurate ground fault location and adaptive adjustment of protection settings, improving the reliability and intelligence level of RMU protection. Attached Figure Description

[0011] Figure 1 This is a flowchart of a method for adjusting the protection settings of a primary and secondary integrated ring network box according to the present invention.

[0012] Figure 2 This is a schematic diagram of a method for adjusting the protection settings of a primary and secondary integrated ring network box according to the present invention.

[0013] Figure 3 This is another schematic diagram of a method for adjusting the protection settings of a primary and secondary integrated ring network box according to the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0015] like Figures 1 to 3 This embodiment of a method for adjusting the protection settings of a primary and secondary integrated ring main unit may specifically include: S101. Obtain zero-sequence current amplitude data from the zero-sequence acquisition unit, obtain sheath potential data from the sheath potential acquisition unit, obtain bus voltage data from the bus voltage acquisition module, and aggregate them into an initial input set.

[0016] Zero-sequence current amplitude data from the outgoing sides of each branch of the ring main unit are acquired from the zero-sequence acquisition unit. The zero-sequence current channels of each branch are synchronously read at a uniform sampling interval to obtain a zero-sequence current amplitude data sequence with branch number identifiers. This sequence is then aligned according to the sampling timestamp to obtain a time-aligned zero-sequence current amplitude record. At the sampling time corresponding to the zero-sequence current amplitude record, sheath potential data at the cable sheath grounding point is acquired from the sheath potential acquisition unit, and three-phase voltage data of the incoming bus of the ring main unit is acquired from the bus voltage acquisition module. The sheath potential data and the three-phase voltage data are associated according to the branch number and sampling timestamp, and then combined with the zero-sequence current amplitude record to form an initial input set for subsequent phase comparison.

[0017] In one embodiment, the zero-sequence acquisition unit employs through-core zero-sequence current transformers arranged on the cable outgoing side of each outgoing switchgear in the ring main unit. One transformer is independently configured for each branch, and the secondary side of the transformer is connected to the zero-sequence current channel of the data acquisition device. The data acquisition device synchronously samples each channel at a uniform sampling rate of 5000 points per second. After synchronously reading the zero-sequence current channel of each branch at the uniform sampling rate, it outputs a zero-sequence current amplitude data sequence with a branch number identifier. The branch number identifier is used to distinguish the affiliation of different outgoing branches within the ring main unit, and based on this identifier, each sequence is aligned according to the sampling timestamp to obtain a time-aligned zero-sequence current amplitude record.

[0018] Specifically, the sheath potential acquisition unit is located on the sheath grounding lead at the cable terminal joint, and acquires the sheath-to-ground potential in real time through a potential sensor mounted on the grounding lead. The bus voltage acquisition module is connected through the secondary winding of the voltage transformer on the incoming side of the ring mains box to obtain the three-phase voltage data of the incoming bus. The sheath potential data and the three-phase voltage data are acquired by the same data acquisition device at the same sampling timestamp, thus providing a basis for time synchronization.

[0019] Understandably, the sheath potential data and the three-phase voltage data are associated with channels based on branch numbers and sampling timestamps. Specifically, based on the physical topology of each branch within the ring network box, a channel index mapping is established within the acquisition device for the sheath potential channel, zero-sequence current channel, and incoming bus voltage channel corresponding to the branch's outgoing line, ensuring a comparable correspondence among the three types of data at the same time. The mapping results, along with the zero-sequence current amplitude record, are aggregated to form an initial input set. This initial input set is organized into multi-channel time-series data frames according to branch numbers for subsequent phase comparison.

[0020] S102. The Fourier transform algorithm is used to calculate the phase difference between the zero-sequence current amplitude data and the sheath potential data in the initial input set. The phase difference is compared with the bus voltage data, and the branch state that exceeds the preset limit is determined to be the active return current state of the far-end branch.

[0021] Zero-sequence current amplitude data and sheath potential data, organized by branch number, are retrieved from the initial input set. For each branch, a Fourier transform algorithm is applied to the two co-originating time-series waveforms to extract the power frequency components, obtaining the power frequency phase angles corresponding to the zero-sequence current amplitude data and the sheath potential data. Subtracting these two yields a branch-by-branch phase difference sequence, which characterizes the leading or lagging relationship between the zero-sequence current and the sheath potential within the power frequency cycle. Using the power frequency phase angle extracted from the zero-sequence voltage component in the bus voltage data via Fourier transform as a common reference, the phase difference sequences of each branch are aligned according to their branch numbers to obtain a normalized phase difference sequence with the bus zero-sequence voltage phase as the zero point. This normalized phase difference sequence eliminates the initial phase angle bias between different branch sampling channels, resulting in a horizontally comparable branch phase difference result. For the normalized phase difference sequence, a pre-established phase difference over-limit judgment window is invoked. The phase difference over-limit judgment window is composed of the same direction interval where the zero-sequence current phase and the sheath potential phase should be under the near-end fault state. If the normalized phase difference of a certain branch exceeds the upper and lower limits of the phase difference over-limit judgment window, the branch is determined to be in the far-end branch return active state. Otherwise, it is maintained as a near-end candidate branch, thus completing the branch state morphology division.

[0022] In one implementation, the multi-channel timing data frames organized by branch number in the initial input set enter the phase comparison stage, and the phase comparison stage applies a Fourier transform algorithm to the zero-sequence current amplitude data and sheath potential data of each branch.

[0023] Specifically, the Fourier transform algorithm performs frequency domain decomposition on the time-series waveform using an integer period of the power frequency as the window length. From the decomposition results, the complex expression corresponding to the fundamental power frequency component is extracted. The argument of this complex expression is the phase angle of the waveform at the power frequency. The zero-sequence current amplitude data, after Fourier transform, outputs the zero-sequence current power frequency phase angle, denoted as φi. The sheath potential data, after Fourier transform, outputs the sheath potential power frequency phase angle, denoted as φu.

[0024] It should be noted that the phase angle acquisition is based on the same full-cycle window of the power frequency. In one possible implementation, the window length is 20 milliseconds, which corresponds to a complete cycle of the power frequency of 50 Hz. This allows the phase angles of the two waveforms to be extracted under the same time reference, avoiding phase deviation caused by window misalignment.

[0025] Specifically, the phase angle φi of the zero-sequence current is subtracted from the phase angle φu of the sheath potential to obtain the branch-by-branch phase difference Δφ = φi - φu. This phase difference Δφ characterizes the leading or lagging relationship of the zero-sequence current relative to the sheath potential. In the case of a near-end real ground fault, the zero-sequence current is mainly provided by the loop formed by the fault point and the earth, and the zero-sequence current and the sheath potential exhibit a fixed relationship of approximately being in phase or out of phase. When backflow occurs in the far-end branch due to inconsistent cable sheath grounding or the presence of branch capacitance, the phase relationship between the zero-sequence current and the sheath potential deviates significantly. The phase difference sequence is output in a rolling fashion according to the sampling time frame, forming a branch-by-branch phase difference timing sequence. Further, the power frequency phase angle extracted from the zero-sequence voltage component in the bus voltage data through Fourier transform is used as a common reference benchmark. The zero-sequence voltage component is obtained by adding the instantaneous values ​​of the three-phase voltages and dividing by 3, and then picking up the power frequency phase angle φ0 after Fourier transform. Subtracting φ0 from the phase difference Δφ of each branch yields the normalized phase difference Δφ' = Δφ - φ0. This eliminates the initial phase angle offset introduced by differences in transformer polarity and wiring direction between sampling channels of different branches, ensuring that the phase difference results of each branch have the same horizontally comparable coordinates. Preferably, the phase difference over-limit judgment window is composed of the same-direction interval where the zero-sequence current phase and the sheath potential phase should be under near-end fault conditions.

[0026] In one embodiment, the determination window takes the sector interval where the normalized phase difference Δφ' falls within ±45 degrees as the same direction determination band, and the branch falling within this band is regarded as the near-end candidate branch; if the normalized phase difference Δφ' exceeds the upper and lower limits of ±45 degrees and falls into the lateral or reverse sector, the branch is marked as the far-end branch return active mode.

[0027] It is understood that the upper and lower limits of the phase difference exceeding the limit judgment window are not fixed values. In another embodiment, for the case of multi-point grounding of the cable sheath, the upper and lower limits of the judgment window can be narrowed to ±30 degrees to cope with the phase drift caused by the superposition of sheath circulating current; for the case of long cable branches and large distributed capacitance, the limits can be appropriately widened to ±60 degrees to avoid misjudging the capacitive return current of normal branches as a fault circuit. The upper and lower limits of the judgment window are formed by sampling and statistically analyzing the phase difference baseline of each branch under no-load and load conditions during the commissioning phase. The specific process is as follows: collect at least 100 phase difference samples, calculate the sample mean μ and standard deviation σ, and then set the upper and lower limits to μ ± 2σ. For example, when the sampling results in μ = 0 degrees and σ = 15 degrees, the limit is -30 degrees to 30 degrees.

[0028] For example, a ring network box contains 4 outgoing branches. In a grounding alarm event, the normalized phase difference Δφ' of branch 1 is measured to be about 10 degrees, which falls within the ±45 degree judgment band. Branch 1 remains a near-end candidate branch. The Δφ' of branches 2, 3, and 4 are measured to be about 110 degrees, 95 degrees, and 130 degrees, respectively, which all exceed the upper limit of the judgment window. Branch 2, 3, and 4 are all marked as far-end branch return active mode.

[0029] It should be noted that the active return current pattern of the remote branch is not directly equivalent to the fault location conclusion. This pattern only indicates that the zero-sequence current of the branch mainly comes from the return current of the branch capacitor or the sheath circulation current. The pattern classification result of the branch serves as the input condition for the subsequent cable branch capacitor identification and zero-sequence current amplitude adjustment, thereby completing the pattern classification of the branch state.

[0030] S103. Identify the cable branch capacitance by the active return current pattern of the remote branch, decompose the cable branch capacitance using wavelet transform algorithm, and substitute the equivalent components obtained from the decomposition into the zero-sequence current amplitude data for adjustment to form updated zero-sequence current amplitude data.

[0031] Zero-sequence current amplitude data of the corresponding branch is retrieved from the set of branches with active return current patterns in the far-end branches. Wavelet transform algorithm is used to decompose the zero-sequence current amplitude data at multiple scales to obtain approximate components in the low-frequency band near the power frequency and detail components in the high-frequency band. The approximate components carry the power frequency main body of the true zero-sequence current of the branch, and the detail components carry the high-frequency return current components excited by the cable branch capacitance during the transient process. The energy proportion of the detail components is used as the basis for identifying the existence and activity level of the cable branch capacitance, and the zero-sequence current decomposition result with frequency band identification is obtained. For the detailed components in the zero-sequence current decomposition result, the power frequency phase angle corresponding to the sheath potential data of the corresponding branch is taken as the reference direction according to the branch number. The detailed components are projected along a direction leading the reference direction by 90 degrees to obtain the capacitive equivalent component, and projected along a direction in the same direction as the reference direction to obtain the resistive equivalent component. The capacitive equivalent component represents the contribution of the return current amplitude caused by the charging and discharging of the branch capacitor, and the resistive equivalent component represents the contribution of the in-phase amplitude caused by the sheath circulating current, thus completing the quantization and decomposition of the equivalent components of the cable branch capacitance. According to the capacitive equivalent component and the resistive equivalent component, the amplitudes of the two are summed according to the branch number to obtain the return current composite amplitude. Then, the corresponding return current composite amplitude is subtracted from the zero-sequence current amplitude data to obtain the remaining amplitude after removing the return current component. The remaining amplitude is used as the updated zero-sequence current amplitude data, reflecting the amplitude level of the actual grounding current of each branch after deducting the influence of the cable branch capacitor return current, thus completing the update of the zero-sequence current amplitude data.

[0032] In one embodiment, in the set of branches with active return current in the remote branch, each branch has been marked as having zero-sequence current whose main source is not from a real ground fault circuit, but is mixed with return current components caused by the inconsistency between cable branch capacitance and sheath grounding.

[0033] It should be noted that this return current component appears in the zero-sequence current amplitude data as a superposition of steady-state contribution and transient disturbance near the power frequency. Directly judging ground faults based on the original zero-sequence current amplitude data will result in misjudgment at the far end. The formation of the updated zero-sequence current amplitude data is achieved by stripping the return current component.

[0034] Specifically, zero-sequence current amplitude data of the corresponding branch is retrieved one by one from the branch set according to the branch number. The zero-sequence current amplitude data exists in the form of a time series, and the sampling rate is based on several thousand sampling points corresponding to a full cycle of power frequency. The zero-sequence current amplitude data is decomposed into multiple scales using a wavelet transform algorithm, and the wavelet transform algorithm selects a mother wavelet with compact support characteristics as the basis function.

[0035] In one possible implementation, the mother wavelet is the db4 wavelet, and the decomposition level is 5.

[0036] It should be noted that the basic principle of the wavelet transform algorithm is to decompose the original time-series waveform layer by layer according to different frequency band scales. Each layer decomposes and outputs a set of approximate components and a set of detail components. The approximate components retain the main energy of the low-frequency band, while the detail components retain the local disturbance energy in the corresponding frequency band. After five layers of decomposition, the low-frequency band near the power frequency is carried by the approximate components of the highest layer, which carry the main power frequency component of the true zero-sequence current of the branch. The frequency bands several to tens of times higher than the power frequency are carried by the detail components of each layer, which carry the high-frequency return current component excited by the cable branch capacitance during the transient process. Furthermore, the energy proportion of the detail components is used as the basis for identifying the existence and activity level of the cable branch capacitance.

[0037] Specifically, the energy percentage is obtained by dividing the sum of squares of the detail components by the sum of squares of the original zero-sequence current amplitude data. If the energy percentage exceeds a preset capacitor return current identification threshold, it is determined that the zero-sequence current amplitude data of that branch contains a significant cable branch capacitor return current component. This threshold is formed by adding a certain margin to the return current energy percentage measured under no-load conditions during the commissioning phase, and preferably 0.15 is taken as a typical threshold value. After the identification is completed, the zero-sequence current decomposition results with frequency band assignment identifiers are output according to the branch number.

[0038] It is understood that the frequency band attribution identifier is used to distinguish between the approximate component value and the detail component value corresponding to each sampling time, so that subsequent projection decomposition can perform targeted processing on the detail component.

[0039] In one embodiment, for the detailed components in the zero-sequence current decomposition result, the power frequency phase angle corresponding to the sheath potential data of the corresponding branch is taken as the reference direction according to the branch number.

[0040] Specifically, taking the reference direction as the positive real axis direction on the complex plane, the detailed components are complex-reconstructed within each full-cycle window of the power frequency. That is, the instantaneous value of the detailed component is taken as the real part, and the orthogonal component obtained by its Hilbert transform is taken as the imaginary part, resulting in a complex vector representation of the detailed components. The complex vector is projected along the imaginary axis direction leading the reference direction by 90 degrees to obtain the capacitive equivalent component, and projected along the real axis direction in the same direction as the reference direction to obtain the resistive equivalent component. The capacitive equivalent component represents the contribution of the return current amplitude caused by the charging and discharging of the branch capacitors, which in a physical sense corresponds to the displacement current component of the cable core capacitance to the sheath and to ground. The resistive equivalent component represents the in-phase amplitude contribution caused by the sheath circulating current, which in a physical sense corresponds to the loop resistive current component formed by multiple grounding points of the sheath. The capacitive equivalent component is denoted as Ic, and the resistive equivalent component is denoted as Ir. The two are output in pairs according to the branch number to complete the quantization and decomposition of the equivalent components of the cable branch capacitance.

[0041] For example, the average amplitude of the complex vector of the detail components of a certain far-end branch after decomposition within the full power frequency cycle is approximately 40 amps. Its projection component Ic on the imaginary axis is approximately 36 amps, and its projection component Ir on the real axis is approximately 17 amps, reflecting that the return current of this branch is mainly contributed by capacitive forces and secondarily by resistive forces. Further, based on the capacitive equivalent component Ic and the resistive equivalent component Ir, the combined return current amplitude is calculated according to the branch number. Where Ic is the imaginary axis projection component and Ir is the real axis projection component, the return current composite amplitude Is represents the total return current component in the zero-sequence current amplitude data of this branch, contributed by the cable branch capacitance and the sheath circulating current. Then, the corresponding return current composite amplitude is subtracted from the zero-sequence current amplitude data, i.e., calculated using the original zero-sequence current amplitude data I0. The remaining amplitude Ie after stripping the reflux components is obtained.

[0042] Preferably, in another embodiment, for branches with single-end grounding of the sheath, the resistive equivalent component Ir is measured to be close to zero, and the return current synthesis amplitude is approximately equal to Ic; for branches with double-end grounding or cross-interconnection grounding of the sheath, the resistive equivalent component Ir increases significantly, and the return current synthesis amplitude is dominated by both capacitive and resistive components. The subtraction process is applicable to both types of operating conditions. The remaining amplitude Ie serves as the updated zero-sequence current amplitude data. It replaces the corresponding items in the original zero-sequence current amplitude data branch by branch number, reflecting the amplitude level of the actual grounding current of each branch after deducting the influence of cable branch capacitance return current. The updated zero-sequence current amplitude data of each branch are summarized to form the updated zero-sequence current amplitude dataset, thus completing the update of the zero-sequence current amplitude data.

[0043] S104. Combine the updated zero-sequence current amplitude data with the sheath potential data to extract the sheath grounding distribution. Identify the nodes in the sheath grounding distribution that match the zero-sequence current amplitude data as the actual grounding branch locations, thus forming the adjustment direction for the grounding fault discrimination conditions.

[0044] The remaining amplitude after stripping the return current component is retrieved from the updated zero-sequence current amplitude data according to the branch number, and the sheath potential data of the corresponding branch is retrieved from the initial input set. For each branch, a point-by-point pairing record of the remaining amplitude and sheath potential is established according to the sampling timestamp, resulting in a joint amplitude potential record for each branch. This joint amplitude potential record carries the point-by-point correspondence between the updated zero-sequence current amplitude and the sheath potential at the same moment. Based on the joint amplitude potential record, the sheath potential amplitude is spatially arranged according to the branch number along the sampling point markers of each sheath potential, resulting in a high-low distribution pattern of the sheath potential along the line. This high-low distribution pattern serves as the sheath grounding distribution. The markers of points where the sheath potential amplitude increases significantly correspond to locations where the sheath grounding impedance increases, thus obtaining the sheath grounding distribution result with node position markers. For the sheath grounding distribution results, the matching degree between the node position identifier and the remaining amplitude is compared one by one. If the sheath potential amplitude rises above the preset potential rise threshold at a certain node position and the remaining amplitude of the branch exceeds the preset amplitude threshold, then the node is determined to be the actual grounding branch position. The branch number corresponding to the actual grounding branch position and the node position identifier together constitute the adjustment direction of the grounding fault discrimination condition.

[0045] In one implementation, the updated zero-sequence current amplitude data is stored branch by branch in the form of the remaining amplitude after removing the return current component, and the remaining amplitude of each branch is a time sequence with sampling timestamps.

[0046] Specifically, the remaining amplitude timing data is retrieved sequentially from the updated zero-sequence current amplitude data according to the branch number, and the sheath potential data timing data of the corresponding branch is retrieved from the initial input set according to the same branch number. Both are obtained with the same sampling interval, which has the basis for timing synchronization.

[0047] It should be noted that the process of establishing point-to-point pairing records of the remaining amplitude and sheath potential for each branch based on the sampling timestamp is implemented using a point-to-point alignment method based on the timestamp. That is, for each sampling moment, the amplitude at that moment is extracted from the remaining amplitude time series, and the potential amplitude at that moment is extracted from the sheath potential data time series. The two are stored as a pair in the branch-specific amplitude-potential joint record. The amplitude-potential joint record uses the branch number as an index and the sampling timestamp as a sub-index, carrying the point-to-point correspondence between the updated zero-sequence current amplitude and the sheath potential at the same moment.

[0048] Specifically, each cable outgoing line under the ring main unit is equipped with a sheath grounding down conductor at the terminal joint, intermediate joint, and opposite end joint. Each down conductor is marked with a sheath potential acquisition point marker. These markers are pre-registered in the dispatch database during the commissioning phase, and the marker content includes the branch number, sheath section number, and cable length from the starting end. Based on the amplitude potential joint record, the sheath potential amplitude is spatially arranged along each sheath potential acquisition point marker according to the branch number. That is, the cable length from the starting end is used as the horizontal axis coordinate, and the sheath potential amplitude is used as the vertical axis coordinate, to plot the high and low distribution of the sheath potential along the line point by point.

[0049] It is understood that the high and low distribution pattern, as the sheath grounding distribution, maintains a relatively low and flat sheath potential along the line under normal operating conditions; where the sheath grounding impedance increases or the grounding point has poor contact, the sheath potential amplitude increases significantly, exhibiting a local peak shape. The locations where the sheath potential amplitude increases significantly in the sheath grounding distribution correspond to the locations where the sheath grounding impedance increases, thus obtaining the sheath grounding distribution result with node location markers. Furthermore, for the sheath grounding distribution result, the consistency between the node location markers and the remaining amplitude is compared one by one.

[0050] Specifically, the matching degree comparison includes two parallel threshold determinations. One is for the rise in the potential amplitude of the sheath, which is defined as the difference between the potential amplitude of the node and the average potential amplitude along the same branch. The rise is compared with a preset potential rise threshold. The other is for the remaining amplitude of the branch, which is compared with a preset amplitude threshold.

[0051] In one embodiment, the preset potential rise threshold can be twice the average potential amplitude along the line, and the preset amplitude threshold can be a certain percentage of the rated zero-sequence protection activation value. Preferably, if the sheath potential amplitude rises above the preset potential rise threshold at a certain node location, and the remaining amplitude of the branch exceeds the preset amplitude threshold, then the node is determined to be a real grounding branch location; if only one of the two thresholds is met or neither is met, then the node remains a suspicious node and is not included in the real grounding branch locations. The real grounding branch location is output in the form of branch number combined with node location identifier. The branch number points to a specific outgoing switch cabinet in the ring network box, and the node location identifier points to a specific sheath grounding point along the outgoing cable. The branch number and node location identifier corresponding to the real grounding branch location together constitute the adjustment direction of the grounding fault discrimination condition. The adjustment direction is stored in a structured record format as the basis for subsequent dynamic adjustment of the grounding protection discrimination criteria.

[0052] S105. Based on the adjustment direction, the grounding protection discrimination criteria are dynamically adjusted. Taking the adjustment direction and bus voltage data as input, a threshold comparison mechanism is used to form an updated grounding protection discrimination criterion, which is used to locate the actual grounding branch.

[0053] The branch number and node location identifier are retrieved from the adjustment direction of the grounding protection discrimination criterion, and the zero-sequence voltage amplitude in the bus voltage data is retrieved from the initial input set. A threshold comparison is performed between the zero-sequence voltage amplitude and a preset zero-sequence voltage activation threshold. If the zero-sequence voltage amplitude exceeds the zero-sequence voltage activation threshold, the branch corresponding to the branch number is marked as a branch to be adjusted. Based on the branch to be adjusted, the zero-sequence current action threshold reference value corresponding to the branch number is retrieved from the pre-registered setting value table in the ring network box protection configuration. Using the location number of the actual grounding branch reflected by the node location identifier in the grounding distribution along the sheath as the basis for selecting the downshift level, the amplitude of the zero-sequence current action threshold reference value is reduced according to the pre-registered downshift level table to obtain the adjusted zero-sequence current action threshold. The adjusted zero-sequence current action threshold and the bus zero-sequence voltage over-limit flag are combined to form an AND gate logic criterion. The AND gate logic criterion determines that when the branch zero-sequence current amplitude exceeds the adjusted zero-sequence current action threshold and the bus zero-sequence voltage exceeds the zero-sequence voltage activation threshold, the branch is determined to be a real grounding branch. The AND gate logic criterion serves as the updated grounding protection discrimination criterion and is used to locate real grounding branches.

[0054] In one embodiment, the adjustment direction of the ground fault discrimination condition is stored in the form of a structured record. The record fields include branch number and node location identifier. The branch number points to a specific outgoing switch cabinet in the ring network box, and the node location identifier points to a specific sheath grounding point along the outgoing cable.

[0055] Specifically, the branch number and node position identifier in each record are retrieved sequentially from the adjustment direction of the ground fault discrimination conditions, and the zero-sequence voltage amplitude in the bus voltage data is retrieved from the initial input set.

[0056] It should be noted that the zero-sequence voltage amplitude is obtained by adding the instantaneous values ​​of the three-phase voltages of the busbar, dividing by 3, and then taking the fundamental frequency amplitude. This reflects the level of zero-sequence voltage rise in the busbar section of the ring mains. Under normal operating conditions, the zero-sequence voltage amplitude remains within a few volts; when a ground fault occurs, the zero-sequence voltage amplitude rises significantly to a large proportion of the rated phase voltage.

[0057] Specifically, a threshold comparison is performed between the zero-sequence voltage amplitude and a preset zero-sequence voltage activation threshold. In one possible implementation, the zero-sequence voltage activation threshold is 0.15 times the rated phase voltage, a value consistent with the engineering practice of low-current grounding line selection activation thresholds in the power industry. If the zero-sequence voltage amplitude exceeds the zero-sequence voltage activation threshold, the branch corresponding to the branch number is marked as a branch to be adjusted, resulting in an adjustment criterion item with a busbar zero-sequence voltage exceeding limit indicator. If the zero-sequence voltage amplitude does not exceed the zero-sequence voltage activation threshold, the original protection setting is maintained and the adjustment process is not initiated. Further, the zero-sequence current action threshold reference value corresponding to the branch number is retrieved from the pre-registered setting value table in the ring main unit protection configuration. This setting value table is formed during the commissioning phase according to the engineering setting principles based on conditions such as the outgoing line length, load current, cable cross-section, and no-load capacitor current of each branch, with each outgoing line corresponding to an action threshold reference value.

[0058] In one embodiment, the actual grounding branch location reflected by the node location identifier has a corresponding location number in the sheath grounding distribution along the line. The location numbers are arranged sequentially from near to far according to the cable length from the beginning, with smaller numbers for the beginning segment and larger numbers for the end segment. Using the location number as the basis for selecting the adjustment level, the zero-sequence current action threshold reference value is adjusted downwards according to a pre-registered adjustment level table. The adjustment level table is a two-dimensional lookup table, with row indexes representing location numbers, column indexes representing cable outgoing type, and table entry values ​​representing adjustment coefficients, which range from 0.5 to 0.9. Preferably, the adjustment coefficient is larger when the location number is closer to the beginning and smaller when the location number is closer to the end, thus tightening the near-end fault threshold and preventing the far-end misjudgment threshold from being excessively lowered. The zero-sequence current action threshold reference value is multiplied by the retrieved adjustment coefficient to obtain the adjusted zero-sequence current action threshold.

[0059] Understandably, the downgrade table is registered based on multiple simulated grounding test data during the commissioning phase. The tests cover artificial grounding conditions at the beginning, middle, and end of each branch, recording the critical current amplitude for correct protection operation at each level, and then using the downgrade coefficient obtained from the reverse lookup to fill in the table entries. An AND gate logic criterion is formed by combining the adjusted zero-sequence current operating threshold with the bus zero-sequence voltage over-limit flag. This AND gate logic criterion stipulates that if the branch zero-sequence current amplitude exceeds the adjusted zero-sequence current operating threshold, and the bus zero-sequence voltage exceeds the zero-sequence voltage activation threshold, the branch is determined to be a true grounded branch when both conditions are met simultaneously. The AND gate logic criterion is implemented using a two-input logic gate: one input is the zero-sequence current amplitude comparison result, and the other input is the bus zero-sequence voltage comparison result; the output is the branch location marker.

[0060] For example, the zero-sequence current operating threshold of branch 2 of a ring main unit is 40 amps. The node position number corresponding to this branch points to the middle section, and the adjustment coefficient is found to be 0.7, resulting in an adjusted zero-sequence current operating threshold of 28 amps. At the same time, the zero-sequence voltage amplitude of the bus is measured to be 0.18 times the rated phase voltage, exceeding the starting threshold by 0.15 times, and the marker is set. The zero-sequence current amplitude of branch 2 is measured to be 32 amps, exceeding the 28 amp threshold. Both inputs of the AND gate criterion are true, the branch positioning marker output is true, and branch 2 is positioned as a real grounded branch. The AND gate logic criterion serves as the updated grounding protection discrimination criterion. The updated grounding protection discrimination criterion is stored branch by branch number and resides in the ring main unit protection device in the form of a criterion entry list, used to locate the real grounded branch.

[0061] S106. Match the updated grounding protection discrimination criteria with the zero-sequence current amplitude data. After the protection setting adjustment conditions are met, generate a protection setting adjustment command, update the ring network box protection setting, and complete the identification and setting adjustment of the actual grounding branch.

[0062] The updated grounding protection discrimination criteria are retrieved by branch number from the AND gate logic criterion entries, and the zero-sequence current amplitude data of the corresponding branch is retrieved from the initial input set. For each branch, the zero-sequence current amplitude data is substituted into the zero-sequence current amplitude comparison input of the AND gate logic criterion entries, and the existing bus zero-sequence voltage comparison result in the AND gate logic criterion entries is read as the voltage comparison input value. The branch location flag is obtained through two-input logic gate operations, resulting in a branch-by-branch criterion matching result. Based on the criterion matching result, the number of consecutive sampling frames for the branch location flag being true is accumulated. If the number of consecutive accumulated frames for the branch location flag exceeds the preset setting adjustment duration frame threshold, the protection setting adjustment condition is determined to be met. The adjusted zero-sequence current action threshold corresponding to the branch is retrieved from the setting value table pre-registered in the ring network box protection configuration as the target setting value. The target setting value and the branch number are encapsulated into a protection setting adjustment instruction to obtain an instruction message for the protection device. The instruction message is sent along the internal data link of the ring network box to the protection device corresponding to the branch. After receiving the instruction message, the protection device replaces the original zero-sequence current action threshold in its setting register with the target setting value, completes the update of the ring network box protection setting value, and writes the branch positioning mark bit into the action record field of the protection device, thus completing the identification and setting adjustment of the actual grounded branch.

[0063] In one implementation, the updated grounding protection discrimination criteria reside in a criterion entry list within the ring main unit protection device, categorized by branch number. Each entry includes the branch number, the adjusted zero-sequence current operating threshold, the zero-sequence voltage activation threshold, and an AND gate logic criterion. The AND gate logic criterion is defined as a combination of conditions: I0 > Ith and U0 > Uth, where I0 is the zero-sequence current, Ith is the operating threshold (e.g., 0.5A), U0 is the zero-sequence voltage, and Uth is the activation threshold (e.g., 10V). The list construction process involves inputting the branch number and real-time measurement data, using a threshold adjustment algorithm (e.g., calculating Ith and Uth based on the average of historical data), and outputting a complete entry list. The initial input set is a set of branch zero-sequence current amplitude data collected in real-time by power system sensors, obtained through sampling 100 times per second by a CT sensor; for example, the set consists of {Branch 1: 5.2A, Branch 2: 3.8A}.

[0064] Specifically, AND gate logic criterion entries are retrieved one by one from the criterion entry list according to the branch number, and the zero-sequence current amplitude data of the corresponding branch is retrieved from the initial input set according to the same branch number.

[0065] It should be noted that the zero-sequence current amplitude data arrives frame by frame in the form of sampling frames, and each sampling frame carries the zero-sequence current amplitude of the branch at the current moment. For each branch, its zero-sequence current amplitude data is substituted into the zero-sequence current amplitude comparison input terminal in the AND gate logic criterion entry, and compared with the adjusted zero-sequence current action threshold to obtain the current comparison result; at the same time, the bus zero-sequence voltage amplitude U0 is collected as input, and the comparison result is calculated by the formula C=U0>5, where U0 is the collected voltage value, 5 is the preset threshold V, and if C is true, the result is 1, which is taken as the voltage comparison input value, and after the AND operation of the two input logic gates, the branch positioning mark bit is output.

[0066] Specifically, the branch location marker is a single-bit state value. A true value indicates that the branch meets the criteria for a true grounded branch in the current sampling frame, while a false value indicates that it does not. The branch location markers for each sampling frame are sequentially written into the criterion matching result queue for each branch according to its branch number. The criterion matching result queue maintains the marker status of several recent frames using a sliding window method. Further, based on the criterion matching results, the true states of the branch location markers are accumulated according to the number of consecutive sampling frames. The specific accumulation method is as follows: a continuous truth counter is set for each branch, and the criterion matching result queue is scanned frame by frame. When the branch location marker is true, the counter is incremented by one; when the branch location marker is false, the counter is reset to zero. The fixed-value adjustment continuous frame number threshold T is determined by the formula T = fs × t. min The calculation yields the result, where fs is the sampling frequency in Hertz, and t... min The minimum system response time, in seconds, for example, fs = 1000 Hz, t min When T=10 at 0.01 seconds, if the cumulative number of frames of the continuous truth counter exceeds T, then the protection setting adjustment condition is determined to be met.

[0067] In one possible implementation, the threshold for the number of frames for the fixed value adjustment is set to 40 frames, which corresponds to a steady-state judgment duration of 80 milliseconds under the power frequency cycle, thus avoiding instantaneous misjudgments caused by transient disturbances such as lightning strikes and load switching.

[0068] It is understood that when the protection setting adjustment condition is met, the adjusted zero-sequence current operating threshold corresponding to the branch is retrieved from the pre-registered setting value table in the ring main unit protection configuration as the target setting value. The target setting value and the branch number are encapsulated together into a protection setting adjustment instruction. The protection setting adjustment instruction is constructed in the form of an instruction message, and the message fields include a message header, target branch number, target setting value field, setting value type code, check code, and timestamp. Preferably, the message header identifies the setting adjustment type instruction, the setting value type code points to the zero-sequence current operating threshold to distinguish it from other protection setting types such as overcurrent setting and instantaneous overcurrent setting, and the check code is generated using cyclic redundancy check to verify whether there are any errors in the message during transmission.

[0069] For example, when the continuous truth counter of a certain outgoing branch reaches 40 frames and the target setting value is 28 amps, the protection setting adjustment command is output in the form of a message. The target branch number in the message is filled with the switch cabinet number of that branch, and the target setting value field is filled with 28 amps. The command message is sent to the protection device corresponding to that branch along the internal data link of the ring network box. The internal data link of the ring network box is constructed in the form of industrial Ethernet or fieldbus. After receiving the command message, the protection device first verifies the message check code. After the verification is successful, it is indexed to the zero-sequence current action threshold storage unit in the setting register according to the setting type code, and the original zero-sequence current action threshold is replaced with the target setting value, thus completing the update of the ring network box protection setting value. Preferably, while completing the setting replacement, the protection device writes the branch positioning mark into the action record field of the protection device. The action record field is added and saved in the form of entries. The entry content includes the action timestamp, target branch number, original zero-sequence current action threshold value, target setting value and trigger reason code. The trigger reason code value points to the setting adjustment caused by the remote return current stripping, so as to distinguish it from manual setting adjustment and periodic inspection adjustment.

[0070] It should be noted that after the fixed value replacement, the original fixed value is entered, and the formula new is used. value =old value Calculate the new value using ×0.8, where new value For the new setting, old value The process involves setting the original value, adjusting it by 0.8, and outputting the result. The action record is written (recording the operation timestamp and type, such as replacement or writing). The actual grounding branch is registered in both the criterion entry list (an array storing grounding criteria) and the action record field (a field in the database recording operation logs), thus completing the identification and setting adjustment of the actual grounding branch.

[0071] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for adjusting the protection settings of a primary and secondary integrated ring main unit, characterized in that, The method includes: Zero-sequence current amplitude data is obtained from the zero-sequence acquisition unit, sheath potential data is obtained from the sheath potential acquisition unit, and bus voltage data is obtained from the bus voltage acquisition module, and then collected into an initial input set. The phase difference between the zero-sequence current amplitude data and the sheath potential data in the initial input set is calculated using the Fourier transform algorithm. The phase difference is compared with the bus voltage data, and the branch state that exceeds the preset limit is determined to be the active return current state of the far-end branch. The cable branch capacitance is identified by the active return current pattern of the far-end branch, and the cable branch capacitance is decomposed by wavelet transform algorithm. The equivalent components obtained by decomposition are substituted into the zero-sequence current amplitude data for adjustment to form updated zero-sequence current amplitude data. The updated zero-sequence current amplitude data is combined with the sheath potential data to extract the sheath grounding distribution. The nodes in the sheath grounding distribution that match the zero-sequence current amplitude data are identified as the actual grounding branch locations, forming the adjustment direction of the grounding fault discrimination conditions. Based on the aforementioned adjustment direction, the grounding protection discrimination criteria are dynamically adjusted. Using the adjustment direction and bus voltage data as input, a threshold comparison mechanism is used to form an updated grounding protection discrimination criterion to locate the actual grounding branch. The updated grounding protection discrimination criteria are matched with the zero-sequence current amplitude data. After the protection setting adjustment conditions are met, a protection setting adjustment command is generated to update the ring network box protection setting, thus completing the identification and setting adjustment of the actual grounding branch.

2. The method for adjusting the protection settings of a primary and secondary integrated ring main unit according to claim 1, characterized in that, The zero-sequence current amplitude data acquired from the zero-sequence acquisition unit, the sheath potential data acquired from the sheath potential acquisition unit, and the bus voltage data acquired from the bus voltage acquisition module are aggregated into an initial input set, including: The zero-sequence current amplitude data of each branch outgoing line side of the ring network box are obtained from the zero-sequence acquisition unit and synchronously read at a uniform sampling interval to obtain a zero-sequence current amplitude data sequence with branch number identifier. The sheath potential data of the cable sheath grounding point is obtained from the sheath potential acquisition unit, and the three-phase voltage data of the incoming bus of the ring network box is obtained from the bus voltage acquisition module. The channels are associated according to the branch number and the sampling timestamp, and the data are collected with the zero-sequence current amplitude data sequence to form the initial input set.

3. The method for adjusting the protection settings of a primary and secondary integrated ring main unit according to claim 1, characterized in that, The step involves using a Fourier transform algorithm to calculate the phase difference between the zero-sequence current amplitude data and the sheath potential data in the initial input set, comparing this with the bus voltage data, and determining the branch state where the phase difference exceeds a preset limit as an active return current state in the far-end branch, including: Zero-sequence current amplitude data and sheath potential data are retrieved from the initial input set according to the branch number. The Fourier transform algorithm is applied to the two co-source time-series waveforms to extract the power frequency components. The corresponding power frequency phase angles are obtained and subtracted to obtain the phase difference sequence of each branch. Using the power frequency phase angle extracted by Fourier transform of the zero-sequence voltage component in the bus voltage data as a common reference benchmark, the phase difference sequence is aligned with the reference benchmark according to the branch number to obtain a normalized phase difference sequence with the zero-sequence voltage phase of the bus as the zero point. If a branch in the normalized phase difference sequence exceeds the upper and lower limits of the phase difference exceeding the limit determination window, it is determined that the branch is in the active return state of the far-end branch.

4. The method for adjusting the protection settings of a primary and secondary integrated ring main unit according to claim 1, characterized in that, The process involves identifying cable branch capacitance through the active return current pattern of the distal branch, decomposing the cable branch capacitance using a wavelet transform algorithm, and substituting the decomposed equivalent components into the zero-sequence current amplitude data for adjustment to form updated zero-sequence current amplitude data, including: Zero-sequence current amplitude data of the corresponding branch is retrieved from the set of branches in the active return current state of the far-end branch. Wavelet transform algorithm is used to decompose the data at multiple scales to obtain the approximate component of the low-frequency band and the detailed component of the high-frequency band near the power frequency. The energy ratio of the detailed component is used as the basis for identifying the activity level of the cable branch capacitance. For the detailed components, the power frequency phase angle corresponding to the sheath potential data of the corresponding branch is taken as the reference direction according to the branch number. The detailed components are projected along a direction that leads the reference direction by 90 degrees to obtain the capacitive equivalent component, and projected along a direction in the same direction as the reference direction to obtain the resistive equivalent component. The return current composite amplitude is obtained by summing the amplitudes of the capacitive equivalent component and the resistive equivalent component according to the branch number. The corresponding return current composite amplitude is then subtracted from the zero-sequence current amplitude data to obtain the remaining amplitude as the updated zero-sequence current amplitude data.

5. The method for adjusting the protection settings of a primary and secondary integrated ring main unit according to claim 1, characterized in that, The process of combining the updated zero-sequence current amplitude data with the sheath potential data to extract the sheath grounding distribution, and identifying the nodes in the sheath grounding distribution that match the zero-sequence current amplitude data as the actual grounding branch locations, forms the adjustment direction for the grounding fault discrimination conditions, including: The remaining amplitude is retrieved from the updated zero-sequence current amplitude data according to the branch number, and a joint amplitude potential record is established with the sheath potential data of the corresponding branch according to the sampling timestamp. According to the amplitude potential joint record, the potential amplitude of the sheath is spatially arranged according to the identification of the potential acquisition point along the sheath, and the high and low distribution pattern of the potential of the sheath along the line is used as the sheath grounding distribution. The node location is compared with the remaining amplitude. If the potential amplitude of the protective layer at the node location rises above a preset potential rise threshold and the remaining amplitude of the corresponding branch exceeds a preset amplitude threshold, then the node is determined to be the location of the actual grounding branch.

6. The method for adjusting the protection settings of a primary and secondary integrated ring main unit according to claim 1, characterized in that, The process involves dynamically adjusting the grounding protection discrimination criteria based on the adjustment direction. Using the adjustment direction and bus voltage data as input, a threshold comparison mechanism is employed to form an updated grounding protection discrimination criterion, locating the actual grounding branch. This includes: The branch number and node position identifier are retrieved from the adjustment direction. The zero-sequence voltage amplitude is retrieved from the bus voltage data and compared with the preset zero-sequence voltage start threshold. If the zero-sequence voltage amplitude exceeds the zero-sequence voltage start threshold, the corresponding branch is marked as a branch to be adjusted. The zero-sequence current action threshold reference value corresponding to the branch to be adjusted is retrieved from the pre-registered setting value table in the ring network box protection configuration. The amplitude is reduced according to the pre-registered reduction level table, using the node position identifier as the basis for selecting the reduction level, and the adjusted zero-sequence current action threshold is obtained. The updated grounding protection discrimination criterion is formed by combining the adjusted zero-sequence current operating threshold with the bus zero-sequence voltage over-limit indicator.

7. The method for adjusting the protection settings of a primary and secondary integrated ring main unit according to claim 1, characterized in that, The process involves matching the updated grounding protection discrimination criteria with the zero-sequence current amplitude data. Once the protection setting adjustment conditions are met, a protection setting adjustment command is generated to update the ring main unit's protection settings, completing the identification and setting adjustment of the actual grounding branch. This includes: From the updated grounding protection discrimination criteria, retrieve the AND gate logic criterion entry according to the branch number, substitute the corresponding branch zero-sequence current amplitude data into the AND gate logic criterion entry, and obtain the branch location mark bit through two input logic gate operations; The branch positioning mark is accumulated as true according to the number of consecutive sampling frames. If the number of consecutive accumulated frames exceeds the preset setting adjustment duration frame threshold, the adjusted zero-sequence current action threshold is taken from the setting value table as the target setting value and encapsulated with the branch number to form the protection setting value adjustment instruction. The protection setting adjustment command is sent to the corresponding protection device along the internal data link of the ring network box, and the original zero-sequence current action threshold in the setting register is replaced with the target setting value.

Images