A power grid fault adaptive disposal method and system based on topology reconfiguration

Abstract

The application discloses a kind of power grid fault adaptive disposal method and system based on topology reconfiguration, it is related to power grid fault disposal technical field, including the following steps: obtaining the operation sequence including target switch, according to the action parameter of target switch and branch transient current, the natural zero-crossing capability of target switch is evaluated;In response to the evaluation result, the maximum injectable current of candidate commutation branch is calculated according to the admittance of power grid, and auxiliary commutation branch is selected in combination with power grid disturbance index;Extract the arc impedance time-varying characteristic in the breaking process of target switch, calculate the resonance frequency offset in combination with the equivalent circuit parameters of auxiliary commutation branch, and generate feedforward control sequence;According to feedforward control sequence, control auxiliary commutation branch injects compensation current to carry out arc extinguishing, and update operation sequence according to the deviation between actual topology and target topology of power grid after arc extinguishing.The application is used to solve the problems that switch is easy to sustain arc in prior art, static compensation mismatch and topology change leads to reconfiguration operation failure.

Description

Technical Field

[0001] This invention relates to the field of power grid fault handling technology, and more specifically, to a power grid fault adaptive handling method and system based on topology reconfiguration. Background Technology

[0002] Existing power grid fault handling and recovery schemes typically involve the dispatch system controlling circuit breakers to disconnect the faulty area and issuing a sequence of switching operations after a fault occurs, in order to isolate the fault and restore power supply to the non-faulty areas.

[0003] However, existing technologies still have some technical problems in practical applications: First, under high-intensity transient current impact conditions, continuous arcing may occur during the fault switch opening process, causing the actual opening conditions to deviate from the expected arc extinguishing conditions, increasing the risk of arc extinguishing failure, and potentially triggering cascading trips or expanding the power outage area; Second, when using active converter methods to assist arc extinguishing, the injection of compensation current may cause additional disturbances such as voltage fluctuations in surrounding nodes, and the preset compensation signal is difficult to adjust according to changes in the arc state and network transient characteristics during the opening process, resulting in a deviation between the compensation current and the actual arc extinguishing conditions, affecting the auxiliary arc extinguishing effect under complex conditions; Finally, existing topology reconfiguration processes are usually executed based on pre-calculated operating steps. If the actual operation state of the switch deviates from the expected state during fault clearing, the subsequent operation sequence may also be mismatched with the current grid state, triggering equipment blocking or operational failure, thereby affecting the continuous execution of the power restoration process. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a power grid fault adaptive handling method and system based on topology reconfiguration. By evaluating the natural zero-crossing capability of the switch and selecting the converter branch, a feedforward control sequence is generated in combination with the time-varying characteristics of the arc impedance, and the operation sequence is updated according to the topology deviation after arc extinguishing, so as to solve the problems of easy continuous arcing of the switch, static compensation mismatch, and failure of reconfiguration operation caused by topology changes in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An adaptive fault handling method for power grids based on topology reconfiguration includes the following steps: The operation sequence of the target switch is obtained, and the natural zero-crossing capability of the target switch is evaluated based on the operating parameters of the target switch and the transient current of the branch. In response to the evaluation results, the maximum injectable current of the candidate converter branch is calculated based on the grid admittance, and the auxiliary converter branch is selected in combination with the grid disturbance index. Extract the time-varying characteristics of the arc impedance during the target switch opening and closing process, combine them with the equivalent circuit parameters of the auxiliary converter branch, calculate the resonant frequency offset caused by the time-varying characteristics of the arc impedance, and generate a feedforward control sequence. Based on the feedforward control sequence, the auxiliary converter branch is controlled to inject compensation current to extinguish the arc, and the operation sequence is updated according to the deviation between the actual grid topology and the target topology after arc extinguishing.

[0006] In a preferred embodiment, evaluating the natural zero-crossing capability of the target switch includes: extrapolating based on the rate of change of the transient current of the branch and combining it with the opening time in the action parameters to obtain the expected current zero-crossing time and the rate of change of zero; when the expected current zero-crossing time meets the preset arcing time window constraint and the rate of change of zero is less than the preset rate of change threshold, it is determined that the target switch has natural zero-crossing capability.

[0007] In a preferred embodiment, the step of calculating the maximum injectable current of the candidate converter branch based on grid admittance includes: obtaining the mutual admittance and transient node voltage difference between the candidate converter branch and the target switching node, and calculating the theoretical injection current waveform based on the frequency-varying characteristics of mutual admittance and the phase shift of the transient node voltage difference; limiting the theoretical injection current waveform based on the maximum output current of the candidate converter branch, and taking the peak value of the current waveform after limiting as the maximum injectable current.

[0008] In a preferred embodiment, the step of selecting the auxiliary converter branch in conjunction with the grid disturbance index includes: calculating the node voltage change and branch power flow change caused by the commissioning of each candidate converter branch as grid disturbance index; weighting the maximum injectable current and the grid disturbance index, and selecting the candidate converter branch with the best score as the auxiliary converter branch.

[0009] In a preferred embodiment, the extraction of the time-varying characteristics of arc impedance during the target switch opening process includes: matching the arc time constant and arc heat dissipation power based on the steady-state operating current of the target switch before the fault, and collecting the arc electrical quantities during the contact separation of the target switch; substituting the arc time constant, arc heat dissipation power and arc electrical quantities into a preset energy balance differential equation, solving for the arc conductance and taking its reciprocal to obtain the time-varying characteristics of arc impedance.

[0010] In a preferred embodiment, generating the feedforward control sequence includes: solving for the equivalent network resonant frequency based on the equivalent inductance and equivalent capacitance of the auxiliary converter branch and the time-varying characteristics of the arc impedance; using the difference between the equivalent network resonant frequency and the power grid frequency as the resonant frequency offset, and performing time integration to obtain the phase compensation angle; and generating a transient compensation waveform based on the phase compensation angle as the feedforward control sequence.

[0011] In a preferred embodiment, controlling the auxiliary converter branch to inject compensation current for arc extinguishing according to the feedforward control sequence includes: controlling the auxiliary converter branch to output an initial compensation current according to the feedforward control sequence; superimposing the initial compensation current and the branch transient current, extracting the fundamental error current component of the superimposed current, and converting it into a frequency compensation term to correct the feedforward control sequence; injecting compensation current based on the corrected feedforward control sequence until a preset interruption confirmation condition is met.

[0012] In a preferred embodiment, satisfying the preset interruption confirmation condition includes: calculating the arc energy integral within a time window when the superimposed current amplitude is less than a preset zero-zone threshold; monitoring the zero-crossing rate of change of the superimposed current and the transient recovery voltage across the target switch contact; and determining that the interruption confirmation condition is satisfied when the arc energy integral, the zero-crossing rate of change, and the transient recovery voltage all satisfy the preset safety threshold condition.

[0013] In a preferred embodiment, updating the operation sequence based on the deviation between the actual topology of the power grid after arc extinguishing and the target topology includes: performing a topology comparison between the actual topology of the power grid after arc extinguishing and the target topology to identify failed branches; removing operation steps associated with the failed branches from the operation sequence; and reordering the remaining operation steps according to preset network constraints to generate an updated operation sequence.

[0014] A topology-reconfiguration-based adaptive fault handling system for implementing any of the methods described herein includes the following modules: a zero-crossing evaluation module, used to acquire an operation sequence including a target switch, and evaluate the natural zero-crossing capability of the target switch based on the operating parameters of the target switch and the transient current of the branch; a branch selection module, used to calculate the maximum injectable current of candidate converter branches based on grid admittance in response to the evaluation result, and select auxiliary converter branches in combination with grid disturbance indicators; a sequence generation module, used to extract the time-varying characteristics of arc impedance during the opening process of the target switch, calculate the resonant frequency offset caused by the time-varying characteristics of arc impedance in combination with the equivalent circuit parameters of the auxiliary converter branch, and generate a feedforward control sequence; and an execution update module, used to control the auxiliary converter branch to inject compensation current for arc extinguishing based on the feedforward control sequence, and update the operation sequence according to the deviation between the actual grid topology and the target topology after arc extinguishing.

[0015] The technical effects and advantages of the adaptive power grid fault handling method and system based on topology reconfiguration of this invention are as follows: 1. This invention assesses the current injection capability of candidate converter branches based on grid admittance, and, in conjunction with the equivalent circuit parameters of the auxiliary converter branch and the time-varying characteristics of the arc impedance during the target switch opening process, determines the resonant frequency offset of the equivalent network and generates a corresponding feedforward control sequence. This allows the compensation current of the auxiliary converter branch to be adjusted according to the changes in the transient characteristics of the network during the opening process, reducing the risk of auxiliary arc extinguishing failure caused by the mismatch between the compensation signal and the actual transient state, and reducing the additional disturbance of the compensation current injection to the node voltage, thereby improving the auxiliary arc extinguishing effect and the operational stability of the handling process under complex transient conditions.

[0016] 2. This invention assesses the natural zero-crossing capability of the target switch in the early stage of fault handling and triggers auxiliary switching based on the assessment results. Simultaneously, after arc extinguishing, it compares the deviation between the actual grid topology and the target topology and updates the operation sequence, enabling subsequent switch operations to be adjusted according to the actual grid conditions. This reduces the risk of arc extinguishing failure caused by direct disconnection under insufficient natural zero-crossing capability, as well as the subsequent operation mismatch or execution failure caused by inconsistency between the actual grid topology and the preset topology, thereby improving the continuity and reliability of fault isolation and subsequent power restoration processes. Attached Figure Description

[0017] Figure 1 This is a schematic flowchart of a power grid fault adaptive handling method based on topology reconfiguration provided in an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram illustrating the time-varying characteristics of arc impedance and the evolution of the equivalent network resonant frequency, provided for an embodiment of the present invention.

[0019] Figure 3 This is a schematic diagram of a power grid fault adaptive handling system module based on topology reconfiguration, provided in an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1, Figure 1 This invention presents an adaptive fault handling method for power grids based on topology reconfiguration, comprising the following steps: Step 1: Obtain the operation sequence of the target switch, and evaluate the natural zero-crossing capability of the target switch based on the operating parameters of the target switch and the transient current of the branch.

[0022] 1.1 Obtain the operation sequence including the target switch, including: When a transient impact fault such as an asymmetric short circuit occurs in an AC / DC hybrid microgrid, in order to achieve fault isolation and power restoration in non-faulty areas, the operation sequence including the target switch is first obtained. The operation sequence is usually automatically issued by the power grid energy management system (EMS) or the monitoring and data acquisition system (SCADA) based on a preset fault isolation strategy (e.g., the near backup protection action logic after the main protection action fails, or the non-faulty area islanding strategy based on topological connectivity). As an optional implementation, the operation sequence can also be generated by calling a pre-trained power big data language model deployed at the dispatch terminal. By using the current real-time topological node status of the power grid and the location of the fault node as input prompt words, the model forward inferences and outputs an operation sequence text composed of multiple switch opening and closing commands arranged in chronological order.

[0023] The mechanical tripping time and rated arcing limit of the target switch are obtained by calling the equipment ledger parameters and used as the action parameters. Simultaneously, the transient current flowing through the branch of the target switch is collected in real time at a sampling frequency of 10kHz using a wideband protection-grade current transformer (response bandwidth upper limit not less than 20kHz) or a Rogowski coil. Considering that the collected raw branch transient current signal contains high-frequency electromagnetic interference noise, the branch transient current is preprocessed; specifically, a Butterworth anti-aliasing low-pass filter (e.g., with a cutoff frequency set to 4kHz) and a one-dimensional standard median filtering algorithm are used. The original discrete sequence of transient current containing noise is taken as input, and by filtering out high-frequency harmonics and glitches, smooth branch transient current sampling points are output.

[0024] 1.2 Extrapolation is performed based on the rate of change of the branch transient current, and combined with the tripping time in the action parameters to obtain the expected zero-crossing time and zero-crossing rate of change of the current, including: Based on the preprocessed branch transient current sampling points, a sliding time window mechanism is introduced, and a second-order Taylor expansion method is used to continuously extrapolate the evolution trajectory of the current transient current. Specifically, as the sliding time window (the window width can be 1 to 2 ms) slides, the measured amplitude of the branch transient current and its multiple derivatives at the current sampling moment are extracted in real time, where the first derivative is the rate of change of the branch transient current. Using the above characteristic quantities, a second-order Taylor series polynomial of the transient current changing with time is constructed, and the current value at the extrapolation moment is set to zero. Based on this, an equation for solving the expected current zero-crossing moment is established, as follows: (1) in, It is a second-order Taylor series polynomial, with time variable. The unit is seconds; For the current moment The measured amplitude of the transient current in the branch circuit, in amperes; The rate of change of the transient current in the branch (i.e., the first derivative) is expressed in A / s. This is the second derivative of the current, with units of A / s. 2 .

[0025] The above second-order Taylor series polynomial equation (i.e., a quadratic equation in one variable) Treating them as unknowns, extract the coefficients of each term in the equation and calculate the discriminant. ;like This indicates that the transient current has no real zero-crossing point under the local decay trajectory of the current sampling time window. At this time, the sliding time window is pushed forward to update the current. And restart the rolling solution; if Then, the quadratic formula can be used to calculate the result. The real solution, and then combined with the current moment Obtain candidate mathematical roots that cross zero.

[0026] During the sliding time window progression, candidate mathematical zero-crossing roots are continuously screened. Specifically, the tripping time in the action parameters is added to the triggering time of the tripping command for the target switch in the operation sequence to obtain the expected contact separation time. When a positive real number solution that is greater than the contact separation time and closest to the contact separation time is extracted within a certain time window, the solution result of that time window is locked as the final expected current zero-crossing time, and the rolling extrapolation of the time window is immediately stopped. It should be noted that for severe DC bias faults in AC / DC hybrid power grids, the transient current may not naturally cross zero for a long time. Therefore, if the time span covered by the sliding time window extrapolation has reached the limit of the preset arcing time window constraint (i.e., the contact separation time plus the rated arcing limit in the action parameters), and the discriminant is still less than 0 or there are no effective zero-crossing roots, the rolling extrapolation is forcibly terminated, and it is directly determined that the expected arcing duration of the target switch does not meet the preset arcing time window constraint.

[0027] Based on the second-order Taylor series polynomial constructed above, time... Differentiate the function and substitute the expected zero-crossing time of the current obtained from the solution into the differentiated function to obtain the expected rate of change at zero crossing. The calculation formula is as follows: (3) in, The expected time of current zero crossing; The expected rate of change at zero is expressed in A / μs.

[0028] 1.3 When the expected current zero-crossing time meets the preset arcing time window constraint and the zero-crossing change rate is less than the preset change rate threshold, the target switch is determined to have natural zero-crossing capability, including: After obtaining the expected zero-crossing time and zero-crossing rate of change, a comprehensive condition judgment is performed. Specifically, the expected arcing duration is obtained by subtracting the contact separation time from the expected zero-crossing time; it is then determined whether the expected arcing duration meets the preset arcing time window constraint; the arcing time window constraint is determined based on the rated arcing limit in the action parameters, and is usually set to half a power frequency cycle, i.e., within 10ms. Simultaneously, it is determined whether the zero-crossing rate of change is less than a preset rate of change threshold, which is a critical safety slope set to prevent arc thermal reignition; for example, for a 10kV voltage level switch, this value can be 15A / μs.

[0029] If the expected arcing duration meets the arcing time window constraint and the zero-crossing change rate is less than the preset change rate threshold, the target switch is determined to have natural zero-crossing capability, and the trip command is directly released according to the original operation sequence. The target switch completes the physical disconnection by itself, and after successful disconnection, it executes the update operation sequence process described in step 4.2 to restore the grid structure after fault isolation. Conversely, if the expected arcing duration exceeds the window constraint, or the expected zero-crossing change rate is greater than or equal to the change rate threshold, it is determined that it does not have natural zero-crossing capability, and the trip command is temporarily intercepted and suspended, triggering the subsequent auxiliary converter arc extinguishing process of actively injecting compensation current.

[0030] This step extrapolates the expected zero-crossing time and zero-crossing rate of the current based on the target switch's operating parameters and transient current change rate, and compares them with the arcing time window and change rate threshold to achieve a pre-assessment of the target switch's breaking capacity. Compared with the traditional passive tripping method, this step can determine in advance whether the target switch has reliable breaking conditions, thereby reducing the risk of continuous arcing and reignition caused by direct tripping due to lack of breaking capacity, and improving the reliability of fault isolation.

[0031] Step 2: In response to the evaluation results, calculate the maximum injectable current of the candidate converter branch based on the grid admittance, and select the auxiliary converter branch in combination with the grid disturbance index.

[0032] If the evaluation result in step 1 indicates that the target switch does not have natural zero-crossing capability, then the trip command of the operation sequence is suspended and the subsequent active intervention process is triggered.

[0033] 2.1 Calculate the maximum injectable current of the candidate converter branch based on the grid admittance, including: 2.1.1 Obtain the mutual admittance and transient node voltage difference between the candidate converter branch and the target switching node, and calculate the theoretical injected current waveform based on the frequency-varying characteristics of the mutual admittance and the phase shift of the transient node voltage difference, as follows: The global node admittance matrix is ​​obtained based on the Energy Management System (EMS). Using Krone elimination, the target switching node and candidate converter branch nodes in the global node admittance matrix are retained, while the remaining network nodes are eliminated. The global node admittance matrix is ​​then divided into blocks, and the nodes to be eliminated are eliminated using Schur complement operations, resulting in a dimension-reduced power frequency static equivalent admittance matrix. The candidate converter branch is a power electronic device electrically adjacent to the target switching node, including flexible DC interconnection devices or energy storage converters, which possess bidirectional rapid power regulation capabilities. Electrical topological proximity means that the number of physical nodes in the grid topology connectivity graph does not exceed 2 hops, or the per-unit value of the equivalent transfer impedance is less than 0.1. The off-diagonal elements between the target switching node and the candidate converter branch nodes in this power frequency static equivalent admittance matrix represent the power frequency equivalent transfer admittance between them, i.e., the power frequency static mutual admittance.

[0034] By deploying broadband voltage sensors or transient waveform recorders at the access points of candidate converter branches and the nodes at both ends of the target switch, the transient node voltage difference between the candidate converter branch and the target switch node is obtained in real time through acquisition and subtraction. The acquired discrete transient node voltage difference signal is then subjected to a Fast Fourier Transform (FFT) to extract the dominant frequency bands and their corresponding transient node voltage phasors in the current signal. The FFT sampling frequency is set to 10kHz, and a Hamming window is introduced to suppress spectral leakage. The dominant frequency band refers to the frequency component whose amplitude exceeds a preset reference value (e.g., 5% of the fundamental frequency amplitude), typically including DC components, the fundamental frequency, and characteristic higher harmonics (e.g., the 3rd and 5th harmonics).

[0035] Because the fault transient process contains multiple frequency components, the skin effect of the transmission line and the parasitic capacitance of the transformer cause the line impedance and admittance to change with frequency. The mutual admittance involved in the subsequent calculation is a complex value with frequency-varying characteristics. To accurately obtain the mutual admittance values ​​corresponding to each dominant frequency band, the J.Marti frequency-varying distributed parameter model built into the Electromagnetic Transient Program (EMTP) is invoked. This model pre-fits the frequency-varying characteristic curves calculated based on line parameters at different frequencies into rational function expressions. During online operation, each dominant frequency band extracted by the aforementioned Fast Fourier Transform is used as the analysis frequency point in sequence. The center frequency value corresponding to the frequency band is used as the actual frequency scalar and substituted into the rational function expression to solve for the dynamic impedance and admittance values ​​at that frequency. The power frequency static parameters of the corresponding branch in the power frequency static equivalent admittance matrix are then replaced, thereby obtaining the updated values ​​of the off-diagonal elements between the target switching node and the candidate converter branch node in the power frequency static equivalent admittance matrix at that frequency, i.e., the complex values ​​of mutual admittance at that frequency.

[0036] For each dominant frequency band, the transient node voltage difference phasor corresponding to that frequency band is multiplied by the complex value of the mutual admittance to obtain the theoretical injected current phasor at that frequency. The magnitude of the mutual admittance determines the current amplitude, and the argument of the mutual admittance determines the phase shift between voltage and current. Using the same sampling frequency and data window length as the aforementioned fast Fourier transform, the theoretical injected current phasor at each dominant frequency band is successively converted into current waveform components in the time domain. All the converted current waveform components are then linearly superimposed on the time axis to generate a composite waveform that continuously varies with time, i.e., the theoretical injected current waveform.

[0037] 2.1.2 The theoretical injected current waveform is limited based on the maximum output current of the candidate commutator branch, and the peak value of the limited current waveform is taken as the maximum injectable current, as follows: Obtain the nameplate parameters of the candidate branch converter's underlying hardware controller and extract the corresponding maximum output current. This maximum output current is the transient overload current limit that the IGBT power module inside the converter can withstand in a short time (e.g., within 10ms). Using this maximum output current, the theoretical injected current waveform is limited. Specifically, the maximum output current is compared with the absolute value of the current at each discrete time point in the theoretical injected current waveform. When the absolute value of the theoretical current at a certain moment exceeds the maximum output current, the current value at that moment is truncated and limited to the maximum output current value; at moments not exceeding the limit, the original value is retained, thus obtaining the limited injected current waveform sequence.

[0038] In the limited-amplification injected current waveform sequence, the single scalar value with the largest value (i.e., the waveform peak value) is extracted as the maximum injectable current for the candidate commutation branch. This limiting process can avoid computational instructions exceeding the hardware tolerance limit at the physical level, thereby effectively preventing the risk of converter burnout due to overcurrent caused by forced commutation.

[0039] 2.2 Based on the maximum injectable current and grid disturbance parameters, the auxiliary converter branch is selected, including: 2.2.1 Calculate the node voltage change and branch power flow change caused by the commissioning of each candidate converter branch, as an indicator of grid disturbance, as follows: For each candidate converter branch, frequency domain analysis is performed on the corresponding theoretical injected current waveform to obtain the phase angle of the power frequency fundamental component; the phase angle is assigned to the maximum injectable current scalar corresponding to the candidate converter branch to construct the maximum injectable current phasor containing its own phase information; the amplitude of the maximum injectable current phasor is the maximum injectable current scalar, and the phase angle is the phase angle of the power frequency fundamental component.

[0040] Invert the aforementioned global node admittance matrix to obtain the global node impedance matrix; extract the main diagonal elements of the global node impedance matrix corresponding to each injection node number as the Thevenin equivalent input impedance for each injection node. Perform a complex multiplication operation between the Thevenin equivalent input impedance and the maximum injectable current phasor, and take the magnitude of the calculation result to deduce the magnitude of the node voltage change caused by the injection operation of the candidate commutation branch.

[0041] Next, based on the global node impedance matrix, a complex matrix of sensitivity coefficients reflecting the influence of node-injected current on the power flow of each branch is derived; specifically, for a certain associated branch (assuming the two end nodes are...),... and ) for nodes The power flow sensitivity coefficient of the injected current is equal to the first element in the global node impedance matrix. Line 1 The mutual impedance of the column and the first Line 1 The product of the difference in mutual impedance of the columns and the inherent admittance of the associated branch is used. From the complex matrix of sensitivity coefficients, the power flow sensitivity coefficients from the injection node of the current candidate converter branch to each surrounding associated branch are extracted. The maximum injectable current phasor is multiplied by each power flow sensitivity coefficient, and the magnitude of the calculation result is taken to obtain the set of transient current change amplitudes for each surrounding associated branch, which characterizes the power flow change. Considering that a single node current injection will cause the power flow of multiple branches in the local power grid to redistribute, for the convenience of unified quantitative evaluation, the largest change amplitude is extracted from the set of transient current change amplitudes and multiplied by the reference rated voltage corresponding to the branch to obtain the apparent power change amplitude, which is used as the most severe power flow disturbance index caused by the candidate converter branch after it is put into the grid, thus characterizing the power flow change of the branch.

[0042] 2.2.2 A weighted score is applied to the maximum injectable current and grid disturbance index, and the candidate converter branch with the best score is selected as the auxiliary converter branch, as follows: Considering the differences in numerical magnitude between current, voltage, and power flow, the maximum-minimum normalization method is adopted to linearly map the maximum injectable current, node voltage change, and branch power flow change to the dimensionless interval [0,1], respectively, to obtain the corresponding normalized index values.

[0043] Based on the three normalized indicators mentioned above, a weighted comprehensive score is calculated for each candidate converter branch, and the candidate converter branch with the highest comprehensive score is selected as the auxiliary converter branch for performing the actual injection operation; the formula for calculating the weighted comprehensive score is as follows: (4) Among them, the score is the overall rating of the candidate converter branch; The normalized maximum injectable current characterizes the arc extinguishing compensation capability; , These are the normalized node voltage change and the normalized branch power flow change, which together characterize the secondary impact on the power grid. , , These are the weight coefficients of the three items mentioned above, and they satisfy... For example, the maximum injectable current weight can be set. Weighting of node voltage changes Branch flow change weight .

[0044] This step calculates the injectable current of candidate converter branches based on grid frequency-varying mutual admittance and determines their actual injection capacity according to hardware limiting. It also combines a weighted score of voltage and power flow changes to select the best auxiliary converter branch. Under the premise of ensuring that the auxiliary converter branch has sufficient arc extinguishing compensation capacity, it can reduce the transient impact of injection intervention on local node voltage and associated power flow, reduce the risk of secondary disturbances and chain reactions in the grid caused by traditional forced converter, and achieve a balance between the auxiliary arc extinguishing effect and the grid safety margin.

[0045] Step 3: Extract the time-varying characteristics of the arc impedance during the target switch opening and closing process, and combine them with the equivalent circuit parameters of the auxiliary converter branch to calculate the resonant frequency offset caused by the time-varying characteristics of the arc impedance and generate a feedforward control sequence.

[0046] 3.1 Extract the time-varying characteristics of the arc impedance during the target switch opening and closing process, including: 3.1.1 Based on the steady-state operating current of the target switch before the fault, the arc time constant and arc heat dissipation power are matched, and the arc electrical quantities during the contact separation of the target switch are collected, as follows: After step 2 completes the optimization of the auxiliary converter branch, the system issues the trip command suspended in step 1, driving the mechanical operating mechanism of the target switch to perform the opening action, causing the contacts to separate and cut off the metallic conductive path. Since an electric arc is generated between the contacts during the high voltage and high current interruption process, and the arc is confirmed by the pre-evaluation in step 1 to be unable to extinguish naturally, after the contacts separate and the arc is ignited, the system injects a compensation current in subsequent steps to force the superimposed transient current to cross zero and extinguish the arc, thereby achieving fault safety isolation with active and passive coordination.

[0047] Within a preset time window before fault triggering (e.g., 2 to 5 power frequency cycles before the fault), three-phase steady-state current signals are collected using a substation measuring-level current transformer. The root mean square (RMS) value of the discrete current amplitudes sampled continuously within this time window is processed to obtain the effective value of the steady-state current, which is then used as the steady-state operating current. Since the arc energy during circuit breaker breaking essentially depends on the transient large current after a short circuit occurs, while the steady-state operating current before the fault only determines the initial thermal equilibrium temperature field before the contacts ignite, to accurately match the arc parameters, after fault triggering and before contact separation, the aforementioned broadband protection-level current transformer (or Rogowski coil) is switched to collect the transient current signal under short-circuit conditions, and the same RMS processing is performed to obtain the effective value of the short-circuit transient breaking current.

[0048] Based on the effective values ​​of the steady-state operating current and the short-circuit transient breaking current, a pre-built historical operating parameter database of the circuit breaker is queried to extract the arc time constant and arc heat dissipation power corresponding to the current operating condition. This database uses a three-dimensional mapped data table format and pre-stores the arc time constant and heat dissipation power calibration values ​​for this type of switch under different combinations of steady-state operating current ranges and different short-circuit transient breaking current ranges. These calibration values ​​are obtained by averaging the short-circuit breaking type test data at the time of equipment delivery. During online operation, the currently calculated effective values ​​of the steady-state operating current and the short-circuit transient breaking current are used as a joint index key to directly retrieve the corresponding parameter values ​​from the database. As an equivalent alternative implementation, a multivariate polynomial empirical function fitted based on factory test data can also be used for calculation: for example, the steady-state current, breaking current, and heat dissipation power can be fitted into a multivariate quadratic polynomial using the least squares method, and then the current steady-state current and breaking current can be substituted into the polynomial to obtain the corresponding parameters.

[0049] Since electric arcs are a high-frequency transient physical phenomenon, and conventional transformers are prone to magnetic saturation during this stage, a capacitive voltage divider (CVD) and a Rogowski coil are arranged on the target switch body to acquire transient arc voltage and arc current data during the separation of the target switch contacts at a high-frequency sampling rate of not less than 100kHz; these are the arc electrical quantities. It should be noted that during the separation of the target switch contacts, the high electric field will break down the insulating medium between the contacts, forming an arc. At this time, the circuit still maintains physical conductivity thanks to the arc plasma channel.

[0050] 3.1.2 Substituting the arc time constant, arc heat dissipation power, and arc electrical quantities into the preset energy balance differential equation, solving for the arc conductance and taking its reciprocal, yields the time-varying characteristics of the arc impedance, as follows: The preset energy balance differential equation is specifically the Mayr arc equation. In this embodiment, this equation is used to track the nonlinear thermodynamic balance process of Joule heat input and heat dissipation within the arc channel, and to solve for the arc conductance. The specific differential equation is as follows: (5) in, Arc conductivity, measured in Siemens units; is the arc time constant, which characterizes the physical inertia of energy loss within the arc channel, and is measured in seconds; This refers to the heat dissipation power of the electric arc, measured in watts. and These are the transient arc voltage (in volts) and transient arc current (in amperes) at both ends of the arc during contact separation, respectively.

[0051] The fourth-order Runge-Kutta method is used to discretize and numerically integrate the above differential equations. Specifically, the arc time constant, arc heat dissipation power, and transient arc voltage and current discretized at a fixed time step (matching the aforementioned high-frequency sampling rate, such as 10 microseconds) are used as algorithm inputs. Starting with the initial conductance value at the moment of arc ignition, four slope prediction values ​​are calculated sequentially within each time step and weighted averaged to iteratively obtain the conductance value at the next moment, thus obtaining the arc conductance that varies with time. Taking the reciprocal of the arc conductance yields the arc impedance value that varies with time, which characterizes the time-varying characteristics of the arc impedance.

[0052] 3.2 Based on the equivalent circuit parameters of the auxiliary converter branch, calculate the resonant frequency shift caused by the time-varying characteristics of the arc impedance and generate a feedforward control sequence, including: 3.2.1 Based on the time-varying characteristics of the equivalent inductance, equivalent capacitance, and arc impedance of the auxiliary commutator branch, the resonant frequency of the equivalent network is determined as follows: When the auxiliary converter branch injects transient current into the target switching node, the distributed parameters of its internal filtering devices (such as LCL filters) and connecting lines can be abstracted into a lumped parameter model, that is, the entire branch is equivalent to a combination of a single inductor and capacitor. Therefore, by directly reading the hardware nameplate parameters of the auxiliary converter branch's underlying controller or the factory-preset filter design parameters, the algebraic sum of the inverter-side inductance and the grid-side inductance in the LCL filter can be taken as the equivalent inductance of the auxiliary converter branch, and its internal parallel filter capacitor can be directly taken as the equivalent capacitance; the equivalent inductance and equivalent capacitance together constitute the equivalent circuit parameters of the auxiliary converter branch.

[0053] Considering the physical distribution network lines typically existing between the target switch and the auxiliary converter branch, based on the node impedance matrix obtained by inverting the power frequency static equivalent admittance matrix in step 2.1.1, the mutual impedance elements between the target switch node and the candidate converter branch node are extracted as the equivalent transfer impedance between the two nodes; the real part of this equivalent transfer impedance is extracted as the equivalent resistance of the intermediate line; the time-varying arc impedance value of the target switch (i.e., the time-varying characteristic of the arc impedance) is connected in series with the equivalent resistance of the intermediate line to form a generalized equivalent resistance. A series equivalent network model composed of this generalized equivalent resistance and the equivalent circuit parameters of the auxiliary converter branch is constructed, and the resonant frequency of the equivalent network is calculated using the following formula: (6) in, The equivalent network resonant frequency is expressed in Hertz (Hz). The equivalent inductance for the auxiliary commutation branch; The equivalent capacitance for the auxiliary commutation branch; The time-varying characteristics of the arc impedance; The intermediate line is the equivalent resistance in ohms. When the target switch and the auxiliary converter branch are connected in parallel on the same bus, the equivalent resistance of the intermediate line can be approximated as 0.

[0054] It should be noted that after the contacts separate, the arc impedance increases sharply with the increase of the arc length. When the expression within the square root of the above equivalent network resonant frequency calculation formula is less than or equal to zero, it indicates that the equivalent network has entered an overdamped or critically damped state. At this time, the high-frequency transient resonance will naturally decay. In order to ensure the stable operation of the control program, the equivalent network resonant frequency is forcibly set to zero.

[0055] like Figure 2 As shown in the figure, the time-varying characteristics of the arc impedance and the evolution curve of the equivalent network resonant frequency are given. As can be seen from the figure, the arc impedance obtained by tracking based on the Mayr equation exhibits a nonlinear time-varying characteristic that rises sharply in the later stage of contact separation. Affected by the drastic change in impedance, the originally fixed equivalent network resonant frequency dynamically shifts and gradually decreases with time.

[0056] 3.2.2 The phase compensation angle is obtained by time integration using the difference between the equivalent network resonant frequency and the power grid frequency as the resonant frequency offset, as detailed below: Since the time-varying characteristics of the arc impedance change with time, the equivalent network resonant frequency is also a dynamic value that changes with time. Subtracting the fixed power grid frequency (e.g., 50Hz) from this dynamic equivalent network resonant frequency yields the resonant frequency offset at the current moment. Integrating this frequency offset over time gives the phase compensation angle, calculated using the following formula: (7) in, For a moment The phase compensation angle, in radians; The moment the target switch contacts separate, in seconds; For the intermediate integration variable used for time integration; For integration variables The corresponding equivalent network resonant frequency, This is the inherent power frequency of the power grid.

[0057] 3.2.3 A transient compensation waveform is generated based on the phase compensation angle and used as a feedforward control sequence, as follows: Based on the space vector pulse width modulation (SVPWM) modulation module at the bottom of the converter, the original fixed reference electrical angle generated by the converter based on the power grid frequency is algebraically superimposed with the aforementioned phase compensation angle to obtain the corrected target vector rotation electrical angle. Subsequently, based on the corrected target vector rotation electrical angle, the action time (i.e., duty cycle) of each basic voltage vector of SVPWM is recalculated, thereby generating a transient compensation waveform that adapts to the frequency offset, which serves as the feedforward control sequence.

[0058] This step extracts the time-varying characteristics of the arc impedance based on the preset energy balance differential equation, calculates the resonant frequency offset by combining the equivalent circuit parameters of the auxiliary commutation branch, and then integrates to generate a feedforward control sequence. By dynamically tracking the nonlinear change trajectory of the arc impedance, the arc impedance is used as a feedforward compensation variable to perform real-time phase correction on the injected signal. This helps to overcome the commutation failure problem caused by the mismatch between the traditional fixed frequency compensation signal and the actual transient physical model, and improves the success rate and reliability of active arc extinguishing.

[0059] Step 4: Based on the feedforward control sequence, control the auxiliary converter branch to inject compensation current to extinguish the arc, and update the operation sequence according to the deviation between the actual grid topology and the target topology after arc extinguishing.

[0060] 4.1 Arc extinguishing is achieved by injecting compensation current into the auxiliary converter branch based on the feedforward control sequence, including: 4.1.1 Based on the feedforward control sequence, control the initial compensation current output of the auxiliary converter branch as follows: The underlying controller of the auxiliary converter branch receives the feedforward control sequence and drives its internal insulated gate bipolar transistor (IGBT) module to perform pulse width modulation, converting the stable voltage on the DC side of the converter into a duty cycle-varying AC high-frequency pulse voltage. After the pulse voltage is smoothed and filtered by the converter's local LCL filter to remove high-frequency switching harmonics, it outputs an initial compensation current that varies continuously with time.

[0061] 4.1.2 The initial compensation current and the branch transient current are superimposed, the fundamental error current component of the superimposed current is extracted, and converted into a frequency compensation term to correct the feedforward control sequence, as follows: According to Kirchhoff's current law, after the initial compensation current is injected into the node where the target switch is located, it is superimposed with the low-frequency, high-amplitude transient fault current in the target switch branch to obtain the superimposed current; the discrete time series of the superimposed current is collected in real time through the aforementioned wideband protection-level current transformer.

[0062] Discrete Fast Fourier Transform (FFT) is used to perform spectral analysis on the discrete-time series of the superimposed current. Specifically, the discrete-time series and a Hamming window function are used as input, and a butterfly operation is performed to complete the time-frequency conversion, extracting the complex result corresponding to the power grid frequency band. The magnitude and argument of this complex result constitute the measured fundamental phasor. A reference zero vector with zero amplitude (i.e., a complex number with both real and imaginary parts zero) is used as the desired fundamental phasor under the ideal arc-extinguishing state. The complex difference between the measured fundamental phasor and the desired fundamental phasor is used to obtain the fundamental error current component characterizing the actual compensation deviation.

[0063] The fundamental frequency error current component is projected onto a synchronous rotating coordinate system to obtain the d-axis active power error component and the q-axis reactive power error component, which are then fed into two digital discrete proportional-integral (PI) regulators at the bottom layer of the converter control. Within each control sampling period (e.g., 100 microseconds), these two PI regulators amplify the current error using a proportional coefficient and accumulate historical errors using an integral coefficient, respectively, and perform step-by-step recursive calculations based on the discrete difference equation, outputting compensation voltage commands for the d-axis and q-axis. In the synchronous rotating coordinate system, the dynamic adjustment of the q-axis voltage component determines the phase deflection of the synthesized voltage vector. This phase deflection rate is equivalent to the dynamic correction of the grid fundamental frequency. Therefore, the aforementioned compensation voltage commands for the d-axis and q-axis constitute the frequency compensation term used to correct the feedforward control sequence.

[0064] Using the space vector pulse width modulation (SVPWM) module of the converter, combined with the target vector rotation electrical angle calculated in step 3.2.3, the compensation voltage commands of the d-axis and q-axis are subjected to inverse Parker transformation and vector synthesis to generate a three-phase pulse duty cycle signal, which serves as the corrected feedforward control sequence.

[0065] 4.1.3 Inject compensation current based on the modified feedforward control sequence until the preset interruption confirmation condition is met, as follows: Based on the corrected feedforward control sequence, the auxiliary commutator branch is re-controlled to inject compensation current. After the compensation current forces the superimposed current into the alternating zero-crossing stage, the system continuously checks whether the preset switching confirmation condition is met, using a control period with the same step size as the aforementioned PI regulator control sampling period. The specific judgment process is as follows: Within a time window when the superimposed current amplitude (i.e., the transient absolute value) is less than a preset zero-zone threshold, the arc energy integral within the zero zone is calculated. The preset zero-zone threshold is typically set to an extremely small current range, for example, ±5A to ±10A. Within this time window, the transient arc voltage, collected in real-time by the capacitive voltage divider arranged on the target switch body, is simultaneously extracted. The product of the transient arc voltage and the superimposed current is discretized and numerically integrated using the trapezoidal rule to obtain the arc energy integral within the zero zone. The formula for calculating the arc energy integral is as follows: (7) in, This is the integral of the electric arc energy, measured in joules (J). The time window during which the superimposed current amplitude remains within the preset zero threshold; This represents the transient arc voltage across the arc during contact separation, measured in volts. The current after superposition is expressed in amperes (A). For discrete sampling point numbers, This is the sampling time step.

[0066] Simultaneously, the zero-crossing rate of change of the superimposed current and the transient recovery voltage across the target switch contacts are monitored. Specifically, under the action of the compensation current, the superimposed current is forced to cross zero. The current values ​​of adjacent discrete sampling points before and after the zero-crossing point are extracted, and the local derivative is calculated by forward difference to obtain the zero-crossing rate of change of the superimposed current. Using the capacitive voltage divider, the transient recovery voltage of the target switch contact gap during the recovery and voltage build-up process of the insulating medium is continuously monitored. This transient recovery voltage refers to the recovery voltage borne by the two ends of the contacts after the superimposed current crosses zero and extinguishes the arc, and the arc plasma channel is physically disconnected.

[0067] When the arc energy integral, zero-crossing rate of change, and transient recovery voltage all meet the preset safety threshold conditions, the interruption confirmation condition is determined to be met. Specifically, the arc energy integral is required to be less than the first safety threshold (e.g., 5J) to prevent excessive absorption of residual Joule heat in the plasma channel before zero crossing, leading to failure of microscopic thermal attenuation and localized ablation of the contact surface; the zero-crossing rate of change is required to be less than the preset rate of change threshold mentioned in step 1 to prevent thermal breakdown caused by residual heat and ionized electrons after zero crossing; the peak value and rate of rise of the transient recovery voltage are required to not exceed the insulation withstand limit given on the equipment nameplate to prevent dielectric breakdown during the inter-electrode recovery period. When all three electrical indicators meet the standards, the interruption confirmation condition is determined to be met, that is, the arc is confirmed to have been safely extinguished.

[0068] 4.2 Update the operation sequence based on the deviation between the actual grid topology and the target topology after arc extinguishing, including: 4.2.1 Perform a topology comparison between the actual power grid topology after arc extinguishing and the target topology to identify failed branches, as follows: By using telemetry data from a miniature synchronous phasor measurement unit (μPMU) and analyzing the status words (i.e., 0 / 1 binary codes representing opening and closing states) of IEC 61850 standard GOOSE messages fed back by the intelligent terminals of each circuit breaker, the actual physical connectivity status of each branch in the entire network is verified in real time to obtain the actual topology of the power grid after arc extinguishing. Based on the operation sequence including the target switch obtained in step 1, the switch change instructions are substituted into the basic network topology before the fault, and the connectivity status of the corresponding branches is modified to obtain the expected network structure after executing the operation sequence under ideal fault-free impact conditions, which is used as the target topology.

[0069] By comparing the node connectivity status of the actual power grid topology with that of the target topology, branches with discrepancies are marked as failed branches. A failed branch is a branch that, after combining remote signaling change time stamps and multiple status confirmations and ruling out abnormal causes in the secondary control circuit such as communication interruption and protection failure, suffers physical structural damage due to the electrodynamic or thermal shock of a short-circuit fault (e.g., welding of moving and stationary contacts, jamming of transmission mechanism, or irreversible breakdown of inter-electrode insulation), thereby losing its ability to transmit power and operate switches.

[0070] 4.2.2 Remove the operation steps that have associated failed branches from the operation sequence, and reorder the remaining operation steps according to the preset network constraints to generate an updated operation sequence, as follows: The instruction set of the operation sequence is traversed, and all opening and closing instructions that target the switch nodes at both ends of the failed branch are identified and eliminated to obtain the remaining operation steps. Subsequently, based on the preset network constraints and the actual power grid topology, the execution sequence of the remaining operation steps is re-verified and sorted. The preset network constraints include radial operation constraints of the distribution network and voltage deviation limit constraints of key nodes. Among them, the radial operation constraints require that the distribution network maintains a tree topology in each intermediate transition stage and the final state of topology reconstruction to avoid short-circuit current overshooting due to instantaneous electromagnetic loop closure. The voltage deviation limit constraints of key nodes require that the voltage of each key node (i.e., load nodes connected to primary and secondary loads and hub nodes where multiple branches intersect) be maintained within the allowable deviation range of the rated voltage, for example, ±7%, to prevent low voltage collapse or overvoltage problems. To ensure the safety of the vulnerable network structure after a fault, the priority of the radial operation constraints is set higher than that of the voltage deviation limit constraints of key nodes.

[0071] During the reordering process, a heuristic depth-first search algorithm with backtracking is used based on the actual power grid topology to virtually execute and optimize the remaining operation steps in the quasi-steady-state simulation system. Specifically, instructions are extracted and virtually executed sequentially according to preset heuristic safety rules. These heuristic safety rules are pre-defined instruction selection priority judgment logics, such as stipulating that the extraction priority of tripping instructions in the candidate instruction set is higher than that of closing instructions, and that among similar closing instructions, instructions whose power supply path points to critical loads (such as hospitals or important industrial nodes) are extracted first. After each virtual execution of an operation step, a power flow calculation program is called for safety verification: if a closed loop appears in the distribution network connectivity diagram after virtual execution, it is determined that the radial operation constraint is violated; if the steady-state power flow calculation results show that the voltage deviation of any node exceeds ±7% of the rated voltage, it is determined that the critical node voltage deviation limit constraint is violated.

[0072] For any operation step that triggers any of the above constraint conflicts during virtual execution, a backtracking mechanism is immediately activated. This means that the virtual execution state of the conflicting operation step is revoked in the simulation environment, and its execution priority is adjusted by timing swapping or queuing backwards. For example, if the "close first, then open" uninterrupted power transfer instruction for a certain tie switch in the original operation sequence triggers a closing loop conflict (i.e., violating the radial operation constraint) during virtual execution, the system will force a timing swap based on the aforementioned constraint priority, adjusting it to a "open first, then close" safe operation sequence. This means that a temporary local load loss is sacrificed to avoid short-circuit circulating current. Then, the next candidate instruction is extracted and re-verified.

[0073] By continuously performing deep searches and iterative adjustments to the timing sequence, a safe execution sequence that does not trigger network constraint conflicts throughout the entire process is selected as the updated operation sequence. If the algorithm still cannot find a safe execution sequence that simultaneously satisfies the above network constraints after traversing all feasible branches, the optimization process is terminated, a safety alarm is output, automatic control permissions are locked, and manual scheduling takes over.

[0074] This step extracts the fundamental frequency error correction feedforward control sequence, confirms arc extinguishing under the premise of meeting multiple electrical safety thresholds, and updates the operation sequence based on topology deviation. Combining micro-level closed-loop verification of electrical quantities with macro-level adaptive correction of network topology helps reduce the risk of compensation failure caused by field interference, avoids subsequent operational errors caused by irreversible equipment damage, and improves the consistency of the entire fault adaptive handling process and the efficiency of power restoration in non-faulty areas.

[0075] Example 2, Figure 3 A topology-reconfiguration-based adaptive fault handling system for implementing any of the methods described herein is presented, comprising the following modules: The zero-crossing evaluation module is used to acquire the operation sequence of the target switch and evaluate the natural zero-crossing capability of the target switch based on the operating parameters of the target switch and the transient current of the branch. The branch selection module is used to calculate the maximum injectable current of candidate converter branches based on grid admittance in response to the evaluation results, and select auxiliary converter branches in combination with grid disturbance indicators. The sequence generation module is used to extract the time-varying characteristics of the arc impedance during the target switch opening and closing process, and, in combination with the equivalent circuit parameters of the auxiliary commutation branch, calculate the resonant frequency offset caused by the time-varying characteristics of the arc impedance and generate a feedforward control sequence. The execution update module is used to control the injection of compensation current into the auxiliary converter branch to extinguish the arc based on the feedforward control sequence, and to update the operation sequence according to the deviation between the actual grid topology and the target topology after arc extinguishing.

[0076] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0077] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0078] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0079] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0080] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A power grid fault adaptive handling method based on topology reconfiguration, characterized in that, Includes the following steps: The operation sequence of the target switch is obtained, and the natural zero-crossing capability of the target switch is evaluated based on the operating parameters of the target switch and the transient current of the branch. In response to the evaluation results, the maximum injectable current of the candidate converter branch is calculated based on the grid admittance, and the auxiliary converter branch is selected in combination with the grid disturbance index. Extract the time-varying characteristics of the arc impedance during the target switch opening and closing process, combine them with the equivalent circuit parameters of the auxiliary converter branch, calculate the resonant frequency offset caused by the time-varying characteristics of the arc impedance, and generate a feedforward control sequence. Based on the feedforward control sequence, the auxiliary converter branch is controlled to inject compensation current to extinguish the arc, and the operation sequence is updated according to the deviation between the actual grid topology and the target topology after arc extinguishing.

2. The method according to claim 1, characterized in that, The evaluation of the natural zero-crossing capability of the target switch includes: Extrapolate based on the rate of change of the transient current of the branch, and combine with the tripping time in the action parameters to obtain the expected time of zero current crossing and the rate of change of zero current crossing. When the expected current crosses zero at a time that satisfies the preset arcing time window constraint and the rate of change of zero is less than the preset rate of change threshold, the target switch is determined to have natural zero-crossing capability.

3. The method according to claim 1, characterized in that, The calculation of the maximum injectable current of the candidate converter branch based on grid admittance includes: Obtain the mutual admittance and transient node voltage difference between the candidate converter branch and the target switching node, and calculate the theoretical injected current waveform based on the frequency-varying characteristics of mutual admittance and the phase shift of the transient node voltage difference; The theoretical injected current waveform is limited based on the maximum output current of the candidate commutator branch, and the peak value of the current waveform after limiting is taken as the maximum injectable current.

4. The method according to claim 1, characterized in that, The selection of auxiliary converter branches based on grid disturbance indicators includes: Calculate the node voltage change and branch power flow change caused by the commissioning of each candidate converter branch, and use them as power grid disturbance indicators. The maximum injectable current and the grid disturbance index are weighted and scored, and the candidate converter branch with the best score is selected as the auxiliary converter branch.

5. The method according to claim 1, characterized in that, The extraction of the time-varying characteristics of the arc impedance during the target switch switching process includes: Based on the steady-state operating current of the target switch before the fault, the arc time constant and arc heat dissipation power are matched, and the arc electrical quantities during the contact separation of the target switch are collected. Substituting the arc time constant, arc heat dissipation power, and arc electrical quantities into the preset energy balance differential equation, the arc conductance is solved and its reciprocal is taken to obtain the time-varying characteristics of the arc impedance.

6. The method according to claim 1, characterized in that, The generation of the feedforward control sequence includes: Based on the equivalent inductance and equivalent capacitance of the auxiliary converter branch, and the time-varying characteristics of the arc impedance, the equivalent network resonant frequency is determined. The difference between the equivalent network resonant frequency and the power grid frequency is taken as the resonant frequency offset, and the phase compensation angle is obtained by time integration. A transient compensation waveform is generated based on the phase compensation angle, which serves as the feedforward control sequence.

7. The method according to claim 1, characterized in that, The step of controlling the injection of compensation current into the auxiliary converter branch for arc extinguishing based on the feedforward control sequence includes: Based on the feedforward control sequence, the auxiliary converter branch is controlled to output the initial compensation current. The initial compensation current and the branch transient current are superimposed, the fundamental error current component of the superimposed current is extracted, and converted into a frequency compensation term to correct the feedforward control sequence. The compensation current is injected based on the modified feedforward control sequence until the preset interruption confirmation condition is met.

8. The method according to claim 7, characterized in that, The conditions for satisfying the preset on / off confirmation include: The arc energy integral is calculated within the time window during which the superimposed current amplitude is less than the preset zero zone threshold. Monitor the zero-crossing rate of change of the superimposed current and the transient recovery voltage across the target switch contacts; When the arc energy integral, zero-crossing rate of change, and transient recovery voltage all meet the preset safety threshold conditions, it is determined that the interruption confirmation condition is met.

9. The method according to claim 1, characterized in that, The step of updating the operation sequence based on the deviation between the actual topology of the power grid after arc extinguishing and the target topology includes: The actual topology of the power grid after arc extinguishing is compared with the target topology to identify failed branches; The operation steps associated with the failed branch are removed from the operation sequence, and the remaining operation steps are reordered according to the preset network constraints to generate an updated operation sequence.

10. A topology-reconfiguration-based adaptive fault handling system for implementing the method as described in any one of claims 1 to 9, characterized in that, Includes the following modules: The zero-crossing evaluation module is used to acquire the operation sequence of the target switch and evaluate the natural zero-crossing capability of the target switch based on the operating parameters of the target switch and the transient current of the branch. The branch selection module is used to calculate the maximum injectable current of candidate converter branches based on grid admittance in response to the evaluation results, and select auxiliary converter branches in combination with grid disturbance indicators. The sequence generation module is used to extract the time-varying characteristics of the arc impedance during the target switch opening and closing process, and, in combination with the equivalent circuit parameters of the auxiliary commutation branch, calculate the resonant frequency offset caused by the time-varying characteristics of the arc impedance and generate a feedforward control sequence. The execution update module is used to control the injection of compensation current into the auxiliary converter branch to extinguish the arc based on the feedforward control sequence, and to update the operation sequence according to the deviation between the actual grid topology and the target topology after arc extinguishing.

Images