New energy power grid line fault reclosing optimization method and device

Abstract

The application discloses a new energy power grid line fault reclosing optimization method and device, relates to the technical field of electric power control, and comprises the following steps: collecting operation electrical quantities of a new energy grid-connected system, and extracting fault characteristic data; analyzing the characteristic data through a fault analysis model to obtain fault discriminant quantities and generate criteria; establishing a reclosing optimization model based on the criteria and solving the reclosing optimization model to obtain a reclosing control strategy; and executing the reclosing control strategy to implement reclosing operation on a power grid line. The application aims at solving the problem of high reclosing failure rate caused by fixed strategy and time delay sensitivity in traditional methods.

Description

Technical Field

[0001] This invention relates to the field of power control technology, and more specifically, to a method and apparatus for optimizing reclosing faults in new energy power grid lines. Background Technology

[0002] Traditional power grids are mainly composed of synchronous generators, whose transient characteristics, short-circuit current, and voltage support capabilities provide reliable data for relay protection and reclosing. However, when large-scale renewable energy sources are connected through power electronic inverters, the stable current output characteristics, insufficient short-circuit current, and response delay significantly increase the risks of inaccurate relay protection measurements, inaccurate fault diagnosis, and reclosing failures.

[0003] In existing technologies, reclosing often employs a fixed time delay strategy, meaning that the circuit breaker waits a preset time after the fault is cleared before closing. This method ignores the dynamic changes in fault current decay, voltage recovery, and system synchronization. Closing the circuit prematurely before the residual current is completely eliminated or the voltage and current have recovered to permissible levels can easily generate inrush currents, leading to relay protection malfunctions, equipment damage, or reclosing failures. Conversely, excessively long waiting times can affect the reliability of the power grid and the sustainability of renewable energy output.

[0004] Furthermore, traditional methods often rely on a single measurement (such as voltage amplitude or current threshold) as a criterion, lacking comprehensive consideration of multi-dimensional characteristics such as residual current, transient current decay, phase angle difference, and frequency deviation. This single criterion is prone to misjudgment and cannot fully reflect the complex dynamic process after a fault in the renewable energy grid. Even though some studies have proposed using optimization models to assist in reclosing decisions, they have failed to fully consider uncertainties such as the execution delay of relay protection and control devices, measurement jitter, and communication link delay, resulting in a high risk of failure for the optimization results in actual operation. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an optimization method for reclosing faults in new energy power grid lines. By quantifying fault characteristic data into multiple criteria and incorporating time delay uncertainty compensation into the optimization model, the reliability of the reclosing strategy under dynamic operating conditions and delay disturbances is ensured, thus solving the problem of high failure rates caused by fixed strategies and time delay sensitivity in traditional methods.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The optimization method for reclosing faults in new energy power grid lines includes the following steps: collecting the operating electrical quantities of the new energy grid-connected system and extracting fault characteristic data; analyzing the characteristic data through a fault analysis model to obtain fault discrimination quantities and generate criteria; establishing a reclosing optimization model based on the criteria and solving it to obtain the reclosing control strategy; and executing the reclosing control strategy to perform reclosing operations on the power grid lines.

[0008] In a preferred embodiment, the specific steps for collecting the operating electrical quantities of the new energy grid-connected system and extracting fault characteristic data are as follows: collecting actual measurements of the three-phase voltage and current on the grid side and preprocessing them to obtain standard electrical quantities; identifying abnormal change ranges of the standard electrical quantities; and calculating the transient current amplitude, voltage drop amplitude, and residual current decay based on the abnormal change ranges as fault characteristic data.

[0009] In a preferred embodiment, the specific steps for constructing the fault analysis model are as follows: constructing a training sample set based on fault feature data and historical labels, and performing fault type labeling and dividing the training set and validation set; using the SMOTE oversampling method to balance the samples; using the XGBoost algorithm to train the model on the training set to obtain the corresponding model parameters; evaluating and solidifying the model parameters on the validation set to obtain the fault analysis model.

[0010] In a preferred embodiment, the specific steps for parsing feature data through a fault analysis model to obtain fault discrimination quantity and generate criterion are as follows: inputting fault feature data into the fault analysis model and outputting a fault type probability distribution; extracting the fault type corresponding to the highest probability according to the probability distribution and quantifying it into a fault discrimination quantity; generating criterion based on the fault discrimination quantity and a preset fault classification rule.

[0011] In a preferred embodiment, the objective function of the reclosing optimization model includes minimizing the closing inrush current and maximizing the reclosing success rate; the constraints are obtained by criteria, including the residual current of the fault phase being below a threshold and lasting for a predetermined duration, the line voltage amplitude recovering to the allowable bandwidth, the phase angle difference and frequency deviation being within the synchronization tolerance, the transient current decaying to a set ratio, and the zero-sequence current not exceeding the limit.

[0012] In a preferred embodiment, the reclosing optimization model is solved using a mixed integer quadratic programming algorithm.

[0013] In a preferred embodiment, the method of solving the reclosing optimization model using a mixed integer quadratic programming algorithm further includes a delay uncertainty compensation step, which specifically involves: estimating the uncertainty interval of the closing execution time based on the mechanical action delay of the circuit breaker, the jitter of the measurement timestamp, and the communication link delay; and constructing a robust constraint set by applying constraint contraction to the constraints of phase angle difference, frequency deviation, and closing inrush current based on the uncertainty interval.

[0014] This invention provides a reclosing optimization device for new energy power grid lines, comprising: a feature extraction module for collecting operating electrical quantities of the new energy grid-connected system and extracting fault feature data; a criterion generation module for parsing the feature data through a fault analysis model to obtain fault discrimination quantities and generate criteria; a strategy generation module for establishing a reclosing optimization model based on the criteria and solving it to obtain a reclosing control strategy; and a reclosing module for executing the reclosing control strategy and performing reclosing operations on the power grid lines.

[0015] A new energy power grid line fault reclosing optimization device includes a memory and a processor: the memory is used to store a program; the processor is used to execute the program to implement the various steps of the new energy power grid line fault reclosing optimization method.

[0016] A readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the various steps of the optimized method for reclosing faults in new energy power grid lines.

[0017] The technical effects and advantages of the new energy power grid line fault reclosing optimization method of the present invention are as follows:

[0018] This invention collects and quantifies multi-dimensional fault characteristic data such as voltage drop amplitude, transient current amplitude, and residual current decay rate, constructs a fault analysis model to generate multiple criteria, and transforms these criteria into constraints in the optimization model, achieving a dynamic mapping from characteristic quantities to control conditions. In the optimization stage, the objective functions are minimizing the closing inrush current and maximizing the reclosing success rate. A mixed-integer quadratic programming method is used to generate the strategy, and a time delay uncertainty compensation mechanism is further introduced to ensure that the closing control remains robust under interference such as circuit breaker mechanical action delay, measurement jitter, and communication delay. Through the above methods, this invention can significantly improve the measurement accuracy and reclosing success rate of relay protection under new energy grid connection conditions, avoiding reclosing failures caused by fixed strategies, single criteria, and time delay sensitivity in traditional methods, and effectively enhancing the safety and reliability of the power grid under complex dynamic conditions. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the optimized reclosing method for new energy power grid lines in an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of a circuit for improving the measurement accuracy of photovoltaic grid fault relay protection provided in an embodiment of the present invention;

[0021] Figure 3 A schematic diagram of a photovoltaic grid fault leakage current elimination device provided in an embodiment of the present invention;

[0022] Figure 4 A schematic diagram of a photovoltaic grid fault transient current generation circuit provided in an embodiment of the present invention;

[0023] Figure 5 A block diagram illustrating the composition of the new energy power grid line fault reclosing optimization device provided in this embodiment of the invention;

[0024] Figure 6 A structural block diagram of an exemplary electronic device provided for implementing embodiments of the present disclosure;

[0025] Figure 7 This is a schematic diagram of an exemplary storage medium that can be used to implement embodiments of the present disclosure, as provided in the embodiments of the present invention. Detailed Implementation

[0026] 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.

[0027] Example 1, Figure 1 The present invention provides an optimized method for reclosing faults in new energy power grid lines, comprising the following steps:

[0028] S1 collects the operating electrical quantities of the new energy grid-connected system and extracts fault characteristic data;

[0029] S2, analyze the feature data through the fault analysis model to obtain the fault discrimination quantity and generate the judgment criterion;

[0030] S3, Based on the criteria, establish a reclosing optimization model and solve it to obtain the reclosing control strategy;

[0031] S4 executes the reclosing control strategy and performs reclosing operations on the power grid lines.

[0032] This embodiment introduces a circuit to improve the accuracy of photovoltaic grid fault relay protection measurements. This allows for rapid disconnection of photovoltaic power input at the moment of a grid fault, avoiding interference from inverter delay control on the collected data and ensuring the authenticity and timeliness of fault voltage and current characteristics. By preprocessing the collected three-phase voltage and current signals, identifying abnormal intervals, and calculating characteristic quantities, key fault feature data, including transient current amplitude, voltage drop amplitude, and residual current decay rate, can be extracted. Compared with traditional measurement methods that rely on delayed actions, this method can more accurately reflect the dynamic characteristics of the initial stage of a fault. Based on this, the residual current decay rate parameter is obtained by fitting an exponential decay model, and the effectiveness of fault elimination is quantified by combining threshold conditions, forming a feature quantity that can be directly applied to criterion generation and optimization constraints. In summary, this embodiment achieves rapid extraction and quantitative characterization of fault features, ensuring the accuracy and reliability of the input data for subsequent criterion generation and optimization model construction, laying the foundation for improving the timeliness of relay protection actions and the effectiveness of reclosing strategies.

[0033] S1 collects the operating electrical quantities of the new energy grid-connected system and extracts fault characteristic data.

[0034] It should be noted that in this embodiment, the new energy grid-connected system is a photovoltaic power generation grid-connected system. Its DC side is connected to the grid via a photovoltaic branch composed of photovoltaic modules, energy storage capacitors, and thyristors, while its AC side is connected to the grid via a three-phase reactor. To ensure the authenticity and timeliness of the collected data, this embodiment introduces a circuit to improve the accuracy of photovoltaic grid fault relay protection measurements, such as... Figure 2 As shown, the system includes several photovoltaic modules, a photovoltaic inverter, a three-phase reactor 1, a trigger transformer 2, an isolation DC chopper step-down module 3, an energy storage capacitor 4, and thyristors 5. This circuit can cut off the photovoltaic power input instantly when a grid fault occurs, preventing the inverter from continuously supplying power. This avoids measurement errors caused by inverter delay control and ensures that the collected voltage and current quantities accurately reflect the grid state at the time of the fault. The operating electrical quantities include three-phase voltage and current signals from the grid side. These signals are collected by voltage transformers (PTs) and current transformers (CTs), and then filtered and isolated before being input to the data acquisition module.

[0035] In this embodiment, the specific steps for collecting the operating electrical quantities of the new energy grid-connected system and extracting fault characteristic data are as follows:

[0036] S11, Collect and preprocess the actual measurements of three-phase voltage and current on the power grid side to obtain standard electrical quantities, specifically:

[0037] At the grid connection point, the collected three-phase voltage and current signals are filtered by a bandpass filter to remove high-frequency noise, and then normalized to obtain standard electrical quantities. Standard electrical quantities are expressed as:

[0038]

[0039] in , Here, represents the system reference voltage and current; a, b, and c represent phase A, phase B, and phase C, respectively. This processing ensures the comparability of measured values ​​under different operating conditions.

[0040] S12 identifies abnormal variation ranges in standard electrical quantities.

[0041] In this embodiment, abnormal change intervals are identified by calculating the first-order difference of standard electrical quantities.

[0042] The specific calculation formula for the first-order differential of standard electrical quantities is as follows:

[0043]

[0044] in, This is the first-order difference value of the voltage; This represents the first-order difference value of the current. When the difference amplitude exceeds a preset threshold, it is determined that the current has entered an abnormal change range.

[0045] It should be noted that the threshold is selected based on the natural fluctuation amplitude of the voltage and current waveforms under normal operating conditions of the power grid. By statistically analyzing a large amount of historical operating data, the mean and variance of the first-order differential under normal fluctuation conditions are obtained. Based on this, the threshold is typically set as the mean plus a certain number of standard deviations to ensure that misjudgments are not made when there are no faults. Simultaneously, the threshold sensitivity is verified using power grid fault test data, ensuring that the threshold can effectively distinguish fault abrupt changes while avoiding excessively low thresholds that could lead to noise interference. Preferably, the threshold can be taken as 3–5 times the mean absolute value of the differential during normal operation; the specific value can be adjusted according to the power grid type and sampling frequency.

[0046] S13, calculate the transient current amplitude, voltage drop amplitude and residual current decay based on the abnormal change range, as fault characteristic data.

[0047] The transient current amplitude is the peak value of the three-phase current at the beginning of the abnormal interval, and its calculation formula is as follows:

[0048]

[0049] in, These represent the instantaneous values ​​of the three-phase currents at the moment the fault occurs.

[0050] The voltage drop is the relative difference in voltage before and after the fault occurs, and its calculation formula is as follows:

[0051]

[0052] in, The reference voltage before the fault occurred. This is the root mean square voltage value at the time of the fault.

[0053] The residual current attenuation is a fitting parameter of the current attenuation curve within the abnormal range, and its calculation formula is as follows:

[0054]

[0055] in, The initial residual current amplitude, The current decay rate is obtained through fitting. It characterizes the decay rate of residual current and is used to quantify whether fault current has been effectively eliminated.

[0056] It should be noted that in this embodiment, the decay process of the residual current is assumed to follow an exponential decay law, that is, the current decays exponentially with time after the fault is cleared. This model can reflect the physical characteristics of the natural dissipation of current after a photovoltaic grid fault and is simple to calculate. The formula in the figure... The residual current amplitude at the moment of the fault occurrence can be obtained from the current peak value at the beginning of the abnormal interval; The current decay rate is obtained by fitting the actual current decay curve using the least squares method.

[0057] This step leverages the accuracy enhancement circuit for photovoltaic grid fault relay protection measurements to rapidly cut off photovoltaic power at the moment a grid fault occurs. This ensures that the collected voltage and current accurately reflect the transient process during a grid fault, avoiding spurious components introduced by inverter delays and thus guaranteeing the accuracy of subsequent judgment and optimization model construction.

[0058] S2 analyzes the feature data through the fault analysis model to obtain the fault discrimination quantity and generate the judgment criteria.

[0059] It should be noted that, in this embodiment, in order to ensure that the fault analysis model can obtain realistic and effective fault feature inputs, a photovoltaic grid fault latent current elimination device is introduced into the system, such as... Figure 3As shown, the device includes several photovoltaic modules, a photovoltaic inverter, solid-state voltage relays KA, KB, and KC, feedback current control resistors RA, RB, and RC, isolation transformers TA, TB, and TC, rectifier bridges DA, DB, and DC, and coupling diode arrays D1 and D2. When a fault occurs in a phase of the power grid, this device can instantaneously block the residual current at the fault point, feeding the power of the faulted phase back to the DC side. This ensures that the input fault characteristic data is not disturbed by the residual current and accurately reflects the fault status.

[0060] In this embodiment, the specific steps for parsing feature data through a fault analysis model to obtain fault discrimination quantities and generate criteria are as follows:

[0061] S21, Input the fault feature data into the fault analysis model and output the fault type probability distribution; the fault analysis model is constructed using the XGBoost algorithm and outputs the probability distribution vector of each fault type, as shown in the following expression:

[0062]

[0063] in, This represents the probability of normal operating conditions. Indicates the probability of a single-phase ground fault. This indicates the probability of a two-phase short-circuit fault. Let represent the probability of a three-phase short-circuit fault, and satisfy . .

[0064] S22, extract the fault type corresponding to the highest probability according to the probability distribution and quantify it into a fault discrimination quantity; for example, when Discriminant This indicates a single-phase ground fault; if the maximum value is... ,but This indicates a fault-free state. The discriminant D can be a discrete integer code representing the fault category, or it can be a continuous probability value used as a weighted judgment criterion.

[0065] S23, Generate criteria based on fault discrimination parameters and preset fault classification rules. Fault classification rules include: if And the corresponding probability value ,in If a preset threshold is set (e.g., 0.8), a "fault confirmed" criterion is generated; if or Then, the "keep monitoring" criterion is generated.

[0066] Furthermore, after generating the "fault confirmed" criterion, the following criteria can be further subdivided and generated:

[0067] 1) Criterion for eliminating residual current: If the residual current attenuation rate is not lower than the preset minimum attenuation rate ,Right now The criterion for eliminating residual current generation is then established. The safety threshold representing the rate of current decay is obtained from historical fault data statistics.

[0068] 2) Voltage recovery criterion: If the line voltage amplitude is within the allowable bandwidth range, i.e. Then, a criterion for voltage recovery to normal is generated, in which... , These are the lower and upper limits of the allowable voltage, respectively;

[0069] 3) Phase angle synchronization criterion: If the voltage phase angle difference before reclosing meets the synchronization tolerance, i.e. Then, a voltage phase angle synchronization criterion is generated, where This is the maximum permissible phase angle deviation value;

[0070] 4) Frequency synchronization criterion: If the frequency deviation before reclosing meets the synchronization tolerance, i.e. Then, a frequency synchronization criterion is generated, where This represents the maximum permissible frequency deviation value.

[0071] 5) Transient current decay criterion: If the transient current amplitude has decayed to a certain percentage below the initial value, i.e. Then the transient current decay criterion is generated, where This is a proportionality coefficient (e.g., 0.1–0.2).

[0072] 6) Zero-sequence current criterion: If the zero-sequence current does not exceed the preset maximum allowable value, i.e. Then, a criterion for zero-sequence current not exceeding the limit is generated, where This is the upper limit threshold for zero-sequence current.

[0073] In this embodiment, the specific steps for constructing the fault analysis model are as follows:

[0074] A training sample set is constructed based on fault feature data and historical labels, and fault type labeling and training set and validation set are performed. The fault feature data includes transient current amplitude, voltage drop amplitude and residual current decay. The historical labels are derived from photovoltaic grid operation records and simulation results. Each sample is labeled with fault type according to "normal operation, single-phase grounding, two-phase short circuit, three-phase short circuit" and divided into training set and validation set in a ratio of 8:2.

[0075] The SMOTE oversampling method is used for sample balancing. By interpolating and synthesizing minority class samples in the feature space, the number of single-phase grounding and three-phase short-circuit class samples is increased, making the number of samples of each class nearly consistent and avoiding deviations during model training.

[0076] The XGBoost algorithm is used to train the model on the training set to obtain the corresponding model parameters; during the training process, the XGBoost algorithm is used to train the model on the training set to obtain the corresponding model parameters. Given an input vector, the output is the probability distribution of each type of fault. The loss function is the multi-class cross-entropy. Model parameters include the learning rate (default is 0.1), the maximum depth of the tree (default is 6), and the number of base learners (default is 500). These parameters can be adjusted based on the performance on the validation set.

[0077] The model parameters are evaluated and solidified on the validation set to obtain the fault analysis model. When the model's F1-score on the validation set is greater than 0.9, the trained model parameters are solidified to obtain the fault analysis model.

[0078] It should be noted that the model training in this embodiment was performed on a computing server configured with an Intel Xeon Gold 6338 processor, 32GB of DDR4 memory, and an NVIDIA Tesla V100 GPU. The software environment consisted of an Ubuntu 20.04 operating system, Python 3.9 and the scikit-learn 0.24 framework, combined with the XGBoost 1.6 open-source library. This environment configuration ensures parallel computing capabilities and efficiency in processing large-scale datasets during the training process.

[0079] This step utilizes the photovoltaic grid fault leakage current elimination device to cut off the leakage current of the faulty phase when a fault occurs, ensuring that the collected feature data does not contain redundant components. This allows the analytical model to output an accurate fault type probability distribution and generate a criterion accordingly, thus providing reliable input for the subsequent reclosing optimization model.

[0080] S3. Based on the criteria, establish a reclosing optimization model and solve it to obtain the reclosing control strategy.

[0081] In this embodiment, the optimization model uses the closing time as the basis. Select binary variables separately (These represent whether phases A / B / C participate) and binary variables for the closing mode. (Single-phase / three-phase) are used as decision variables; the constraints are obtained by quantifying the subdivision criteria generated by S2; the objective function is used to balance the closing impact and success probability under the premise of satisfying the safety criteria; the model form is mixed integer quadratic programming (MIQP).

[0082] It should be noted that in this embodiment, the system incorporates a photovoltaic grid fault transient current generation circuit, such as... Figure 4 As shown, the circuit includes several photovoltaic modules, a photovoltaic inverter, a permanent magnet synchronous motor 1, a pulse transformer 2, a DC step-down module 3, a farad capacitor 4, a thyristor 5, a motor excitation winding 6, a flexible drive shaft 7, a flywheel 8, and other components. This circuit generates a transient component equivalent to that of a synchronous generator at the moment of a fault, resulting in the transient characteristics obtained in S1 (such as the transient current amplitude). Attenuation rate It is observable, quantifiable, and reproducible; these characteristics directly contribute to the calculation of objectives and constraints in this step.

[0083] In this embodiment, the objective function of the reclosing optimization model includes minimizing the closing inrush current and maximizing the reclosing success rate; for ease of implementation and evaluation, the objective function can be expressed as:

[0084]

[0085] in To estimate the closing inrush current, a quadratic approximation can be used. (coefficient (Originated from offline calibration) The probability of successful reclosing can be estimated by mapping the probability output of S2 with historical samples through logistic regression. For the phase angle difference before overlap, For frequency difference, Due to voltage deviation, all three are within Prediction; weighting and .

[0086] The constraints are derived from the criteria, including the residual current of the faulted phase being below a threshold for a predetermined duration, the line voltage amplitude recovering to the allowable bandwidth, the phase angle difference and frequency deviation being within the synchronization tolerance, the transient current decaying to a set proportion, and the zero-sequence current not exceeding the limit. Their mathematical expressions are as follows:

[0087] 1) Residual current constraint:

[0088]

[0089] in The residual current threshold, For the duration of continuously meeting the threshold, To maintain the lower limit of time;

[0090] 2) Voltage recovery constraint:

[0091]

[0092] in , , This refers to the allowable deviation coefficient;

[0093] 3) Phase angle / frequency synchronization constraints:

[0094]

[0095] 4) Transient current decay constraint:

[0096]

[0097] 5) Zero-sequence current constraint:

[0098]

[0099] in .

[0100] To be compatible with discrete decision-making, it also includes logical constraints for phase / mode selection (such as...). and Consistency constraints).

[0101] The reclosing optimization model is solved using a mixed integer quadratic programming algorithm.

[0102] It should be noted that this model includes a quadratic objective term and Discrete variables, belonging to the MIQP problem in the prior art; a solver can be used to obtain the global optimum or feasible near-optimal solution within a given time limit, and it is suitable for online / quasi-online operation.

[0103] Solving the reclosing optimization model using a mixed-integer quadratic programming algorithm also includes a time delay uncertainty compensation step, specifically:

[0104] Based on the mechanical action delay of the circuit breaker, the jitter of the measurement timestamp, and the communication link delay, the uncertainty interval of the closing execution time is estimated as follows:

[0105]

[0106] in Delay due to mechanical action of circuit breaker Measure timestamp jitter With communication link delay Obtained by superimposing the upper bound;

[0107] Based on the constraints on phase angle difference, frequency deviation, and closing inrush current within the uncertain interval, a robust constraint set is constructed by applying constraint contraction. Specifically:

[0108] Within the uncertainty interval The constraints are implemented through conservative contraction, for example, by using a bound function. , Derivation of robust tolerance, and with By tightening the boundary of the impact current, a robust constraint set is constructed to ensure that the solution still meets the safety requirements under time delay perturbation.

[0109] The robustness tolerance expression is as follows:

[0110]

[0111] In the formula, This is the upper bound of the rate of change of phase angle difference. The upper bound of the rate of change of frequency deviation is given by the measurement or prediction model of the historical synchronization phasor measurement device (PMU), that is, the first-order difference upper bound of phase angle difference and frequency deviation is calculated from a large number of operating samples. For the uncertainty of the impulse current at time The sensitivity coefficient was obtained by deriving the sensitivity curve of the impact current to the triggering time offset through simulation and experimental measurement, and then obtaining an approximate coefficient through linear fitting. This is a robust upper bound used within the uncertain interval to tighten the boundary of the impulse current constraint.

[0112] The nominal threshold will be determined through the above steps. and shrink to and And the nominal estimate of the impact current is expanded to This allows for the construction of a robust constraint set, ensuring that safety constraints are still met within the actual execution delay perturbation range.

[0113] This step introduces a photovoltaic grid fault transient current generation circuit. and By using physically verifiable features, the "criteria" of S2 are precisely quantified into computable optimization constraints. Within the MIQP framework, objectives and constraints are combined to generate a reclosing control strategy, thereby reducing closing impact and improving success rate while ensuring synchronization and safety boundaries.

[0114] S4 executes the reclosing control strategy and performs reclosing operations on the power grid lines.

[0115] It should be noted that, in this embodiment, the reclosing control strategy, after being generated by the optimization model, is transformed into specific control commands and issued to the circuit breaker actuator. The closing time, phase selection, and reclosing mode (single-phase or three-phase) output by the optimization model are used by the control device to generate corresponding logic control signals, which are transmitted to the circuit breaker operating mechanism via the protection and control terminal through the communication link. To ensure accurate operation, the control commands are synchronized with the unified timing system of the power grid, thereby ensuring that the circuit breaker completes the closing operation at the planned time.

[0116] Upon receiving a control command, the circuit breaker's closing coil excites and triggers the mechanical actuator, causing the contacts to close. During the closing process, voltage and current quantities are collected in real time by voltage and current transformers and a synchronous phasor measurement device, and transmitted to the monitoring system to determine whether the closing action meets expectations. If the line voltage and current return to normal after closing, the reclosing is considered successful; if voltage drops, current exceeding limits, or other abnormalities occur, the reclosing is considered a failure, and the system will immediately trigger subsequent protection measures to prevent equipment damage and fault propagation.

[0117] After the operation is completed, key data from the entire execution process, including action time, control commands, circuit breaker response delay, success or failure results, and abnormal measurement data, are recorded in the monitoring device for subsequent analysis and optimization. By reprocessing this data, the criterion thresholds and optimization model parameters can be calibrated, enabling adaptive iteration and continuous improvement of the method.

[0118] Furthermore, in photovoltaic grid-connected scenarios, to avoid excessive inrush current during reclosing, inverters typically implement a certain delay after reclosing before gradually restoring power output. This embodiment also considers this detail, enabling the grid to smoothly transition to normal operation after the reclosing operation.

[0119] Example 2: To verify the applicability and effectiveness of the method of the present invention, the operating conditions of a 35 kV photovoltaic grid-connected line were selected as a case study. The method described in Example 1 was used to extract fault features, generate criteria, and model and solve the reclosing optimization control strategy.

[0120] 1) Description of the case study scenario

[0121] The example circuit connects two distributed photovoltaic (PV) power stations with a total installed capacity of 50 MW, and is connected to the grid via a 35 kV distribution network. The circuit configuration includes a 35 kV high-voltage circuit breaker and PV grid auxiliary devices, including a circuit to improve the accuracy of PV grid fault relay protection measurements, a device to eliminate PV grid fault current leakage, and a circuit to generate PV grid fault transient current.

[0122] Three typical operating conditions are defined: single-phase ground fault, two-phase short-circuit fault, and three-phase short-circuit fault. Each type of fault is described in... The fault occurs instantly and is cleared within 100 ms. After the fault, line voltage and current signals are acquired using CT, PT, and PMU devices to extract fault characteristic data, including transient current amplitude, voltage drop amplitude, and residual current decay rate.

[0123] It should be noted that this embodiment uses a 35 kV photovoltaic grid-connected line as an example, but the method of the present invention is not limited to this voltage level, and is also applicable to new energy grid scenarios with different voltage levels such as 10 kV, 110 kV, and 220 kV.

[0124] 2) Model calculation and optimization

[0125] According to steps S1–S2 of Example 1, fault feature data are extracted, input into the fault analysis model, and the probability distribution of various faults is output. The results are shown in Table 1.

[0126] Table 1

[0127]

[0128] In the criterion generation stage, by combining the detailed criteria of residual current elimination, voltage recovery, phase angle and frequency synchronization, transient current decay and zero-sequence current not exceeding the limit, the judgment result of whether the reclosing conditions are met under each operating condition is obtained.

[0129] Based on these criteria, a reclosing optimization model is established, with the objective function being to minimize the closing inrush current and maximize the success rate. Constraints are quantified from the criteria and include residual current decay, voltage recovery, phase angle and frequency synchronization, transient current decay ratio, and zero-sequence current limitation. The model is solved using a mixed-integer quadratic programming (MIQP) algorithm, taking into account uncertainties caused by circuit breaker delay, measurement jitter, and communication delay.

[0130] The optimized solution results are shown in Table 2.

[0131] Table 2

[0132]

[0133] 3) Results Comparison and Analysis

[0134] This method is compared and analyzed with the traditional fixed-delay reclosing method. The results show that:

[0135] Under single-phase ground fault conditions, this method reduces the inrush current from 4.5 kA in the traditional method to 2.9 kA, a reduction of approximately 36%.

[0136] Under both two-phase and three-phase short-circuit conditions, the success rate is improved by about 6–8%, and the inrush current is reduced by about 15–20%.

[0137] Traditional methods are prone to relay protection failure or reclosing failure due to interference from inrush current and transient current. However, this method effectively eliminates the above-mentioned adverse factors by introducing photovoltaic grid auxiliary devices and optimization models.

[0138] In summary, this example demonstrates that the method of the present invention can achieve fault feature extraction and criterion generation on 35 kV photovoltaic grid-connected lines. By combining the optimization model to solve the reclosing control strategy, it not only reduces the closing inrush current but also improves the reclosing success rate, verifying the applicability and effectiveness of the method in new energy grid scenarios.

[0139] Example 3, Figure 5 An optimized reclosing device for fault-prone power grid lines is presented, including:

[0140] The feature extraction module is used to collect the operating electrical quantities of the new energy grid-connected system and extract fault feature data;

[0141] The criterion generation module is used to analyze feature data through the fault analysis model, obtain fault discrimination quantities, and generate criteria.

[0142] The strategy generation module is used to establish a reclosing optimization model based on criteria and solve it to obtain the reclosing control strategy.

[0143] The reclosing module is used to execute the reclosing control strategy and perform reclosing operations on the power grid lines.

[0144] Example 4,

[0145] A new energy power grid line fault reclosing optimization device, such as Figure 6 As shown, it includes a memory and a processor: the memory is used to store a program; the processor is used to execute the program to implement any of the embodiments in Example 1.

[0146] Since the new energy power grid line fault reclosing optimization device introduced in this embodiment is the same device used to implement the method in Embodiment 1 of this invention, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the method described in Embodiment 1 of this application. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any device used by those skilled in the art to implement the method in this application embodiment falls within the scope of protection of this application.

[0147] Example 5,

[0148] A readable storage medium having a computer program stored thereon, such as Figure 7 As shown, when the computer program is executed by the processor, it implements any of the embodiments in Example 1.

[0149] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0150] 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.

[0151] 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.

[0152] 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.

[0153] 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.

[0154] 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. An optimized method for fault reclosing in a new energy power grid, characterized in that, Includes the following steps: Collect the operating electrical quantities of the new energy grid-connected system and extract fault characteristic data; The feature data is analyzed by the fault analysis model to obtain the fault discrimination quantity and generate the judgment criterion; Based on the criteria, an optimization model for reclosing is established and the reclosing control strategy is obtained by solving the model. Implement the reclosing control strategy to perform reclosing operations on the power grid lines; The specific steps for constructing the fault analysis model are as follows: A training sample set is constructed based on fault feature data and combined with historical labels, and fault type labeling and training set and validation set are performed. Sample balancing was performed using the SMOTE oversampling method. The XGBoost algorithm is used to train the model on the training set to obtain the corresponding model parameters; The model parameters are evaluated and solidified on the validation set to obtain the fault analysis model; The objective function of the reclosing optimization model includes minimizing the closing inrush current and maximizing the reclosing success rate, as shown in the following formula: in To estimate the closing inrush current, a quadratic approximation can be used. ,coefficient Obtained from offline calibration; The probability of successful reclosing can be estimated by mapping the probability output of the fault analysis model with historical samples through logistic regression. For the closing time, Choose binary variables for each phase. This is a binary variable representing the closing mode. For the phase angle difference before overlap, For frequency difference, Due to voltage deviation, all three are within Prediction; weighting and ; The constraints include the residual current of the faulted phase being below the threshold and lasting for a predetermined duration, the line voltage amplitude recovering to the allowable bandwidth, the phase angle difference and frequency deviation being within the synchronization tolerance, the transient current decaying to a set ratio, and the zero-sequence current not exceeding the limit.

2. The optimized reclosing method for new energy power grid lines according to claim 1, characterized in that, The specific steps for collecting the operating electrical quantities of the new energy grid-connected system and extracting fault characteristic data are as follows: The actual measurements of three-phase voltage and current on the power grid side are collected and preprocessed to obtain standard electrical quantities; Identify abnormal variation ranges in standard electrical quantities; The transient current amplitude, voltage drop amplitude, and residual current decay are calculated based on the abnormal change range and used as fault characteristic data.

3. The optimized reclosing method for new energy power grid lines according to claim 2, characterized in that, The specific steps for analyzing feature data through a fault analysis model to obtain fault discrimination quantities and generate judgment criteria are as follows: Input the fault characteristic data into the fault analysis model and output the probability distribution of fault types; Extract the fault type corresponding to the highest probability based on the probability distribution and quantify it as a fault discrimination quantity; Criteria are generated based on fault discrimination parameters and preset fault classification rules.

4. The optimized reclosing method for new energy power grid lines according to claim 3, characterized in that, The reclosing optimization model is solved using a mixed integer quadratic programming algorithm.

5. The optimized reclosing method for new energy power grid lines according to claim 4, characterized in that, Solving the reclosing optimization model using a mixed-integer quadratic programming algorithm also includes a time delay uncertainty compensation step, specifically: Based on the mechanical action delay of the circuit breaker, the jitter of the measurement timestamp, and the communication link delay, the uncertainty interval of the closing execution time is estimated; Based on the constraints of the uncertain interval on the phase angle difference, frequency deviation and closing inrush current, a robust constraint set is constructed by applying constraint contraction.

6. A fault reclosing optimization device for new energy power grid lines, characterized in that, include: The feature extraction module is used to collect the operating electrical quantities of the new energy grid-connected system and extract fault feature data; The criterion generation module is used to analyze feature data through the fault analysis model, obtain fault discrimination quantities, and generate criteria. The specific construction steps of the fault analysis model are as follows: a training sample set is constructed based on fault feature data and combined with historical labels, and fault type labeling and division of training set and validation set are performed; the SMOTE oversampling method is used to balance the samples; the XGBoost algorithm is used to train the model on the training set to obtain the corresponding model parameters; the model parameters are evaluated and solidified on the validation set to obtain the fault analysis model. The strategy generation module is used to establish a reclosing optimization model based on criteria and solve it to obtain the reclosing control strategy. The objective function of the reclosing optimization model includes minimizing the closing inrush current and maximizing the reclosing success rate, as shown in the following formula: in To estimate the closing inrush current, a quadratic approximation can be used. ,coefficient Obtained from offline calibration; The probability of successful reclosing can be estimated by mapping the probability output of the fault analysis model with historical samples through logistic regression. The closing time, Choose binary variables for each phase. This is a binary variable representing the closing mode; For the phase angle difference before overlap, For frequency difference, Due to voltage deviation, all three are within Prediction; weighting and ; The constraints include the residual current of the faulted phase being below the threshold and lasting for a predetermined duration, the line voltage amplitude recovering to the allowable bandwidth, the phase angle difference and frequency deviation being within the synchronization tolerance, the transient current decaying to a set ratio, and the zero-sequence current not exceeding the limit. The reclosing module is used to execute the reclosing control strategy and perform reclosing operations on the power grid lines.

7. A fault reclosing optimization device for new energy power grid lines, characterized in that, Including memory and processor: The memory is used to store programs; The processor is used to execute the program to implement each step of the new energy power grid line fault reclosing optimization method as described in any one of claims 1-5.

8. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements each step of the new energy power grid line fault reclosing optimization method as described in any one of claims 1-5.

Images