Voltage sag duration evaluation method, system, equipment, medium and product

Abstract

The invention relates to the technical field of power distribution network systems, and discloses a voltage sag duration evaluation method, system and device, a medium and a product. According to the method, real-time influence factor data are recognized through a relay protection action cooperation mode recognition model, and a current relay protection action cooperation mode is determined; updating the voltage sag duration of the duration rule base according to the obtained predicted fusing time of the fuse, and determining the voltage sag duration according to the current relay protection action cooperation mode through the duration rule base; therefore, the duration time of the voltage sag is determined by utilizing real-time updating of the relay protection action cooperation mode identification model and the duration time rule base, and the accuracy and the real-time performance of evaluation of the duration time of the voltage sag are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network systems, and particularly to a method, system, device, medium and product for evaluating the duration of voltage sags. Background Art

[0002] With the rapid development of modern industrial production and power systems towards high reliability and high power quality, voltage sag (VS) has become one of the core issues of concern in the field of power systems due to its frequent occurrence and serious harm to sensitive equipment. Voltage sags are usually caused by events such as grid short-circuit faults, lightning strikes, access of distributed power sources or startup of large-capacity equipment. Its essence is that the effective value of the voltage drops short-term to 10% - 90% of the rated value, and the duration is usually from 10 milliseconds to several minutes. Although the duration is short, problems such as industrial production interruption, damage to precision equipment and downtime of digital systems caused by it may cause significant economic losses. <L

[0003] As a core index to measure the severity of voltage sag events, the accurate evaluation of the duration of voltage sags has multiple engineering values. First, this parameter can provide a key action timing benchmark for dynamic voltage compensation devices, significantly reducing the risk of downtime of industrial sensitive equipment through millisecond-level response; second, through the spatio-temporal coupling analysis with fault location information, potential weak nodes such as abnormal system impedance and improper protection configuration can be accurately exposed, providing a targeted governance basis for improving the resilience of the power grid; in addition, the quantitative evaluation based on the probability distribution of the duration can not only support the risk modeling of power quality insurance pricing on the user side, but also guide the differentiated investment strategy of the power grid, and give priority to strengthening the active defense ability of high voltage sag risk areas.

[0004] In the prior art, the evaluation of the duration of voltage sags in the distribution network depends on accurate parameters, with a large amount of calculation, and the accuracy and real-time performance of the evaluation of the duration of voltage sags are poor. Summary of the Invention

[0005] In view of this, the present invention provides a method, system, device, medium and product for evaluating the duration of voltage sags, which solves the technical problems that the evaluation of the duration of voltage sags in the distribution network depends on accurate parameters, with a large amount of calculation, and the accuracy and real-time performance of the evaluation of the duration of voltage sags are poor.

[0006] The first aspect of the present invention provides a method for evaluating the duration of voltage sags, including:

[0007] Through a preset relay protection action coordination mode recognition model, according to the pre-input real-time influencing factor data, determine the current relay protection action coordination mode corresponding to the real-time influencing factor data; wherein, the relay protection action coordination mode recognition model is obtained by training a classifier with a plurality of historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each of the historical samples of relay protection action coordination modes.

[0008] Update the voltage sag duration of the duration rule base according to the predicted fuse melting time obtained; wherein, the duration rule base includes the mapping relationship between the relay protection action coordination mode and the voltage sag duration.

[0009] Based on the duration rule base, according to the current relay protection action coordination mode, determine the voltage sag duration corresponding to the current relay protection action coordination mode.

[0010] Preferably, the method further includes:

[0011] Obtain a plurality of historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each of the historical samples of relay protection action coordination modes, and form a historical training sample set.

[0012] Preprocess the historical training sample set.

[0013] Train a classifier according to the preprocessed historical training sample set to obtain the relay protection action coordination mode recognition model.

[0014] Preferably, the method further includes:

[0015] According to the relay protection action coordination mode and the voltage sag influence time corresponding to the relay protection action coordination mode, determine the voltage sag duration under multiple different fault positions; wherein, the voltage sag influence time includes the fault clearing time, the fuse melting time and the main protection action time; the voltage sag duration is divided into the voltage sag duration of the branch where the fault is located, the voltage sag duration of other branches of the feeder where the fault is located, and the voltage sag duration of other feeders on the same bus where the fault is located according to the fault position.

[0016] Construct the duration rule base according to the relay protection action coordination mode, the voltage sag influence time corresponding to the relay protection action coordination mode, and the voltage sag duration.

[0017] Preferably, the method further includes: predicting the predicted fuse melting time; the predicting the predicted fuse melting time includes:

[0018] Obtain the time-series data of the fuse melting time;

[0019] Perform normalization processing on the time-series data of the fuse melting time;

[0020] Based on the long short-term memory neural network, predict through the normalized time-series data of the fuse melting time to obtain the predicted melting time of the fuse at the current moment.

[0021] Preferably, the method further includes:

[0022] Under the physical constraints of the thermodynamic equation of the fuse, calculate the cumulative thermal energy according to the current time-series integral;

[0023] Determine the theoretical melting time according to the cumulative thermal energy and the heat capacity of the fuse;

[0024] According to the predicted melting time, the actual melting time and the theoretical melting time, determine the loss value through the loss function;

[0025] Judge whether the predicted melting time reaches the optimal prediction according to the loss value;

[0026] In the case where it is judged that the predicted melting time does not reach the optimal prediction, go to the step of obtaining the time-series data of the fuse melting time until it is judged that the predicted melting time reaches the optimal prediction, and output the predicted melting time.

[0027] Preferably, the update of the voltage sag duration of the duration rule base according to the predicted melting time of the obtained fuse includes:

[0028] Update the melting time of the fuse in the duration rule base according to the predicted melting time, and determine the voltage sag duration according to the melting time of the fuse.

[0029] In a second aspect, the present invention also provides a voltage sag duration evaluation system, including:

[0030] A protection action recognition module, configured to identify the model through a preset relay protection action coordination method, and determine the current relay protection action coordination method corresponding to the real-time influence factor data according to the pre-input real-time influence factor data; wherein, the relay protection action coordination method recognition model is obtained by training a classifier with multiple historical samples of relay protection action coordination methods and the influence factor historical sample data corresponding to each of the historical samples of relay protection action coordination methods;

[0031] A rule base update module for updating the voltage sag duration of the duration rule base according to the predicted fuse blowing time obtained; wherein, the duration rule base includes the mapping relationship between the relay protection action coordination mode and the voltage sag duration;

[0032] A duration evaluation module for determining the voltage sag duration corresponding to the current relay protection action coordination mode based on the duration rule base according to the current relay protection action coordination mode.

[0033] In a third aspect, the present invention further provides an electronic device, which includes a memory and a processor. A computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of the voltage sag duration evaluation method as described in the first aspect.

[0034] In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the steps of the voltage sag duration evaluation method as described in the first aspect are implemented.

[0035] In a fifth aspect, the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the steps of the voltage sag duration evaluation method as described in the first aspect.

[0036] As can be seen from the above technical solutions, the present invention identifies the real-time influencing factor data through the relay protection action coordination mode identification model to determine the current relay protection action coordination mode, and then updates the voltage sag duration of the duration rule base according to the predicted fuse blowing time obtained. According to the current relay protection action coordination mode, the duration of the voltage sag is determined through the duration rule base, so as to determine the duration of the voltage sag by using the real-time update of the relay protection action coordination mode identification model and the duration rule base, improving the accuracy and real-time performance of the evaluation of the duration of the voltage sag. Description of the Drawings

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to these drawings without creative efforts.

[0038] Figure 1An application environment diagram of a voltage sag duration evaluation method provided by an embodiment of the present invention;

[0039] Figure 2 A flowchart of a voltage sag duration evaluation method provided by an embodiment of the present invention;

[0040] Figure 3 An example logic diagram of a voltage sag duration evaluation method provided by an embodiment of the present invention;

[0041] Figure 4 A schematic structural diagram of a voltage sag duration evaluation system provided by an embodiment of the present invention;

[0042] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0043] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0044] The voltage sag duration evaluation method provided by the embodiments of the present application can be applied to an application environment as Figure 1 shown. Among them, the terminal 101 communicates with the server 102 through a network. The data storage system can store the data that the server 102 needs to process. The data storage system can be integrated on the server 102, or placed in the cloud or other network servers. The terminal 101 or the server 102 determines the current relay protection action coordination method corresponding to the real-time influencing factor data through a preset relay protection action coordination method recognition model according to the pre-input real-time influencing factor data; among them, the relay protection action coordination method recognition model is obtained by training a classifier with a plurality of historical samples of relay protection action coordination methods and the influencing factor historical sample data corresponding to each historical sample of the relay protection action coordination method; the voltage sag duration of the duration rule base is updated according to the predicted fuse blowing time obtained; among them, the duration rule base includes the mapping relationship between the relay protection action coordination method and the voltage sag duration; based on the duration rule base, the voltage sag duration corresponding to the current relay protection action coordination method is determined according to the current relay protection action coordination method.

[0045] The terminal 101 can be but is not limited to various personal computers, laptop computers, smart phones, tablet computers, etc.

[0046] The server 102 can be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0047] As Figure 2 shown, an embodiment of the present application provides a method for evaluating the duration of voltage sags. Taking the application of this method to the Figure 1 terminal 101 or server 102 as an example, it includes the following steps S1 to S3. Among them:

[0048] Step S1: Through a preset relay protection action coordination mode recognition model, according to the pre-input real-time influencing factor data, determine the current relay protection action coordination mode corresponding to the real-time influencing factor data; among them, the relay protection action coordination mode recognition model is obtained by training a classifier with multiple historical samples of relay protection action coordination modes and the corresponding historical sample data of influencing factors for each relay protection action coordination mode historical sample.

[0049] Among them, the relay protection action coordination mode recognition model is constructed based on a machine learning algorithm. By learning and training a large amount of historical data, it can accurately identify the mapping relationship between the real-time influencing factor data and the relay protection action coordination mode. This model can handle the complex and changeable operation environment of the power system, quickly respond and determine the current relay protection action coordination mode.

[0050] Among them, the relay protection action coordination mode is the action sequence and coordination relationship of various relay protection devices in the power system when a fault occurs, so as to ensure the safe and stable operation of the power system. Different relay protection action coordination modes will have different impacts on the duration of voltage sags. Therefore, accurately identifying the current relay protection action coordination mode is a key step in evaluating the duration of voltage sags.

[0051] At the same time, the influencing factor data of the relay protection action coordination mode is the influencing factor data associated with the relay protection action coordination mode, including but not limited to fault type, fault location, power grid topology structure, line impedance, protection device configuration, pre-steady-state voltage, fault current time series, fuse characteristic parameters, user access point impedance, etc. These factor data can be obtained through real-time monitoring and collection of the power system and used as input information for the relay protection action coordination mode recognition model to analyze and process.

[0052] Step S2: Update the voltage sag duration in the duration rule base according to the predicted fuse blowing time obtained; among them, the duration rule base contains the mapping relationship between the relay protection action coordination mode and the voltage sag duration.

[0053] Among them, the fusing time of the fuse can be used to accurately predict the fusing time of the fuse, update the relevant data in the duration rule base in real time, and thus more accurately evaluate the voltage sag duration. Through the update mechanism, it is ensured that the duration rule base can provide timely and reliable voltage sag duration evaluation results in the face of different fault scenarios.

[0054] Step S3: Based on the duration rule base, determine the voltage sag duration corresponding to the current relay protection action coordination mode according to the current relay protection action coordination mode.

[0055] It can be understood that since the duration rule base contains the mapping relationship between the relay protection action coordination mode and the voltage sag duration, it is easy to obtain the voltage sag duration corresponding to the current relay protection action coordination mode through matching with the current relay protection action coordination mode.

[0056] It should be noted that in the embodiment of the present application, the real-time influencing factor data is identified through the relay protection action coordination mode recognition model to determine the current relay protection action coordination mode, and then the voltage sag duration of the duration rule base is updated according to the predicted fusing time of the obtained fuse. According to the current relay protection action coordination mode, the duration of the voltage sag is determined through the duration rule base, so as to determine the duration of the voltage sag by using the real-time update of the relay protection action coordination mode recognition model and the duration rule base, improving the accuracy and real-time performance of the evaluation of the voltage sag duration.

[0057] In some embodiments, the process of constructing the relay protection action coordination mode recognition model includes:

[0058] Step S11: Obtain a plurality of historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each historical sample of the relay protection action coordination mode respectively, and form a historical training sample set.

[0059] Among them, the relay protection system of the power grid directly affects the duration of the voltage sag event. Circuit breakers (with reclosing function) and sectionalizers have different effects on upstream and downstream nodes during instantaneous and permanent faults, and the action characteristics of different protection types also affect the voltage sag duration.

[0060] Typical distribution network protection action coordination modes are shown in Table 1.

[0061] Table 1

[0062]

[0063] Among them, 1) Coordination method 1 (No. 1): The fault is a transient fault, the fault current does not reach the thermal fusing limit of the fuse, the fault is self-cleared, and the main protection does not operate. At this time, the users on the feeder where the fault is located will suffer a voltage sag with a duration of the fault clearing time T Q : T1 = T Q . The users on other branches of the feeder will suffer a voltage sag with a duration of the fault clearing time: T2 = T Q . The users on other feeders under the same bus will also suffer a voltage sag with a duration of the fault clearing time T3 = T Q .

[0064] 2) Coordination method 2 (No. 2): The fault is a transient fault. Before the fault is cleared, the fault current reaches the thermal fusing limit of the fuse, and the fuse disconnects the faulty branch at time T R , and the main protection does not operate. At this time, the users on the feeder where the fault is located will suffer a long interruption. The users on other branches of the feeder will suffer a voltage sag with a duration of the fuse disconnection time: T2 = T R . The users on other feeders under the same bus will also suffer a voltage sag with a duration of the fuse disconnection time T3 = T R .

[0065] 3) Coordination method 3 (No. 3): The fault is a transient fault. Before the fault is cleared, it reaches the main protection operation time, the fault current value does not reach the thermal fusing limit of the fuse, the fault is self-cleared during the main protection tripping, and the main protection reclosing is successful. At this time, the users on the feeder where the fault is located will suffer a deep sag with a duration of the main protection operation time T D , a short interruption T1 = T D . The users on other branches of the feeder will also suffer a deep sag with a duration of the main protection operation time T D , a short interruption T2 = T D . The users on other feeders under the same bus will suffer a voltage sag with a duration of the main protection operation time T3 = T D .

[0066] 4) Coordination method 4 (No. 4): The fault is a permanent fault, the fault current reaches the thermal fusing limit of the fuse, and the fuse disconnects the branch at time T R , cutting off the faulty branch. At this time, the users on the feeder where the fault is located will suffer a long interruption. The users on other branches of the feeder will also suffer a voltage sag with a duration of the fuse disconnection time: T2 = T R . The users on other feeders under the same bus will also suffer a voltage sag with a duration of the fuse disconnection time T3 = T R .

[0067] 5) Coordination mode 5 (No. 5): The fault is a permanent fault. Before the fault current value reaches the fuse heating and fusing limit, the main protection operates, the main protection reclosing fails, and the feeder where the fault is located is cut off. At this time, the users on the feeder where the fault is located will successively experience a deep voltage sag with a duration of T D , a short-term interruption, a deep voltage sag with a duration of T D , and a long-term interruption: T1 = 2T D . The users on other branches of the feeder will also suffer a deep voltage sag with a duration of T D , a short-term interruption, a deep voltage sag with a duration of T D , and a long-term interruption: T2 = 2T D . The users on other feeders under the same bus will suffer 2 voltage sags with a duration of T D : T3 = 2T D .

[0068] Step S12. Preprocess the historical training sample set.

[0069] Among them, the preprocessing is to normalize continuous features (such as impedance, current value) and encode categorical features (such as fault type, protection device configuration, etc.).

[0070] Step S13. Train the classifier according to the preprocessed historical training sample set to obtain a relay protection action coordination mode recognition model.

[0071] Exemplarily, the construction process of the relay protection action coordination mode recognition model is as follows:

[0072] 1) Data preparation and feature engineering.

[0073] Input features: fault type, fault location, power grid topology, line impedance, protection device configuration, pre-fault steady-state voltage, fault current time series, fuse characteristic parameters, user access point impedance, etc.

[0074] Labels: 5 protection action modes (corresponding to 5 numbers).

[0075] Data preprocessing: Normalize continuous features (such as impedance, current value) and encode categorical features (such as fault type, protection device configuration, etc.).

[0076] 2) Selection and training of the base learner (Level-0 model).

[0077] Base model selection: Adopt a deep neural network (DNN) machine learning model, as shown in the following formula.

[0078]

[0079] In the formula, X(i) is the feature input of the i-th deep neural network layer, X (i-1) W is the feature input of the i-1th deep neural network layer. (i) is the random initialization weight, b (i) is the learnable bias matrix, is the activation function.

[0080] Training method: Use categorical cross entropy to verify the probabilities of the five protective action modes predicted by the generated base model. The loss function is:

[0081]

[0082] Where: N is the number of samples; C is the total number of categories (5 protection action modes); y n,c is the true label of the nth sample. If the nth sample belongs to the cth category, then y n,c =1, otherwise y n,c =0. p n,c is the model's predicted probability that the nth sample belongs to the cth category.

[0083] 3) Meta-learner (Level-1 model) training.

[0084] Input features: The prediction results of the base learner, that is, the probability output of each model for the five protection action modes.

[0085] Model selection: XGBoost integrates multiple weak classifiers (decision trees) through the gradient boosting framework, and its loss function is also:

[0086]

[0087] The objective function is defined as:

[0088]

[0089] in, is the regularization term of the tree. T is the number of leaf nodes, is the leaf weight.

[0090] Objective: To synthesize the output of the base model and generate the final protection action mode classification results (numbered 1 to 5).

[0091] In some embodiments, the process of building a duration rule base includes:

[0092] Step S21: Determine the voltage sag durations at multiple different fault locations according to the relay protection action coordination mode and the voltage sag influence time corresponding to the relay protection action coordination mode. Among them, the voltage sag influence time includes the fault clearing time, the fusing time of the fuse, and the main protection action time. The voltage sag durations are divided into the voltage sag duration of the branch where the fault is located, the voltage sag durations of other branches of the feeder where the fault is located, and the voltage sag durations of other feeders on the same bus where the fault is located according to the fault location.

[0093] Among them, the fault clearing time is the time required for the fault to be cleared, which depends on factors such as the type of fault, the fault location, and the protection configuration of the power system. For example, in the case of a transient fault, when the fault current does not reach the heat fusing limit of the fuse, the fault may be self-cleared, and at this time the fault clearing time is the time required for the fault to be self-cleared.

[0094] The fusing time of the fuse is the time required for the fuse to heat up and fuse under the action of the fault current. The length of the fusing time depends on the characteristic parameters of the fuse, such as the rated current of the fuse, the fusing characteristic curve, etc., as well as the magnitude and waveform of the fault current.

[0095] The main protection action time is the time required for the main protection device to issue a tripping command and cut off the fault after detecting the fault. The length of the main protection action time depends on factors such as the response speed of the protection device, the time constant of the tripping circuit, and the action time of the circuit breaker.

[0096] Step S22: Construct a duration rule library according to the relay protection action coordination mode, the voltage sag influence time corresponding to the relay protection action coordination mode, and the voltage sag duration.

[0097] Among them, when determining the voltage sag duration, factors such as the fault clearing time, the fusing time of the fuse, and the main protection action time need to be comprehensively considered, as well as the influence of the fault location on the voltage sag duration. By constructing a detailed duration rule library, the voltage sag durations in different fault scenarios can be accurately evaluated, as shown in Table 2.

[0098] Table 2

[0099]

[0100] In some embodiments, this method further includes: predicting the predicted fusing time of the fuse; predicting the predicted fusing time of the fuse, including:

[0101] Step S31: Obtain the time series data of the fusing time of the fuse.

[0102] Among them, the fuse time series data are the actual fuse time records of the fuse under different fault currents, different ambient temperatures and other conditions. These data can be obtained through laboratory tests or monitoring during actual operation, and are used to train the fuse time prediction model.

[0103] Step S32: Standardize the fuse time series data.

[0104] The standardization process is to perform Z-Score standardization on the fuse time series data, that is:

[0105]

[0106] In the formula, and are the mean and standard deviation of the fuse time series data respectively, is the fuse time series data before standardization, is the fuse time series data after standardization.

[0107] Step S33: Based on the long short-term memory neural network, predict through the standardized fuse time series data to obtain the predicted fuse time of the fuse at the current moment.

[0108] Among them, the long short-term memory neural network (Long Short-Term Memory, LSTM) is a deep learning model for time series prediction, which can capture the long-term dependencies in time series data. In the fuse time prediction task, the long short-term memory neural network can predict the fuse time of the fuse under the current fault conditions by learning the standardized fuse time series data. By introducing the fuse time prediction model, the embodiments of the present application can update the predicted fuse time of the fuse in real time, thereby further improving the accuracy and real-time performance of the voltage sag duration evaluation.

[0109] LSTM controls the transmission of time series information through a gating mechanism (forget gate, input gate, output gate).

[0110] Forget Gate: Determines how much of the historical cell state to retain:

[0111]

[0112] Input Gate: Determines how much new information to update:

[0113]

[0114]

[0115] Cell state update:

[0116]

[0117] Output Gate: Determines the output feature:

[0118]

[0119]

[0120] Where: is the input at the current time, is the hidden state at the previous time, carrying historical time series information, is the Sigmoid function, is the hyperbolic tangent activation function, is the element-wise multiplication, are the trainable weight matrix and bias term. is the output of the forget gate, controlling the retention ratio of the historical cell state of. is the output of the input gate, controlling the update ratio of the candidate state of. is the updated cell state, fusing historical information and current input information. is the output of the output gate, determining the content of the current hidden state of. is the hidden state at the current time, used to be passed to the next time or output the prediction result.

[0121] In some embodiments, the method further includes:

[0122] Step S34, under the physical constraints of the thermodynamic equation of the fuse, calculate the cumulative thermal energy according to the current time series integral;

[0123] Among them, the thermodynamic equation of the fuse is a physical process describing the temperature change of the fuse under the action of a fault current. By solving the thermodynamic equation, the temperature distribution and change trend of the fuse during the fault process can be obtained. The cumulative thermal energy is the total heat absorbed by the fuse from the start of the fault to the current time, and it can be obtained by integrating the fault current time series data. When calculating the cumulative thermal energy, physical parameters such as the heat capacity and thermal resistance of the fuse, as well as factors such as the magnitude and waveform of the fault current, need to be considered. By introducing the physical constraints of the thermodynamic equation, the accuracy of the fuse time prediction can be further improved.

[0124] The thermodynamic equation of the fuse is:

[0125]

[0126] In the formula, k is the fuse time coefficient, is the thermal capacity of the fuse, t is the time, and I t is the current.

[0127] For each sample, the cumulative heat energy is calculated by integrating according to the current time series as follows:

[0128]

[0129] In the formula, E is the cumulative heat energy.

[0130] Step S35: Determine the theoretical melting time according to the cumulative heat energy and the thermal capacity of the fuse;

[0131] Among them, the theoretical melting time is calculated through the cumulative heat energy as follows:

[0132]

[0133] In the formula, is the theoretical melting time.

[0134] Step S36: Determine the loss value through the loss function according to the predicted melting time, the actual melting time and the theoretical melting time.

[0135] Among them, the mean square error between the melting time predicted by LSTM and the actual time and the difference from the theoretical melting time are added as constraint terms to the loss function, as shown in the following formula:

[0136]

[0137] In the formula, L3 is the model loss function, M is the number of samples, is the predicted value of the melting time, is the actual value of the melting time, is the theoretical value of the melting time, is the weight coefficient to balance data-driven and physical constraints.

[0138] Step S37: Judge whether the predicted melting time reaches the optimal prediction according to the loss value.

[0139] Among them, by setting a fixed loss threshold, when the loss value is less than the loss threshold, it is determined that the predicted melting time reaches the optimal prediction, and when the loss value is not less than the loss threshold, it is determined that the predicted melting time does not reach the optimal prediction.

[0140] Step S38: In the case where it is determined that the predicted melting time does not reach the optimal prediction, go back to the step of obtaining the melting time time series data of the fuse until it is determined that the predicted melting time reaches the optimal prediction, and output the predicted melting time.

[0141] In some embodiments, the voltage sag duration of the duration rule base is updated according to the predicted melting time of the obtained fuse, including:

[0142] Update the fusing time of the fuse in the duration rule library according to the predicted fusing time, and determine the voltage sag duration according to the fusing time of the fuse.

[0143] Exemplarily, a voltage sag duration evaluation method provided by an embodiment of the present application is as Figure 3 shown, and the specific process is as follows:

[0144] 1) Data construction: Integrate characteristic data such as power grid topology, fault type, and protection device configuration, and construct a training set in combination with historical voltage sag records.

[0145] 2) Model building: Through the Stacking integration framework, hierarchically train the deep neural network (DNN) base learner and the XGBoost meta-learner to achieve protection action mode classification.

[0146] 3) Duration mapping: Based on the protection action classification result and the duration rule library, calculate the voltage sag duration. And dynamically correct the voltage sag duration in combination with physical rules, i.e., fusing time prediction.

[0147] 4) Effect verification: Input the test set data, compare the predicted duration with the actual value, and quantitatively evaluate the model accuracy through indicators such as MAE (Mean Absolute Error) and RMSE (Root Mean Square Error).

[0148] Based on the same inventive concept, an embodiment of the present application also provides a voltage sag duration evaluation system for implementing the voltage sag duration evaluation method involved above.

[0149] The implementation solution provided by this system to solve the problem is similar to the implementation solution described in the above method. Therefore, the specific limitations in one or more embodiments of the voltage sag duration evaluation system provided below can refer to the limitations on the voltage sag duration evaluation method in the above text and will not be elaborated here.

[0150] As Figure 4 shown, an embodiment of the present application provides a voltage sag duration evaluation system, including:

[0151] The protection action recognition module 100 is used to determine the current relay protection action coordination mode corresponding to the real-time influencing factor data according to the preset relay protection action coordination mode recognition model and the pre-input real-time influencing factor data. Among them, the relay protection action coordination mode recognition model is obtained by training a classifier with multiple historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each historical sample of relay protection action coordination mode.

[0152] The rule base update module 200 is used to update the voltage sag duration of the duration rule base according to the predicted fuse melting time obtained. Among them, the duration rule base contains the mapping relationship between the relay protection action coordination mode and the voltage sag duration.

[0153] The duration evaluation module 300 is used to determine the voltage sag duration corresponding to the current relay protection action coordination mode based on the duration rule base according to the current relay protection action coordination mode.

[0154] In some embodiments, the system further includes: an identification model construction module, which is used to:

[0155] Obtain multiple historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each historical sample of relay protection action coordination mode, and form a historical training sample set.

[0156] Preprocess the historical training sample set.

[0157] Train a classifier according to the preprocessed historical training sample set to obtain a relay protection action coordination mode recognition model.

[0158] In some embodiments, the system further includes: a rule base construction module, which is used to:

[0159] Determine the voltage sag duration under multiple different fault locations according to the relay protection action coordination mode and the voltage sag influence time corresponding to the relay protection action coordination mode. Among them, the voltage sag influence time includes the fault clearing time, the fuse melting time, and the main protection action time. The voltage sag duration is divided into the voltage sag duration of the branch where the fault is located, the voltage sag duration of other branches of the feeder where the fault is located, and the voltage sag duration of other feeders on the same bus where the fault is located according to the fault location.

[0160] Construct a duration rule base according to the relay protection action coordination mode, the voltage sag influence time corresponding to the relay protection action coordination mode, and the voltage sag duration.

[0161] In some embodiments, the system further includes: a time prediction module, configured to: predict the predicted fusing time of the fuse; predicting the predicted fusing time of the fuse includes:

[0162] Obtain the time series data of the fusing time of the fuse;

[0163] Perform normalization processing on the time series data of the fusing time;

[0164] Based on the long short-term memory neural network, predict through the normalized time series data of the fusing time to obtain the predicted fusing time of the fuse at the current moment.

[0165] In some embodiments, the system further includes: a time update module, configured to:

[0166] Under the physical constraints of the thermodynamic equation of the fuse, calculate the cumulative thermal energy according to the current time series integral;

[0167] Determine the theoretical fusing time according to the cumulative thermal energy and the heat capacity of the fuse;

[0168] Determine the loss value through the loss function according to the predicted fusing time, the actual fusing time, and the theoretical fusing time;

[0169] Judge whether the predicted fusing time reaches the optimal prediction according to the loss value;

[0170] In the case where it is judged that the predicted fusing time does not reach the optimal prediction, go to the step of obtaining the time series data of the fusing time of the fuse until it is judged that the predicted fusing time reaches the optimal prediction, and output the predicted fusing time.

[0171] In some embodiments, the duration evaluation module 300 is configured to:

[0172] Update the fusing time of the fuse in the duration rule library according to the predicted fusing time, and determine the voltage sag duration according to the fusing time of the fuse.

[0173] As Figure 5 shown, an embodiment of the present application provides an electronic device. An electronic device, the electronic device 10 includes a memory 20 and a processor 30. A computer program is stored in the memory 20. When the computer program is executed by the processor 30, the processor 30 is caused to execute the steps of the voltage sag duration evaluation method in the above embodiments.

[0174] An embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the steps of the voltage sag duration evaluation method in the above embodiments are implemented.

[0175] An embodiment of the present application provides a computer program product. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer is caused to execute the steps of the voltage sag duration evaluation method described in the above embodiments.

[0176] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, electronic devices, computer storage media, and computer program products described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0177] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0178] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the arrows, these steps do not necessarily have to be executed in the order indicated by the arrows. Unless otherwise clearly stated in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages do not necessarily have to be executed at the same time, but can be executed at different times. The execution order of these steps or stages does not necessarily have to be sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0179] In several embodiments provided by the present invention, it should be understood that the disclosed systems, electronic devices, computer storage media, computer program products, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, indirect couplings or communication connections of devices or units, and can be in electrical, mechanical, or other forms.

[0180] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed across multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0181] In addition, each functional unit in various embodiments of the present invention may be integrated in a processing unit, may exist separately as individual physical units, or two or more units may be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0182] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (English full name: Read-Only Memory, English abbreviation: ROM), random access memories (English full name: Random Access Memory, English abbreviation: RAM), magnetic disks or optical discs and other various media that can store program codes.

[0183] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. And these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of various embodiments of the present invention.

Claims

1. A method for evaluating the duration of voltage sags, characterized in that, Including: Through a preset identification model for relay protection action coordination modes, based on the pre-input real-time influencing factor data, determine the current relay protection action coordination mode corresponding to the real-time influencing factor data; wherein, the relay protection action coordination mode identification model is obtained by training a classifier with multiple historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each of the historical samples of relay protection action coordination modes; Update the voltage sag duration of the duration rule base according to the predicted fusing time of the fuse obtained; wherein, the duration rule base includes the mapping relationship between the relay protection action coordination mode and the voltage sag duration; Based on the duration rule base, according to the current relay protection action coordination mode, determine the voltage sag duration corresponding to the current relay protection action coordination mode.

2. The voltage sag duration evaluation method according to claim 1, wherein Also including: Obtain multiple historical samples of relay protection action coordination modes and the historical sample data of influencing factors corresponding to each of the historical samples of relay protection action coordination modes, and form a historical training sample set; Preprocess the historical training sample set; Train a classifier according to the preprocessed historical training sample set to obtain the relay protection action coordination mode identification model.

3. The voltage sag duration evaluation method according to claim 1, wherein Also including: According to the relay protection action coordination mode and the voltage sag influence time corresponding to the relay protection action coordination mode, determine the voltage sag durations under multiple different fault locations; wherein, the voltage sag influence time includes the fault clearing time, the fusing time of the fuse, and the main protection action time; the voltage sag durations are divided according to the fault location into the voltage sag duration of the branch where the fault is located, the voltage sag durations of other branches of the feeder where the fault is located, and the voltage sag durations of other feeders on the same bus where the fault is located; Construct the duration rule base according to the relay protection action coordination mode, the voltage sag influence time corresponding to the relay protection action coordination mode, and the voltage sag duration.

4. The voltage sag duration evaluation method according to claim 3, wherein Also including: Predict the predicted fusing time of the fuse; The prediction of the predicted fusing time of the fuse includes: Obtain the time series data of the fusing time of the fuse; Perform standardization processing on the time series data of the fusing time; Based on a long short-term memory neural network, predict through the standardized time series data of the fusing time to obtain the predicted fusing time of the fuse at the current moment.

5. The voltage sag duration evaluation method according to claim 4, wherein Also including: Under the physical constraints of the thermodynamic equation of the fuse, calculate the cumulative heat energy according to the time series integral of the current; Determine the theoretical fusing time according to the cumulative heat energy and the heat capacity of the fuse; According to the predicted fusing time, the actual fusing time, and the theoretical fusing time, determine the loss value through a loss function; Judge whether the predicted fusing time reaches the optimal prediction according to the loss value; In the case of judging that the predicted fusing time does not reach the optimal prediction, go to the step of obtaining the time series data of the fusing time of the fuse until it is judged that the predicted fusing time reaches the optimal prediction, and output the predicted fusing time.

6. The method for evaluating the voltage sag duration according to claim 4 or 5, characterized in that, Updating the voltage sag duration of the duration rule base according to the predicted fuse blowing time obtained includes: Updating the fuse blowing time of the fuse in the duration rule base according to the predicted fuse blowing time, and determining the voltage sag duration according to the fuse blowing time of the fuse.

7. A voltage sag duration evaluation system, characterized in that, Including: A protection action identification module, configured to identify a model through a preset relay protection action coordination method, and determine the current relay protection action coordination method corresponding to the real-time influence factor data according to the pre-input real-time influence factor data; wherein, the relay protection action coordination method identification model is obtained by training a classifier with multiple historical samples of relay protection action coordination methods and the influence factor historical sample data respectively corresponding to each of the historical samples of relay protection action coordination methods; A rule base update module, configured to update the voltage sag duration of the duration rule base according to the predicted fuse blowing time obtained; wherein, the duration rule base includes a mapping relationship between the relay protection action coordination method and the voltage sag duration; A duration evaluation module, configured to determine the voltage sag duration corresponding to the current relay protection action coordination method based on the duration rule base according to the current relay protection action coordination method.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor. When the computer program stored in the memory is executed by the processor, the processor executes the steps of the voltage sag duration evaluation method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, the steps of the voltage sag duration evaluation method according to any one of claims 1-6 are implemented.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the steps of the voltage sag duration evaluation method according to any one of claims 1-6.

Images