Power simulation data generation method and device, equipment and storage medium

Abstract

The invention discloses an electric power simulation data generation method and device, equipment and a storage medium, and the method comprises the steps: responding to an electric power simulation fault type inputted by a user, and determining a power utilization scene and corresponding fault equipment according to the fault type; calling a corresponding power grid simulation topological structure according to the power consumption scene and the corresponding fault equipment; according to the fault type, the power consumption scene and the fault equipment, inputting the fault type, the power consumption scene and the fault equipment into a preset fault prediction model to generate fault parameters; and integrating the fault parameters into the power grid simulation topological structure to carry out power fault simulation so as to complete generation of power data simulation data. The technical problems that in the prior art, power simulation data lacks dynamism and real-time performance, scenes are lacked, interactivity is insufficient, and data authenticity is low are solved.

Description

Technical Field

[0001] This invention relates to the field of power data simulation technology, and in particular to a method, apparatus, device and storage medium for generating power simulation data. Background Technology

[0002] With the rapid development of the power industry and continuous technological advancements, the demand for power professionals is constantly increasing. To meet this demand, power grid companies, universities, and training institutions are strengthening the practical training component of power engineering education. These programs require students to master the basic principles, operating characteristics, and operational methods of power systems, as well as possess fault handling and data analysis skills. However, due to limitations in facilities, equipment, and faculty, current distribution network training suffers from problems such as a lack of dynamic and real-time data, missing scenarios, insufficient interactivity, and low data accuracy.

[0003] Traditional teaching and training often use fixed data or simple random generation of training data from case studies, which lacks dynamism and real-time performance and cannot simulate the complex operating conditions of power distribution networks (such as the reproduction of multiple types of faults and load fluctuations). At the same time, existing simulation systems are unable to cover coupled scenarios of multiple time scales (hourly, daily, monthly) and multiple spatial levels (power supply lines, different types of dedicated transformers, public transformers, and residential users) from a time and space perspective. Moreover, the randomly generated data does not take into account physical constraints (such as power balance, voltage stability), resulting in a large deviation between the simulation results and the real system. Summary of the Invention

[0004] This invention provides a method, apparatus, device, and storage medium for generating power simulation data, in order to solve the technical problems of existing power simulation data, such as lack of dynamism and real-time performance, missing scenarios, insufficient interactivity, and low data authenticity.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a method for generating power simulation data, comprising: In response to the power simulation fault type input by the user, the power consumption scenario and its corresponding faulty equipment are determined according to the fault type; Based on the described power consumption scenario and its corresponding faulty equipment, the corresponding power grid simulation topology is invoked; Based on the fault type, power consumption scenario, and faulty equipment, the data is input into a preset fault prediction model to generate fault parameters. The fault parameters are integrated into the power grid simulation topology to perform power fault simulation and generate power data simulation data.

[0006] As a preferred embodiment, the method for constructing the preset fault prediction model includes: Collect fault sample data; The fault sample data is preprocessed, and the preprocessed fault sample data is labeled and paired to obtain training data; The training data is feature-standardized, and derived features are created; Based on the standardized training data, a corresponding regression algorithm is selected, and the preset initial model is iteratively trained using the standardized training data and the derived features according to the regression algorithm, resulting in a preset fault prediction model.

[0007] As a preferred embodiment, the standardized training data includes a training set, a validation set, and a test set; the step of iteratively training a preset initial model using the standardized training data and the derived features to obtain a preset fault prediction model specifically includes: Build a pre-defined initial model; The derived features are injected into the preset initial model, and the training set is input into the preset initial model for iterative training; In each preset iteration cycle, the validation set is input into the preset initial model trained in the current iteration, and the results of the preset initial model trained in the current iteration are verified and evaluated with the validation set to obtain an evaluation score; based on the evaluation score, the structure or learning parameters of the preset initial model are adjusted until a preset number of iterations is reached or the evaluation score meets a preset condition, and the trained prediction model is obtained. The test set is input into the prediction model for verification and evaluation, and the prediction model that passes the verification and evaluation is used as the preset fault prediction model.

[0008] As a preferred embodiment, the step of selecting a corresponding regression algorithm based on the feature-standardized training data specifically includes: When the amount of training data after feature standardization is greater than a preset number, a deep neural network algorithm is selected. When the amount of training data after feature standardization is less than or equal to a preset number, the random forest algorithm is selected.

[0009] As a preferred embodiment, the creation of derived features specifically includes: Based on the training data, identify the original features corresponding to each training data point; Based on the original features, a relational expression is constructed, and the constructed relational expression is used as a derived feature; the derived features include: node power balance relational expression, power balance relational expression, and regional load correlation relational expression.

[0010] As a preferred embodiment, the method for constructing the node power balance relationship includes: obtaining the total power between each node in the power grid simulation topology, and constructing the node power balance relationship by combining the impedance of conductors of different diameters and lengths in the distribution network. The method for constructing the power balance relationship includes: constructing the power balance relationship by obtaining the power consumption and loss of metering nodes in various power consumption scenarios in the power grid simulation topology; The method for constructing the regional load correlation equation includes: constructing the regional load correlation equation based on the relationship between each node in the power grid simulation topology.

[0011] As a preferred embodiment, the fault types include: line loss faults, metering and data acquisition faults, electricity theft faults, photovoltaic power generation faults, and power distribution equipment faults.

[0012] Accordingly, the present invention also provides a power simulation data generation device, comprising: a fault type module, a calling module, a fault parameter module, and a simulation module; The fault type module is used to respond to the power simulation fault type input by the user, and determine the power consumption scenario and its corresponding faulty equipment according to the fault type. The calling module is used to call the corresponding power grid simulation topology structure according to the power consumption scenario and its corresponding faulty equipment; The fault parameter module is used to generate fault parameters by inputting the fault type, power consumption scenario and faulty equipment into a preset fault prediction model. The simulation module is used to integrate the fault parameters into the power grid simulation topology to perform power fault simulation and generate power data simulation data.

[0013] Accordingly, the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the power simulation data generation method as described above.

[0014] Accordingly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the power simulation data generation method as described above.

[0015] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: The technical solution of this invention determines the corresponding power consumption scenario and its corresponding faulty equipment by responding to the power simulation fault type input by the user. This ensures the interactivity between fault data and avoids the problem of low realism caused by detachment from reality and missing scenarios. By calling the corresponding power grid simulation topology, it ensures that the power simulation data can be used for fault simulation within the topology structure, and that the generation of fault data is dynamic and real-time. The corresponding fault type, power consumption scenario, and faulty equipment are input into the preset fault prediction model to generate corresponding fault parameters. These fault parameters are then integrated into the power grid simulation topology to control the execution of the corresponding power fault simulation. This ensures the coupling relationship between fault occurrence, parameter change, and stability during the simulation process, thus ensuring overall interactivity. Attached Figure Description

[0016] Figure 1 : A flowchart illustrating the steps of a power simulation data generation method provided in an embodiment of the present invention; Figure 2 : A diagram illustrating the simulation data generation method and control system provided in this embodiment of the invention; Figure 3 : A diagram showing the composition of a substation feeder model provided in an embodiment of the present invention; Figure 4 : This is a diagram illustrating the composition of the virtual load voltage and current source provided in an embodiment of the present invention; Figure 5 : A diagram illustrating the composition of the special transformer model provided in an embodiment of the present invention; Figure 6 : A diagram showing the composition of the common variable model provided in the embodiments of the present invention; Figure 7 : This is a flowchart of the process for establishing a fault prediction model provided in an embodiment of the present invention; Figure 8 : This is a structural diagram of a power simulation data generation device provided in an embodiment of the present invention. Detailed Implementation

[0017] 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 Please refer to Figure 1 The present invention provides a method for generating power simulation data, comprising the following steps S101-S104: S101: In response to the power simulation fault type input by the user, determine the power consumption scenario and its corresponding faulty equipment according to the fault type; As a preferred embodiment, the fault types include: line loss faults, metering and data acquisition faults, electricity theft faults, photovoltaic power generation faults, and power distribution equipment faults.

[0019] In this embodiment, line loss fault refers to a situation where the difference between the total power supply and the total electricity sales exceeds a specified threshold. This case involves two forms: one is as follows... Figure 2 The feeder line loss in the substation is caused by two main issues: Firstly, the difference between the total metering value of the feeder line and the total metering value of the individual power consumption scenarios (such as dedicated transformers and public transformers connected to the feeder line) exceeds the standard, resulting in feeder line loss faults. Figure 2 The line loss of Zhonggong transformer is due to the difference between the total meter readings of the transformer and those of residential and commercial storefront users exceeding the standard, resulting in a line loss fault in the transformer area.

[0020] It should be noted that, Figure 2 This document presents a teaching and training simulation data generation method and control system composition diagram based on a power model. The substation feeder model 1, referencing the actual 10kV feeder topology of a substation, consists of a simulated high-voltage feeder cabinet 11, a simulated overhead line 12, a simulated underground cable 13, and a virtual load voltage and current source 14. Its detailed construction modules are as follows: Figure 3 As shown.

[0021] The simulated high-voltage feeder cabinet 11 consists of a cabinet, a high-voltage vacuum circuit breaker, a voltage transformer, a current transformer, a zero-sequence current transformer, a live indicator, a grounding switch, a surge arrester, an energy meter, a data acquisition terminal, and a fault simulation module. The fault simulation module can receive control commands sent by the local control console 4 of the substation feeder model to complete the fault occurrence and recovery of the switching circuit, metering circuit, energy meter, and data acquisition terminal.

[0022] The simulated overhead line 12 is composed of simulated overhead line segments. The simulated line segments can receive control commands sent by the local control console 4 of the substation feeder model to set parameters such as wire diameter, material, and length to simulate different types of overhead lines.

[0023] The simulated underground cable 13 is composed of simulated underground cable segments. The simulated underground cable segments can receive control commands sent by the local control console 4 of the substation feeder model to set parameters such as wire diameter, material, and length to simulate different types of underground cable lines.

[0024] Virtual load voltage and current source 14, as follows: Figure 4 As shown, it consists of a power supply module, a CPU core control module, a signal output module, a power amplification module, and a communication module. The communication module can receive control commands sent by the local control console 4 of the substation feeder model to set parameters such as the amplitude, frequency, and phase angle of the voltage and current signals, simulating the output power of the high-voltage feeder under different operating conditions.

[0025] Model 2 for dedicated transformers simulates actual dedicated transformer power consumption scenarios by building a platform transformer, a box-type transformer, and a distribution room. Each scenario is configured with corresponding transformers, switches, voltage and current transformers, energy meters, data acquisition terminals, and virtual load voltage and current sources, based on actual scenarios, as detailed below. Figure 5 As shown.

[0026] Model 3, referencing actual dedicated transformer power consumption scenarios, constructs simulated public transformer power consumption scenarios in the form of benchtop transformers, box-type transformers, and distribution rooms. Each scenario is configured with corresponding transformers, switches, metering voltage and current transformers, energy meters, data acquisition terminals, virtual load voltage and current sources, and also includes different configurations for three-phase power users, single-phase residential users, distributed photovoltaic access households, and electric vehicle charging piles belonging to the public transformer, as detailed below. Figure 6 As shown.

[0027] The local control console 4 for the substation feeder model, the local control console 5 for the dedicated transformer model, and the local control console 6 for the public transformer model, through their built-in control software system, complete the control and information acquisition of parameters such as voltage, current, frequency, and phase angle, as well as operating modes, at various measuring points in the substation feeder, dedicated transformer, and public transformer model scenarios, and generate simulation data by simulating and setting line parameters. Simultaneously, they can receive unified scheduling and control commands issued by the control system master station 7.

[0028] The main control station 7 uses built-in control system software to simulate line loss faults, metering and acquisition faults, electricity theft faults, photovoltaic power generation faults, and power distribution equipment faults in various model scenarios, as well as generate, collect, and display data for different operating modes such as light load, heavy load, over-compensation, and under-compensation.

[0029] In this embodiment, metering acquisition failure refers to abnormalities in the collection of metering and electricity consumption information in various scenarios. Figure 3 Medium and high voltage metering cabinets Figure 5 Centralized platform transformer, box-type transformer, distribution room transformer Figure 6 Such faults can occur in the central transformer substation, the platform transformer substation, the power distribution room transformer, and its subordinate users.

[0030] In this embodiment, the simulated electricity theft fault generally occurs Figure 3 Medium and high voltage metering cabinets Figure 5 Dedicated transformers for central platform transformers, box-type transformers, and power distribution room transformer scenarios and Figure 6 Subordinate users of various types of public transformers.

[0031] In this embodiment, the photovoltaic power generation failure occurs in the distribution network area with distributed photovoltaic power generation users.

[0032] In this embodiment, the power distribution equipment failure occurs in the distribution cabinets, distribution boxes, and reactive power compensation cabinets of dedicated transformer and public transformer users.

[0033] S102: Based on the described power consumption scenario and its corresponding faulty equipment, call the corresponding power grid simulation topology; In this embodiment, the power distribution network simulation topology model can be adaptively modified and constructed according to the actual power consumption scenario and its corresponding faulty equipment. For example, it can be constructed and called as follows: Figure 3 The topology shown.

[0034] S103: Based on the fault type, power consumption scenario and faulty equipment, input the data into the preset fault prediction model to generate fault parameters; As a preferred embodiment, the method for constructing the preset fault prediction model includes: Collect fault sample data; The fault sample data is preprocessed, and the preprocessed fault sample data is labeled and paired to obtain training data; The training data is feature-standardized, and derived features are created; Based on the standardized training data, a corresponding regression algorithm is selected, and the preset initial model is iteratively trained using the standardized training data and the derived features according to the regression algorithm, resulting in a preset fault prediction model.

[0035] In this embodiment, the fault sample data includes, but is not limited to: a multi-level model (distribution network feeder, transformer, dedicated transformer load, three-phase load under the public transformer, single-phase load) based on the above-mentioned distribution network topology, defining a database of equipment parameters (such as line impedance Z, transformer capacity S, load rate λ). Simulation data includes the three-phase voltage (Ua, Ub, Uc), current (Ia, Ib, Ic), phase angle between voltage and current, power factor (total power factor, A-phase power factor, B-phase power factor, C-phase power factor), active power (total active power, A-phase active power, B-phase active power, C-phase active power), reactive power (total reactive power, A-phase reactive power, B-phase reactive power, C-phase reactive power), active and reactive power data of each node generated by the accumulation of power over time, and also includes the working status information of power distribution equipment such as switch positions on distribution cabinets, reactive power compensation, and signal indicator lights.

[0036] In this embodiment, a supervised learning method is employed to construct an intelligent identification model for abnormal states in the distribution network. This model addresses the problem of collecting fault sample data related to line losses, metering, electricity theft, photovoltaic power generation, and abnormal operation of distribution equipment in the distribution network. It systematically integrates multi-source data from the electricity marketing system and the metering automation system. The training samples cover various equipment in the 10kV distribution network, including feeders, dedicated transformers, public transformers, residential users, and power users.

[0037] In this embodiment, each training sample consists of the following parameter features and identification labels: Static equipment parameters: describe the inherent properties of the equipment, such as the material, length, cross-sectional area, and resistance of the cable.

[0038] Dynamic fault parameters: Record the real-time operating status when a fault occurs, such as current, voltage, active power, reactive power, power factor, meter readings (forward and reverse), temperature, and the status of power distribution equipment switching signals.

[0039] Identification label (fault type): Indicates the specific problem corresponding to the sample, such as "line loss fault", "metering and data acquisition fault", "electricity theft fault", "photovoltaic power generation fault", "power distribution equipment fault", etc.

[0040] By training with a large number of "parameter feature-label" paired samples, the model can learn and master the mapping relationship from multidimensional data to specific failure modes.

[0041] In this embodiment, the inputs to the preset fault prediction model include: equipment static parameters—scene type name, capacity, wire diameter, length, resistance, etc.—which are the foundation of the model and are inherent attributes of the equipment that are determined before the fault occurs. The outputs include: fault parameters—fault type, voltage, current, power, power factor, electrical quantity, temperature, switch status, indicator light status, etc. The target of the model's prediction is a quantifiable representation of the fault phenomenon. Through feature standardization, the fundamental assumption of a physical relationship between "equipment parameters" and "fault parameters" is established, which is the theoretical basis for the model's validity.

[0042] It's important to note that feature standardization eliminates the influence of different units on parameters, allowing the model to learn each feature more fairly. Take line length and line resistance as examples: length is measured in meters and can range from hundreds to thousands or tens of thousands, while resistance is measured in ohms and can range from a few tenths of a digit to a single digit. Without standardization, the model might perceive "length" as larger and more volatile, thus focusing more on "length" and neglecting "resistance," similar to using both "centimeters" and "meters" without a unified unit, leading to misjudgments. By first calculating the feature mean, then the feature standard deviation, and finally applying a standardization formula, the original data is transformed into a new data distribution with a mean of 0 and a standard deviation of 1. This allows features with different units to be compared fairly on the same scale, providing numerical stability to the model, significantly accelerating convergence, and making the optimization process smoother, ultimately leading to a better and more reliable final model.

[0043] As a preferred embodiment, the standardized training data includes a training set, a validation set, and a test set; the step of iteratively training a preset initial model using the standardized training data and the derived features to obtain a preset fault prediction model specifically includes: Build a pre-defined initial model; The derived features are injected into the preset initial model, and the training set is input into the preset initial model for iterative training; In each preset iteration cycle, the validation set is input into the preset initial model trained in the current iteration, and the results of the preset initial model trained in the current iteration are verified and evaluated with the validation set to obtain an evaluation score; based on the evaluation score, the structure or learning parameters of the preset initial model are adjusted until a preset number of iterations is reached or the evaluation score meets a preset condition, and the trained prediction model is obtained. The test set is input into the prediction model for verification and evaluation, and the prediction model that passes the verification and evaluation is used as the preset fault prediction model.

[0044] In this embodiment, the data is divided into a training set and a test set to ensure that the evaluation is of the model's generalization ability, rather than its ability to memorize the training data.

[0045] After randomly shuffling the entire dataset, it was divided into three non-overlapping parts: a training set (70% of the total data) on which the model learns and adjusts its internal parameters; a validation set (15% of the total data) used to adjust the model's hyperparameters and select the best model during training, similar to a midterm exam to test learning outcomes and adjust the learning method in a timely manner; and a test set (15% of the total data) used only after the entire model training and tuning process is completely finished to perform a one-time, final evaluation of the model's capabilities, ensuring the absolute fairness of the evaluation results.

[0046] The first step is learning. The model learns on the training set.

[0047] Step 2: Simulation Evaluation. Evaluate the current model using a validation set, and adjust the model structure or learning parameters based on the results.

[0048] Step 3: Repeat. Repeat steps 1 and 2 until the model performs optimally and stably on the validation set.

[0049] Step 4: Final Evaluation. After confirming the model is properly tuned, use the previously unused, archived test set to conduct a final evaluation of the model. The resulting scores (such as accuracy and mean squared error) represent an unbiased estimate of the model's performance in the real world.

[0050] As a preferred embodiment, the step of selecting a corresponding regression algorithm based on the feature-standardized training data specifically includes: When the amount of training data after feature standardization is greater than a preset number, a deep neural network algorithm is selected. When the amount of training data after feature standardization is less than or equal to a preset number, the random forest algorithm is selected.

[0051] In this embodiment, the most suitable nonlinear regression algorithm is selected based on the amount of training data and specific requirements. Initially, the Random Forest algorithm can be chosen, corresponding to 3000-10000 historical fault data points. Because its tree model has relatively low data requirements, it can efficiently learn complex nonlinear relationships and feature interactions from small to medium-sized datasets. It is less prone to overfitting (especially through ensemble methods) and has a fast training speed, facilitating rapid iteration and validation of ideas. It can be quickly trained on ordinary servers and is likely to yield a model with considerable accuracy in a short time, meeting initial business needs. If massive historical data (hundreds of thousands of data points) from large power supply companies is available and resources permit, the Deep Neural Network (DNN) algorithm can be used. This algorithm can automatically extract deep, abstract feature patterns from massive amounts of data. When the data volume is large enough, its performance ceiling is usually very high, potentially discovering extremely complex combinations of fault symptoms that are difficult for the human brain and simple models to detect, thereby achieving ultimate prediction accuracy.

[0052] As a preferred embodiment, the creation of derived features specifically includes: Based on the training data, identify the original features corresponding to each training data point; Based on the original features, a relational expression is constructed, and the constructed relational expression is used as a derived feature; the derived features include: node power balance relational expression, power balance relational expression, and regional load correlation relational expression.

[0053] As a preferred embodiment, the method for constructing the node power balance relationship includes: obtaining the total power between each node in the power grid simulation topology, and constructing the node power balance relationship by combining the impedance of conductors of different diameters and lengths in the distribution network. The method for constructing the power balance relationship includes: constructing the power balance relationship by obtaining the power consumption and loss of metering nodes in various power consumption scenarios in the power grid simulation topology; The method for constructing the regional load correlation equation includes: constructing the regional load correlation equation based on the relationship between each node in the power grid simulation topology.

[0054] In this embodiment, by combining or calculating the original features, prior physical knowledge is injected into the model, strongly guiding the model to learn the correct rules. This typically significantly improves the model's performance and generalization ability. The following derived features are introduced: Node power balance equations: In the equation, the total power of a node is equal to the power of all branch nodes P. i1 P i2 ...P in The sum. In the power calculation at each node, equipment parameters are taken into account. For example, in the power distribution network, conductors of different diameters and lengths have different impedances Z, resulting in different power losses P. L In each node scenario, the transformer capacity parameter S must be considered as a constraint, and different load rates λ must be combined to complete simulation scenarios of different working conditions such as light load, economic operation, heavy load, and overload.

[0055] Electricity balance formula: ∑P gen =∑P load +∑P loss , where P gen For metering nodes in various power supply scenarios, such as high-voltage feeder metering points and public transformer substation metering points, P load For metering nodes in various scenarios consuming electricity, such as metering nodes for dedicated transformers in platform-mounted substations, dedicated transformers in box-type substations, and public transformer metering nodes in distribution rooms, P loss Total losses include line losses, transformer losses, etc. The main metering node for the public transformer serves as both the load on the feeder and the metering node for the power supplied to the public transformer area. Calculations must consider line impedance parameters to determine line losses, and transformer copper and iron losses based on transformer capacity and load rate. Regional load correlation: Following the steps and methods described above, combined with Figure 2 , Figure 3 , Figure 5 , Figure 6 The simulation area shown is based on the power distribution network topology in a real-world scenario. Figure 2 The feeder model, dedicated transformer model, and public transformer model described in the text Figure 3 Different types of power distribution network line models, combined Figure 5 Different types of special transformer models in China Figure 6 The public transformer model and its subordinate models such as residential households, three-phase power households, distributed photovoltaic users, and charging piles are combined into a distribution network model in a hierarchical and classified manner. Based on the total power supply of the feeders, the power is distributed to the dedicated transformer branches / public transformer branches, and the power is further distributed to the subdivided power consumption models such as residential / power / distributed photovoltaic / charging piles under the public transformer. Based on the aforementioned power balance, power balance and other constraints, regional load correlation simulation data is generated.

[0056] By creating derived features, the learning difficulty of the model is greatly reduced, helping it to capture the profound physical laws behind the data more quickly and accurately.

[0057] S104: Integrate the fault parameters into the power grid simulation topology to perform power fault simulation and generate power data simulation data.

[0058] In this embodiment, the fault parameters output by the model are finally deeply integrated with the constructed and invoked physical distribution network simulation topology model and the background control system.

[0059] Implementing the above embodiments has the following effects: The technical solution of this invention determines the corresponding power consumption scenario and its corresponding faulty equipment by responding to the power simulation fault type input by the user. This ensures the interactivity between fault data and avoids the problem of low realism caused by detachment from reality and missing scenarios. By calling the corresponding power grid simulation topology, it ensures that the power simulation data can be used for fault simulation within the topology structure, and that the generation of fault data is dynamic and real-time. The corresponding fault type, power consumption scenario, and faulty equipment are input into the preset fault prediction model to generate corresponding fault parameters. These fault parameters are then integrated into the power grid simulation topology to control the execution of the corresponding power fault simulation. This ensures the coupling relationship between fault occurrence, parameter change, and stability during the simulation process, thus ensuring overall interactivity.

[0060] Example 2 Please see Figure 8 The present invention provides a power simulation data generation device, comprising: a fault type module 201, a calling module 202, a fault parameter module 203, and a simulation module 204; The fault type module 201 is used to respond to the power simulation fault type input by the user and determine the power consumption scenario and its corresponding faulty equipment according to the fault type. The calling module 202 is used to call the corresponding power grid simulation topology structure according to the power consumption scenario and its corresponding faulty equipment; The fault parameter module 203 is used to generate fault parameters by inputting the fault type, power consumption scenario and faulty equipment into a preset fault prediction model. The simulation module 204 is used to integrate the fault parameters into the power grid simulation topology to perform power fault simulation and generate power data simulation data.

[0061] As a preferred embodiment, the method for constructing the preset fault prediction model includes: Collect fault sample data; The fault sample data is preprocessed, and the preprocessed fault sample data is labeled and paired to obtain training data; The training data is feature-standardized, and derived features are created; Based on the standardized training data, a corresponding regression algorithm is selected, and the preset initial model is iteratively trained using the standardized training data and the derived features according to the regression algorithm, resulting in a preset fault prediction model.

[0062] As a preferred embodiment, the standardized training data includes a training set, a validation set, and a test set; the step of iteratively training a preset initial model using the standardized training data and the derived features to obtain a preset fault prediction model specifically includes: Build a pre-defined initial model; The derived features are injected into the preset initial model, and the training set is input into the preset initial model for iterative training; In each preset iteration cycle, the validation set is input into the preset initial model trained in the current iteration, and the results of the preset initial model trained in the current iteration are verified and evaluated with the validation set to obtain an evaluation score; based on the evaluation score, the structure or learning parameters of the preset initial model are adjusted until a preset number of iterations is reached or the evaluation score meets a preset condition, and the trained prediction model is obtained. The test set is input into the prediction model for verification and evaluation, and the prediction model that passes the verification and evaluation is used as the preset fault prediction model.

[0063] As a preferred embodiment, the step of selecting a corresponding regression algorithm based on the feature-standardized training data specifically includes: When the amount of training data after feature standardization is greater than a preset number, a deep neural network algorithm is selected. When the amount of training data after feature standardization is less than or equal to a preset number, the random forest algorithm is selected.

[0064] As a preferred embodiment, the creation of derived features specifically includes: Based on the training data, identify the original features corresponding to each training data point; Based on the original features, a relational expression is constructed, and the constructed relational expression is used as a derived feature; the derived features include: node power balance relational expression, power balance relational expression, and regional load correlation relational expression.

[0065] As a preferred embodiment, the method for constructing the node power balance relationship includes: obtaining the total power between each node in the power grid simulation topology, and constructing the node power balance relationship by combining the impedance of conductors of different diameters and lengths in the distribution network. The method for constructing the power balance relationship includes: constructing the power balance relationship by obtaining the power consumption and loss of metering nodes in various power consumption scenarios in the power grid simulation topology; The method for constructing the regional load correlation equation includes: constructing the regional load correlation equation based on the relationship between each node in the power grid simulation topology.

[0066] As a preferred embodiment, the fault types include: line loss faults, metering and data acquisition faults, electricity theft faults, photovoltaic power generation faults, and power distribution equipment faults.

[0067] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0068] Implementing the above embodiments has the following effects: The technical solution of this invention determines the corresponding power consumption scenario and its corresponding faulty equipment by responding to the power simulation fault type input by the user. This ensures the interactivity between fault data and avoids the problem of low realism caused by detachment from reality and missing scenarios. By calling the corresponding power grid simulation topology, it ensures that the power simulation data can be used for fault simulation within the topology structure, and that the generation of fault data is dynamic and real-time. The corresponding fault type, power consumption scenario, and faulty equipment are input into the preset fault prediction model to generate corresponding fault parameters. These fault parameters are then integrated into the power grid simulation topology to control the execution of the corresponding power fault simulation. This ensures the coupling relationship between fault occurrence, parameter change, and stability during the simulation process, thus ensuring overall interactivity.

[0069] Example 3 Accordingly, the present invention also provides a terminal device, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the power simulation data generation method as described in any of the above embodiments.

[0070] The terminal device in this embodiment includes a processor, a memory, and a computer program and computer instructions stored in the memory and executable on the processor. When the processor executes the computer program, it implements the various steps described in Embodiment 1 above, for example... Figure 1The steps S101 to S104 are shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above-described device embodiment, such as the fault parameter module 203.

[0071] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the terminal device. For example, the fault parameter module 203 is used to generate fault parameters by inputting the fault type, power consumption scenario, and faulty equipment into a preset fault prediction model.

[0072] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the schematic diagram is merely an example of a terminal device and does not constitute a limitation on the terminal device. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, etc.

[0073] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0074] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function, etc.; the data storage area may store data created based on the use of the mobile terminal, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital card (SD card), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0075] Wherein, if the modules / units integrated in the terminal device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, it can implement the steps of the various method embodiments described above. Wherein, the computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content contained in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0076] Example 4 Accordingly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the power simulation data generation method as described in any of the above embodiments.

[0077] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for generating power simulation data, characterized in that, include: In response to the power simulation fault type input by the user, the power consumption scenario and its corresponding faulty equipment are determined according to the fault type; Based on the described power consumption scenario and its corresponding faulty equipment, the corresponding power grid simulation topology is invoked; Based on the fault type, power consumption scenario, and faulty equipment, the data is input into a preset fault prediction model to generate fault parameters. The fault parameters are integrated into the power grid simulation topology to perform power fault simulation and generate power data simulation data.

2. The method for generating power simulation data as described in claim 1, characterized in that, The method for constructing the preset fault prediction model includes: Collect fault sample data; The fault sample data is preprocessed, and the preprocessed fault sample data is labeled and paired to obtain training data. The training data is feature-standardized, and derived features are created; Based on the standardized training data, a corresponding regression algorithm is selected, and the preset initial model is iteratively trained using the standardized training data and the derived features according to the regression algorithm, resulting in a preset fault prediction model.

3. The method for generating power simulation data as described in claim 2, characterized in that, The standardized training data includes a training set, a validation set, and a test set; the iterative training of the preset initial model using the standardized training data and the derived features to obtain a preset fault prediction model specifically includes: Build a pre-defined initial model; The derived features are injected into the preset initial model, and the training set is input into the preset initial model for iterative training; In each preset iteration cycle, the validation set is input into the preset initial model trained in the current iteration, and the results of the preset initial model trained in the current iteration are verified and evaluated with the validation set to obtain an evaluation score; based on the evaluation score, the structure or learning parameters of the preset initial model are adjusted until a preset number of iterations is reached or the evaluation score meets a preset condition, and the trained prediction model is obtained. The test set is input into the prediction model for verification and evaluation, and the prediction model that passes the verification and evaluation is used as the preset fault prediction model.

4. The method for generating power simulation data as described in claim 2, characterized in that, The step of selecting the corresponding regression algorithm based on the feature-standardized training data specifically includes: When the amount of training data after feature standardization is greater than a preset number, a deep neural network algorithm is selected; When the amount of training data after feature standardization is less than or equal to a preset number, the random forest algorithm is selected.

5. A method for generating power simulation data as described in any one of claims 2-4, characterized in that, The creation of derived features specifically includes: Based on the training data, identify the original features corresponding to each training data point; Based on the original features, a relational expression is constructed, and the constructed relational expression is used as a derived feature; the derived features include: node power balance relational expression, power balance relational expression, and regional load correlation relational expression.

6. The method for generating power simulation data as described in claim 5, characterized in that, The method for constructing the node power balance relationship includes: obtaining the total power between each node in the power grid simulation topology, and combining the impedance of conductors of different diameters and lengths in the distribution network to construct the node power balance relationship; The method for constructing the power balance relationship includes: constructing the power balance relationship by obtaining the power consumption and loss of metering nodes in various power consumption scenarios in the power grid simulation topology; The method for constructing the regional load correlation equation includes: constructing the regional load correlation equation based on the relationship between each node in the power grid simulation topology.

7. A method for generating power simulation data as described in any one of claims 1-4, characterized in that, The fault types include: line loss faults, metering and data acquisition faults, electricity theft faults, photovoltaic power generation faults, and power distribution equipment faults.

8. A power simulation data generation device, characterized in that, include: Fault type module, calling module, fault parameter module, and simulation module; The fault type module is used to respond to the power simulation fault type input by the user, and determine the power consumption scenario and its corresponding faulty equipment according to the fault type. The calling module is used to call the corresponding power grid simulation topology structure according to the power consumption scenario and its corresponding faulty equipment; The fault parameter module is used to generate fault parameters by inputting the fault type, power consumption scenario and faulty equipment into a preset fault prediction model. The simulation module is used to integrate the fault parameters into the power grid simulation topology to perform power fault simulation and generate power data simulation data.

9. A terminal device, characterized in that, The device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the power simulation data generation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the power simulation data generation method as described in any one of claims 1 to 7.

Images