Method and device for determining influence parameters of transient stability of power system and electronic equipment

Abstract

The invention discloses a method and device for determining influence parameters of transient stability of a power system and electronic equipment. The method comprises the following steps: acquiring an operation scene of a power system; determining a transient stability index and a plurality of target power parameters corresponding to the power system in the operation scene; determining a first causal relationship and a second causal relationship corresponding to the power system according to the plurality of target power parameters and the transient stability index; determining a plurality of influence chains according to the plurality of target power parameters, the first causal relationship and the second causal relationship; and determining transient stability influence parameters corresponding to the power system according to the plurality of influence chains. According to the method and the device, the technical problem that the influence parameters of the transient stability of the power system are not accurately determined when the influence parameters of the transient stability of the power system are determined in related technologies is solved.

Description

Technical Field

[0001] This invention relates to the field of power systems, and more specifically, to a method, apparatus, and electronic device for determining the influence parameters of transient stability in a power system. Background Technology

[0002] In related technologies, power system transient stability measures the ability of a power system to transition to a new steady state or return to its original steady-state operation after being disturbed. Determining the influencing parameters of power system transient stability provides crucial information for designing protection and control strategies, optimizing system structure, and adjusting operating modes. However, in these technologies, there is a technical problem of inaccurate determination of these influencing parameters.

[0003] There is currently no effective solution to the above problems. Summary of the Invention

[0004] This invention provides a method, apparatus, and electronic device for determining the influence parameters of power system transient stability, in order to at least solve the technical problem in the related art where the determination of the influence parameters of power system transient stability is inaccurate.

[0005] According to one aspect of the present invention, a method for determining the impact parameters of transient stability of a power system is provided, comprising: acquiring an operating scenario of the power system; determining a transient stability index and a plurality of target power parameters corresponding to the power system under the operating scenario, wherein the transient stability index is used to represent the degree of operational stability of the power system after a disturbance, and the plurality of target power parameters are parameters used to reflect the operating state of the power system; determining a first causal relationship and a second causal relationship corresponding to the power system based on the plurality of target power parameters and the transient stability index, wherein the first causal relationship is used to represent the causal association relationship between any two of the plurality of target power parameters, and the second causal relationship is used to represent the causal association relationship between the plurality of target power parameters and the transient stability index respectively; determining a plurality of influence chains based on the plurality of target power parameters, the first causal relationship, and the second causal relationship; and determining transient stability impact parameters corresponding to the power system based on the plurality of influence chains.

[0006] Optionally, determining multiple target power parameters corresponding to the power system under the operating scenario includes: determining multiple initial power parameters corresponding to the power system under the operating scenario; determining a first correlation index corresponding to each of the multiple initial power parameters, and multiple second correlation indices corresponding to each of the multiple initial power parameters, wherein the corresponding first correlation index is used to represent the degree of correlation between the corresponding initial power parameter and the transient stability index, and the corresponding multiple second correlation indices are used to represent the degree of correlation between the corresponding initial power parameter and other initial power parameters respectively; and determining multiple target power parameters corresponding to the power system based on the first correlation index corresponding to each of the multiple initial power parameters and the multiple second correlation indices corresponding to each of the multiple initial power parameters.

[0007] Optionally, determining multiple target power parameters corresponding to the power system based on the first correlation index corresponding to each of the multiple initial power parameters and the multiple second correlation indices corresponding to each of the multiple initial power parameters includes: determining a first candidate parameter from the multiple initial power parameters, wherein the first candidate parameter is the initial power parameter with the largest first correlation index among the multiple initial power parameters; and determining, according to the execution order of performing multiple screenings on the multiple initial power parameters, a first candidate parameter from the multiple first remaining parameters based on the first correlation index corresponding to each of the multiple first remaining parameters and the second correlation index between each of the multiple first remaining parameters and the first candidate parameter. The second candidate parameter is defined as follows: the plurality of first remaining parameters are other initial power parameters besides the first candidate parameter among the plurality of initial power parameters; based on the first correlation index corresponding to the plurality of second remaining parameters and the second correlation index between the plurality of second remaining parameters and the second candidate parameter, a third candidate parameter is determined from the plurality of second remaining parameters, wherein the plurality of second remaining parameters are other initial power parameters besides the second candidate parameter among the plurality of first remaining parameters, until multiple screening operations are completed, resulting in a plurality of candidate parameters corresponding to the power system; based on the plurality of candidate parameters, a plurality of target power parameters corresponding to the power system are determined.

[0008] Optionally, determining multiple target power parameters corresponding to the power system based on the multiple candidate parameters includes: when the multiple candidate parameters include active power and reactive power, determining the power ratio corresponding to the power system based on the active power and the reactive power; and determining the multiple target power parameters corresponding to the power system based on the power ratio and the multiple candidate parameters.

[0009] Optionally, determining the transient stability impact parameters corresponding to the power system based on the plurality of impact chains includes: determining, according to the execution order of the plurality of impact chains, multiple target nodes corresponding to any one of the plurality of impact chains, wherein the plurality of target nodes include first nodes corresponding to the plurality of target power parameters and second nodes corresponding to the transient stability index; determining parent nodes corresponding to the plurality of target nodes; determining sub-impact parameters corresponding to any one of the impact chains based on the target node data corresponding to the plurality of target nodes and the parent node data corresponding to the parent nodes corresponding to the plurality of target nodes, wherein the sub-impact parameters are used to reflect the transient stability parameters of the power system under any one of the impact chains; until the plurality of impact chains are executed, obtaining the sub-impact parameters corresponding to the plurality of impact chains; and determining the transient stability impact parameters corresponding to the power system based on the sub-impact parameters corresponding to the plurality of impact chains.

[0010] Optionally, determining the sub-influence parameter corresponding to any influence chain based on the target node data corresponding to the plurality of target nodes and the parent node data corresponding to the parent nodes corresponding to the plurality of target nodes includes: determining the predicted data corresponding to the plurality of target nodes based on the parent node data corresponding to the parent nodes corresponding to the plurality of target nodes; determining the data deviation value corresponding to the plurality of target nodes based on the target node data and the predicted data; and determining the sub-influence parameter corresponding to any influence chain based on the data deviation value corresponding to the plurality of target nodes and the parent node data corresponding to the parent nodes corresponding to the parent nodes corresponding to the plurality of target nodes.

[0011] Optionally, determining the transient stability index corresponding to the power system under the operating scenario includes: determining the power angle difference index corresponding to the multiple generator sets under the operating scenario, wherein the corresponding power angle difference index represents the degree of difference in power angle between the two generators included in the corresponding generator set, and the power system includes the multiple generator sets; and determining the transient stability index corresponding to the power system based on the power angle difference index corresponding to the multiple generator sets.

[0012] According to one aspect of the present invention, an apparatus for determining the impact parameters of transient stability of a power system is provided, comprising: an acquisition module for acquiring an operating scenario of the power system; a first determination module for determining a transient stability index and a plurality of target power parameters corresponding to the power system under the operating scenario, wherein the transient stability index is used to represent the degree of operational stability of the power system after a disturbance, and the plurality of target power parameters are parameters used to reflect the operating state of the power system; a second determination module for determining a first causal relationship and a second causal relationship corresponding to the power system based on the plurality of target power parameters and the transient stability index, wherein the first causal relationship is used to represent the causal association relationship between any two of the plurality of target power parameters, and the second causal relationship is used to represent the causal association relationship between the plurality of target power parameters and the transient stability index respectively; a third determination module for determining a plurality of influence chains based on the plurality of target power parameters, the first causal relationship, and the second causal relationship; and a fourth determination module for determining transient stability impact parameters corresponding to the power system based on the plurality of influence chains.

[0013] According to one aspect of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the method for determining the influence parameters of power system transient stability as described in any of the preceding claims.

[0014] According to one aspect of the present invention, a computer-readable storage medium is provided, wherein when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the method for determining the influence parameters of transient stability of a power system as described above.

[0015] In this embodiment of the invention, the following steps are taken: First, an operating scenario of the power system is obtained. Then, a transient stability index and multiple target power parameters corresponding to the power system are determined within the operating scenario. The transient stability index represents the degree of stability of the power system after a disturbance, and the multiple target power parameters reflect the operating state of the power system. Based on the multiple target power parameters and the transient stability index, a first causal relationship and a second causal relationship corresponding to the power system are determined. The first causal relationship represents the causal association between any two of the multiple target power parameters, and the second causal relationship represents the causal association between each of the multiple target power parameters and the transient stability index. Based on the multiple target power parameters, the first causal relationship, and the second causal relationship, multiple influence chains are determined. Finally, based on the multiple influence chains, transient stability influence parameters corresponding to the power system are determined. Based on the operating scenarios of power systems, the transient stability index and multiple target power parameters are determined, which can accurately capture the stability and operating state of the power system after disturbances. By clarifying the primary and secondary causal relationships based on these parameters and indices, the intrinsic correlation between target power parameters and the direct correlation between parameters and the stability index can be clarified. By constructing multiple influence chains based on these causal relationships, the transmission path from parameters to the stability index can be systematically sorted out. Based on multiple influence chains, the transient stability influence parameters of the power system can be accurately determined, so as to achieve precise control of the operation of the power system. This solves the technical problem in related technologies where the influence parameters of power system transient stability are not accurately determined. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0017] Figure 1 This is a flowchart of a method for determining the influence parameters of power system transient stability according to an embodiment of the present invention;

[0018] Figure 2 This is a schematic flowchart of the method for determining the influence parameters of power system transient stability in an optional embodiment of the present invention;

[0019] Figure 3 This is a schematic diagram of the system structure for determining the influence parameters of power system transient stability in an optional embodiment of the present invention;

[0020] Figure 4 This is a structural block diagram of a device for determining the influence parameters of transient stability of a power system according to an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0023] First, some nouns or terms that appear in the description of the embodiments of this application shall be interpreted as follows:

[0024] Adam Algorithm: The Adam algorithm is a gradient descent-based optimization algorithm used for training deep learning models. It dynamically adjusts the learning rate of each parameter by combining the advantages of momentum and adaptive learning rate.

[0025] SHAP method: The SHAP method is a game theory-based model interpretation technique used to interpret the output of machine learning models.

[0026] SHAP value: The SHAP value is used to interpret the prediction results of the algorithm model and helps to understand the contribution of each feature to the model output.

[0027] Maximum Correlation-Minimum Redundancy Method: The maximum correlation-minimum redundancy method is a feature selection method whose core objective is to select a subset of features from high-dimensional data that are highly correlated with the target variable and have the lowest redundancy among them.

[0028] Peter-Clark Algorithm: The Peter-Clark (PC) algorithm is a constraint-based causal discovery algorithm. Its core is to infer the causal structure between variables from observed data through conditional independence tests.

[0029] Example 1

[0030] According to an embodiment of the present invention, an embodiment of a method for determining the influence parameters of transient stability of a power system is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0031] Figure 1 This is a flowchart of a method for determining the influence parameters of power system transient stability according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0032] S102, Obtain the operating scenario of the power system;

[0033] This involves the power system, which is a system related to power operation and includes at least one of the following components: power generation, power transmission, power transformation, power distribution, and power consumption, which are respectively used for the production, transmission, distribution, and use of electrical energy.

[0034] This includes operational scenarios, which reflect the operating status and environmental conditions of the power system, including but not limited to load levels, power generation output, network topology, equipment status, and weather conditions, describing the specific situations in which the power system operates in practice.

[0035] Obtaining the operating scenarios of the power system helps to understand the specific situations in actual operation of the power system, providing an analytical basis for subsequent transient stability analysis.

[0036] S104, determine the transient stability index and multiple target power parameters corresponding to the power system under the operating scenario. The transient stability index is used to represent the degree of stability of the power system after a disturbance, and the multiple target power parameters are parameters used to reflect the operating state of the power system.

[0037] This includes the transient stability index, which quantifies the ability or extent to which a power system can recover to a stable operating state after being subjected to external disturbances (such as short circuits or fault clearing). Transient stability (also known as power system transient stability) refers to the ability of a power system to maintain synchronous operation and transition to a new steady state or restore its original steady-state operation after experiencing large disturbances such as short-circuit faults or the disconnection of transmission lines. Transient stability index It can be determined using the following formula:

[0038]

[0039] in, This represents the maximum absolute value of the power angle difference between any two generators in the k-th operating scenario.

[0040] This involves multiple target power parameters, which are power parameters used to reflect the specific operating status of the power system under different operating scenarios. These target power parameters include generator active power, reactive power, generator terminal voltage, load active power, reactive power, and line transmission active power, reactive power, etc.

[0041] Since the transient stability index of the power system under the determined operating scenario can quantitatively assess the system's stability recovery capability after being disturbed, and the selected multiple target power parameters (such as generator active power, reactive power, etc.) can comprehensively reflect the real-time operating status of the system, it can accurately capture the dynamic characteristics of the system and identify key influencing factors.

[0042] S106. Based on multiple target power parameters and transient stability index, determine the first causal relationship and the second causal relationship corresponding to the power system. The first causal relationship is used to represent the causal relationship between any two target power parameters among the multiple target power parameters, and the second causal relationship is used to represent the causal relationship between the multiple target power parameters and the transient stability index respectively.

[0043] This involves the first causal relationship, which is the causal relationship between any two parameters among multiple target power parameters, that is, how the change of one power parameter affects the change of another power parameter.

[0044] This involves a second causal relationship, which is the causal relationship between multiple target power parameters and the transient stability index, that is, how the changes of each power parameter affect the degree of stability recovery of the power system after a disturbance.

[0045] This involves causal relationships, which are causal connections between two or more variables, meaning that a change in one variable is the cause or result of a change in another variable.

[0046] By clarifying the primary causal relationship between multiple target power parameters, the interaction mechanism between parameters can be revealed. At the same time, determining the secondary causal relationship between parameters and transient stability index can quantify the degree of influence of each parameter on system stability. Therefore, a complete parameter influence path can be constructed to ensure the accurate identification of key influencing factors and the traceability of influence chains in transient stability analysis.

[0047] S108, based on multiple target power parameters, a first causal relationship, and a second causal relationship, determine multiple influence chains;

[0048] This involves multiple influence chains, which describe how target electrical parameters ultimately affect the transient stability index through interactions. For any one of these influence chains, starting with the corresponding target electrical parameter and ending with the transient stability index, a directed connection (corresponding to causal relationships) clearly shows the complete process by which one or more target electrical parameters, through interactions and layer-by-layer transmission, ultimately affect the transient stability index.

[0049] By clarifying the primary causal relationship between multiple target power parameters and the secondary causal relationship between the parameters and the transient stability index, a directed transmission path from parameters to stability can be constructed. Therefore, it is possible to systematically sort out the layer-by-layer influence mechanism of parameter changes, ensure accurate identification of key transmission paths, and quantify the degree of influence of each link on system stability.

[0050] Specifically, a causal directed acyclic graph (DAG) can be constructed based on multiple target power parameters, a first causal relationship, and a second causal relationship. Multiple influence chains can then be determined based on the DAG. This DAG is used to represent the causal relationship between features and transient stability quantification indicators. The construction of the DAG can be implemented based on the Peter-Clark algorithm.

[0051] S110, based on multiple influence chains, determines the transient stability influence parameters corresponding to the power system.

[0052] This involves transient stability impact parameters, which are parameters that reflect and affect the transient stability of the power system. These include the transient stability assessment results of the power system, the impact characteristics of each impact chain on the transient stability of the power system, and the target power parameters that have a significant impact on the transient stability of the power system. These parameters are used to reflect which target power parameters can improve the operational stability of the power system by adjusting them.

[0053] Because multiple influence chains can be used to trace the transmission path from target power parameters to transient stability indices, clarify the causal relationships between parameters, quantify the comprehensive impact of each target power parameter on transient stability, and identify the specific characteristics of its impact on transient stability, key target power parameters that have a significant impact on the transient stability of the power system can be accurately identified. This avoids the problems of inaccurate analysis or lack of focus in regulation caused by a large number of parameters, thereby helping to achieve effective assessment and precise regulation of the transient stability of the power system and improve the system's operational stability and reliability.

[0054] Specifically, it can be determined through a neural network prediction model. This neural network prediction model is a multilayer perceptron neural network with a residual structure, including an input layer, three residual blocks, and an output layer, wherein the input layer has a dimension of [missing information]. Each residual block includes a primary channel and extended channels, with the primary channel having dimensions of [dimensions to be filled in]. The extended channel dimensions are respectively Each residual block consists of a main channel (containing a fully connected layer, a normalization layer, and a non-linear activation function) and a shortcut channel. The fully connected main channel layers map the main channel dimension to the extended channel dimension and then back to the main channel dimension, maintaining dimensionality consistency. The shortcut channel directly adds the main channel input to the output. For example... =60, , .

[0055] Furthermore, the neural network prediction model can be trained using a sample dataset, which includes a training set, a validation set, and a test set. Specifically, the training process involves using mean squared error loss as the loss function, employing the Adam algorithm to iteratively update the model parameters of the prediction model, and calculating the loss of the prediction model on the validation set in each iteration. When the loss on the validation set does not decrease within 50 consecutive iterations, the iteration process is stopped, and the prediction model with the highest accuracy on the validation set is selected as the final transiently stable evaluation artificial intelligence model. This completes the training of the neural network prediction model.

[0056] Through the above steps S102-S110, the operating scenario of the power system is obtained; under the operating scenario, the transient stability index and multiple target power parameters corresponding to the power system are determined, wherein the transient stability index is used to represent the degree of stability of the power system after a disturbance, and the multiple target power parameters are parameters used to reflect the operating state of the power system; based on the multiple target power parameters and the transient stability index, a first causal relationship and a second causal relationship corresponding to the power system are determined, wherein the first causal relationship is used to represent the causal relationship between any two target power parameters, and the second causal relationship is used to represent the causal relationship between each of the multiple target power parameters and the transient stability index; based on the multiple target power parameters, the first causal relationship, and the second causal relationship, multiple influence chains are determined; based on the multiple influence chains, the transient stability influence parameters corresponding to the power system are determined. Based on the operating scenarios of power systems, the transient stability index and multiple target power parameters are determined, which can accurately capture the stability and operating state of the power system after disturbances. By clarifying the primary and secondary causal relationships based on these parameters and indices, the intrinsic correlation between target power parameters and the direct correlation between parameters and the stability index can be clarified. By constructing multiple influence chains based on these causal relationships, the transmission path from parameters to the stability index can be systematically sorted out. Based on multiple influence chains, the transient stability influence parameters of the power system can be accurately determined, so as to achieve precise control of the operation of the power system. This solves the technical problem in related technologies where the influence parameters of power system transient stability are not accurately determined.

[0057] As an optional embodiment, determining multiple target power parameters corresponding to the power system under an operating scenario includes: determining multiple initial power parameters corresponding to the power system under the operating scenario; determining a first correlation index corresponding to each of the multiple initial power parameters, and multiple second correlation indices corresponding to each of the multiple initial power parameters, wherein the corresponding first correlation index is used to represent the degree of correlation between the corresponding initial power parameter and the transient stability index, and the corresponding multiple second correlation indices are used to represent the degree of correlation between the corresponding initial power parameter and other initial power parameters respectively; and determining multiple target power parameters corresponding to the power system based on the first correlation index corresponding to each of the multiple initial power parameters and the multiple second correlation indices corresponding to each of the multiple initial power parameters.

[0058] This involves multiple initial power parameters, which are power parameters that are initially selected in the power system operation scenario analysis and can reflect the operating status of the power system.

[0059] This involves the first correlation index, which is used to quantify the degree of correlation between the corresponding initial power parameters and the transient stability index. The larger the value of the first correlation index, the more significant the impact of the initial power parameters on the transient stability of the power system, and it is an important basis for screening target parameters.

[0060] This involves multiple second correlation indices, which are used to quantify the degree of correlation between the corresponding initial power parameter and other initial power parameters. Each initial power parameter corresponds to a set of second correlation indices, reflecting the strength of its interaction with other initial power parameters in the power system, and is used to identify the coupling relationship between parameters.

[0061] By identifying multiple initial power parameters reflecting the operating state of the power system under the operating scenario, basic data support is provided for subsequent parameter selection. Then, by calculating the first correlation index corresponding to each initial power parameter, the correlation strength with the transient stability index can be quantified, clarifying the significance of each parameter's impact on transient stability. Simultaneously, calculating multiple second correlation indices corresponding to each initial power parameter can capture the coupling relationship and interaction strength between parameters. Based on these correlation indices, target power parameter selection prioritizes key parameters that significantly impact transient stability and have low redundancy with other parameters, avoiding interference from irrelevant or highly correlated parameters. This achieves accurate positioning of key parameters affecting power system transient stability, providing a reliable basis for subsequent targeted optimization of system stability.

[0062] The first correlation index corresponding to each of the multiple initial power parameters can be determined in the following way:

[0063]

[0064] in, The first correlation index; Let be the i-th initial power parameter; Y be the transient stability index; and d be the total number of initial power parameters.

[0065] The multiple second correlation indices corresponding to the multiple initial power parameters can be determined in the following way:

[0066]

[0067] in, For the i-th initial power parameter With the j-th initial power parameter The second correlation index between them.

[0068] As an optional embodiment, multiple target power parameters corresponding to the power system are determined based on a first correlation index corresponding to multiple initial power parameters and multiple second correlation indices corresponding to the multiple initial power parameters. This includes: determining a first candidate parameter from the multiple initial power parameters, wherein the first candidate parameter is the initial power parameter with the largest first correlation index among the multiple initial power parameters; and, according to the execution order of multiple screenings of the multiple initial power parameters, selecting from the multiple first remaining parameters based on the first correlation index corresponding to the multiple first remaining parameters and the second correlation index between the multiple first remaining parameters and the first candidate parameter. In the process, a second candidate parameter is determined, wherein multiple first residual parameters are other initial power parameters besides the first candidate parameter among multiple initial power parameters; based on the first correlation index corresponding to each of the multiple second residual parameters, and the second correlation index between each of the multiple second residual parameters and the second candidate parameter, a third candidate parameter is determined from the multiple second residual parameters, wherein multiple second residual parameters are other initial power parameters besides the second candidate parameter among multiple first residual parameters, until multiple screening operations are completed, resulting in multiple candidate parameters corresponding to the power system; based on the multiple candidate parameters, multiple target power parameters corresponding to the power system are determined.

[0069] This involves a first candidate parameter, which is the initial power parameter selected from all initial power parameters that has the largest first correlation index value (reflecting the greatest impact on transient stability). In other words, it is the first power parameter that has the most significant impact on the transient stability of the power system, and it is the first target power parameter selected (which is the starting point for the selection of target parameters).

[0070] This involves the execution order of multiple screening processes. This order involves screening multiple initial power parameters in sequence to achieve a progressive selection of these parameters. Specifically, the execution order can be as follows: first, screen the initial power parameter with the highest first correlation index; then, sequentially screen the remaining parameters that have a weak correlation with the selected initial power parameter but a high first correlation index, gradually narrowing the screening range to ensure that each screening is based on the current optimal parameter combination.

[0071] This involves multiple first residual parameters, which are the remaining initial power parameters after excluding the first candidate parameters from all initial power parameters. These first residual parameters are the candidate parameter set for the second round of screening.

[0072] This involves a second candidate parameter, which is a parameter selected from multiple first residual parameters based on the first correlation index of each of the multiple first residual parameters and the second correlation index between each first residual parameter and the first candidate parameter. This second candidate parameter has a significant impact on transient stability and avoids redundancy with the first candidate parameter.

[0073] This involves multiple second residual parameters, which are the remaining initial power parameters after excluding the second candidate parameters from the multiple first residual parameters. These are the candidate parameter set for the third round of screening.

[0074] This involves a third candidate parameter, which is a parameter selected from multiple second remaining parameters based on the first correlation index of each of the multiple second remaining parameters and the second correlation index between each second remaining parameter and the second candidate parameter, continuing the screening logic of high impact and low redundancy.

[0075] This involves multiple candidate parameters, which are the first, second, and third candidate parameters determined sequentially after multiple rounds of screening, as well as all candidate parameters selected in subsequent rounds of screening. This is the basic set for finally determining the target power parameters.

[0076] First, the first candidate parameter with the largest first correlation index is selected from multiple initial power parameters. This identifies the core parameter that has the most significant impact on the transient stability of the power system, providing a key parameter basis for subsequent selection. Then, following the execution order of multiple selections, the second candidate parameter is determined based on the first correlation index of each of the multiple remaining first parameters and the second correlation index with the first candidate parameter. This retains parameters with high impact on transient stability while avoiding redundancy with the selected parameters. This process is repeated to gradually select key initial power parameters and eliminate irrelevant or redundant initial power parameters. This ensures that the selected target power parameters have a significant impact on the transient stability of the power system while avoiding redundancy between parameters, thus improving the accuracy of subsequent stability analysis and control.

[0077] Specifically, based on the first correlation index corresponding to each of the multiple first residual parameters, and the second correlation index between each of the multiple first residual parameters and the first candidate parameter, the second candidate parameter is determined from the multiple first residual parameters. This can be determined using the maximum correlation-minimum redundancy (mRMR) score, as shown in the following formula:

[0078]

[0079]

[0080] in, The mRMR score for the j-th initial electrical parameter. This represents the newly added initial power parameter with the highest score (i.e., the first residual parameter with the largest first correlation index and the smallest second correlation index). This is the selected feature set (i.e., the selected set of initial power parameters). This is the remaining feature set (i.e., the remaining set of initial electrical parameters).

[0081] The same methods can be used for subsequent screening, so I will not repeat them here.

[0082] As an optional embodiment, multiple target power parameters corresponding to the power system are determined based on multiple candidate parameters, including: when the multiple candidate parameters include active power and reactive power, determining the power ratio corresponding to the power system based on active power and reactive power; and determining multiple target power parameters corresponding to the power system based on the power ratio and the multiple candidate parameters.

[0083] This includes active power, which is the power in the power system that actually does work and is converted into other forms of energy (such as mechanical energy and thermal energy). It is an electrical parameter that maintains the normal operation of power equipment and ensures the effective use of electrical energy.

[0084] This includes reactive power, which is the power in a power system used to establish and maintain electric and magnetic fields and ensure the normal operation of power equipment, but does not perform external work. It is mainly used for energy exchange in electromagnetic equipment such as transformers and motors, and is an important parameter for maintaining voltage stability and ensuring power quality in the power system.

[0085] This involves the power ratio, which is the ratio between corresponding powers in a power system, including the power ratio of generators, the power ratio of loads, and the power ratio of lines. This power ratio includes the ratio of the active power of the corresponding equipment (such as generators, loads, and lines) to the sum of all active power, the ratio of reactive power to the sum of all reactive power, and the ratio of active power to reactive power.

[0086] Since multiple candidate parameters include active power and reactive power, determining the power ratio corresponding to the power system based on active power and reactive power can further quantify the relative relationship between power in the power system. This can further reflect the operating status of equipment and the stability characteristics of the system. By combining the power ratio with the candidate parameters to comprehensively screen target power parameters, the energy distribution and operation of the power system can be more comprehensively reflected. This will help to achieve a comprehensive and accurate positioning of key control parameters for transient stability of the power system.

[0087] Specifically, active power includes the active power of generators (including the active power of each generator and the sum of the active power of all generators), the active power of loads (including the active power of each load and the sum of the active power of all loads), and the active power of lines; reactive power includes the reactive power of loads (including the reactive power of each load and the sum of the reactive power of all loads) and the reactive power of lines. Therefore, the power ratio includes the power ratio of generators, the power ratio of loads, and the power ratio of lines.

[0088] As an optional embodiment, determining the transient stability impact parameters corresponding to the power system based on multiple impact chains includes: determining multiple target nodes corresponding to any one of the multiple impact chains according to the execution order of the multiple impact chains, wherein the multiple target nodes include first nodes corresponding to multiple target power parameters and second nodes corresponding to transient stability indices; determining parent nodes corresponding to each of the multiple target nodes; determining sub-impact parameters corresponding to any one impact chain based on the target node data corresponding to each of the multiple target nodes and the parent node data corresponding to the parent nodes corresponding to each of the multiple target nodes, wherein the sub-impact parameters are used to reflect the transient stability parameters of the power system under any one impact chain; until the multiple impact chains are executed, obtaining the sub-impact parameters corresponding to each of the multiple impact chains; and determining the transient stability impact parameters corresponding to the power system based on the sub-impact parameters corresponding to each of the multiple impact chains.

[0089] This involves the execution order of multiple impact chains. The execution order of these multiple impact chains represents the sequential logic of analyzing and processing multiple transient stability impact chains in the power system one by one, ensuring that all impact chains are fully analyzed.

[0090] This involves any one of multiple influence chains.

[0091] This involves multiple target nodes, which are the key nodes that need to be analyzed in any impact chain. These include the first node reflecting changes in power parameters and the second node reflecting transient stability, which together constitute the core of the impact chain assessment.

[0092] This involves the first node, which is the node corresponding to the target power parameter among multiple target nodes, and is the starting point or intermediate carrier for the transmission of causal relationships.

[0093] This involves a second node, which is the node corresponding to the transient stability index among multiple target nodes and is the endpoint of causal relationship transmission.

[0094] This involves parent node data, which consists of specific numerical values ​​or characteristic information corresponding to the parent node of the target node in the influence chain. The parent node is the preceding node that has a direct causal influence on the target node, and its data is an important basis for calculating the sub-influence parameters.

[0095] This involves sub-influence parameters, which are calculated based on the target node data and the corresponding parent node data in any influence chain. These parameters are used to quantify the specific influence characteristics of power parameters on transient stability under that influence chain.

[0096] By analyzing each influence chain sequentially according to its execution order, and identifying multiple target nodes, including the first node corresponding to the target power parameter and the second node corresponding to the transient stability index, we can focus on the core causal elements in the influence chain and clarify the key carriers for parameter transmission and stability assessment. Further identifying the parent nodes corresponding to multiple target nodes allows us to identify the direct causal relationships between each core node, providing a logical basis for quantifying influence characteristics. Calculating the sub-influence parameters of any influence chain based on target node and parent node data accurately captures the specific effects of power parameters on transient stability within that chain, avoiding the limitations of single-link analysis. After analyzing all influence chains and summarizing all sub-influence parameters, we can integrate the influence patterns of multiple links, comprehensively covering the key influence dimensions of power system transient stability, achieving a systematic and accurate determination of transient stability influence parameters, and providing a reliable causal basis for subsequent targeted optimization of system stability.

[0097] Specifically, determining the prediction data corresponding to each of the multiple target nodes based on the parent node data of the parent nodes corresponding to the target nodes can be achieved in the following way:

[0098] For any target node, construct a mapping function that takes all parent nodes of that target node as input and the target node itself as output. , is represented as:

[0099]

[0100] in, For predictive data; A node in a DAG; for The parent node, that is, the node pointing to All nodes.

[0101] As an optional embodiment, based on the target node data corresponding to each of the multiple target nodes and the parent node data corresponding to the parent nodes corresponding to each of the multiple target nodes, the sub-influence parameter corresponding to any influence chain is determined, including: determining the predicted data corresponding to each of the multiple target nodes based on the parent node data corresponding to the parent nodes corresponding to each of the multiple target nodes; determining the data deviation value corresponding to each of the multiple target nodes based on the target node data and the predicted data; and determining the sub-influence parameter corresponding to any influence chain based on the data deviation value corresponding to each of the multiple target nodes and the parent node data corresponding to the parent nodes corresponding to each of the multiple target nodes.

[0102] This involves prediction data, which is an estimated value of the target node based on the parent node data of the parent nodes of multiple target nodes. It is used to reflect the theoretical expected state of the target node under the influence of the parent node data.

[0103] This includes target node data, which consists of real values ​​or measured characteristic information recorded by multiple target nodes in the actual operation scenario of the power system. It is the basic data reflecting the actual state of the target nodes.

[0104] This includes the data deviation value, which is the difference between the target node data and the predicted data.

[0105] By determining the predicted data based on the parent node data of the respective parent nodes of multiple target nodes, the theoretical expected state of the target nodes can be obtained through the causal relationship of the parent nodes, providing a benchmark for measuring the actual state. By combining the actual measured data of the target nodes with the predicted data to calculate the data deviation value, the difference between the theoretical expectation and the actual operation can be accurately captured, reflecting fluctuations or disturbances not covered by the parent node data. Then, based on these data deviation values ​​and the parent node data, the sub-influence parameters can be determined. This integrates the theoretical influence of causal relationships with the deviation characteristics of actual operation, comprehensively and accurately quantifying the effect of any influence chain on the transient stability of the power system, providing reliable underlying support for subsequently summarizing the influence laws of multiple influence chains and determining the final transient stability influence parameters.

[0106] As an optional embodiment, determining the transient stability index corresponding to the power system under the operating scenario includes: determining the power angle difference index corresponding to multiple generator sets under the operating scenario, wherein the corresponding power angle difference index represents the degree of difference in power angle between two generators included in the corresponding generator set, and the power system includes multiple generator sets; and determining the transient stability index corresponding to the power system based on the power angle difference index corresponding to the multiple generator sets.

[0107] Among them, the power angle difference index is involved. This power angle difference index is used to quantify the degree of power angle difference between two generators in a single generator set. The power angle reflects the balance state of the electromagnetic torque and mechanical torque of the generator rotor. This power angle difference index directly reflects the synchronous operation consistency of the generators in the generator set. The larger the difference, the worse the synchronization of the unit, and the more likely it is to affect the transient stability of the power system.

[0108] This involves multiple generator sets, which are sets of power generation units formed by combining all the generators in the power system in pairs. For example, if the power system includes generator 1, generator 2, and generator 3, then the multiple generator sets include: (generator 1, generator 2), (generator 2, generator 3), and (generator 1, generator 3).

[0109] This involves generators, which are devices used to convert mechanical energy (such as heat, wind power, etc.) into electrical energy.

[0110] By determining the power angle difference index of multiple generator sets under the operating scenario, the synchronous operation consistency between any two generators can be quantified. This helps to accurately capture the potential impact of power angle differences within generator sets on the transient stability of the power system. The larger the difference, the worse the synchronization and the higher the stability risk. Furthermore, by comprehensively determining the transient stability index of the power system based on the power angle difference index of all generator sets, the synchronous operation status information of the generators in the power system can be fully quantified. This provides a data foundation for achieving a comprehensive and accurate quantification of the transient stability level of the power system.

[0111] Based on the above embodiments and optional embodiments, an optional implementation method is provided, which is described in detail below.

[0112] In related technologies, power system transient stability measures the ability of a power system to transition to a new steady state or return to its original steady-state operation after being disturbed. Determining the influencing parameters of power system transient stability provides crucial information for designing protection and control strategies, optimizing system structure, and adjusting operating modes. However, in these technologies, there is a technical problem of inaccurate determination of these influencing parameters.

[0113] There is currently no effective solution to the above problems.

[0114] In view of this, an optional embodiment of the present invention provides a method for determining the influence parameters of transient stability of a power system, which can effectively solve the above-mentioned technical problems.

[0115] Figure 2This is a schematic flowchart of a method for determining the influence parameters of power system transient stability in an optional embodiment of the present invention, as shown below. Figure 2 As shown, a detailed description follows.

[0116] S1, Obtain the operating scenario of the power system;

[0117] Specifically, it can be determined based on the historical operating conditions of the power system and load forecasts.

[0118] S2, determine the transient stability index and multiple target power parameters corresponding to the power system under the operating scenario. The transient stability index is used to represent the degree of stability of the power system after a disturbance, and the multiple target power parameters are parameters used to reflect the operating state of the power system.

[0119] Optionally, determining the transient stability index corresponding to the power system under the operating scenario includes: determining the power angle difference index corresponding to multiple generator sets under the operating scenario, wherein the corresponding power angle difference index represents the degree of difference in power angle between two generators included in the corresponding generator set, and the power system includes multiple generator sets; and determining the transient stability index corresponding to the power system based on the power angle difference index corresponding to the multiple generator sets.

[0120] Specifically, determining the power angle difference index corresponding to multiple generator sets under the operating scenario includes: identifying multiple operating characteristics and preset faults corresponding to the power system, where the multiple operating characteristics include the active power, reactive power, and terminal voltage of the generators, the active power and reactive power of the loads, and the active power and reactive power transmitted through the lines; simulating the power system based on the multiple operating characteristics and preset faults to obtain the power angle difference index corresponding to each generator set; and determining the transient stability index corresponding to the power system based on the largest power angle difference index among the power angle difference indices corresponding to each generator set. The power system simulation can be based on power flow calculations and transient stability simulations. The transient stability index includes transient stability quantification indicators under the operating scenario. Taking a power system with N generators and M loads as an example, it includes:

[0121] Based on the historical operation and load forecast of the power system, K operating scenarios are set (e.g., using a 36-node system, with N=8 generators, M=10 loads, L=33 lines, and K set to 30000). Initial features under the K operating scenarios are obtained, resulting in a feature set F, including steady-state features (i.e., multiple operating features) and preset faults (i.e., multiple preset faults). The steady-state features include: the active power, reactive power, and terminal voltage of N generators; the active power and reactive power of M loads; and the active power and reactive power transmitted by L lines. There are a total of 110 initial features (i.e., 110 initial power parameters), and the preset faults include 33 types of faults.

[0122] To address this, based on multiple operating characteristics and multiple preset faults, the power system is simulated to obtain the power angle difference index corresponding to multiple generator units. This can be used to simulate the transient stability of the power system under K operating scenarios after a preset fault occurs, based on power flow calculations and transient stability simulation calculations. The transient stability quantification index is calculated based on the maximum relative power angle difference between generators during the simulation period. The transient stability quantification index under the k-th operating scenario is... :

[0123]

[0124] in, It represents the maximum absolute value of the power angle difference between any two generators in the k-th operating scenario (i.e., the maximum power angle difference index).

[0125] Optionally, determining multiple target power parameters corresponding to the power system under the operating scenario includes: determining multiple initial power parameters corresponding to the power system under the operating scenario; determining a first correlation index corresponding to each of the multiple initial power parameters, and multiple second correlation indices corresponding to each of the multiple initial power parameters, wherein the corresponding first correlation index is used to represent the degree of correlation between the corresponding initial power parameter and the transient stability index, and the corresponding multiple second correlation indices are used to represent the degree of correlation between the corresponding initial power parameter and other initial power parameters respectively; and determining multiple target power parameters corresponding to the power system based on the first correlation index corresponding to each of the multiple initial power parameters and the multiple second correlation indices corresponding to each of the multiple initial power parameters.

[0126] Taking a power system with N generators and M loads as an example, we will quantify the feature set F and the transient stability index under the operating scenario. Merge them to obtain a labeled dataset D, which contains K samples and d features (i.e., multiple initial power parameters).

[0127] The first correlation index corresponding to each of the multiple initial power parameters can be determined in the following way:

[0128] For dataset D, calculate the mutual information between all d features and the target variable (i.e., the transient stability index). (That is, the first correlation index). Wherein, the i-th feature... (That is, the i-th initial power parameter) and the objective variable Y=( ,…, ,…, Mutual information between them is represented as:

[0129]

[0130] The multiple second correlation indices corresponding to the multiple initial power parameters can be determined in the following way:

[0131] For dataset D, calculate the mutual information between all feature pairs. (That is, the second correlation index), where the i-th feature With the j-th feature The mutual information between target variables is denoted as:

[0132]

[0133] The above steps can be implemented based on the maximum correlation-minimum redundancy (mRMR) method.

[0134] Optionally, based on the first correlation index corresponding to the multiple initial power parameters and the multiple second correlation indices corresponding to the multiple initial power parameters, multiple target power parameters corresponding to the power system are determined, including: determining a first candidate parameter from the multiple initial power parameters, wherein the first candidate parameter is the initial power parameter with the largest first correlation index among the multiple initial power parameters; and determining, according to the execution order of multiple screenings of the multiple initial power parameters, a first candidate parameter from the multiple first remaining parameters based on the first correlation index corresponding to the multiple first remaining parameters and the second correlation index between the multiple first remaining parameters and the first candidate parameter. The second candidate parameter is defined as follows: multiple first residual parameters are other initial power parameters besides the first candidate parameter among multiple initial power parameters; based on the first correlation index corresponding to each of the multiple second residual parameters, and the second correlation index between each of the multiple second residual parameters and the second candidate parameter, a third target power parameter is determined from the multiple second residual parameters, wherein the multiple second residual parameters are other initial power parameters besides the second candidate parameter among multiple first residual parameters, until multiple screening operations are completed, resulting in multiple candidate parameters corresponding to the power system; based on the multiple candidate parameters, multiple target power parameters corresponding to the power system are determined.

[0135] Taking a power system with N generators and M loads as an example, the feature with the highest mutual information with the target variable is selected as the first selected feature, and the selected feature set is updated. The formula is:

[0136]

[0137] in, The feature with the greatest mutual information with the target variable is selected as the first feature, i.e., the first candidate parameter.

[0138]

[0139] in, The selected feature set; This is the remaining feature set.

[0140] The selected feature set includes all candidate parameters determined in each of the multiple screening processes performed on multiple initial power parameters, such as the first candidate parameter, the second candidate parameter, etc., until all screening processes are completed. The remaining feature set includes the other initial power parameters besides the candidate parameters in each of the multiple screening processes performed on multiple initial power parameters. For example, after the first screening, the remaining feature set includes multiple first remaining parameters; after the second screening, the remaining feature set includes multiple second remaining parameters.

[0141] Based on the first correlation index corresponding to each of the multiple first residual parameters, and the second correlation index between each of the multiple first residual parameters and the first candidate parameter, the second candidate parameter is determined from the multiple first residual parameters. This can be achieved as follows: Based on the first correlation index corresponding to each of the multiple first residual parameters, and the second correlation index between each of the multiple first residual parameters and the first candidate parameter, the first residual parameter with the largest first correlation index and the smallest second correlation index is determined as the second candidate parameter. It should be noted that the process of performing multiple screenings for multiple initial power parameters is a cyclical process, and the above method can be used for each screening, which will not be elaborated further.

[0142] For example, for a loop process that performs multiple screenings for multiple initial power parameters, let the loop number w (i.e., the number of screenings) be set, and the following process be repeated w times for the remaining features:

[0143] For each feature in the remaining feature set, calculate the maximum relevance-minimum redundancy (mRMR) score for each feature in turn. Select the feature with the highest mRMR score (i.e., the first remaining parameter with the largest first correlation index and the smallest second correlation index) and add it to the selected feature set. The formula is as follows:

[0144]

[0145]

[0146] in, For mRMR score, This represents the newly added feature with the highest score (i.e., the first residual parameter with the largest first correlation index and the smallest second correlation index).

[0147] For example, the value of w can be set to 50, meaning that 50 features will be selected as the feature set for subsequent analysis.

[0148] Optionally, based on multiple candidate parameters, multiple target power parameters corresponding to the power system are determined, including: when the multiple candidate parameters include active power and reactive power, determining the power ratio corresponding to the power system based on active power and reactive power; and determining multiple target power parameters corresponding to the power system based on the power ratio and the multiple candidate parameters.

[0149] Specifically, active power includes the active power of generators (including the active power of each generator and the sum of the active power of all generators), the active power of loads (including the active power of each load and the sum of the active power of all loads), and the active power of lines; reactive power includes the reactive power of loads (including the reactive power of each load and the sum of the reactive power of all loads) and the reactive power of lines. Therefore, the power ratio includes the power ratio of generators, the power ratio of loads, and the power ratio of lines.

[0150] The power ratio of a generator can be determined using the following formula:

[0151]

[0152] in, The power ratio of the generator can be expressed as the output percentage of different generators; The active power of the t-th generator can be represented by the active power output of the generator under the feature index t; The sum of the active power of all generators can be expressed as the sum of the output of all generators.

[0153] The power ratio of a load includes the active power ratio and the reactive power ratio.

[0154] The active power ratio of the load can be determined in the following way:

[0155]

[0156] in, The active power ratio of the load can be expressed as the active power ratio of different loads; Let be the active power of the t-th load, that is, the active power of the load with characteristic index t; It is the active power of all loads, that is, the sum of the active power of all loads.

[0157] The reactive power ratio of a load can be determined in the following way:

[0158]

[0159] in, The active power ratio of the load can be expressed as the reactive power ratio of different loads; Let t be the reactive power of the t-th load, that is, the reactive power of the load with characteristic index t. It represents the reactive power of all loads, that is, the sum of the reactive power of all loads.

[0160] The power ratio of a line can be determined using the following formula:

[0161]

[0162] in, The power ratio of the line can be expressed as the ratio of active to reactive power. The active power of the line can be represented by the active power flow of the line with the characteristic index t. The reactive power of the line can be represented by the reactive power flow of the line with the characteristic index t.

[0163] Furthermore, based on the power ratio, several new combined characteristics are obtained: the power ratio of the generator, the power ratio of the load, and the power ratio of the line.

[0164] S3. Based on multiple target power parameters and transient stability index, determine the first causal relationship and the second causal relationship corresponding to the power system. The first causal relationship is used to represent the causal relationship between any two target power parameters among the multiple target power parameters, and the second causal relationship is used to represent the causal relationship between the multiple target power parameters and the transient stability index respectively.

[0165] S4. Based on multiple target power parameters, a first causal relationship, and a second causal relationship, multiple influence chains are determined;

[0166] Specifically, based on multiple target power parameters, a first causal relationship, and a second causal relationship, a causal directed acyclic graph (DAG) is constructed. Multiple influence chains are then determined based on the DAG. This DAG is used to represent the causal relationship between features and transient stability quantification indicators. The construction of the DAG can be implemented using the Peter-Clark algorithm.

[0167] S5 determines the transient stability impact parameters corresponding to the power system based on multiple influence chains.

[0168] Optionally, based on multiple influence chains, the transient stability impact parameters corresponding to the power system are determined, including: according to the execution order of the multiple influence chains, for any one of the multiple influence chains, determining multiple target nodes corresponding to any one influence chain, wherein the multiple target nodes include first nodes corresponding to multiple target power parameters and second nodes corresponding to transient stability indices; determining parent nodes corresponding to each of the multiple target nodes; based on the target node data corresponding to each of the multiple target nodes and the parent node data corresponding to the parent nodes corresponding to each of the multiple target nodes, determining sub-influence parameters corresponding to any one influence chain, wherein the sub-influence parameters are used to reflect the transient stability parameters of the power system under any one influence chain; until the multiple influence chains are executed, the sub-influence parameters corresponding to each of the multiple influence chains are obtained; and based on the sub-influence parameters corresponding to each of the multiple influence chains, determining the transient stability impact parameters corresponding to the power system.

[0169] Optionally, based on the target node data corresponding to each of the multiple target nodes and the parent node data corresponding to each of the multiple target nodes, the sub-influence parameters corresponding to any influence chain are determined, including: determining the predicted data corresponding to each of the multiple target nodes based on the parent node data corresponding to each of the multiple target nodes; determining the data deviation value corresponding to each of the multiple target nodes based on the target node data and the predicted data; and determining the sub-influence parameters corresponding to any influence chain based on the data deviation value corresponding to each of the multiple target nodes and the parent node data corresponding to each of the multiple target nodes.

[0170] For example, for each node (i.e., the target node) in a causal directed acyclic graph, construct a mapping function with the remaining parent nodes as input and the target node as output, where the mapping function for the s-th node is... :

[0171]

[0172] in, These are the model's predicted values ​​(i.e., the predicted data). A node in a DAG; for The parent node, that is, the node pointing to All nodes.

[0173] Based on the mapping function of each node, calculate the residual between its predicted value and the actual value. (That is, data bias value), to obtain a set of residual values ​​corresponding to each sample { },in:

[0174]

[0175] The obtained residual value { Treating { as independent and identically distributed noise samples, kernel density estimation (KDE) is used to evaluate { The probability distribution of} is estimated to obtain the noise variable. probability distribution , To represent noise, Gaussian noise can be used, and the mapping function can be fitted using the random forest regression method.

[0176] Based on mapping function and noise variables The structural causal model is obtained, and the formula is:

[0177]

[0178] Based on this structural causal model, the sub-influence parameters corresponding to any influence chain are determined.

[0179] Optionally, based on multiple influence chains, the transient stability influence parameters corresponding to the power system can be determined using a neural network prediction model. Training the neural network prediction model includes:

[0180] Taking a power system with N generators and M loads as an example, we will quantify the feature set F and the transient stability index under the operating scenario. The combined features and the selected features are used as input features (i.e., multiple target power parameters) to obtain the dataset for training the neural network. .

[0181] The mean-variance normalization method was used to normalize the dataset. Normalization is performed to convert features of different dimensions into features of a unified dimension, resulting in the processed dataset. ; the processed dataset The set is divided into training set, validation set and test set according to a set ratio, which can be 6:2:2.

[0182] A transient stability margin prediction model (i.e., a neural network prediction model, hereinafter referred to as the prediction model) is constructed based on the normalized training set. The prediction model is a multilayer perceptron neural network with an introduced residual structure, including an input layer, three residual blocks, and an output layer. The input layer has a dimension of [missing information]. Each residual block includes a primary channel and extended channels, with the primary channel having dimensions of [dimensions to be filled in]. The extended channel dimensions are respectively Each residual block consists of a main channel (containing a fully connected layer, a normalization layer, and a non-linear activation function) and a shortcut channel. The fully connected layer of the main channel maps the main channel dimension to the extended channel dimension and then back to the main channel dimension, maintaining dimensionality consistency. The shortcut channel directly adds the main channel input to the output. Optionally, =60, , .

[0183] Next, model training is performed, using mean squared error loss as the loss function. The Adam algorithm is used to iteratively update the model parameters of the prediction model, and the loss of the prediction model on the validation set is calculated in each iteration. When the loss on the validation set does not decrease within 50 consecutive iterations, the iteration process is stopped, and the prediction model with the highest accuracy on the validation set is selected as the final transiently stable evaluation artificial intelligence model. This completes the training of the neural network prediction model.

[0184] Furthermore, controllable features (such as generator output, voltage, active and reactive power of load) are selected from the screened features as features to be explained. Actual power grid operation data are obtained, and the features to be explained are selected to construct the sample to be explained. For example, 16 features such as active power output and terminal voltage of 8 generators are selected as features to be explained.

[0185] For the selected features to be explained, a feature combination intervention set is constructed. This includes selecting and combining features from the selected features to obtain multiple feature combinations, such as the active power output of two generators, or the active power of a generator and load at a certain node. This allows for joint intervention on multiple variables in the combination during subsequent intervention. For example, all possible combinations of no more than three features to be explained can be constructed. Then, based on the acquired data sample, the output of the transient stability assessment AI model after inputting that sample is obtained, which serves as the model output before intervention. .

[0186] Based on the aforementioned structural causal model, causal intervention is applied to multiple feature combinations, i.e., the features to be explained are perturbed. Formal tools for causal intervention are used to obtain the intervention distribution of variables. The data after intervention is input into a trained transient stable evaluation AI model, and the model output after intervention is calculated using the following formula:

[0187]

[0188] in, The expected result of the transient stability assessment model output after causal intervention for feature combination S; This indicates that the average value of the results within the parentheses is taken; The initial values ​​corresponding to the feature combination S; Let be the applied perturbation vector; This is a causal intervention operation used to apply perturbations; This is a trained artificial intelligence model for transient stability assessment, used to output the corresponding transient stability assessment results based on the feature vector x of the input power system.

[0189] The improved SHAP value of feature combinations is calculated, whereby the SHAP value measures the contribution of perturbations in the feature combination as input to the model output, thereby explaining the influence of different input features on the prediction model's decision output and achieving interpretability analysis of the impact of different input features on decision-making in the prediction model. The formula for calculating the improved SHAP value of combined features is as follows:

[0190]

[0191] in, The contribution of the perturbation of the feature combination S to the output of the transient stability assessment model; The output of the model after intervention. This is the output before any intervention.

[0192] The above steps can be used to determine the system's implementation through the influence parameters of power system transient stability. Figure 3 This is a schematic diagram of the system structure for determining the influence parameters of power system transient stability in an optional embodiment of the present invention, as shown below. Figure 3 As shown, it includes:

[0193] Data acquisition module: used to acquire operational scenario data of the target system, including input features and output metrics;

[0194] Feature selection module: Used to select important and influential features using the mRMR method and construct combined features for subsequent prediction model training and causal model calculation;

[0195] Model training module: Used to train the neural network model using the obtained dataset and select the optimal model on the validation set;

[0196] Causal Feature Calculation Module: Used to calculate the causal graph of candidate feature sets and output indicators using causal discovery algorithms, and extract the causal feature set;

[0197] Interpretive Analysis Module: This module is used to intervene in features based on the structural causal model, calculate output changes, obtain causal contribution results, and thus complete the interpretability analysis of the model prediction results.

[0198] The above optional implementation methods can achieve at least the following beneficial effects:

[0199] (1) Compared with related technologies, the present invention introduces the mRMR method in the feature selection module. This method not only considers the correlation strength between features and transient stability quantification indicators, but also considers the correlation between features, so that the model input avoids two highly linear features, such as the active power flow between two adjacent nodes on the line, thereby improving the model prediction effect and interpretability.

[0200] (2) Compared with related technologies, new statistical features are introduced as inputs when constructing the transient stability prediction model, such as the generator output allocation ratio and the active and reactive power ratio of the line, which are statistical features strongly correlated with transient stability quantitative indicators, thereby enhancing the model training effect. In addition, a residual network is introduced into the hidden layer of the model to alleviate the gradient vanishing problem during deep network training, enhance the model's expressive power and generalization ability, thereby improving the ability to characterize the transient dynamic characteristics of the power system, and helping to improve the prediction accuracy and robustness under different operating scenarios.

[0201] (3) Compared with related technologies, this invention introduces causal inference technology on the basis of traditional SHAP interpretation method. A large amount of data is obtained through scenario construction and feature screening for PC algorithm to construct causal graph, and a structural causal model (SCM) is constructed based on the causal graph, and causal intervention is performed on features, thereby taking into account the correlation and causality between features. This consideration is particularly important for power flow calculation features with high correlation in power system.

[0202] (4) Compared with related technologies, this invention improves the limitations of the traditional SHAP method in calculating the weighted average contribution of a single feature in all feature combinations. Instead, it adds the contribution calculation of feature combination intervention, linearly combining the features to be explained to obtain the contribution of each feature combination. This solves the problem that traditional model interpretation methods cannot explain the effect of multiple features acting simultaneously. It is of great significance for the linkage control of generators and the joint regulation of voltage and active power in power systems.

[0203] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0204] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0205] Example 2

[0206] According to embodiments of the present invention, an apparatus for implementing the above-described method for determining the influence parameters of power system transient stability is also provided. Figure 4 This is a structural block diagram of a device for determining the influence parameters of transient stability of a power system according to an embodiment of the present invention, such as... Figure 4 As shown, the device includes: an acquisition module 402, a first determination module 404, a second determination module 406, a third determination module 408, and a fourth determination module 410. The device will be described in detail below.

[0207] The module 402 is used to acquire the operating scenario of the power system; the first determining module 404, connected to the acquisition module 402, is used to determine the transient stability index and multiple target power parameters corresponding to the power system under the operating scenario, wherein the transient stability index is used to represent the degree of stability of the power system after a disturbance, and the multiple target power parameters are parameters used to reflect the operating state of the power system; the second determining module 406, connected to the first determining module 404, is used to determine the first causal relationship and the second causal relationship corresponding to the power system based on the multiple target power parameters and the transient stability index, wherein the first causal relationship is used to represent the causal relationship between any two target power parameters, and the second causal relationship is used to represent the causal relationship between the multiple target power parameters and the transient stability index respectively; the third determining module 408, connected to the second determining module 406, is used to determine multiple influence chains based on the multiple target power parameters, the first causal relationship, and the second causal relationship; the fourth determining module 410, connected to the third determining module 408, is used to determine the transient stability influence parameters corresponding to the power system based on the multiple influence chains.

[0208] It should be noted that the above-mentioned acquisition module 402, first determination module 404, second determination module 406, third determination module 408, and fourth determination module 410 correspond to steps S102 to S110 in the method for determining the influence parameters of transient stability of power system. The multiple modules and the corresponding steps are the same in terms of implementation examples and application scenarios, but are not limited to the content disclosed in the above embodiment 1.

[0209] Example 3

[0210] According to another aspect of the present invention, an electronic device is also provided, comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to execute instructions to implement the method for determining the influence parameters of power system transient stability as described above.

[0211] Example 4

[0212] According to another aspect of the present invention, a computer-readable storage medium is also provided, which, when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform the power system transient stability influence parameter determination method described above.

[0213] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0214] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0215] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0216] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0217] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0218] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the 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 to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0219] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for determining the influence parameters of transient stability in a power system, characterized in that, include: Obtain the operating scenario of the power system; Under the aforementioned operating scenario, a transient stability index and multiple target power parameters corresponding to the power system are determined. The transient stability index is used to represent the degree of operational stability of the power system after a disturbance occurs, and the multiple target power parameters are parameters used to reflect the operating state of the power system. Based on the plurality of target power parameters and the transient stability index, a first causal relationship and a second causal relationship corresponding to the power system are determined, wherein the first causal relationship is used to represent the causal relationship between any two target power parameters among the plurality of target power parameters, and the second causal relationship is used to represent the causal relationship between the plurality of target power parameters and the transient stability index respectively; Based on the multiple target power parameters, the first causal relationship, and the second causal relationship, multiple influence chains are determined; Based on the multiple influence chains, the transient stability influence parameters corresponding to the power system are determined.

2. The method according to claim 1, characterized in that, Determine multiple target power parameters corresponding to the power system under the stated operating scenario, including: Determine multiple initial power parameters corresponding to the power system under the aforementioned operating scenario; A first correlation index corresponding to each of the plurality of initial power parameters is determined, and a plurality of second correlation indices corresponding to each of the plurality of initial power parameters are determined, wherein the corresponding first correlation index is used to represent the degree of correlation between the corresponding initial power parameter and the transient stability index, and the corresponding plurality of second correlation indices are used to represent the degree of correlation between the corresponding initial power parameter and other initial power parameters respectively; Based on the first correlation index corresponding to the plurality of initial power parameters, and the plurality of second correlation indices corresponding to the plurality of initial power parameters, a plurality of target power parameters corresponding to the power system are determined.

3. The method according to claim 2, characterized in that, The step of determining multiple target power parameters corresponding to the power system based on the first correlation index corresponding to each of the multiple initial power parameters and the multiple second correlation index corresponding to each of the multiple initial power parameters includes: From the plurality of initial power parameters, a first candidate parameter is determined, wherein the first candidate parameter is the initial power parameter with the largest first correlation index among the plurality of initial power parameters; According to the execution order of performing multiple screenings on the plurality of initial power parameters, based on the first correlation index corresponding to the plurality of first remaining parameters and the second correlation index between the plurality of first remaining parameters and the first candidate parameter, a second candidate parameter is determined from the plurality of first remaining parameters, wherein the plurality of first remaining parameters are other initial power parameters besides the first candidate parameter among the plurality of initial power parameters; Based on the first correlation index corresponding to each of the multiple second remaining parameters, and the second correlation index between each of the multiple second remaining parameters and the second candidate parameter, a third candidate parameter is determined from the multiple second remaining parameters. The multiple second remaining parameters are other initial power parameters in the multiple first remaining parameters besides the second candidate parameter. This process is repeated until multiple screening operations are completed, resulting in multiple candidate parameters corresponding to the power system. Based on the multiple candidate parameters, multiple target power parameters corresponding to the power system are determined.

4. The method according to claim 3, characterized in that, The step of determining multiple target power parameters corresponding to the power system based on the multiple candidate parameters includes: When the plurality of candidate parameters include active power and reactive power, the power ratio corresponding to the power system is determined based on the active power and the reactive power. Based on the power ratio and the plurality of candidate parameters, a plurality of target power parameters corresponding to the power system are determined.

5. The method according to claim 1, characterized in that, The determination of transient stability impact parameters corresponding to the power system based on the multiple influence chains includes: According to the execution order of the multiple impact chains, for any one of the multiple impact chains, multiple target nodes corresponding to any one impact chain are determined, wherein the multiple target nodes include first nodes corresponding to the multiple target power parameters respectively, and second nodes corresponding to the transient stability index; Determine the parent node corresponding to each of the plurality of target nodes; Based on the target node data corresponding to the plurality of target nodes and the parent node data corresponding to the parent nodes corresponding to the plurality of target nodes, the sub-influence parameters corresponding to any influence chain are determined, wherein the sub-influence parameters are used to reflect the transient stability parameters of the power system under any influence chain; Until all the multiple influence chains have been executed, the sub-influence parameters corresponding to each of the multiple influence chains are obtained; Based on the sub-influence parameters corresponding to the multiple influence chains, the transient stability influence parameters corresponding to the power system are determined.

6. The method according to claim 5, characterized in that, The step of determining the sub-influence parameter corresponding to any influence chain based on the target node data corresponding to the plurality of target nodes and the parent node data corresponding to the parent nodes corresponding to the plurality of target nodes includes: Based on the parent node data of the parent nodes corresponding to the multiple target nodes, predictive data corresponding to the multiple target nodes are determined. Based on the target node data and predicted data corresponding to the multiple target nodes, determine the data deviation values ​​corresponding to the multiple target nodes respectively; Based on the data deviation values ​​corresponding to the multiple target nodes and the parent node data corresponding to the parent nodes of the multiple target nodes, the sub-influence parameters corresponding to any influence chain are determined.

7. The method according to any one of claims 1 to 6, characterized in that, Determining the transient stability index corresponding to the power system under the operating scenario includes: Under the aforementioned operating scenario, the power angle difference index corresponding to each of the multiple generator sets is determined, wherein the corresponding power angle difference index represents the degree of difference in power angle between the two generators included in the corresponding generator set, and the power system includes the multiple generator sets; Based on the power angle difference index corresponding to each of the multiple generator sets, the transient stability index corresponding to the power system is determined.

8. A device for determining the influence parameters of transient stability in a power system, characterized in that, include: The acquisition module is used to acquire the operating scenario of the power system; The first determining module is used to determine the transient stability index and multiple target power parameters corresponding to the power system under the operating scenario, wherein the transient stability index is used to represent the degree of stability of the power system after a disturbance, and the multiple target power parameters are parameters used to reflect the operating state of the power system; The second determining module is used to determine a first causal relationship and a second causal relationship corresponding to the power system based on the plurality of target power parameters and the transient stability index. The first causal relationship is used to represent the causal relationship between any two target power parameters among the plurality of target power parameters, and the second causal relationship is used to represent the causal relationship between the plurality of target power parameters and the transient stability index, respectively. The third determining module is used to determine multiple influence chains based on the multiple target power parameters, the first causal relationship, and the second causal relationship; The fourth determining module is used to determine the transient stability influence parameters corresponding to the power system based on the multiple influence chains.

9. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the method for determining the influence parameters of power system transient stability as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the method for determining the influence parameters of power system transient stability as described in any one of claims 1 to 7.

Images