Power grid performance index determination method and device and electronic equipment

Abstract

The invention discloses a power grid performance index determination method and apparatus, and an electronic device. The method comprises the following steps: receiving a performance index determination request for a target power grid system; in response to the performance index determination request, calling operation response data and power data corresponding to the to-be-tested performance item according to the performance identifier; determining an operation characteristic value and an operation response characteristic value of the target power grid system according to the power data and the operation response data; determining a fusion feature value according to the operation feature value, a first weight value corresponding to the operation feature value, the operation response feature value and a second weight value corresponding to the operation response feature value; and according to the fusion feature value, determining a target performance index value of the target power grid system corresponding to the to-be-tested performance item. The technical problem that in the prior art, due to the fact that the determined fusion data cannot reflect the main operation characteristics of the power grid system, the determined performance index result is not accurate is solved.

Description

Technical Field

[0001] The present invention relates to the field of data analysis, and in particular to a method, device and electronic equipment for determining power grid performance indicators. Background Art

[0002] Determining power system performance indicators is crucial for accurately assessing system operating status, ensuring safe and stable power supply, optimizing resource allocation, and improving overall efficiency. Currently, simple superposition or linear combination methods are primarily used to process power data and operational response data to generate fused data, which is then used to determine power grid system performance indicators. However, the fused data generated by these methods fails to reflect the key operational characteristics of the power grid system, resulting in inaccurate performance indicators.

[0003] To address the above-mentioned problems, no effective solutions have been proposed so far. Summary of the Invention

[0004] The embodiments of the present invention provide a method, device and electronic device for determining power grid performance indicators, so as to at least solve the technical problem in the prior art that the determined fusion data cannot reflect the main operating characteristics of the power grid system, resulting in inaccurate performance indicator results.

[0005] According to one aspect of an embodiment of the present invention, a method for determining a power grid performance indicator is provided, comprising: receiving a performance indicator determination request for a target power grid system, wherein the performance indicator determination request carries a performance identifier of a performance item to be measured; in response to the performance indicator determination request, retrieving operation response data and power data corresponding to the performance item to be measured based on the performance identifier; determining an operation characteristic value and an operation response characteristic value of the target power grid system based on the power data and the operation response data, wherein the operation characteristic value represents the operation status of multiple power devices in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to an operation instruction in the target power grid system; determining a fusion characteristic value based on the operation characteristic value and a first weight value corresponding to the operation characteristic value, and the operation response characteristic value and a second weight value corresponding to the operation response characteristic value; and determining a target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fusion characteristic value.

[0006] Optionally, based on the power data and the operation response data, the operation characteristic values ​​and operation response characteristic values ​​of the target power grid system are determined, including: determining an initial network diagram corresponding to the target power grid system, wherein the initial network diagram includes multiple device nodes and multiple transmission edges, and the multiple device nodes are respectively provided with corresponding node data, and the corresponding node data include status data corresponding to multiple operation status items of corresponding power equipment, and the multiple transmission edges are used to represent the power transmission direction of multiple power equipment in the target power grid system; based on the power data, determining multiple real-time operation status values ​​corresponding to the multiple power equipment; based on the initial network diagram and the multiple real-time operation status values ​​corresponding to the multiple power equipment, determining the target network diagram; based on the target network diagram, determining the operation characteristic values.

[0007] Optionally, based on the fusion characteristic value, determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured includes: calling a performance indicator determination model corresponding to the performance item to be measured, wherein the performance indicator determination model is provided with target model parameters, and the target model parameters are determined based on an initial indicator determination model provided with initial model parameters and sample data corresponding to the performance item to be measured, and the sample data includes sample fusion characteristic values ​​and sample performance indicator values ​​of the target power grid system; based on the performance indicator determination model and the fusion characteristic value, determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured.

[0008] Optionally, before calling the performance indicator determination model corresponding to the performance item to be measured, it also includes: determining an initial indicator determination model corresponding to the performance item to be measured based on sample data, wherein the initial indicator determination model is provided with initial model parameters; determining a first error value corresponding to the initial indicator determination model; when the first error value is greater than a first error threshold, updating the initial model parameters until the target model parameters are obtained, wherein the target error value corresponding to the target model parameters is less than the first error threshold; determining the performance indicator determination model corresponding to the performance item to be measured based on the target model parameters and the initial indicator determination model.

[0009] Optionally, determining the first error value corresponding to the initial indicator determination model includes: determining the test performance indicator value based on the test fusion characteristic value and the initial indicator determination model; calling the error function corresponding to the performance item to be tested, wherein the error function is used to determine the error value between the test performance indicator value and the actual performance indicator; determining the first error value based on the test performance indicator value, the error function and the actual performance indicator value corresponding to the test fusion characteristic value.

[0010] Optionally, based on the fusion characteristic value, the target performance index value corresponding to the target power grid system and the performance item to be measured is determined, including: in the case where the fusion characteristic item includes multiple sub-fusion characteristic items, determining the characteristic item contribution values ​​corresponding to the multiple sub-fusion characteristic items to the performance item to be measured respectively; determining the corresponding fusion characteristic item whose corresponding characteristic item contribution value is greater than a predetermined threshold as the target characteristic item; based on the sub-fusion characteristic value corresponding to the target characteristic item, determining the target performance index value corresponding to the target power grid system and the performance item to be measured.

[0011] Optionally, after determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fusion characteristic value, it also includes: determining the true performance indicator value corresponding to the performance item to be measured; determining a second error value between the true performance indicator value and the target performance indicator value; when the second error value is greater than a second error threshold, updating the first weight value and the second weight value to obtain a first updated weight and a second updated weight, so as to determine the performance indicator value corresponding to the target power grid system and the performance item to be measured based on the first updated weight and the second updated weight.

[0012] According to one aspect of an embodiment of the present invention, a device for determining a power grid performance indicator is provided, comprising: a receiving module for receiving a performance indicator determination request for a target power grid system, wherein the performance indicator determination request carries a performance identifier of a performance item to be measured; a response module for, in response to the performance indicator determination request, retrieving operation response data and power data corresponding to the performance item to be measured based on the performance identifier; a first determination module for determining an operation characteristic value and an operation response characteristic value of the target power grid system based on the power data and the operation response data, wherein the operation characteristic value represents the operation status of multiple power devices in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to an operation instruction in the target power grid system; a second determination module for determining a fused characteristic value based on the operation characteristic value and a first weight value corresponding to the operation characteristic value, and the operation response characteristic value and a second weight value corresponding to the operation response characteristic value; and a third determination module for determining a target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fused characteristic value.

[0013] According to one aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the method for determining any one of the above-mentioned power grid performance indicators.

[0014] According to one aspect of an embodiment of the present invention, a computer-readable storage medium is provided. When the computer-readable storage medium is executed by a processor of an electronic device, the electronic device can execute any of the above-mentioned methods for determining a power grid performance indicator.

[0015] In an embodiment of the present invention, a performance indicator determination request for a target power grid system is received, wherein the performance indicator determination request carries a performance identifier of the performance item to be measured; in response to the performance indicator determination request, operation response data and power data corresponding to the performance item to be measured are retrieved based on the performance identifier; based on the power data and the operation response data, an operation characteristic value and an operation response characteristic value of the target power grid system are determined, wherein the operation characteristic value represents the operation status of multiple power devices in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to the operation instruction in the target power grid system; based on the operation characteristic value and a first weight value corresponding to the operation characteristic value, and the operation response characteristic value and a second weight value corresponding to the operation response characteristic value, a fusion characteristic value is determined; based on the fusion characteristic value, the target performance corresponding to the target power grid system and the performance item to be measured is determined. The method of using the energy index value as the basis, by determining the fusion characteristic value according to the operation characteristic value and the first weight value corresponding to the operation characteristic value, the operation response characteristic value and the second weight value corresponding to the operation response characteristic value, achieves the purpose of determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured according to the fusion characteristic value, determines the operation characteristics and operation response characteristics of the target power grid system, and their corresponding weight values ​​respectively, can determine the interaction between the power data characteristics and the operation response characteristics, thereby determining the fusion characteristics that can reflect the main characteristics of the target power grid system, and determines the target performance indicator value corresponding to the performance item to be measured according to the fusion characteristics, thereby improving the accuracy of the performance indicator results, thereby solving the technical problem in the prior art that the determined fusion data cannot reflect the main operation characteristics of the power grid system, resulting in inaccurate performance indicator results. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0017] Figure 1 is a flow chart of a method for determining a power grid performance indicator according to an embodiment of the present invention;

[0018] Figure 2 is a flow chart of a method for determining a power grid performance indicator provided by an optional embodiment of the present invention;

[0019] Figure 3 It is a structural block diagram of a device for determining power grid performance indicators according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] Example 1

[0023] According to an embodiment of the present invention, an embodiment of a method for determining power grid performance indicators is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a group of computer executables, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0024] Figure 1 FIG. 1 is a flow chart of a method for determining a power grid performance indicator according to an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0025] Step S102: receiving a performance indicator determination request for a target power grid system, wherein the performance indicator determination request carries a performance identifier of a performance item to be measured.

[0026] In step S102 provided in the present application, a request for determining a performance indicator of a target power grid system is received.

[0027] Among them, a target power grid system is involved, and the target power grid system refers to the power network system whose performance indicators of the performance items to be tested need to be evaluated.

[0028] Among them, a performance indicator determination request is involved, and the performance indicator determination request refers to a request for evaluating or monitoring specific performance indicators of the performance items to be measured of the target power grid system.

[0029] Among them, the performance items to be measured are involved. The performance items to be measured refer to the specific aspects or indicators of the target power grid system to be monitored or evaluated, such as predicting the future state of the power grid, classifying the operation mode or fault type of the power grid.

[0030] Among them, the performance identifier is involved. The performance identifier refers to the code or name used to uniquely identify the performance item to be measured in the performance indicator determination request.

[0031] In this step, a performance indicator determination request is received, intended to determine or evaluate a specific performance indicator of the target power grid system. The performance indicator identification of the performance item to be measured carried in the performance indicator determination request is used to specify the specific performance item to be monitored or evaluated, thereby determining the performance indicator of the performance item to be measured. This is a fundamental step in determining the performance indicator of the performance item to be measured in the target power grid system. It helps to quickly focus on the problem and conduct targeted analysis of the power grid operating status without the need for comprehensive and indiscriminate monitoring of the entire system.

[0032] Step S104 , in response to the performance indicator determination request, retrieve the operation response data and power data corresponding to the performance item to be measured according to the performance identifier.

[0033] In step S104 provided in the present application, operation response data and power data corresponding to the performance item to be measured are retrieved.

[0034] Among them, operation response data is involved, which refers to the system response or equipment status change data after executing a certain operation instruction in the target power grid system.

[0035] Among them, power data is involved. Power data refers to various types of data generated during the operation of the target power grid system, such as the real-time operating status of the target power grid system (such as voltage, current, power, etc.), historical operation records, equipment status information, etc.

[0036] In this step, when a performance indicator determination request for evaluating the performance item to be tested is received, the relevant data resources, including operation response data and power data, will be automatically called according to the performance identifier (i.e., the specific performance indicator type) contained in the performance indicator determination request to perform a detailed performance evaluation.

[0037] Through this step, the performance identifier is identified and the required data is automatically retrieved, realizing the automation of the evaluation process. The performance evaluation based on actual operation response data and power data provides a solid data foundation for power system management and optimization, enhances the scientific nature and reliability of decision-making, and improves the efficiency and accuracy of the evaluation.

[0038] Step S106: Determine the operating characteristic value and operating response characteristic value of the target power grid system based on the power data and the operating response data, wherein the operating characteristic value represents the operating status of multiple power devices in the target power grid system within a predetermined time period, and the operating response characteristic value represents the response status corresponding to the operating instruction in the target power grid system.

[0039] In step S106 provided in the present application, the operating characteristic values ​​and the operation response characteristic values ​​of the target power grid system are determined.

[0040] Among them, operating characteristic values ​​are involved. Operating characteristic values ​​refer to a series of numerical values ​​or vectors determined based on power data, such as average power, maximum voltage, load rate, equipment health index, etc. They can summarize the operating status of multiple power equipment in the target power grid system within a certain period of time.

[0041] Among them, the operation response characteristic value is involved. The operation response characteristic value refers to a quantitative description of the effect of the operation instruction converted from the operation response data, such as the speed of the dispatch instruction response, the degree of equipment response, the change in system status after the operation, etc. It can be used to evaluate the effectiveness of the operation and the responsiveness of the power grid.

[0042] In this step, the target grid system's operational characteristic values ​​are determined by analyzing its operational characteristic values, and its operational response characteristic values ​​are determined by analyzing its operational response data. The operational characteristic values ​​characterize the grid's operational state, encompassing the operating status of multiple power devices within a predetermined time period. The operational response characteristic values ​​focus on the grid system's response to operational commands, reflecting the effectiveness of the operations and the dynamic response capabilities of the grid devices. This step provides a deeper understanding of the grid's operational patterns, quantifies the effects of operational commands, and provides data support for optimizing grid dispatch strategies and establishing predictive models.

[0043] Step S108 : determining a fusion feature value according to the operation feature value and the first weight value corresponding to the operation feature value, the operation response feature value and the second weight value corresponding to the operation response feature value.

[0044] In step S108 provided in this application, a fusion feature value is determined.

[0045] Herein, a first weight value is involved, which refers to a weight associated with an operating characteristic value, and is used to indicate the importance of each operating characteristic value in the process of evaluating or deciding the performance item to be measured.

[0046] Among them, the second weight value is involved, which refers to the weight associated with the operation response characteristic value, and is used to quantify the degree of influence of each operation response characteristic value on the overall state of the system or the prediction result in the process of evaluating or deciding the performance item to be tested.

[0047] Among them, the fusion characteristic value is involved. The fusion characteristic value refers to a comprehensive indicator that integrates the operation characteristic value and the operation response characteristic value, and can comprehensively evaluate the status and operation of the target power grid system.

[0048] Through this step, the fusion characteristic value is determined based on the operating characteristic value and the first weight value corresponding to the operating characteristic value, the operation response characteristic value and the second weight value corresponding to the operation response characteristic value. The first weight value and the second weight value reflect the relationship between the operating characteristic value and the operation response characteristic value in the fusion process, as well as their relative importance for evaluating the performance item to be measured, thereby determining the fusion characteristics that can reflect the main characteristics of the target power grid system, comprehensively evaluating the state of the power grid system and the impact of operation, so that the model can better capture complex dynamics and significantly improve the prediction accuracy of the model.

[0049] It should be noted that the first weight value and the second weight value can be determined by introducing an attention mechanism to perform weighted aggregation on the neighborhood of the device node to highlight the influence of important neighborhood nodes; higher weights can also be given to specific time steps in the time series to capture changes at key time points; they can also be determined by designing a multimodal fusion mechanism, such as feature splicing, weighted summation, or attention-based fusion methods, which can effectively combine the features of the two modalities. For example, feature splicing can be used to combine data stream features and business flow features into a comprehensive feature vector.

[0050] Step S110 : determining target performance indicator values ​​corresponding to the target power grid system and the performance items to be measured based on the fused characteristic values.

[0051] In step S110 provided in the present application, target performance indicator values ​​corresponding to the target power grid system and the performance items to be measured are determined.

[0052] Among them, the target performance indicator value is involved, and the target performance indicator value refers to a numerical value that specifically describes the performance or status of the target power grid system under the performance item to be measured.

[0053] In this step, based on the determined fusion feature values, machine learning models or other analytical tools are used to determine the target power grid system's performance under specific performance items to be measured, resulting in a target performance indicator value. This step considers multiple relevant factors to determine the target performance indicator value, making the predicted or evaluated target performance indicator value more comprehensive and accurate, reflecting the actual operating status of the power grid system and providing a data basis for power dispatch and operation and maintenance decisions.

[0054] Through the above steps S102-S110, a performance indicator determination request for the target power grid system can be received, wherein the performance indicator determination request carries a performance identifier of the performance item to be measured; in response to the performance indicator determination request, the operation response data and power data corresponding to the performance item to be measured are retrieved based on the performance identifier; based on the power data and the operation response data, the operation characteristic value and the operation response characteristic value of the target power grid system are determined, wherein the operation characteristic value represents the operation status of multiple power equipment in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to the operation instruction in the target power grid system; based on the operation characteristic value and the first weight value corresponding to the operation characteristic value, the operation response characteristic value and the second weight value corresponding to the operation response characteristic value, a fusion characteristic value is determined; based on the fusion characteristic value, the target power grid system and the performance item to be measured corresponding to the operation characteristic value are determined. The target performance indicator value is obtained by determining the fusion characteristic value based on the operation characteristic value and the first weight value corresponding to the operation characteristic value, the operation response characteristic value and the second weight value corresponding to the operation response characteristic value, so as to achieve the purpose of determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fusion characteristic value, and determining the operation characteristics and operation response characteristics of the target power grid system, as well as their corresponding weight values ​​respectively, and determining the interaction between the power data characteristics and the operation response characteristics, so as to determine the fusion characteristics that can reflect the main characteristics of the target power grid system, and determining the target performance indicator value corresponding to the performance item to be measured based on the fusion characteristics, thereby improving the accuracy of the performance indicator results, thereby solving the technical problem in the prior art that the determined fusion data cannot reflect the main operation characteristics of the power grid system, resulting in inaccurate performance indicator results.

[0055] As an optional embodiment, the operation characteristic values ​​and operation response characteristic values ​​of the target power grid system are determined based on the power data and the operation response data, including: determining an initial network diagram corresponding to the target power grid system, wherein the initial network diagram includes multiple device nodes and multiple transmission edges, and the multiple device nodes are respectively provided with corresponding node data, and the corresponding node data include status data corresponding to multiple operation status items of the corresponding power equipment, and the multiple transmission edges are used to represent the power transmission direction of the multiple power equipment in the target power grid system; based on the power data, determining multiple real-time operation status values ​​corresponding to the multiple power equipment; based on the initial network diagram and the multiple real-time operation status values ​​corresponding to the multiple power equipment, determining the target network diagram; based on the target network diagram, determining the operation characteristic values.

[0056] In this embodiment, specific steps for determining the operating characteristic value are described.

[0057] Among them, the initial network diagram is involved, which refers to a graph model established based on the physical topology structure of the target power grid system.

[0058] This involves device nodes, which are nodes in the initial network diagram and represent specific equipment in the target power grid system, such as generators, transformers, transmission lines, and substations. Each device node has its own attributes and status data, reflecting the role and real-time operating status of the corresponding power equipment in the target power grid system.

[0059] Among them, the transmission edge is involved. The transmission edge refers to the edge connecting the device nodes, indicating the direction of power flow from one device to another, and reflecting the transmission path and connection relationship of power in the power grid.

[0060] Among them, node data is involved. Node data refers to the status data of operating status items related to power equipment, such as voltage level, power output, equipment temperature, etc.

[0061] Among them, the real-time operating status value is involved. The real-time operating status value refers to the value collected by the monitoring system that reflects the current operating status of the power grid equipment, such as real-time power output value, voltage fluctuation value, current value, etc.

[0062] This involves the target network diagram, which is a new network diagram that is updated based on the initial network diagram and incorporates real-time operating status values. The target network diagram more accurately reflects the current operating status of the power grid and the direction of power transmission, and is a key basis for in-depth analysis and decision support.

[0063] In this step, an initial network diagram is first constructed, reflecting the physical connections of the target power grid system and the operating status of power equipment. Then, using real-time power data, the operating status values ​​of each device node are updated, generating a target network diagram that more closely reflects the current operating conditions of the target power grid system. Finally, the operating characteristic values ​​of the target power grid system are extracted from the target network diagram.

[0064] By defining device nodes and transmission edges, an initial network diagram that can reflect the network architecture and power transmission direction of the target power grid system is determined. Combined with power data, the network diagram is updated in real time and operating characteristic values ​​are extracted to determine the target network diagram that can reflect the real-time operating status of each power device in the target power grid system and the real-time transmission status of the lines, thereby improving the efficiency of subsequent determination of the operating status value and the accuracy of the determined operating status value.

[0065] As an optional embodiment, the target performance indicator value corresponding to the target power grid system and the performance item to be measured is determined based on the fusion characteristic value, including: calling the performance indicator determination model corresponding to the performance item to be measured, wherein the performance indicator determination model is provided with target model parameters, and the target model parameters are determined based on an initial indicator determination model provided with initial model parameters and sample data corresponding to the performance item to be measured, and the sample data includes sample fusion characteristic values ​​and sample performance indicator values ​​of the target power grid system; based on the performance indicator determination model and the fusion characteristic value, the target performance indicator value corresponding to the target power grid system and the performance item to be measured is determined.

[0066] In this embodiment, specific steps of determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured are described.

[0067] Among them, the performance indicator determination model is involved. The performance indicator determination model refers to a model used to predict or determine the performance indicators corresponding to the performance items to be measured in the power grid system. By learning existing sample data, a mapping relationship between characteristic values ​​and performance indicators is established.

[0068] This involves target model parameters, which are key parameters in the performance indicator determination model. They determine the model's structure and function. Target model parameters are trained and optimized to ensure that the model can accurately predict the performance indicator values ​​of the target power grid system.

[0069] Among them, the initial indicator determination model is involved. The initial indicator determination model refers to the starting point of the performance indicator determination model learning process. Subsequent training and optimization will be based on the initial indicator determination model.

[0070] Among them, the initial model parameters are involved. The initial model parameters refer to a set of parameters set at the beginning of the initial indicator determination model training, which are usually obtained through random initialization or other heuristic methods.

[0071] In this step, a trained performance indicator determination model is first called. This model contains target model parameters, which are derived by analyzing a training dataset containing sample fusion feature values ​​and sample performance indicator values. The current fusion feature values ​​are input into the performance indicator determination model, which then outputs the target performance indicator value corresponding to the current target power grid system and the performance item to be measured based on the learned mapping relationship.

[0072] Through this step, the model is determined based on the performance indicators corresponding to the performance items to be measured, which can accurately predict and evaluate the performance indicators of the target power grid model, realize the automation of performance indicator prediction, and immediately give the prediction results based on real-time data input, providing a scientific basis for scheduling decisions.

[0073] As an optional embodiment, before calling the performance indicator determination model corresponding to the performance item to be measured, it also includes: determining the initial indicator determination model corresponding to the performance item to be measured based on sample data, wherein the initial indicator determination model is provided with initial model parameters; determining a first error value corresponding to the initial indicator determination model; when the first error value is greater than a first error threshold, updating the initial model parameters until the target model parameters are obtained, wherein the target error value corresponding to the target model parameters is less than the first error threshold; determining the performance indicator determination model corresponding to the performance item to be measured based on the target model parameters and the initial indicator determination model.

[0074] In this embodiment, the specific steps of determining a performance indicator determination model corresponding to the performance item to be measured are described.

[0075] Among them, the first error value is involved. The first error value refers to a numerical value used to measure the difference between the performance prediction value determined by the initial indicator determination model and the actual performance indicator value in the sample data, such as the mean square error (MSE), mean absolute error (MAE) or the results of other loss functions.

[0076] Among them, the first error threshold is involved. The first error threshold refers to a preset error limit. If the error corresponding to the initial indicator determination model exceeds the first error threshold, it is considered that the model performance does not meet the requirements and needs further adjustment.

[0077] In this step, first, an initial indicator determination model is trained using sample data, and the initial model parameters corresponding to the initial indicator determination model are in a randomly initialized state. The error (i.e., the first error value) between the prediction result of the initial indicator determination model and the actual performance indicator in the sample data is determined. If the first error value exceeds the preset first error threshold, this indicates that the prediction effect of the initial indicator determination model is not good, and the model needs to be optimized by updating the initial model parameters. The model parameters are iteratively adjusted until the prediction error of the model (target error value) is lower than the first error threshold. The model parameters at this time are called target model parameters. Through this step, model training and parameter optimization can significantly improve the accuracy of the model prediction performance indicators.

[0078] As an optional embodiment, determining a first error value corresponding to the initial indicator determination model includes: determining a test performance indicator value based on a test fusion characteristic value and the initial indicator determination model; calling an error function corresponding to the performance item to be tested, wherein the error function is used to determine the error value between the test performance indicator value and the actual performance indicator; determining a first error value based on the test performance indicator value, the error function and the actual performance indicator value corresponding to the test fusion characteristic value.

[0079] In this embodiment, the specific steps of determining the first error value are described.

[0080] Among them, the test performance index value is involved. The test performance index value refers to the specific power grid performance index value predicted by the initial index determination model based on the test fusion characteristic value, which is used to verify the accuracy of the initial index determination model.

[0081] Among them, the error function is involved. The error function refers to a function used to measure the degree of difference between the model prediction result and the true value. For example, if the goal is to predict the future state of the power grid, the mean square error (MSE) can be used as the loss function; if the goal is to classify the operating mode or fault type of the power grid, the cross entropy loss function can be used.

[0082] In this step, the test fusion feature value is first used as input, and the initial indicator determination model is used to make a prediction to obtain the test performance indicator value. Next, the corresponding error function is called based on the characteristics of the performance item to be measured. The choice of error function should match the type of indicator to be measured. For example, the mean squared error is suitable for regression prediction tasks, while the cross-entropy loss may be used for classification tasks. Finally, the test performance indicator value predicted by the initial indicator determination model is compared with the actual performance indicator value, and the first error value is determined using the error function. This error value reflects the accuracy of the model prediction and is an important feedback signal in the model optimization process.

[0083] Through this step, the first error value is determined, providing a quantitative indicator of the gap between the model prediction and the actual state, so that the performance of the model can be objectively and quantitatively evaluated, which facilitates the identification of model deficiencies, guides the model optimization process, and improves the model's prediction accuracy.

[0084] As an optional embodiment, the target performance indicator value corresponding to the target power grid system and the performance item to be measured is determined based on the fusion characteristic value, including: when the fusion characteristic item includes multiple sub-fusion characteristic items, determining the characteristic item contribution values ​​corresponding to the multiple sub-fusion characteristic items and the performance item to be measured respectively; determining the corresponding fusion characteristic item whose corresponding characteristic item contribution value is greater than a predetermined threshold as the target characteristic item; and determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the sub-fusion characteristic value corresponding to the target characteristic item.

[0085] In this embodiment, specific steps of determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured are described.

[0086] Among them, sub-fusion feature items are involved, and sub-fusion feature items refer to sub-feature items subdivided within the fusion feature items.

[0087] Among them, the feature item contribution value is involved. The feature item contribution value refers to a numerical value used to measure the degree of influence or importance of the sub-fusion feature item on the overall performance item to be measured.

[0088] This involves a predetermined threshold, which is a pre-set numerical standard used to screen sub-fusion features that have a significant impact on overall performance. Only when the corresponding feature contribution value exceeds the predetermined threshold will the corresponding sub-fusion feature be considered as a target feature and included in subsequent analysis.

[0089] In this step, when the fusion characteristic value contains multiple sub-fusion characteristic items, the contribution of each sub-fusion characteristic item to the overall performance item to be measured is first evaluated separately to obtain multiple characteristic item contribution values. Then, according to the preset threshold standard, the sub-fusion characteristic items with characteristic item contribution values ​​higher than the predetermined threshold are screened out as target characteristic items. Finally, based on the sub-fusion characteristic values ​​of these target characteristic items, the specific performance values ​​of the target power grid system on the performance item to be measured, that is, the target performance index values, are determined through appropriate analysis or model prediction.

[0090] Through this step, the contribution value of the feature item is determined to quantify the impact of the sub-fusion feature item in the fusion feature on the performance index, which helps to identify which data points or operation details are critical and which are negligible noise or minor factors, effectively reducing the dimension of the model input and improving the model training speed and prediction accuracy.

[0091] As an optional embodiment, after determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fusion characteristic value, it also includes: determining the true performance indicator value corresponding to the performance item to be measured; determining a second error value between the true performance indicator value and the target performance indicator value; when the second error value is greater than the second error threshold, updating the first weight value and the second weight value to obtain the first updated weight and the second updated weight, so as to determine the performance indicator value corresponding to the target power grid system and the performance item to be measured based on the first updated weight and the second updated weight.

[0092] In this embodiment, specific steps of obtaining the first updated weight and the second updated weight are described.

[0093] Among them, the real performance index value is involved. The real performance index value refers to the actual performance value of the performance item to be measured obtained through direct measurement or system recording during the actual operation of the target power grid system.

[0094] This involves a second error threshold, which is a pre-set standard value used to determine whether the second error value is within an acceptable range. If the second error value exceeds the second error threshold, it indicates that the performance indicator value predicted by the model deviates significantly from the actual situation, and the first and second weight values ​​need to be adjusted.

[0095] In this step, first, the performance of the target power grid system in the performance item to be measured, that is, the target performance index value, is predicted based on the fused characteristic value. Then, by comparing it with the actual performance index value obtained by actual measurement, a second error value is determined to evaluate the accuracy of the prediction. If the second error value is greater than the pre-set second error threshold, it means that there is a significant difference between the prediction result and the actual operating status. The first weight value and the second weight value are updated, that is, the relative importance of the operating characteristic value and the operation response characteristic value in the model is adjusted to obtain the first updated weight and the second updated weight. These updated weight values ​​will be used for retraining or optimization of the model, in order to more accurately reflect the operating status and performance of the power grid system in future predictions.

[0096] Based on the above embodiment and optional embodiment, an optional implementation manner is provided, which is described in detail below.

[0097] Traditional methods rely heavily on manual experience or simple statistical analysis, making them incapable of effectively processing the massive amounts of multi-source, heterogeneous data generated by power grid operations. As power systems become increasingly complex, these methods are exhibiting significant limitations in data processing capabilities and the depth of association mining. First, traditional methods often struggle to fully capture the spatiotemporal correlations within data streams. Power grid data not only exhibits dynamic temporal variations but also complex spatial topological relationships. Existing methods often employ time-based, segmented, or localized analysis, neglecting the holistic and interrelated nature of data across spatiotemporal dimensions. This results in an incomplete understanding of the grid's operating status and makes it difficult to accurately predict its complex dynamic behavior. Second, existing methods struggle with the integration of control and regulation flows with data streams. Control and regulation flows often involve complex decision-making logic and operational processes, while data streams contain extensive real-time measurements and historical records. Organically integrating these two streams and establishing effective association models presents a challenge. Current methods often employ simple superposition or linear combination approaches, failing to fully exploit the nonlinear relationships and deep-level interaction patterns between the two, limiting the accuracy and reliability of association analysis results. Furthermore, existing methods lack adaptability to the real-time and dynamic nature of data. The operating status of the power grid changes in real time, and traditional methods are mostly based on static data or fixed time windows for analysis, making it difficult to quickly respond to the dynamic changes of the power grid. In the face of emergency situations such as sudden failures or load changes, these methods often cannot provide effective decision support in a timely manner, affecting the emergency response capability of the power grid. In summary, the existing methods for associating business flows and data flows in power grid regulation have obvious deficiencies in terms of spatiotemporal correlation modeling, business and data fusion, real-time dynamic adaptability, and model scalability. It is urgent to introduce more advanced technical means and analysis methods to improve the scientific nature and accuracy of power grid regulation and meet the complex and changing operation requirements of modern power systems.

[0098] In view of this, an optional embodiment of the present invention provides a method for associating power grid control business flows and data flows based on a spatiotemporal graph convolutional network, which can solve the problems of insufficient spatiotemporal correlation modeling and difficulty in fusing business flows and data flows in existing methods, improve the accuracy and real-time performance of power grid control decisions, provide strong support for dynamic monitoring and prediction of power grid operating status, and play a role in ensuring the safe and stable operation of the power grid.

[0099] Figure 2 is a flow chart of a method for determining a power grid performance index provided by an optional embodiment of the present invention, such as Figure 2As shown, it mainly includes data collection and preprocessing: data collection, data cleaning, data normalization, time series alignment, data labeling and classification, and data storage and management; building a dynamic power grid spatiotemporal graph: determining the definitions of nodes and edges, building an initial power grid spatiotemporal graph, real-time data access and dynamic updating of the graph, spatiotemporal feature extraction of the graph, and graph verification and optimization; feature extraction and fusion: node feature extraction, spatiotemporal feature fusion, business flow feature extraction, data flow and business flow feature association, and feature dimensionality reduction and selection; deep learning model construction: spatiotemporal graph convolutional network design, introduction of attention mechanism and multimodal fusion mechanism design; model training and optimization: defining loss function and selecting optimization algorithm, data set partitioning and cross-validation, hyperparameter adjustment, model training and iteration, and performance evaluation and optimization.

[0100] S1. Receive a request for determining performance indicators of a target power grid system.

[0101] S2. In response to the performance indicator determination request, retrieve the operation response data and power data corresponding to the performance item to be measured according to the performance identifier.

[0102] Data collection and preprocessing ensures the comprehensiveness and integrity of the research by collecting grid operation data (similar to the power data mentioned above) and regulation business data (similar to the operational response data mentioned above). Data cleaning, normalization, and time series alignment eliminate noise, outliers, and inconsistencies in the data, improving its accuracy and usability. This stage lays a solid data foundation for subsequent spatiotemporal graph construction, feature extraction, and model training, and is a key prerequisite for the effectiveness of the entire method. It specifically includes the following steps:

[0103] Data collection: The purpose of data collection is to obtain various types of data generated during the operation of the power grid and related control business data. Specifically, data needs to be collected from multiple links in the power system, including the power generation side (such as the power output and temperature of the generator set), the transmission side (such as the current, voltage, and power of the transmission line), the substation side (such as the switch status of the substation, bus voltage, etc.), and the power consumption side (such as user load data, demand response information, etc.). At the same time, it is also necessary to collect control business flow data, such as dispatch instruction records (including instruction type, time, target device, etc.) and fault handling records (including the time, location, and handling process of the fault). This data is usually stored in different systems, such as data acquisition and monitoring control systems (SCADA systems), energy management systems (EMS systems), fault recording systems, etc. During the data collection process, it is necessary to ensure the integrity, accuracy, and timeliness of the data to provide a reliable data foundation for subsequent analysis.

[0104] Data cleaning: The goal of data cleaning is to remove noise, outliers, and erroneous data. Power grid data comes from a wide range of sources, and data quality varies. It may contain measurement errors, erroneous recordings due to equipment failures, and missing data. The first step in data cleaning is to identify and address outliers. This can be achieved by setting reasonable thresholds to detect data points outside the normal range and, depending on the specific situation, removing or correcting these outliers. For example, for voltage data, if a measurement is significantly above or below the normal operating range, it may be due to equipment failure or measurement error, requiring further inspection and processing. Data cleaning also addresses missing data. Missing data can be caused by equipment failure, communication interruptions, and other reasons. Missing data can be filled using interpolation methods such as linear interpolation and polynomial interpolation. Data cleaning improves data accuracy and reliability, providing high-quality data support for subsequent analysis and modeling.

[0105] Data normalization: Data normalization is the process of converting data of different dimensions and magnitudes to the same numerical range. Its purpose is to eliminate dimensional and magnitude differences between the data to facilitate subsequent model training and calculations. Power grid data contains various types of electrical quantities and state variables, such as voltage (in volts), current (in amperes), and power (in watts), each with varying dimensions and magnitudes. Normalization maps this data to a uniform range such as [0, 1] or [-1, 1], enabling the model to process it more efficiently. Data normalization can improve the model's convergence speed and training effectiveness, enabling the model to better learn the characteristics and patterns in the data.

[0106] Time series alignment: The goal of time series alignment is to align time series data from different data sources to a consistent time interval and time axis. Different devices and systems in the power grid may have different data sampling frequencies. For example, a SCADA system may sample every few seconds, while a PMU (Phasor Measurement Unit) may sample dozens of times per second. This difference in sampling frequency leads to data misalignment in the temporal dimension, impacting subsequent analysis and modeling. Time series alignment can be achieved through time interpolation or sampling rate adjustment. Time interpolation uses mathematical methods to insert reasonable data values ​​at missing time points. Common interpolation methods include linear interpolation and spline interpolation. For example, for data with a low sampling frequency, linear interpolation can be used to insert new data points between adjacent sampling points to align the time intervals with the high-frequency data. Sampling rate adjustment unifies the data sampling frequency through downsampling or upsampling. Downsampling reduces the sampling frequency of high-frequency data, preserving key information through methods such as averaging and maximum values. Upsampling increases the sampling frequency of low-frequency data, filling in missing data through methods such as interpolation. Through time series alignment, the uniformity and consistency of data in the time dimension can be ensured, providing a reliable data basis for subsequent spatiotemporal correlation analysis.

[0107] Data Labeling and Classification: The purpose of data labeling and classification is to provide clear labels and classification information for subsequent model training and analysis. Grid control service flow data typically includes dispatch instructions and fault handling records. This data needs to be correlated with grid operation data (such as equipment status and load changes) to analyze the impact of control operations on grid operation. Data labeling can be performed using expert knowledge or automated algorithms. For example, dispatch instruction data can be annotated based on instruction type (such as power generation plan adjustment, equipment switching), and execution time to identify corresponding grid operation changes. For fault handling records, grid operation data before and after the fault can be annotated based on fault type (such as short circuit, equipment failure), and handling process. Data classification categorizes the labeled data according to different service scenarios or operating states, allowing the model to better learn the characteristics and patterns of each scenario. For example, data can be categorized into normal operation, fault state, and peak load state. Data labeling and classification provide clear goals and directions for model training, improving model accuracy and generalization.

[0108] Data storage and management: The goal of data storage and management is to efficiently store and manage cleaned, normalized, and annotated data so that subsequent analysis and modeling can quickly access and utilize this data. Power grid data is large and complex, requiring an appropriate storage solution to ensure efficient data storage and rapid retrieval. Common storage solutions include relational databases (such as MySQL and Oracle) and non-relational databases (such as MongoDB and Hadoop). Relational databases are suitable for storing and managing structured data, enabling efficient data queries and operations using the Structured Query Language (SQL). Non-relational databases are more suitable for storing large amounts of unstructured data, such as log data and image data. During data storage, data must be properly organized and indexed to improve retrieval efficiency. For example, data can be indexed by fields such as timestamps and device identifiers to quickly retrieve the required data for model training and analysis. Furthermore, data backup and recovery mechanisms must be established to ensure data security and integrity. Efficient data storage and management provide reliable data support for subsequent spatiotemporal graph convolutional network modeling and association analysis, ensuring the smooth progress of the entire research process.

[0109] S3. Determine the operation characteristic value and operation response characteristic value of the target power grid system based on the power data and the operation response data.

[0110] First, a dynamic grid spatiotemporal graph (similar to the target network graph) is constructed. By defining nodes (similar to the device nodes and edges) and transmission edges, grid devices and their connections are abstracted into a graph structure that intuitively reflects the topological characteristics of the grid and the interactions between devices. A dynamic update mechanism ensures that the graph can reflect grid operation changes in real time, providing a dynamic spatiotemporal context for spatiotemporal correlation analysis. This stage provides a structured data framework for subsequent feature extraction and deep learning model construction, and is an important foundation for exploring grid operation patterns. Specifically, it includes the following steps:

[0111] Determine the definition of nodes and edges: The purpose of defining nodes and edges is to abstract the physical devices and electrical connections in a power grid into a graph structure. In a power grid, nodes can represent various key devices, such as power plants, substations, transformers, and the ends of transmission lines. Each node has unique electrical properties and operating status. For example, a power plant node may contain attributes such as power output and fuel consumption; a substation node may contain information such as voltage level and equipment status. Edges represent the electrical connections between nodes, such as transmission lines and busbar connections. Edge attributes can include line resistance, reactance, and transmission power. By clearly defining nodes and edges, complex power grid topologies can be transformed into a clear graph model, providing a structural foundation for subsequent spatiotemporal correlation analysis. For example, for a regional power grid consisting of multiple substations and transmission lines, each substation can be considered a node, and each transmission line an edge. This allows for a topological graph of the regional power grid, intuitively reflecting the grid's connectivity and device distribution.

[0112] Constructing an initial grid spatiotemporal graph (similar to the initial network graph described above): Constructing an initial grid spatiotemporal graph involves combining defined nodes and edges into a complete graph structure and assigning initial attribute values ​​to each node and edge. Based on the grid's physical topology and device parameters, all nodes and edges are connected according to their actual connectivity relationships, forming a static initial graph. When constructing the initial graph, initial attribute values ​​must be assigned to each node and edge based on grid operational data. For example, initial node attribute values ​​can include electrical quantities such as voltage and power at the current moment, while initial edge attribute values ​​can include line current and power transmission direction. The construction of the initial graph must ensure that it accurately reflects the grid's initial operational state, providing a foundation for subsequent dynamic updates and spatiotemporal correlation analysis. For example, during normal grid operation, an initial graph constructed based on real-time measurement data can be used to analyze the grid's steady-state operating characteristics. In the event of a grid fault, an initial graph constructed based on pre-fault operational data can be used for fault diagnosis and analysis.

[0113] Real-time data access and dynamic graph updates: Real-time data access and dynamic graph updates are key steps in ensuring that the grid's spatiotemporal graph reflects the grid's real-time operational status. Grid operational status is dynamic, and equipment operating parameters and topology are subject to change at any time. Therefore, real-time access to grid operational data is necessary, including changes in node electrical quantities, equipment switching operations, and fault information. The grid's spatiotemporal graph can then be dynamically updated based on this real-time data. For example, when a transmission line is disconnected due to a fault, the edges in the graph need to be updated promptly, changing from a connected state to a disconnected state. Similarly, when the load at a substation changes, the power attributes of the corresponding nodes need to be updated. Dynamic updates can be achieved by acquiring data from a real-time monitoring system and rapidly updating the graph's structure and attribute values ​​using data stream processing techniques. Through real-time data access and dynamic graph updates, the grid's spatiotemporal graph can reflect changes in the grid's operational status in real time, providing accurate model support for real-time monitoring, fault diagnosis, and control decisions. For example, in the event of a grid fault, the dynamically updated graph can quickly locate the fault and analyze its impact on grid operation, providing timely decision support for dispatchers.

[0114] Graph Verification and Optimization: Graph verification and optimization are critical steps in ensuring the accuracy and validity of power grid spatiotemporal graphs. Constructed power grid spatiotemporal graphs require verification to ensure they accurately reflect the actual operating status of the power grid and provide reliable support for subsequent analysis. Verification can be accomplished by comparing the graph with actual grid operating data. For example, the electrical quantities of nodes in the graph can be compared with actual measured data to check for significant deviations. Simulating grid operating scenarios can also be used to verify that dynamic updates to the graph accurately reflect changes in the grid status. If deviations or deficiencies are found in the graph, the graph structure or attribute values ​​need to be optimized. For example, if the power transmission direction of a line is found to be inconsistent with the actual situation, the line connections or the accuracy of data access may need to be re-examined. Graph verification and optimization ensure the quality of the power grid spatiotemporal graph, enabling it to better serve the correlation analysis of power grid regulation business flows and data flows. For example, after a power grid upgrade, optimizing the graph structure and attribute values ​​can ensure that the graph accurately reflects the new operating status of the grid, providing strong support for optimized grid dispatch.

[0115] Next, feature vectors reflecting the grid's operating status are extracted from the grid's spatiotemporal graph to provide input for subsequent model training and analysis. In the grid's spatiotemporal graph, each node and edge possesses rich spatiotemporal characteristics. A node's spatiotemporal characteristics include the temporal variations of its own electrical quantities (such as voltage and power time series) and its interactions with neighboring nodes (such as power transfer relationships). Edge spatiotemporal characteristics include changes in line current and power transfer direction. By designing appropriate feature extraction methods, this complex spatiotemporal information can be transformed into concise feature vectors. For example, a sliding window technique can be used to extract statistical features such as a node's average power and maximum power over a period of time. Alternatively, the dynamic characteristics of edges can be extracted by calculating the rate of change of power transfer between nodes. Spatiotemporal feature extraction requires full consideration of the grid's physical characteristics and operating patterns to ensure that the extracted features effectively reflect the grid's operating status and changing trends. For example, during peak load periods, extracted spatiotemporal features can highlight load growth trends and the operating stress of key equipment, providing important insights for grid scheduling decisions. The specific steps involved are:

[0116] Node feature extraction (similar to the operational feature values ​​described above): Node feature extraction involves extracting feature vectors from each node in the grid's spatiotemporal graph that reflect its operational status. These features serve as the basic input for subsequent model training. Node features typically include electrical quantity characteristics (such as voltage amplitude, current magnitude, active power, reactive power, etc.) and operational status characteristics (such as equipment temperature and pressure). For example, for a generator node, features might include output power, fuel consumption rate, and equipment operating temperature; for a substation node, features might include bus voltage and equipment load factor. These features provide a visual representation of the node's current operational status. Furthermore, time series features can be extracted from historical node data, such as the cyclical changes in the daily load curve and the power fluctuation range. Extracting these rich node features provides detailed and comprehensive node information for subsequent spatiotemporal correlation analysis, helping the model better understand the operational characteristics of each node in the grid and its role in the overall grid operation.

[0117] Extraction of business flow features (same as the above-mentioned operation response feature values): Business flow feature extraction is the extraction of feature vectors from the control business flow data that can reflect the control operations and decision-making process. These features will be correlated with the data flow features for analysis. Control business flow data typically includes dispatch instruction records (such as power generation plan adjustments, equipment switching operations, etc.), fault handling records (such as fault occurrence time, handling process, etc.), etc. For example, for dispatch instructions, information such as instruction type, instruction time, and target equipment can be extracted as features; for fault handling records, information such as fault type, fault location, and handling time can be extracted as features. These features can reflect the intention and execution process of the control business operations, providing business background information for subsequent correlation analysis. By extracting business flow features, the control business logic can be combined with power grid operation data, helping the model understand the impact of control operations on the power grid operation status, thereby providing a basis for optimizing the control strategy.

[0118] S4. Determine a fusion feature value according to the operation feature value and the first weight value corresponding to the operation feature value, the operation response feature value and the second weight value corresponding to the operation response feature value.

[0119] Feature extraction and fusion extracts node electrical quantity and time series features and integrates data stream features with control traffic features to comprehensively reflect the grid's operating status and the impact of control operations. This stage not only enriches the model's input information but also improves model training efficiency and generalization capabilities through feature dimensionality reduction and selection. The fused features provide richer semantic information for deep learning models and are a key step in improving model performance.

[0120] Spatiotemporal feature fusion combines the electrical quantity characteristics of a node with time series characteristics to capture the spatiotemporal correlation of power grid data. The operating state of a power grid not only changes over time but is also affected by its spatial topology. Therefore, it is necessary to fuse the electrical quantity characteristics of a node with the time series characteristics to form a more comprehensive spatiotemporal feature representation. For example, a time series model (such as LSTM or GRU) can be used to model the historical electrical quantity data of a node and extract dynamic features along the time dimension. Simultaneously, graph convolution operations can be combined to aggregate the features of neighboring nodes and enhance the spatial feature representation of the node. Through this fusion approach, the model can simultaneously consider the changes in the node's own characteristics at different times as well as the spatial correlation between the node and its neighboring nodes. For example, during peak load periods on the power grid, spatiotemporal feature fusion can more accurately identify which nodes have the fastest load growth and how the power transmission relationship between these nodes changes. This spatiotemporal feature fusion can provide more precise decision support for power grid regulation and control, helping dispatchers better understand the dynamic operating characteristics of the power grid. Specifically, it includes the following steps:

[0121] Data Stream and Service Flow Feature Correlation: Data stream and service flow feature correlation involves fusing extracted grid operation data features with control service flow features to uncover potential correlations between the two. By combining data stream features (such as node spatiotemporal characteristics) with service flow features (such as dispatch instruction features), a comprehensive feature representation can be constructed for subsequent model training and analysis. For example, the timing of dispatch instructions can be aligned with the timing of grid operation data to analyze changes in grid operation status before and after the dispatch instruction execution. By designing specific fusion mechanisms, such as weighted summation, feature concatenation, or an attention mechanism, the significant impact of service operations on grid operation data can be highlighted. For example, using the attention mechanism, the model can automatically learn which service operations have the greatest impact on grid operation status, thereby providing more accurate decision support for grid control. This data stream and service flow feature correlation helps the model better understand the complex relationship between control service operations and grid operation status, providing strong support for optimizing grid control strategies.

[0122] Introduction of the Attention Mechanism: The attention mechanism is introduced to enhance the model's ability to focus on key information and improve its sensitivity to important nodes, edges, or time steps. In the correlation analysis of power grid control business flows and data streams, certain nodes (such as critical substations), certain time steps (such as peak load periods), or certain business operations (such as emergency fault handling) may have a more significant impact on the grid's operating status. By introducing the attention mechanism into the Spatiotemporal Graph Convolutional Network (STGCN), the model can automatically learn the importance weights of this key information, giving it greater attention during the modeling process. For example, the attention mechanism can perform weighted aggregation on the neighborhood of a node to highlight the influence of important neighboring nodes; it can also assign higher weights to specific time steps in a time series to capture changes at key points in time. This mechanism not only improves the model's prediction accuracy but also enhances its interpretability, helping dispatchers understand the basis for model decisions and thus better apply it to practical power grid control scenarios.

[0123] Designing a multimodal fusion mechanism: Designing a multimodal fusion mechanism is a key step in deeply integrating the characteristics of power grid data streams with those of control traffic flows. Grid operational data (such as voltage, current, and power) and control traffic data (such as dispatch instructions and fault handling records) reflect the physical operating state of the power grid and the intended control, respectively, and there exists a complex interactive relationship between the two. By designing multimodal fusion mechanisms, such as feature concatenation, weighted summation, or attention-based fusion methods, the characteristics of these two modalities can be effectively combined. For example, feature concatenation can be used to combine data stream and traffic flow features into a comprehensive feature vector, which is then input into a spatiotemporal graph convolutional network (STGCN model). Alternatively, an attention mechanism can be used to dynamically adjust the weights of the two modal features to highlight the more important features in specific scenarios. This multimodal fusion mechanism can more comprehensively explore the correlations between power grid control traffic flows and data streams, providing a more accurate basis for optimizing power grid control strategies and enhancing the intelligent level of power grid operation.

[0124] Feature Dimensionality Reduction and Selection: Feature Dimensionality Reduction and Selection involves selecting the most representative and effective feature subset from the large number of extracted features to improve model training efficiency and generalization. The feature extraction process may generate a large number of features, some of which may contain redundant information or contribute little to model training. Therefore, feature dimensionality reduction and selection are necessary. Feature dimensionality reduction can be achieved through methods such as principal component analysis (PCA) and linear discriminant analysis (LDA), mapping the high-dimensional feature space to a lower-dimensional space while retaining the most important information. Feature selection can be achieved through statistical methods (such as correlation analysis), model-based methods (such as feature importance scoring), or search-based methods (such as genetic algorithms). For example, correlation analysis can eliminate features with low correlation with the target variable, while feature importance scoring can select features that contribute most to model predictions. Feature Dimensionality Reduction and Selection can reduce the model's input dimensionality, reduce model complexity, improve model training speed and generalization, and avoid overfitting. For example, in the power grid control model, feature selection can retain the features most relevant to the changes in the power grid operating status, thereby improving the model's ability to detect abnormal power grid conditions and the accuracy of control decisions.

[0125] S5. Determine the target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fused characteristic value.

[0126] Using a deep learning model and fusion feature values, the target performance indicator values ​​corresponding to the target power grid system and the performance items to be measured are determined. The deep learning model is constructed by designing a spatiotemporal graph convolutional network (STGCN, the same as the performance indicator determination model mentioned above), combined with an attention mechanism and a multimodal fusion mechanism, which can simultaneously capture the spatiotemporal correlation of power grid data and the complexity of the control business. The model constructed at this stage can learn the inherent laws of power grid operation and explore the deep connection between control operations and power grid status. The training and optimization of the model further improves its performance and adaptability, providing a powerful decision support tool for power grid control. The specific steps include:

[0127] Design of Spatiotemporal Graph Convolutional Network: The design of the STGCN is the core step in building the model. Its purpose is to capture the spatiotemporal correlation of power grid data through graph convolution operations and time series modeling modules. STGCN combines graph convolutional networks (GCN) and time series models (such as LSTM or GRU) to simultaneously process the spatial dependencies and temporal dynamics of power grid data. The graph convolution layer learns the spatial feature representation of nodes by aggregating the features of nodes and their neighboring nodes; the time series module processes the time series characteristics of node features to capture the changing patterns of data in the time dimension. For example, in power grid operation, the power transmission relationship between substations can be analyzed through the graph convolution layer, while the time series module can capture the changing trends of substation loads over time. By rationally designing the network structure of the STGCN, including hyperparameters such as the number of layers, convolution kernel size, and time step, the complex spatiotemporal dynamics of the power grid can be better modeled, providing a strong model foundation for subsequent correlation analysis.

[0128] Model training and optimization: By defining appropriate loss functions and optimization algorithms, rationally partitioning the data set and performing cross-validation, and adjusting hyperparameters, model training and optimization enable the model to learn effective features and patterns on the training data and demonstrate good generalization capabilities on the validation data. This phase, through continuous training and optimization, gradually improves the model's prediction accuracy and reliability, ensuring it accurately reflects the relationship between power grid regulation business flows and data flows, providing reliable model support for practical applications. Specifically, the following steps are included:

[0129] Defining a loss function and selecting an optimization algorithm: Defining a loss function and selecting an optimization algorithm are fundamental to model training. The loss function (similar to the error function described above) measures the difference between the model's predicted value and the true value, while the optimization algorithm adjusts the model parameters to minimize the loss function. For correlation analysis of power grid control business flows and data streams, the choice of loss function depends on the specific task. For example, if the goal is to predict the future state of the power grid, the mean squared error (MSE) loss function can be used; if the goal is to classify the power grid's operating mode or fault type, the cross-entropy loss function can be used. The choice of optimization algorithm is also crucial. Common optimization algorithms include stochastic gradient descent (SGD), the adaptive moment estimation optimizer (Adam optimizer), and root mean square propagation (RMSprop). The Adam optimizer is widely used due to its adaptive learning rate and excellent convergence performance. By properly defining the loss function and selecting the optimization algorithm, the model can effectively learn the characteristics and patterns in the data during training, providing a foundation for subsequent optimization adjustments.

[0130] Dataset Partitioning and Cross-Validation: Dataset partitioning (similar to the sample data described above) and cross-validation are crucial steps in evaluating model performance. By dividing the dataset into training, validation, and test sets, we ensure that the model's performance evaluation on different data subsets is representative and reliable. The training set is used to train the model, the validation set is used to adjust the model's hyperparameters and prevent overfitting, and the test set is used to ultimately evaluate the model's generalization ability. Cross-validation methods (such as k-fold cross-validation) can further improve the accuracy of model evaluation. For example, in k-fold cross-validation, the dataset is partitioned into k subsets, with k-1 subsets used for training each time. The remaining subset is used for validation. This is repeated k times, and the average value is taken as the final validation result. Through appropriate dataset partitioning and cross-validation, we can ensure stable model performance on different data subsets, avoid performance fluctuations caused by random data partitioning, and thus provide a reliable basis for model optimization and adjustment.

[0131] Hyperparameter adjustment (same as the model parameters described above): Hyperparameter adjustment is a key step in optimizing model performance. Hyperparameters include learning rate, batch size, number of network layers, and convolution kernel size. These parameters have a significant impact on the model's training results and convergence speed. Hyperparameter adjustment is typically performed through methods such as grid search, random search, or Bayesian optimization. For example, grid search exhaustively searches for the optimal parameter combination within a predefined hyperparameter grid; random search randomly selects hyperparameter combinations for evaluation. During the adjustment process, the optimal hyperparameters need to be selected based on performance metrics on the validation set (such as loss function value and accuracy). For example, adjusting the learning rate can control the model's convergence speed. Excessively high learning rates may prevent the model from converging, while excessively low learning rates may result in excessively slow training. Through reasonable hyperparameter adjustment, model performance can be significantly improved, enabling it to better adapt to the complex relationship between power grid regulation business flows and data flows.

[0132] Model training and iteration: Model training and iteration is the process by which a model learns data features and patterns. Through multiple iterations of training, the model can gradually optimize its parameters and improve prediction accuracy. During training, the model performs forward and backward propagation on the training set data according to the defined loss function and optimization algorithm, updating the model parameters. Model performance improves with each iteration. During training, it is necessary to monitor model convergence by observing changes in the loss function value. If the loss function value no longer decreases significantly, the model has converged and training can be stopped. In addition, to prevent overfitting, an early stopping mechanism can be implemented. This involves monitoring model performance on a validation set and terminating training early if performance on the validation set stops improving after a certain number of iterations. Through effective model training and iteration, the model can learn effective features and patterns on the training set and generalize well to the validation set, providing reliable decision support for power grid control.

[0133] Performance evaluation and optimization: Performance evaluation and optimization are crucial steps in ensuring that a model meets practical application requirements. By evaluating the model's performance on a test set, the model's generalization capability and practical application value can be verified. Performance evaluation metrics typically include prediction accuracy, recall, and mean squared error (MSE). The choice of specific metrics depends on the task type. For example, in power grid fault prediction tasks, recall is a crucial metric as it reflects the model's ability to detect faults. In power grid status prediction tasks, mean squared error measures the deviation between the model's predicted value and the true value. Based on the performance evaluation results, the model can be further optimized. For example, if the model's performance is found to be poor in certain scenarios, relevant data features can be added or the model structure can be adjusted. Through continuous performance evaluation and optimization, the model's performance can be gradually improved, enabling it to better adapt to the complex interrelationships between power grid regulation business flows and data streams, providing strong support for the safe and stable operation of the power grid.

[0134] After completing the above five steps, a method for associating business flows and data flows in power grid control based on spatiotemporal graph convolutional networks is formed. This method solves the problems of insufficient spatiotemporal correlation modeling and difficulty in fusing business flows and data flows in existing methods, improves the accuracy and real-time performance of power grid control decisions, provides strong support for dynamic monitoring and prediction of power grid operating status, and plays a role in ensuring the safe and stable operation of the power grid.

[0135] Through the above optional implementation, at least the following beneficial effects can be achieved:

[0136] (1) The method provided by the optional embodiment of the present invention has a powerful spatiotemporal correlation modeling capability. When processing power grid data, traditional methods often find it difficult to fully capture the spatiotemporal correlation characteristics in the data stream. They often adopt time-segmented or local analysis methods, ignoring the integrity and correlation of the data in the spatiotemporal dimension. The method provided by the optional embodiment of the present invention is based on a spatiotemporal graph convolutional network. By constructing a dynamic power grid spatiotemporal graph, the power grid equipment is used as nodes, and the physical connections and electrical relationships are used as edges. It can intuitively reflect the topological structure of the power grid and the relationship between the devices. On this basis, the spatiotemporal graph convolutional network combines graph convolution operations and time series modeling modules, such as long short-term memory networks (LSTMs) or gated recurrent units (GRUs), and can simultaneously process the spatiotemporal features of nodes. The graph convolution layer can capture the spatial dependencies between nodes and aggregate the features of neighboring nodes to enhance the spatial feature expression of the nodes; the time series module can process the time series characteristics of node features and capture the dynamic changes of data in the time dimension. This spatiotemporal correlation modeling capability enables the new method to more accurately understand the spatiotemporal evolution of power grid operating conditions and predict potential abnormalities in the grid, such as equipment failures and load overloads. This provides strong support for fault warning and preventive maintenance, effectively reducing grid operation risks and improving power supply reliability.

[0137] (2) The method provided by the optional embodiment of the present invention can realize efficient fusion analysis of business flow and data flow. The existing methods have obvious deficiencies in processing the fusion of control business flow and data flow. They usually adopt simple superposition or linear combination methods, which cannot fully explore the nonlinear relationship and deep interaction mode between the two, resulting in limited accuracy and reliability of the correlation analysis results. The method provided by the optional embodiment of the present invention realizes efficient fusion analysis of business flow and data flow by introducing attention mechanism and multimodal fusion mechanism. First, the control business flow data (such as dispatch instructions, fault handling records, etc.) are converted into feature vectors and combined with data flow features. Then, the attention mechanism is added to the spatiotemporal graph convolutional network, which can highlight the key nodes and time steps, so that the model pays more attention to the important impact of business operations on specific data indicators. In addition, the multimodal fusion mechanism can deeply fuse data flow features and business flow features in the spatiotemporal graph convolutional network and fully explore the correlation relationship between them. This efficient fusion analysis enables the new method to more accurately identify the complex correlation between control business operations and power grid operation data, and provide more comprehensive and accurate information support for the optimization of power grid control strategies. For example, by analyzing the changes in the operating status of the power grid under different business decisions, potential risks and problems can be discovered in advance, so as to formulate more reasonable and efficient scheduling plans and fault response measures, and improve the operating efficiency, stability and reliability of the power grid.

[0138] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0139] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, or of course 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, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several methods for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0140] Example 2

[0141] According to an embodiment of the present invention, a device for implementing the above-mentioned method for determining the power grid performance index is also provided. Figure 3 FIG. 1 is a structural block diagram of a device for determining a power grid performance index according to an embodiment of the present invention. Figure 3 As shown, the apparatus includes: a receiving module 302, a responding module 304, a first determining module 306, a second determining module 308 and a third determining module 310. The apparatus will be described in detail below.

[0142] A receiving module 302 is configured to receive a performance indicator determination request for a target power grid system, wherein the performance indicator determination request carries a performance identifier of the performance item to be measured; a response module 304 is connected to the above-mentioned receiving module 302, and is configured to respond to the performance indicator determination request and, based on the performance identifier, retrieve operation response data and power data corresponding to the performance item to be measured; a first determination module 306 is connected to the above-mentioned response module 304, and is configured to determine an operation characteristic value and an operation response characteristic value of the target power grid system based on the power data and the operation response data, wherein the operation characteristic value represents the operation status of multiple power devices in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to the operation instruction in the target power grid system; a second determination module 308 is connected to the above-mentioned first determination module 306, and is configured to determine a fused characteristic value based on the operation characteristic value and a first weight value corresponding to the operation characteristic value, and the operation response characteristic value and a second weight value corresponding to the operation response characteristic value; a third determination module 310 is connected to the above-mentioned second determination module 308, and is configured to determine a target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fused characteristic value.

[0143] It should be noted here that the above-mentioned receiving module 302, response module 304, first determination module 306, second determination module 308 and third determination module 310 correspond to steps S102 to S110 in the method for determining the power grid performance indicators. The examples and application scenarios implemented by the multiple modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned embodiment 1.

[0144] Example 3

[0145] According to another aspect of an embodiment of the present invention, an electronic device is provided, including: a processor; and a memory for storing processor-executable data, wherein the processor is configured to execute to implement any of the above methods for determining a power grid performance indicator.

[0146] Example 4

[0147] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. When the computer-readable storage medium is executed by a processor of an electronic device, the electronic device can execute any of the above-mentioned methods for determining a power grid performance indicator.

[0148] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0149] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0150] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0151] The units described as separate components may or may not be physically separate, and 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 may be selected according to actual needs to achieve the purpose of the present embodiment.

[0152] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0153] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, 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. The computer software product is stored in a storage medium and includes several methods for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0154] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for determining a power grid performance index, characterized in that: include: receiving a performance indicator determination request for a target power grid system, wherein the performance indicator determination request carries a performance identifier of a performance item to be measured; In response to the performance indicator determination request, and based on the performance identifier, retrieve operation response data and power data corresponding to the performance item to be measured; Determining, based on the power data and the operation response data, an operation characteristic value and an operation response characteristic value of the target power grid system, wherein the operation characteristic value represents the operating status of multiple power devices in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to the operation instruction in the target power grid system; Determining a fusion feature value according to the operation feature value and a first weight value corresponding to the operation feature value, the operation response feature value and a second weight value corresponding to the operation response feature value; According to the fused characteristic value, a target performance indicator value corresponding to the target power grid system and the performance item to be measured is determined.

2. The method according to claim 1, characterized in that The determining, based on the power data and the operation response data, the operation characteristic value and the operation response characteristic value of the target power grid system includes: Determining an initial network graph corresponding to the target power grid system, wherein the initial network graph includes a plurality of device nodes and a plurality of transmission edges, the plurality of device nodes are respectively provided with corresponding node data, the corresponding node data includes status data corresponding to a plurality of operating status items of the corresponding power equipment, and the plurality of transmission edges are used to represent power transmission directions of the plurality of power equipment in the target power grid system; Determining, based on the power data, a plurality of real-time operating status values ​​corresponding to the plurality of power devices; Determining a target network diagram according to the initial network diagram and a plurality of real-time operating status values ​​corresponding to the plurality of power devices; The operation characteristic value is determined according to the target network diagram.

3. The method according to claim 1, characterized in that Determining, based on the fused characteristic value, a target performance indicator value corresponding to the target power grid system and the performance item to be measured includes: Retrieving a performance indicator determination model corresponding to the performance item to be measured, wherein the performance indicator determination model is provided with target model parameters, the target model parameters are determined based on an initial indicator determination model provided with initial model parameters and sample data corresponding to the performance item to be measured, the sample data including a sample fusion feature value and a sample performance indicator value of the target power grid system; The target performance indicator value corresponding to the target power grid system and the performance item to be measured is determined according to the performance indicator determination model and the fusion characteristic value.

4. The method according to claim 3, characterized in that Before calling the performance indicator determination model corresponding to the performance item to be measured, the method further includes: Determining an initial indicator determination model corresponding to the performance item to be measured based on the sample data, wherein the initial indicator determination model is provided with initial model parameters; determining a first error value corresponding to the initial indicator determination model; When the first error value is greater than a first error threshold, updating the initial model parameters until target model parameters are obtained, wherein a target error value corresponding to the target model parameters is less than the first error threshold; A performance indicator determination model corresponding to the performance item to be measured is determined based on the target model parameters and the initial indicator determination model.

5. The method according to claim 4, characterized in that The determining of a first error value corresponding to the initial indicator determination model includes: Determine the test performance index value based on the test fusion feature value and the initial index determination model; Retrieving an error function corresponding to the performance item to be measured, wherein the error function is used to determine an error value between the test performance indicator value and the actual performance indicator; The first error value is determined according to the test performance indicator value, the error function and the actual performance indicator value corresponding to the test fusion feature value.

6. The method according to claim 1, characterized in that Determining, based on the fused characteristic value, a target performance indicator value corresponding to the target power grid system and the performance item to be measured includes: In the case where the fusion feature item includes multiple sub-fusion feature items, determining feature item contribution values ​​corresponding to the multiple sub-fusion feature items and the performance item to be measured respectively; Determining a corresponding fusion feature item whose corresponding feature item contribution value is greater than a predetermined threshold as a target feature item; According to the sub-fusion characteristic value corresponding to the target characteristic item, the target performance indicator value corresponding to the target power grid system and the performance item to be measured is determined.

7. The method according to any one of claims 1 to 6, characterized in that After determining the target performance indicator value corresponding to the target power grid system and the performance item to be measured based on the fused characteristic value, the method further includes: Determining a true performance indicator value corresponding to the performance item to be measured; Determining a second error value between the actual performance indicator value and the target performance indicator value; When the second error value is greater than the second error threshold, the first weight value and the second weight value are updated to obtain a first updated weight and a second updated weight, so as to determine the performance indicator value corresponding to the target power grid system and the performance item to be measured based on the first updated weight and the second updated weight.

8. A device for determining a power grid performance index, characterized in that: include: A receiving module, configured to receive a performance indicator determination request for a target power grid system, wherein the performance indicator determination request carries a performance identifier of a performance item to be measured; a response module, configured to respond to the performance indicator determination request and retrieve, based on the performance identifier, operation response data and power data corresponding to the performance item to be measured; a first determining module, configured to determine an operation characteristic value and an operation response characteristic value of the target power grid system based on the power data and the operation response data, wherein the operation characteristic value represents the operation status of multiple power devices in the target power grid system within a predetermined time period, and the operation response characteristic value represents the response status corresponding to the operation instruction in the target power grid system; a second determining module, configured to determine a fusion feature value based on the operation feature value and the first weight value corresponding to the operation feature value, the operation response feature value and the second weight value corresponding to the operation response feature value; The third determination module is configured to determine, based on the fusion characteristic value, a target performance indicator value corresponding to the target power grid system and the performance item to be measured.

9. An electronic device, characterized in that: include: processor; a memory for storing executables of the processor; The processor is configured to execute the method for determining a power grid performance indicator according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that When the computer-readable storage medium is executed by a processor of an electronic device, the electronic device is enabled to execute the method for determining a power grid performance indicator according to any one of claims 1 to 7.

Images