Power grid operation strategy determination method and device, and electronic equipment

Abstract

This invention discloses a method, apparatus, and electronic device for determining the operation strategy of a power grid. The method includes: determining factor matrices corresponding to multiple operational dimensions in the previous time period, and the current operation strategy and current operational efficiency of the power grid in the current time period; determining a predicted operation strategy and predicted operational efficiency of the power grid in the current time period based on the factor matrices corresponding to the multiple operational dimensions; determining an efficiency impact index of the power grid based on the strategy deviation between the current operation strategy and the predicted operation strategy, and the efficiency deviation between the current operational efficiency and the predicted operational efficiency; and updating the current operation strategy based on the efficiency impact index, the current operational efficiency, and the operational constraints of the power grid to obtain an updated operation strategy for the power grid in the next time period. This invention solves the technical problem in related technologies where unreasonable determination of the power grid operation strategy leads to low power grid operational efficiency.

Description

Technical Field

[0001] This invention relates to the field of power systems, and more specifically, to a method, apparatus, and electronic equipment for determining the operation strategy of a power grid. Background Technology

[0002] In related technologies, the power grid's operational efficiency can be ensured to meet operational requirements by determining the power grid's operating strategy. However, in these technologies, there is a technical problem where an unreasonable determination of the power grid's operating strategy leads to low power grid operational efficiency.

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

[0004] This invention provides a method, apparatus, and electronic device for determining the operation strategy of a power grid, in order to at least solve the technical problem in the related art where the determination of the operation strategy of a power grid is unreasonable, resulting in low power grid operation efficiency.

[0005] According to one aspect of the present invention, a method for determining the operation strategy of a power grid is provided, comprising: determining factor matrices corresponding to multiple operation dimensions in a previous time period, and the current operation strategy and current operation efficiency of the power grid in the current time period, wherein the factor matrices are used to characterize the correlation between power grid parameters within the corresponding operation dimensions; determining a predicted operation strategy and predicted operation efficiency of the power grid in the current time period based on the factor matrices corresponding to the multiple operation dimensions; determining an efficiency impact index of the power grid based on the strategy deviation between the current operation strategy and the predicted operation strategy, and the efficiency deviation between the current operation efficiency and the predicted operation efficiency, wherein the efficiency impact index is used to characterize the degree of influence of the strategy deviation on the efficiency deviation; and updating the current operation strategy based on the efficiency impact index, the current operation efficiency, and the operation constraints of the power grid to obtain an updated operation strategy of the power grid in the next time period.

[0006] Optionally, determining the predicted operation strategy of the power grid in the current time period based on the factor matrices corresponding to the multiple operating dimensions includes: determining a multidimensional feature tensor of the power grid in the previous time period, wherein the multidimensional feature tensor is used to characterize the correlation between the multiple operating dimensions; determining the predicted operating state of the power grid in the current time period based on the multidimensional feature tensor and the factor matrices corresponding to the multiple operating dimensions; and determining the predicted operation strategy of the power grid in the current time period based on the predicted operating state.

[0007] Optionally, determining the predicted operating state of the power grid in the current time period based on the multidimensional feature tensor and the factor matrices corresponding to the multiple operating dimensions includes: when the multiple operating dimensions include equipment dimension, spatial dimension, state dimension, operating condition dimension, and time dimension, determining the spatial coupling characteristics of the power grid based on the factor matrices of the equipment dimension, the spatial dimension, the state dimension, and the operating condition dimension, wherein the spatial coupling characteristics are the coupling characteristics of the operating states between devices in the power grid in the spatial dimension; determining the temporal dependency characteristics of the power grid based on the factor matrices of the time dimension and the state dimension, wherein the temporal dependency characteristics are the dependency characteristics of the operating states of the power grid in the time dimension; and determining the predicted operating state of the power grid in the current time period based on the spatial coupling characteristics and the temporal dependency characteristics.

[0008] Optionally, determining the spatial coupling characteristics of the power grid based on the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension includes: determining the device adjacency matrix of the power grid, wherein the device adjacency matrix is ​​used to characterize the topological connection relationship between any two devices in the power grid; and determining the spatial coupling characteristics of the power grid based on the device adjacency matrix, the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension.

[0009] Optionally, determining the power grid's performance impact index based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the performance deviation between the current operating performance and the predicted operating performance, includes: determining a control attenuation parameter corresponding to the current operating strategy, wherein the control attenuation parameter characterizes the attenuation characteristics of the influence of the current operating strategy on the power grid's operating state in the spatiotemporal dimension; and determining the power grid's performance impact index based on the control attenuation parameter, the strategy deviation between the current operating strategy and the predicted operating strategy, and the performance deviation between the current operating performance and the predicted operating performance.

[0010] Optionally, updating the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid to obtain an updated operating strategy for the power grid in the next time period includes: updating the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid to obtain a pending operating strategy for the power grid in the next time period; performing operating simulations of the power grid in the next time period under multiple scenarios based on the pending operating strategy to obtain simulated operating states corresponding to the multiple scenarios, wherein the multiple scenarios include equipment aging scenarios, operating cost fluctuation scenarios, power grid load fluctuation scenarios, and operating state fluctuation scenarios; and adjusting the pending operating strategy based on the simulated operating states corresponding to the multiple scenarios to obtain an updated operating strategy.

[0011] Optionally, updating the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid to obtain an updated operating strategy for the power grid in the next time period includes: when the current operating efficiency includes output adaptation efficiency, operation and maintenance adaptation efficiency, and environmental adaptation efficiency, constructing a fusion adaptation efficiency of the power grid based on the output adaptation efficiency, the operation and maintenance adaptation efficiency, and the environmental adaptation efficiency; and updating the current operating strategy based on the efficiency impact index, the fusion adaptation efficiency, and the operating constraints of the power grid to obtain an updated operating strategy for the power grid in the next time period.

[0012] According to one aspect of the present invention, a power grid operation strategy determination apparatus is provided, comprising: a first determination module, configured to determine factor matrices corresponding to multiple operation dimensions in a previous time period, and the current operation strategy and current operation efficiency of the power grid in the current time period, wherein the factor matrices are used to characterize the correlation between power grid parameters within the corresponding operation dimensions; a second determination module, configured to determine a predicted operation strategy and predicted operation efficiency of the power grid in the current time period based on the factor matrices corresponding to the multiple operation dimensions; a third determination module, configured to determine an efficiency impact index of the power grid based on the strategy deviation between the current operation strategy and the predicted operation strategy, and the efficiency deviation between the current operation efficiency and the predicted operation efficiency, wherein the efficiency impact index is used to characterize the degree of influence of the strategy deviation on the efficiency deviation; and a fourth determination module, configured to update the current operation strategy based on the efficiency impact index, the current operation efficiency, and the operation constraints of the power grid, to obtain an updated operation strategy of the power grid in the next time period.

[0013] According to one aspect of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the power grid operation strategy determination method described in any of the preceding embodiments.

[0014] According to one aspect of the present invention, a computer-readable storage medium is provided, wherein when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the power grid operation strategy determination method described in any of the preceding claims.

[0015] In this embodiment of the invention, by determining the factor matrix of multiple operating dimensions in the previous period, the inherent correlation law of power grid parameters can be characterized. Based on the factor matrix, the predicted operating strategy and predicted operating efficiency for the current period can be predicted, and an operating benchmark can be established when no adaptive intervention to deal with random disturbances is applied. By comparing the current operating state with the predicted benchmark state, the strategy deviation and efficiency deviation are obtained, and the efficiency impact index is determined accordingly. The efficiency deviation change caused by each corresponding change in the operating strategy deviation is quantified. Then, the operating strategy is updated by combining the efficiency impact index, the current operating efficiency, and the power grid operating constraints. This quantified causal relationship can be directly integrated into the strategy correction process to obtain an optimized operating strategy that adapts to the actual operating needs of the power grid in the next period. This solves the technical problem in related technologies where the determination of the power grid operating strategy is unreasonable, resulting in low power grid operating efficiency. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and are part of the functional claims, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0017] Figure 1 This is a flowchart of a method for determining the operation strategy of a power grid according to an embodiment of the present invention;

[0018] Figure 2 This is a flowchart of a method for determining the operation strategy of a power grid in an optional embodiment of the present invention;

[0019] Figure 3 This is a structural block diagram of a power grid operation strategy determination device according to an embodiment of the present invention. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] Example 1

[0023] According to an embodiment of the present invention, an embodiment of a method for determining the operation strategy of a power grid is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0024] Figure 1 This is a flowchart of a method for determining the operation strategy of a power grid according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0025] S102, determine the factor matrices corresponding to multiple operating dimensions in the previous period, as well as the current operating strategy and current operating efficiency of the power grid in the current period. The factor matrices are used to characterize the correlation between power grid parameters within the corresponding operating dimensions.

[0026] This involves the previous time period, which is the historical time interval of the power grid that has been completed before the current moment (e.g., the previous hour, the previous day, the previous scheduling cycle, etc.). This is the basic time interval used to extract the historical operation patterns of the power grid and construct parameter correlation relationships.

[0027] This involves multiple operational dimensions, which are classification dimensions that characterize the power grid's operational characteristics from different perspectives. These dimensions include equipment, space, state, operating conditions, and time, and are used to comprehensively represent the various attributes of power grid operation.

[0028] This involves a factor matrix, which is a matrix constructed based on power grid operation data of each operational dimension. It is used to characterize the correlation, dependency and coupling characteristics between various power grid parameters within the corresponding operational dimension.

[0029] This includes the current time period, which is the period during which the power grid is currently in operation.

[0030] This includes the current operating strategy, which is the operating strategy that the power grid has adopted or is currently implementing during the current time period.

[0031] This includes the current operating performance, which is the actual operating performance of the power grid obtained based on the current operating strategy in the current time period (such as power supply stability, equipment adaptability, state security, etc.), and is used to reflect the actual support level of the current operating strategy for the power grid operating state.

[0032] By determining the factor matrices corresponding to multiple operational dimensions in the previous period, we can extract the inherent correlation patterns of historical power grid operating parameters from multiple dimensions. By obtaining the current operating strategy and current operating efficiency of the power grid in the current period, we can provide a real and comparable basis for subsequent analysis, and thus provide complete data support and feature foundation for predicting operating status, quantifying the impact of strategies, and optimizing operating strategies.

[0033] S104, based on the factor matrices corresponding to multiple operational dimensions, determine the predictive operation strategy and predictive operation efficiency of the power grid in the current time period;

[0034] Optionally, based on the factor matrices corresponding to multiple operating dimensions, the predicted operating strategy of the power grid in the current period is determined, including: determining the multidimensional feature tensor of the power grid in the previous period, wherein the multidimensional feature tensor is used to characterize the correlation between multiple operating dimensions; determining the predicted operating state of the power grid in the current period based on the multidimensional feature tensor and the factor matrices corresponding to multiple operating dimensions; and determining the predicted operating strategy of the power grid in the current period based on the predicted operating state.

[0035] This involves a predictive operation strategy, which is the operation strategy for the current period predicted based on the multidimensional feature tensor of the power grid in the previous period and the factor matrices corresponding to the multiple operation dimensions. This predictive operation strategy is the operation strategy under the condition of no additional intervention in the current period. In actual operation in the current period, unforeseen situations may occur, and the multidimensional feature tensor of the power grid in the previous period and the factor matrices corresponding to the multiple operation dimensions in the previous period cannot be predicted in advance. Therefore, it can serve as a benchmark reference without random disturbances and strategy intervention, and can be used for comparative analysis with the actual current operation strategy.

[0036] This involves predicted operational efficiency, which is the theoretical operational performance of the power grid for the current period obtained based on the predicted operational strategy. This includes, for example, power supply stability without intervention, equipment adaptability, and state security, and is used as a benchmark for comparing efficiency deviations.

[0037] This involves a multidimensional feature tensor, a high-dimensional feature structure constructed by fusing operational data from multiple operational dimensions. The multidimensional feature tensor is used to characterize the relationships between multiple operational dimensions, uniformly depicting the coupling patterns of multidimensional operational features. For example, this multidimensional feature tensor includes a five-dimensional tensor constructed from equipment, spatial, state, operating condition, and time dimensions. Its main function is to fuse multi-source heterogeneous parameters within a unified framework, providing a joint feature representation for predicting operational states.

[0038] This involves predicting the operating status, which is the current operating status of the power grid based on the multidimensional feature tensor of the power grid in the previous period and the factor matrices corresponding to multiple operating dimensions. For example, the predicted operating status includes node voltage status, line current status, equipment load status, and power grid operating condition status.

[0039] By determining the multidimensional feature tensor of the power grid in the previous period, the correlation between multiple operational dimensions can be integrated to form a unified feature representation, which can characterize the correlation and coupling between multiple operational dimensions. Based on the multidimensional feature tensor and the factor matrix of each operational dimension, the predicted operational state can be determined (evolving only according to the power grid's own operational trend). This can accurately predict the power grid state without intervention, and thus determine the predicted operational strategy based on the predicted operational state. It can accurately obtain the benchmark strategy for deviation comparison, providing a reliable reference for subsequent quantification of the real impact of the strategy on efficiency.

[0040] Optionally, based on a multidimensional feature tensor and factor matrices corresponding to multiple operational dimensions, the predicted operational state of the power grid in the current time period is determined, including: with multiple operational dimensions including equipment dimension, spatial dimension, state dimension, operating condition dimension, and time dimension, determining the spatial coupling characteristics of the power grid based on the factor matrices of the equipment dimension, spatial dimension, state dimension, and operating condition dimension, wherein the spatial coupling characteristics are the coupling characteristics of the operational states between devices in the power grid in the spatial dimension; determining the temporal dependency characteristics of the power grid based on the factor matrices of the time dimension and the factor matrices of the state dimension, wherein the temporal dependency characteristics are the dependency characteristics of the operational states of the power grid in the time dimension; and determining the predicted operational state of the power grid in the current time period based on the spatial coupling characteristics and the temporal dependency characteristics.

[0041] This includes the equipment dimension, which is used to characterize the operating attributes of various equipment in the power grid. For example, it includes the relevant attributes of power grid equipment such as transformers, lines, switches, and compensation devices, which are used to reflect the operating parameter characteristics of the equipment itself.

[0042] This involves a spatial dimension, which characterizes the physical topology connections and relative positions of different devices within the power grid. This includes the locations and layouts of substations, feeders, distribution transformers, and line sections, reflecting the interrelationships and spatial propagation characteristics of the power grid's operating status at different physical locations. Specifically, voltage fluctuations at one node in the power grid propagate to adjacent nodes through topological connections, thus quantifying the spatial coupling characteristics of the power grid in a spatial dimension.

[0043] This involves a state dimension, which is used to characterize the operating state of the power grid (including operating state parameters such as voltage, current, load rate, and health index), and is used to reflect the real-time operating state of the power grid.

[0044] This includes the operating condition dimension, which is used to characterize the operating conditions of the power grid (such as environmental and operating condition parameters including meteorological conditions, fluctuations in new energy output, and load levels), and to reflect the characteristics of the external operating conditions of the power grid.

[0045] This involves the time dimension, which is used to describe the dynamic evolution and memory effect of the power grid's operating state over time. The operating state of the power grid (such as load and voltage) has continuity and trend on the time axis. The current state often depends on the state of the previous moment or several moments ago. In the time dimension, the temporal dependence characteristics of the power grid's operating state can be captured.

[0046] Dimensions characterizing the evolution of power grid operating status over time include time series information such as historical time series, current time, and future time periods.

[0047] This involves spatial coupling characteristics, which are used to characterize the interrelationship and mutual influence of the operating states of various devices in the power grid in terms of spatial distribution, reflecting the spatial linkage relationship of the operating states of devices in different locations.

[0048] This involves time-dependent features, which are used to characterize the correlation between the past and present operating states of the power grid in the time dimension and the influence of the past on the present. They reflect the continuity and dependence of the power grid operating state over time.

[0049] By extracting spatial coupling features of the power grid based on factor matrices of equipment, space, state, and operating conditions, the mutual propagation and linkage of different equipment operating states in the topology can be quantified. At the same time, by extracting temporal dependency features based on factor matrices of time and state, the historical continuity and evolution of operating states on the time axis can be quantified. Then, by combining spatial coupling features and temporal dependency features to predict the operating state of the current period, it is possible to take into account both the mutual influence of power grid operating states in the spatial dimension and the dynamic evolution in the temporal dimension (that is, to take into account both spatial correlation and temporal evolution characteristics), thereby achieving accurate prediction of the overall operating state of the power grid.

[0050] Optionally, the spatial coupling characteristics of the power grid are determined based on the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension, including: determining the device adjacency matrix of the power grid, wherein the device adjacency matrix is ​​used to characterize the topological connection relationship between any two devices in the power grid; and determining the spatial coupling characteristics of the power grid based on the device adjacency matrix, the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension.

[0051] This involves a device adjacency matrix, which is a matrix used to describe the physical connection relationship between any two devices in a power grid. For example, an element with a value of 1 indicates that the two devices are directly connected, and an element with a value of 0 indicates that there is no direct connection. Its main function is to quantify the topology and spatial relationship of power grid devices.

[0052] This involves a device-level factor matrix, which is a matrix constructed based on the operating parameters of power grid equipment. It is used to characterize the correlation between the operating parameters of the equipment itself, such as the correlation between parameters like load and health status of equipment like transformers, lines, and switches.

[0053] This involves a spatial dimension factor matrix, which is a matrix constructed based on the spatial distribution information of the power grid and is used to characterize the spatial relationships between different spatial locations and different topological regions.

[0054] This involves a state-dimensional factor matrix, which is a matrix constructed based on power grid operation state data and is used to characterize the correlation between state parameters such as voltage, current, and load rate.

[0055] This includes a factor matrix for the operating condition dimension, which is a matrix constructed based on power grid operating condition data and is used to characterize the correlation between operating condition parameters such as meteorological conditions, renewable energy output, and load levels.

[0056] This involves spatial coupling characteristics, which are used to characterize the mutual influence and transmission of the operating states of power grid equipment in space, reflecting the state linkage law between equipment caused by topological connections.

[0057] By determining the device adjacency matrix to characterize the topological connections between devices, the physical connections and spatial layout of power grid devices can be quantified. Based on the joint calculation of the device adjacency matrix and the multi-dimensional factor matrix, device parameters, spatial topology, operating status and external conditions can be integrated into the feature extraction process. This ensures that the spatial coupling features simultaneously reflect topological constraints, device attributes and operating conditions, avoiding the descriptive distortion caused by a single dimension. This allows for a true depiction of the state linkage and propagation laws between devices, thereby improving the accuracy and comprehensiveness of the spatial coupling features.

[0058] S106. Based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, the efficiency impact index of the power grid is determined. The efficiency impact index is used to characterize the degree of influence of strategy deviation on efficiency deviation.

[0059] Optionally, based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, the efficiency impact index of the power grid is determined, including: determining the control attenuation parameter corresponding to the current operating strategy, wherein the control attenuation parameter is used to characterize the attenuation characteristics of the degree of influence of the current operating strategy on the operating state of the power grid in the spatiotemporal dimension; and determining the efficiency impact index of the power grid based on the control attenuation parameter, the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency.

[0060] This includes strategy bias, which characterizes the difference between the current operating strategy and the predicted operating strategy, and is used to reflect the degree of deviation of the actual executed strategy from the baseline strategy without intervention.

[0061] This includes performance deviation, which characterizes the difference between current operational performance and predicted operational performance, and is used to reflect the degree of deviation of actual operational results from the baseline effect without intervention.

[0062] This includes the effectiveness impact index, which is an indicator used to quantify the degree of impact of strategy deviation on effectiveness deviation, and is used to reflect the actual impact of strategy changes on the grid operation.

[0063] This involves a control attenuation parameter, which is a characteristic parameter used to describe the degree of influence of the current operating strategy on the power grid operating state gradually weakens over time and space, and is used to reflect the spatiotemporal attenuation characteristics of the strategy's influence.

[0064] By determining the control attenuation parameters corresponding to the current operating strategy, the attenuation characteristics of the strategy's impact in the spatiotemporal dimension can be accurately characterized, eliminating distortion interference caused by the spread of the strategy's impact over time and space. Based on the control attenuation parameters, strategy deviation, and performance deviation, the performance impact index can be determined, which can truly quantify the actual impact of strategy changes on the grid's operating performance, avoid performance quantification errors caused by not considering attenuation characteristics, and improve the accuracy and reliability of the performance impact index.

[0065] S108, based on the efficiency impact index, current operating efficiency, and grid operating constraints, updates the current operating strategy to obtain the updated operating strategy for the grid in the next time period.

[0066] Optionally, based on the efficiency impact index, current operating efficiency, and grid operating constraints, the current operating strategy is updated to obtain the updated operating strategy for the grid in the next time period. This includes: updating the current operating strategy based on the efficiency impact index, current operating efficiency, and grid operating constraints to obtain the grid's operational strategy to be adjusted in the next time period; performing grid operation simulations for the next time period under multiple scenarios based on the operational strategy to be adjusted to obtain the simulated operating states corresponding to each scenario, including equipment aging scenarios, operating cost fluctuation scenarios, grid load fluctuation scenarios, and operating state fluctuation scenarios; and adjusting the operational strategy to be adjusted based on the simulated operating states corresponding to each scenario to obtain the updated operating strategy.

[0067] This involves the operational constraints of the power grid, which are the state and parameter limitations that the power grid must meet in actual operation, such as voltage limits, current limits, load rate limits, and safe operating ranges of equipment, in order to ensure the stable and safe operation of the power grid.

[0068] This involves an operational strategy to be adjusted, which is an intermediate strategy obtained based on the performance impact index, current operational performance, and operational constraints. This strategy will be used for subsequent multi-scenario simulation verification and further optimization.

[0069] This involves updating the operation strategy, which is an optimized operation strategy that is applicable to the next time period after efficiency quantification and constraint correction. This includes operation scheduling, load distribution, equipment control, etc., to improve the grid operation efficiency in the next time period.

[0070] This involves operational simulation, which uses simulation methods such as digital twins to simulate the dynamic response of the power grid over time, taking the operational strategy to be adjusted as the control input. For example, this operational simulation includes power flow calculation, transient stability simulation, and equipment state evolution simulation. Its main function is to assess the operational effectiveness of the operational strategy in advance without actually executing it in the real power grid. A digital twin is a virtual mirror model that is mapped one-to-one with the power grid and synchronized in real time. It achieves bidirectional data interaction between the physical power grid and the virtual model through a system of differential-algebraic equations. For example, this digital twin includes equipment state mapping, topology mapping, and operating parameter mapping. Its main function is to reproduce the operating state of the real power grid in virtual space, providing a high-fidelity virtual entity for operational simulation.

[0071] This involves simulated operating status, which is the simulated operating status of the power grid under various scenarios obtained through simulation, including voltage, current, equipment load, safety status, etc., used to evaluate the effectiveness of the operating strategy to be adjusted.

[0072] This involves multiple scenarios, which simulate typical operating environments that a power grid may face, including equipment aging scenarios, operating cost fluctuation scenarios, power grid load fluctuation scenarios, and operating status fluctuation scenarios, to comprehensively test the adaptability of the strategy.

[0073] By combining the performance impact index, current operating performance, and updated operating constraints to obtain the operating strategy to be adjusted, the actual impact of the strategy on performance and safety limitations can be incorporated into the initial optimization. Based on the operating strategy to be adjusted, operation simulations are carried out in multiple typical scenarios to verify the adaptability and stability of the strategy under different operating conditions in advance. The final updated operating strategy is obtained by adjusting the operating status of multiple scenario simulations, which can ensure that the updated strategy takes into account performance improvement, constraint compliance, and scenario robustness, and avoids problems such as low performance or abnormal status in actual operation.

[0074] Optionally, based on the efficiency impact index, current operating efficiency, and grid operating constraints, the current operating strategy is updated to obtain the updated operating strategy for the grid in the next time period. This includes: assuming the current operating efficiency includes output adaptation efficiency, operation and maintenance adaptation efficiency, and environmental adaptation efficiency, constructing the grid's integrated adaptation efficiency based on these three factors; and updating the current operating strategy based on the efficiency impact index, integrated adaptation efficiency, and grid operating constraints to obtain the updated operating strategy for the grid in the next time period.

[0075] This includes output matching efficiency, which is an operational efficiency indicator that characterizes the degree of matching between the power grid's power output and load demand. Examples include power supply matching, load supply stability, and voltage and current compliance. Its main function is to reflect the reliable support level of the power grid's power output.

[0076] This includes operation and maintenance adaptation efficiency, which is an operation efficiency indicator that characterizes the degree of matching between the operating status of power grid equipment and operation and maintenance control requirements. Examples include reasonable equipment load, good health status, and adaptability of operation and maintenance actions. Its main function is to reflect the safe and stable operation level of power grid equipment.

[0077] Among them, environmental adaptability is involved. Environmental adaptability is an operational efficiency indicator that characterizes the degree of matching between the power grid's operating status and the external working environment. For example, it includes the adaptability of new energy output, meteorological conditions, and load environment. Its main function is to reflect the power grid's ability to adapt to the external environment.

[0078] This includes the integration and adaptation efficiency, which is a comprehensive operational efficiency indicator built by integrating output adaptation efficiency, operation and maintenance adaptation efficiency, and environmental adaptation efficiency. It is used to uniformly quantify the multi-dimensional operational adaptation level of the power grid and provide a comprehensive efficiency basis for strategy updates.

[0079] By integrating output adaptation efficiency, operation and maintenance adaptation efficiency, and environmental adaptation efficiency to construct integrated adaptation efficiency, the multi-dimensional operational adaptation level of the power grid can be comprehensively quantified, avoiding the one-sidedness of single-dimensional evaluation. Based on the efficiency impact index, integrated adaptation efficiency, and operational constraints, the operation strategy can be updated to take into account both comprehensive efficiency improvement and safe operation limitations, ensuring that the updated operation strategy in the next period is more in line with the actual operation needs of the power grid and improving the overall operation efficiency of the power grid.

[0080] Through the steps S102-S108 above, by determining the factor matrix of multiple operating dimensions in the previous period, the inherent correlation law of power grid parameters can be characterized. Based on the factor matrix, the predicted operating strategy and predicted operating efficiency for the current period can be predicted, and an operating benchmark without adaptive intervention to deal with random disturbances can be established. By comparing the current operating state with the predicted benchmark state, the strategy deviation and efficiency deviation are obtained, and the efficiency impact index is determined accordingly. The efficiency deviation change caused by each corresponding change in the operating strategy deviation is quantified. Then, the operating strategy is updated by combining the efficiency impact index, the current operating efficiency, and the power grid operating constraints. This quantified causal relationship can be directly integrated into the strategy correction process to obtain an optimized operating strategy that adapts to the actual operating needs of the power grid in the next period.

[0081] Based on the above embodiments and optional embodiments, an optional implementation method is provided, which is described in detail below.

[0082] In related technologies, the power grid's operational efficiency can be ensured to meet operational requirements by determining the power grid's operating strategy. However, in these technologies, there is a technical problem where an unreasonable determination of the power grid's operating strategy leads to low power grid operational efficiency.

[0083] There is currently no effective solution to the above problems.

[0084] In view of this, an optional embodiment of the present invention provides a method for determining the operation strategy of a power grid, which can effectively solve the above-mentioned technical problems.

[0085] Figure 2 This is a flowchart of a method for determining the operation strategy of a power grid in an optional embodiment of the present invention, such as... Figure 2 As shown, a detailed description follows.

[0086] S1, determine the factor matrices corresponding to multiple operating dimensions in the previous period, and the current operating strategy and current operating efficiency of the power grid in the current period. The factor matrices are used to characterize the correlation between power grid parameters in the corresponding operating dimensions.

[0087] S2, based on the factor matrices corresponding to multiple operational dimensions, determines the predictive operation strategy and predictive operation efficiency of the power grid in the current time period;

[0088] Specifically, based on the factor matrices corresponding to multiple operational dimensions, the predicted operation strategy of the power grid in the current period is determined, including: determining the multidimensional feature tensor of the power grid in the previous period, wherein the multidimensional feature tensor is used to characterize the correlation between multiple operational dimensions; determining the predicted operation state of the power grid in the current period based on the multidimensional feature tensor and the factor matrices corresponding to multiple operational dimensions; and determining the predicted operation strategy of the power grid in the current period based on the predicted operation state.

[0089] Determine the factor matrices corresponding to multiple operational dimensions in the previous time period and determine the multidimensional feature tensor of the power grid in the previous time period, including:

[0090] The process involves acquiring raw power grid operation data, constructing multidimensional operation indicators based on this data, obtaining a power grid state sequence based on these indicators, establishing an initial feature tensor for the power grid based on the state sequence, and decomposing and reducing the dimensionality of the initial feature tensor to obtain factor matrices corresponding to the multiple operation dimensions of the multidimensional feature tensor.

[0091] The power grid state sequence is represented as follows:

[0092]

[0093] in, Let t be the power grid state sequence at time t; Let p be the power grid state characteristic component under the p-th operating index at time t. The total dimension of multidimensional operational indicators.

[0094] For example, the multi-dimensional operation indicators cover the equipment operation domain, power supply quality domain, operation cost domain, operation and maintenance efficiency domain, and operation environment domain in the power grid. There are a total of 18 multi-dimensional operation indicators, including:

[0095] The equipment operation domain includes four items: the average load rate of the main transformer (%), which reflects the equipment utilization efficiency; the dynamic capacity margin of the line (%) calculated based on the N-1 criterion; the equipment health index (0-1); and the comprehensive line loss rate (%). Among them, the N-1 criterion is involved. The N-1 criterion is a topology verification criterion for the safe operation of the power grid. It refers to the judgment criterion that when any single component such as a single line or a single main transformer is out of service, the remaining power grid topology can still maintain normal power supply operation and all operating parameters remain within the allowable range. It is mainly used to verify the fault tolerance capability and operational redundancy level of the power grid structure.

[0096] The power quality domain includes four items: voltage qualification rate (%), power supply reliability rate (%), harmonic distortion rate (%), and cumulative frequency deviation duration (seconds).

[0097] The operating cost domain includes four items: unit operation and maintenance cost, resource turnover rate, power supply response cycle, and resource utilization efficiency ratio.

[0098] The operational efficiency domain includes three items: average fault repair time (minutes), anomaly feedback response rate (%), and operational work order closure rate (%).

[0099] The operating environment domain includes three items: the intensity of environmental impact based on text mining (0-10), the meteorological risk index that integrates real-time data such as temperature and humidity (0-1), and the penetration rate of new energy (%).

[0100] Among them, the initial characteristic tensor of the power grid Represented as:

[0101]

[0102] in, For real number tensors; From the device perspective; Spatial dimension; For the state dimension; From the perspective of working conditions; The time dimension.

[0103] By constructing an initial feature tensor of the power grid, parameters at different levels can be quantified and modeled. For example, the equipment dimension includes power parameters corresponding to transformers, lines, switches, and compensation devices; the spatial dimension includes power parameters corresponding to areas of different sizes; the state dimension includes multi-dimensional operating indicators; the operating condition dimension includes operating condition parameters corresponding to environmental, meteorological, and new energy factors; and the time dimension includes timestamps, such as minute-level timestamps.

[0104] The initial feature tensor is decomposed and dimensionality reduced to obtain factor matrices corresponding to the multiple operating dimensions of the multidimensional feature tensor, including:

[0105] Construct the objective function:

[0106]

[0107] in, This is the initial characteristic tensor of the power grid, also known as the observation tensor; For multidimensional feature tensors, , For the rank of the device dimension, For the rank of the spatial dimension, Let be the rank of the state dimension. For the rank of the working condition dimension, For the rank of the time dimension; This is a factor matrix representing the device dimension; The factor matrix represents the spatial dimension; The factor matrix represents the state dimension; The factor matrix represents the working condition dimension; A factor matrix with time dimension; is the weight coefficient of the i-th regularization term, used to control the strength of the corresponding regularization constraint; The i-th regularization term is used to constrain the properties of the corresponding dimension factor matrix (such as sparsity, smoothness, or orthogonality) to avoid overfitting.

[0108] Based on the objective function, the initial feature tensor is decomposed and dimensionality reduced to obtain the factor matrices corresponding to the multiple operating dimensions of the multidimensional feature tensor.

[0109] Specifically, based on the objective function, the initial feature tensor is decomposed and dimensionality reduced to obtain a multidimensional feature tensor and a device-dimensional factor matrix, including:

[0110] The factor matrix for the device dimension is determined using the following formula:

[0111]

[0112] in, This is the factor matrix for the device dimension after the (k+1)th iteration update; The initial characteristic tensor of the power grid Matrix-based expansion in the first dimension (device dimension); Let be the multidimensional feature tensor under the k-th iteration; Let be the factor matrix of the spatial dimension in the k-th iteration; Let be the factor matrix representing the state dimension under the k-th iteration; Let be the factor matrix of the working condition dimension under the k-th iteration; Let be the time-dimension factor matrix for the k-th iteration; The square of the Frobenius norm is used to measure the error between tensors; This means calculating each row of the matrix first. The norm is then used to sum the results of all rows, resulting in a scalar.

[0113] The Frobenius norm is a matrix norm used to measure the size of matrix elements; it is a way to measure the overall size of a matrix. The norm is the Euclidean distance of a vector from the origin to the endpoint, and is used to measure the length or size of a vector.

[0114] By using the factor matrix obtained from the equipment dimension, we can obtain the factor matrices corresponding to the spatial dimension, state dimension, operating condition dimension, and time dimension, which will not be elaborated further.

[0115] The multidimensional feature tensor is determined using the following formula:

[0116]

[0117] in, This is the multidimensional feature tensor after the (k+1)th iteration update.

[0118] Specifically:

[0119] Regularization term This indicates the selection of key equipment parameters;

[0120] Regularization term This indicates the capture of hierarchical relationships within regions;

[0121] Regularization term This indicates that the continuity of the state dimension is maintained;

[0122] Regularization term This indicates the extraction of key environmental factors;

[0123] Regularization term This indicates a constraint on the smoothness of the time dimension, to avoid abrupt changes.

[0124] For example, the initial feature tensor is decomposed and dimensionality reduced to obtain the factor matrices corresponding to the multiple operating dimensions of the multidimensional feature tensor. This can be achieved using Tucker decomposition, and a time dimension regularization term is introduced to perform tensor decomposition and dimensionality reduction, retaining 95% of the energy, thereby reducing the dimensionality of the parameter space and lowering the computational complexity. Finally, the adaptive alternating direction multiplier method (ADMM) is used to iteratively solve the problem, decomposing the complex optimization problem of the above objective function into two sub-problems: updating the factor matrix and updating the core tensor.

[0125] Specifically, based on a multidimensional feature tensor and factor matrices corresponding to multiple operational dimensions, the predicted operational state of the power grid in the current time period is determined, including: Given multiple operational dimensions including equipment dimension, spatial dimension, state dimension, operating condition dimension, and time dimension, the spatial coupling characteristics of the power grid are determined based on the factor matrices of the equipment dimension, spatial dimension, state dimension, and operating condition dimension, where the spatial coupling characteristics are the coupling characteristics of the operational states between devices in the power grid in the spatial dimension; the temporal dependency characteristics of the power grid are determined based on the factor matrices of the time dimension and the factor matrices of the state dimension, where the temporal dependency characteristics are the dependency characteristics of the operational states of the power grid in the time dimension; and the predicted operational state of the power grid in the current time period is determined based on the spatial coupling characteristics and the temporal dependency characteristics.

[0126] Specifically, based on the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension, the spatial coupling characteristics of the power grid are determined, including: determining the device adjacency matrix of the power grid, wherein the device adjacency matrix is ​​used to characterize the topological connection relationship between any two devices in the power grid; and determining the spatial coupling characteristics of the power grid based on the device adjacency matrix, the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension.

[0127] Specifically, determining the spatial coupling characteristics of the power grid based on the device adjacency matrix, device-dimensional factor matrix, spatial-dimensional factor matrix, state-dimensional factor matrix, and operating condition-dimensional factor matrix includes: constructing a spatiotemporal joint input based on the device-dimensional factor matrix, spatial-dimensional factor matrix, and operating condition-dimensional factor matrix; determining the enhanced adjacency matrix based on the state-dimensional factor matrix and the device adjacency matrix; and determining the spatial coupling characteristics of the power grid based on the enhanced adjacency matrix and the spatiotemporal joint input.

[0128] Among them, a spatiotemporal joint input is constructed based on the factor matrix of the equipment dimension, the factor matrix of the spatial dimension, and the factor matrix of the operating condition dimension, as shown in the formula:

[0129]

[0130] in, This is a tensor splicing operation.

[0131] For example, a two-layer hybrid intelligent model is employed to determine the spatial coupling characteristics and temporal dependency characteristics of the power grid. Based on these characteristics, the predicted operating state of the power grid in the current time period is determined. The two-layer hybrid intelligent model is constructed using a Long Short-Term Memory (LSTM) network and a Spatiotemporal Graph Convolutional Network (ST-GCN) to dynamically model the power grid's operating parameters and predict their spatiotemporal evolution. The network architecture of the two-layer hybrid intelligent model consists of three layers: a feature recombination layer, a spatiotemporal joint modeling layer, and a dynamic fusion layer.

[0132] The feature recombination layer transforms the decomposition results, i.e., the multidimensional feature tensor and the factor matrices corresponding to the multiple operational dimensions, into a spatiotemporal feature map. For example, the input factor matrix obtained from the decomposition includes the factor matrix of the device dimension. Spatial dimension factor matrix Factor matrix of working conditions Device-based factor matrix Spatial dimension factor matrix Factor matrix of working conditions Multi-source features are fused along the feature dimension to construct a spatiotemporal joint input. The spatiotemporal joint modeling layer contains two channels: a spatial dependency modeling channel (corresponding to ST-GCN) and a temporal evolution modeling channel (corresponding to LSTM) to capture features in parallel. Specifically, the spatial dependency modeling channel captures spatial coupling features, while the temporal evolution modeling channel captures temporal dependency features.

[0133] The enhanced adjacency matrix is ​​determined based on the state-dimensional factor matrix and the device adjacency matrix, using the following formula:

[0134]

[0135] in, To enhance the adjacency matrix, This is the device adjacency matrix, also known as the device connection adjacency matrix (0-1 binarization). This is used to generate a diagonal matrix that reflects the inherent correlation between multidimensional operational indicators corresponding to the state dimension.

[0136] For example, based on the above, in the ST-GCN channel, the correlation between device topology and indicators is fused to obtain an enhanced adjacency matrix.

[0137] Among them, based on the enhanced adjacency matrix and the spatiotemporal joint input, the spatial coupling characteristics of the power grid are determined by the following formula:

[0138]

[0139] in, For the first The spatial coupling characteristics of the time-space power grid, i.e., the output feature matrix of the spatial graph convolutional network; It is a non-linear activation function; This is the diffusion order, which is also the maximum order of graph convolution; It is a neighborhood order index; To enhance the adjacency matrix The corresponding degree matrix; To enhance the adjacency matrix of the first Order topological association; For the first Spatiotemporal joint input of the layer; For the first The learnable weight matrix of the layer.

[0140] For example, based on the above, a hierarchical diffusion mechanism is used to capture spatial dependence and obtain the spatial coupling characteristics of the power grid.

[0141] The determination of the time-series dependency characteristics of the power grid based on the time-dimensional factor matrix and the state-dimensional factor matrix includes: concatenating the time-dimensional factor matrix and the state-dimensional factor matrix into a time-series input, and inputting the time-series input into a long short-term memory network to obtain the time-series dependency characteristics of the power grid.

[0142] The timing input is represented as follows:

[0143]

[0144] in, For timing input, Let be the factor matrix in the time dimension at time t; The factor matrix represents the state dimension at time t.

[0145] The sequence representation corresponding to the time-series dependency characteristics of the power grid is as follows:

[0146]

[0147] in, These represent the time-series dependency characteristics of the power grid at the corresponding time points. The time dimension length is the total number of consecutive sampling moments.

[0148] Based on spatial coupling characteristics and temporal dependence characteristics, the predicted operating state of the power grid in the current time period is determined by the following formula:

[0149]

[0150] in, For the predicted operating status of the power grid; This is a spatial coupling characteristic; This is a time-dependent feature, corresponding to the time-series feature of the last time step; To activate the function, a nonlinearity is introduced; This is the transpose of the factor matrix for the working condition dimension; This is a learnable weight matrix for the working condition dimension.

[0151] For example, based on the above, a multidimensional feature tensor is utilized through a dynamic fusion layer. Multidimensional feature tensors, based on The importance of the quantitative operation condition dimension is determined, and ultimately the two-layer hybrid intelligent model achieves refined quantitative analysis of the power grid operation status through structured feature recombination, spatiotemporal joint modeling, and dynamic fusion mechanism.

[0152] S4. Based on the strategy deviation between the current operating strategy and h, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, determine the power grid efficiency impact index, whereby the efficiency impact index is used to characterize the degree of influence of strategy deviation on efficiency deviation.

[0153] Specifically, based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, the power grid efficiency impact index is determined, including: determining the control attenuation parameter corresponding to the current operating strategy, wherein the control attenuation parameter is used to characterize the attenuation characteristics of the degree of influence of the current operating strategy on the operating state of the power grid in the spatiotemporal dimension; and determining the power grid efficiency impact index based on the control attenuation parameter, the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency.

[0154] Based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, the power grid efficiency impact index is determined using the following formula:

[0155]

[0156] in, The effectiveness impact index characterizes time-varying causal effects; Let be the operating efficiency of the power grid at time t; Let be the predicted operating efficiency of the power grid at time t; Let t be the power grid's operating strategy at time t; The predictive operation strategy of the power grid at time t; For performance deviation; This is a strategy bias; To control the attenuation parameter, representing the spatiotemporal attenuation coefficient, These are learnable parameters.

[0157]

[0158] in, Including basic attenuation rate Environmental sensitivity coefficient ; From time At that time The integral is used to calculate the time accumulation of the attenuation intensity; For integration time; for Factor matrix of working conditions at any given time.

[0159]

[0160] in, For spatiotemporal graph convolutional networks; , These are factor matrices corresponding to multiple operational dimensions; It is a long short-term memory network; For time lag windows (e.g., 24 hours).

[0161] For the above strategy prediction model and The factor matrices can be the same or different, for example, Including is Factor matrices for the device dimension, spatial dimension, state dimension, and operating condition dimension at any given time. It includes a factor matrix with a time dimension and a factor matrix with a state dimension.

[0162]

[0163] in, It is a spacetime tensor mapping function; Let be the factor matrix of the device dimension at time t; The factor matrix represents the spatial dimension; This is a factor matrix representing the working condition dimension.

[0164] Based on the above outcome prediction model, the outcome variables are estimated to obtain the predicted operational efficiency. This outcome prediction model is consistent with or shares parameters with the strategy prediction model.

[0165] Specifically, based on the efficiency impact index, current operating efficiency, and grid operating constraints, the current operating strategy is updated to obtain the updated operating strategy for the grid in the next time period. This includes: assuming the current operating efficiency includes output adaptation efficiency, operation and maintenance adaptation efficiency, and environmental adaptation efficiency, constructing the grid's integrated adaptation efficiency based on these three factors; and updating the current operating strategy based on the efficiency impact index, integrated adaptation efficiency, and grid operating constraints to obtain the updated operating strategy for the grid in the next time period.

[0166] For example, after updating the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid to obtain the updated operating strategy of the power grid for the next time period, the process further includes: based on the strategy optimization function, continuously adjusting the updated operating strategy of the power grid for the next time period, and controlling the power grid based on the adjusted operating strategy until the end of the next time period. Here, the strategy optimization function is a function that aims to maximize the operating efficiency gain of the operating strategy, that is, to maximize the improvement of power grid operating efficiency, including power supply reliability, equipment safety, stable state, and reduced losses.

[0167] The policy optimization function is expressed as:

[0168]

[0169] in, To maximize the performance gain of strategy X; To integrate and adapt performance, it can be characterized in the next time period. The performance gain of the execution strategy X at any given time; To perform an averaging operation; This is the risk penalty coefficient; This is the penalty coefficient for strategy changes; As a power grid operation risk index; This is the strategy offset index; This represents the total length of the next time period.

[0170]

[0171] in, To produce an appropriate level of efficiency, that is, a parameter for resource utilization efficiency; To adapt performance for operation and maintenance, i.e., operation and maintenance response parameters; Environmental adaptation performance, also known as operating environment parameters.

[0172]

[0173] in, For the first Operating status characteristics of the equipment; Weighted curvature penalty to suppress policy mutations on high-risk equipment; The total number of devices in the power grid; For the first Real-time risk characteristics of the equipment.

[0174]

[0175] in, KL divergence is used to measure the degree of difference between two probability distributions / policy vectors. This represents the total number of historical operating strategies of the power grid. For the first The weighting coefficients of the historical operating strategies of the power grid; For the first Historical operating strategies of the power grid.

[0176]

[0177] in, It is a normalized exponential function.

[0178] Based on the policy optimization function, the updated operation policy of the power grid for the next time period is continuously adjusted within the next time period. This can be achieved using the Spatiotemporally Extended Deep Deterministic Policy Gradient (ST-DDPG). The Spatiotemporally Extended Deep Deterministic Policy Gradient (ST-DDPG) is a power grid continuous control policy optimization algorithm that extends the Deep Deterministic Policy Gradient (DDPG) reinforcement learning algorithm by incorporating spatiotemporal features (spatial topology and temporal dependencies). The training of the Spatiotemporally Extended Deep Deterministic Policy Gradient (ST-DDPG) is as follows:

[0179]

[0180] in, Gradient operators for policy network parameters; The optimization objective for the spatiotemporally extended depth deterministic strategy gradient represents the expected long-term operating efficiency of the power grid. This refers to the number of samples in the experience playback batch, which is also the total number of historical power grid state-policy-performance samples used in one training session. Let be the action gradient operator, representing the gradient of the action of the policy. Calculate the gradient to evaluate how sensitive the network's computational actions are to performance. It is a spatiotemporal evaluation network used to take the power grid state and policy actions as input and output the long-term operational performance score of the policy. For the first The operating status of each sample power grid; For power grid operation strategy actions (continuous control quantities), such as voltage regulation, frequency regulation, power distribution, equipment control commands, etc. For spatiotemporal evaluation networks Network parameters; For the policy network, input the first... The operating status of individual power grids Then, a deterministic optimal power grid operation strategy is output; For policy networks Network parameters.

[0181]

[0182] in, The spatiotemporal characteristics of the feature reconstructing layer at time t. For causal effect time window, action This indicates that the execution strategy is being implemented. The control action vectors required at that time.

[0183] For example, It can integrate spatial coupling features and temporal dependency features and operating strategies , represented as:

[0184]

[0185] Based on the above, by establishing a time-varying causal effect calculation model and long-term policy optimization, the causal effect of the policy can be evaluated and optimized in the long term, thereby improving the long-term performance and stability of the policy. Specifically, the aforementioned time-varying causal effect can quantify the strength of causal effects at different time points and assess the impact of the policy at different time points; this can be achieved using a machine learning model. Furthermore, based on the aforementioned policy optimization function, long-term policy optimization is achieved, balancing the long-term performance gain of the policy with short-term volatility risk.

[0186] S5 updates the current operating strategy based on the efficiency impact index, current operating efficiency, and grid operating constraints to obtain the updated operating strategy for the grid in the next time period.

[0187] Specifically, based on the efficiency impact index, current operating efficiency, and grid operating constraints, the current operating strategy is updated to obtain the updated operating strategy for the grid in the next time period. This includes: updating the current operating strategy based on the efficiency impact index, current operating efficiency, and grid operating constraints to obtain the grid's operational strategy to be adjusted in the next time period; based on the operational strategy to be adjusted, performing grid operation simulations for the next time period under multiple scenarios to obtain the simulated operating states corresponding to each scenario, including equipment aging scenarios, operating cost fluctuation scenarios, grid load fluctuation scenarios, and operating state fluctuation scenarios; and adjusting the operational strategy to be adjusted based on the simulated operating states corresponding to each scenario to obtain the updated operating strategy.

[0188] For example:

[0189] 1) Multi-dimensional dynamic aging modeling and physical-virtual mapping.

[0190] A multi-stage dynamic aging model is constructed, coupling the equipment aging process with the operating strategy. The verification scenarios are extended to four core operating scenarios: equipment aging, operating cost fluctuations, grid load fluctuations, and operating status fluctuations (such as renewable energy access scenarios), covering the entire lifecycle verification requirements. The coupling model between equipment aging rate and operating strategy is as follows:

[0191]

[0192] in, Let t be the time-varying aging coefficient of the equipment at time t; This refers to the initial aging factor of the equipment. As a load mitigation factor; for Real-time device load current; The rated current of the equipment; The strategy influence weighting coefficient represents the strength of the impact of the power grid operation strategy on the aging rate of equipment. The better the strategy, the more controllable the aging process. for The power grid operation strategy vector at any given time.

[0193] Achieving bidirectional interaction between the physical system and the virtual twin (digital twin virtual model) through a system of differential-algebraic equations:

[0194]

[0195] in, To take into account the dynamic state evolution characteristics of the power grid, including real-time load of equipment and time-varying aging of equipment; These are power grid state variables, such as node voltage and line current; For control inputs, such as generator output and load adjustment commands; This is a correction factor; Operational constraints refer to the physical constraints, power flow constraints, equipment constraints, and security constraints that the power grid must meet.

[0196] Through simulation accuracy verification, the physical-virtual state deviation must be less than a certain threshold; otherwise, adaptive calibration of model parameters is triggered.

[0197]

[0198] in, For learning rate, These are model parameters; These are the new parameters after adaptive calibration of the model; These are the old parameters before model calibration; This is a parametric gradient operator used to correct parameters along the direction that reduces the physical-virtual bias. This refers to the physical-to-virtual state deviation.

[0199] Define the loss function for parameter calibration:

[0200]

[0201] in, The overall error between the digital twin model and the real physical power grid; To calibrate the time window length; Let t be the actual state vector of the physical power grid at time t; Output the state vector of the virtual twin model at time t.

[0202] 2) Stress test index space.

[0203] Design a scenario parameter generator that covers four types of verification scenarios.

[0204] The scenario simulating operating cost fluctuations using operating cost volatility is represented as follows:

[0205]

[0206] in, The degree of fluctuation in power grid operating costs at time t; The degree of fluctuation in power grid operating costs at time t-1; This represents the average fluctuation of operating costs. The standard deviation of operating cost fluctuations; For time step; The operating cost is a random disturbance term that follows a standard normal distribution.

[0207] By injecting random pulse loads, a load change scenario (i.e., a power grid load fluctuation scenario) is simulated, where the pulse probability... The results are dynamically adjusted based on the LSTM predictions, as follows:

[0208]

[0209] in, This refers to the power of the pulsed load. This is the pulse load amplitude coefficient; As the baseline load; This is the Bernoulli distribution (0–1 distribution). The Bernoulli distribution is a discrete probability distribution that describes a single random trial with only two possible outcomes (success or failure). This represents the probability of pulse occurrence.

[0210] Generating multi-stage aging coefficients based on aging models This data is then mapped to a device health index to simulate device aging scenarios.

[0211] The operational status fluctuation scenario can specifically be a renewable energy access scenario. Specifically, the renewable energy access scenario can be simulated using the volatility of distributed energy access, as follows:

[0212]

[0213] in, Let t be the total fluctuating power of distributed energy resources at time t; This represents the total number of distributed energy resources connected to the grid. For time t, the first Taiwan's distributed energy base output power; This refers to the fluctuation range coefficient of new energy sources. For the first Taiwan's new energy fluctuating angular frequency; For the first The initial phase of Taiwan's new energy fluctuations.

[0214] After updating the current operating strategy to obtain the updated operating strategy for the power grid in the next time period, the process also includes scoring the updated operating strategy. Based on the scoring results, it is determined whether further adjustments / re-determination of the updated operating strategy are needed. The scoring formula is as follows:

[0215]

[0216] in, A comprehensive score for power grid operation strategies; For the first The weighting coefficients of the scoring indicators; For the first The score of each scoring indicator; For the first The minimum score for each scoring indicator; For the first The maximum score of each scoring indicator.

[0217] in, Reflecting the power grid's performance in resource utilization, To measure the resource utilization efficiency of the power grid during time t. This is the period for resource utilization assessment.

[0218] Reflecting the performance of the power grid in terms of operational reliability, among which Measure in the The probability that a power supply system will perform its function under specified conditions and time in a given device or area. Total number of devices;

[0219] Reflecting the operation and maintenance response efficiency of power grid companies, among which Indicates the fault repair time. This is a preset standard time used to measure the timeliness of fault repair;

[0220] Reflecting the penetration rate of new energy Relative to target value Performance.

[0221] Weight The objective is dynamically allocated based on strategy optimization. A verification feedback mechanism is established based on the above settings, and the score... If the value falls below a set threshold, a policy re-optimization loop is triggered.

[0222]

[0223] in, For the re-optimized new power grid operation strategy; This refers to the power grid operation strategy updated in the previous round; Optimize the gradient step size for the strategy.

[0224]

[0225] in, The learning rate is used to optimize the strategy and control the magnitude of each adjustment to the strategy. For policy gradient operators; It is a very small positive number (excluding zero constant).

[0226] By avoiding oscillations through gradient normalization, the re-optimized strategy is re-injected into the differential equation for iterative verification, thus forming a closed loop.

[0227] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0228] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0229] Example 2

[0230] According to embodiments of the present invention, an apparatus for implementing the above-described method for determining the operation strategy of a power grid is also provided. Figure 3 This is a structural block diagram of a power grid operation strategy determination device according to an embodiment of the present invention, such as... Figure 3 As shown, the device includes: a first determining module 302, a second determining module 304, a third determining module 306, and a fourth determining module 308. The device will be described in detail below.

[0231] The first determining module 302 is used to determine the factor matrices corresponding to multiple operating dimensions in the previous time period, as well as the current operating strategy and current operating efficiency of the power grid in the current time period. The factor matrices are used to characterize the correlation between power grid parameters within the corresponding operating dimensions.

[0232] The second determining module 304, connected to the first determining module 302, is used to determine the predicted operation strategy and predicted operation efficiency of the power grid in the current time period based on the factor matrices corresponding to multiple operation dimensions.

[0233] The third determining module 306, connected to the second determining module 304, is used to determine the power grid efficiency impact index based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency. The efficiency impact index is used to characterize the degree of influence of the strategy deviation on the efficiency deviation.

[0234] The fourth determining module 308, connected to the third determining module 306, is used to update the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid, so as to obtain the updated operating strategy of the power grid in the next time period.

[0235] It should be noted that the first determining module 302, the second determining module 304, the third determining module 306, and the fourth determining module 308 mentioned above correspond to steps S102 to S108 in the method for determining the operation strategy of the power grid. The multiple modules and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in the above embodiment 1.

[0236] Example 3

[0237] According to another aspect of the present invention, an electronic device is also provided, comprising: a processor; and a memory for storing processor-executable instructions, wherein the processor is configured to execute instructions to implement the power grid operation strategy determination method of any of the above embodiments.

[0238] Example 4

[0239] According to another aspect of the present invention, a computer-readable storage medium is also provided, which, when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform the power grid operation strategy determination method described above.

[0240] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0241] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0242] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0243] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0244] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0245] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0246] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for determining the operation strategy of a power grid, characterized in that, include: Determine the factor matrices corresponding to multiple operating dimensions in the previous period, and the current operating strategy and current operating efficiency of the power grid in the current period. The factor matrices are used to characterize the correlation between power grid parameters within the corresponding operating dimensions. Based on the factor matrices corresponding to the multiple operational dimensions, the predictive operational strategy and predictive operational efficiency of the power grid in the current time period are determined. Based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, the efficiency impact index of the power grid is determined, wherein the efficiency impact index is used to characterize the degree of influence of strategy deviation on efficiency deviation. Based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid, the current operating strategy is updated to obtain the updated operating strategy of the power grid for the next time period.

2. The method according to claim 1, characterized in that, The step of determining the predictive operation strategy of the power grid for the current time period based on the factor matrices corresponding to the multiple operational dimensions includes: Determine the multidimensional feature tensor of the power grid in the previous time period, wherein the multidimensional feature tensor is used to characterize the correlation between the multiple operating dimensions; Based on the multidimensional feature tensor and the factor matrices corresponding to the multiple operating dimensions, the predicted operating state of the power grid in the current time period is determined. Based on the predicted operating status, the predicted operating strategy of the power grid for the current time period is determined.

3. The method according to claim 2, characterized in that, The step of determining the predicted operating state of the power grid in the current time period based on the multidimensional feature tensor and the factor matrices corresponding to the multiple operating dimensions includes: Given that the multiple operating dimensions include device dimension, spatial dimension, state dimension, operating condition dimension, and time dimension, the spatial coupling characteristics of the power grid are determined based on the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension, wherein the spatial coupling characteristics are the coupling characteristics of the operating states between devices in the power grid in the spatial dimension. Based on the factor matrix of the time dimension and the factor matrix of the state dimension, the time-series dependency characteristics of the power grid are determined, wherein the time-series dependency characteristics are the time-series dependency characteristics of the operating state of the power grid. Based on the spatial coupling characteristics and the temporal dependence characteristics, the predicted operating state of the power grid in the current time period is determined.

4. The method according to claim 3, characterized in that, The determination of the spatial coupling characteristics of the power grid based on the factor matrix of the device dimension, the factor matrix of the spatial dimension, the factor matrix of the state dimension, and the factor matrix of the operating condition dimension includes: Determine the device adjacency matrix of the power grid, wherein the device adjacency matrix is ​​used to characterize the topological connection relationship between any two devices in the power grid; Based on the device adjacency matrix, the device-dimensional factor matrix, the spatial-dimensional factor matrix, the state-dimensional factor matrix, and the operating condition-dimensional factor matrix, the spatial coupling characteristics of the power grid are determined.

5. The method according to claim 1, characterized in that, The determination of the power grid's performance impact index based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the performance deviation between the current operating performance and the predicted operating performance, includes: Determine the control attenuation parameter corresponding to the current operating strategy, wherein the control attenuation parameter is used to characterize the attenuation characteristics of the influence of the current operating strategy on the operating state of the power grid in the spatiotemporal dimension; Based on the control attenuation parameters, the strategy deviation between the current operating strategy and the predicted operating strategy, and the efficiency deviation between the current operating efficiency and the predicted operating efficiency, the efficiency impact index of the power grid is determined.

6. The method according to claim 1, characterized in that, The process of updating the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid to obtain the updated operating strategy of the power grid for the next time period includes: Based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid, the current operating strategy is updated to obtain the operating strategy to be adjusted for the power grid in the next time period. Based on the operation strategy to be adjusted, the power grid is simulated for the next time period under multiple scenarios to obtain the simulation operation status corresponding to each of the multiple scenarios. The multiple scenarios include equipment aging scenario, operating cost fluctuation scenario, power grid load fluctuation scenario, and operating status fluctuation scenario. Based on the simulation running states corresponding to the multiple scenarios, the running strategy to be adjusted is obtained to obtain an updated running strategy.

7. The method according to any one of claims 1 to 6, characterized in that, The process of updating the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid to obtain the updated operating strategy of the power grid for the next time period includes: Given that the current operational efficiency includes output adaptation efficiency, operation and maintenance adaptation efficiency, and environmental adaptation efficiency, the integrated adaptation efficiency of the power grid is constructed based on the output adaptation efficiency, the operation and maintenance adaptation efficiency, and the environmental adaptation efficiency. Based on the efficiency impact index, the fusion adaptation efficiency, and the power grid's operational constraints, the current operating strategy is updated to obtain the updated operating strategy for the power grid in the next time period.

8. A device for determining the operation strategy of a power grid, characterized in that, include: The first determining module is used to determine the factor matrices corresponding to multiple operating dimensions in the previous time period, as well as the current operating strategy and current operating efficiency of the power grid in the current time period. The factor matrices are used to characterize the correlation between power grid parameters within the corresponding operating dimensions. The second determining module is used to determine the predicted operation strategy and predicted operation efficiency of the power grid in the current time period based on the factor matrices corresponding to the multiple operation dimensions respectively. The third determining module is used to determine the power grid's performance impact index based on the strategy deviation between the current operating strategy and the predicted operating strategy, and the performance deviation between the current operating performance and the predicted operating performance. The performance impact index is used to characterize the degree of influence of the strategy deviation on the performance deviation. The fourth determining module is used to update the current operating strategy based on the efficiency impact index, the current operating efficiency, and the operating constraints of the power grid, so as to obtain the updated operating strategy of the power grid in the next time period.

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

10. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by the processor of the electronic device, the electronic device is able to perform the power grid operation strategy determination method as described in any one of claims 1 to 7.

Images