An ice disaster scenario power grid active toughening scheduling method, device, equipment and storage medium

Abstract

The application discloses an ice disaster scenario active toughening scheduling method and device, equipment and storage medium, relates to the power system scheduling technical field, and the method comprises the following steps: acquiring source network load basic data and flexibility resource parameters in the ice disaster scenario; constructing an energy storage dynamic value coefficient according to the source network load basic data, and performing day-ahead distribution robust pre-scheduling based on the energy storage dynamic value coefficient to obtain an energy storage state robust safety envelope; performing intra-day adaptive trajectory tracking according to the energy storage state robust safety envelope and real-time working conditions to obtain real-time scheduling instructions; controlling the flexibility resources to perform stepwise response according to the real-time scheduling instructions to obtain toughening scheduling results; and through the day-ahead pre-scheduling and intra-day adaptive trajectory tracking guided by the energy storage dynamic value coefficient, the active toughening capability and power supply reliability of the power grid in the ice disaster scenario are improved.

Description

Technical Field

[0001] This application relates to the field of power system dispatching technology, and in particular to a method, device, equipment and storage medium for active resilience dispatching of power grids under ice disaster scenarios. Background Technology

[0002] With the increasing frequency of extreme weather events and the high proportion of renewable energy connected to the grid, the power system faces increasingly severe challenges in terms of climate resilience. Ice storms, as typical long-duration extreme disasters, are characterized by a dual squeeze of rapidly decreasing renewable energy output and surging heating loads, posing a serious threat to the power grid's ability to maintain power supply. Enhancing the power grid's proactive resilience capabilities under long-duration ice storm scenarios has become an urgent need to ensure the safe and stable operation of the power system.

[0003] However, existing grid dispatching methods for ice storm scenarios mostly focus on intraday instantaneous rolling balancing or static day-ahead planning, lacking a global forward-looking mechanism across time periods. Existing models fail to fully exploit the energy time-shifting potential of flexible resources such as energy storage, and cannot effectively guide energy storage to reserve energy in advance before the disaster intensifies. This leads to the premature depletion of the system's adjustment margin when encountering extreme source-load mismatch, making it difficult to support the continuous supply needs under long-duration ice storm scenarios. Summary of the Invention

[0004] The main objective of this application is to provide a method, apparatus, equipment, and storage medium for proactively resilient power grid scheduling in ice storm scenarios, aiming to solve the technical problem of how to improve the proactive resilience capability of power grid scheduling in long-duration ice storm scenarios.

[0005] To achieve the above objectives, this application provides a method for proactively resilient power grid scheduling under ice storm scenarios, comprising the following steps: Acquire basic data on source, network, and load, as well as flexibility resource parameters, under ice storm scenarios; Based on the source-grid-load basic data, a dynamic value coefficient for energy storage is constructed, and day-ahead distributed bar pre-scheduling is performed based on the dynamic value coefficient for energy storage to obtain a robust safety envelope for energy storage status. Based on the robust safety envelope of the energy storage state and the real-time operating conditions, intraday adaptive trajectory tracking is performed to obtain real-time scheduling instructions; Based on the real-time scheduling instructions, the flexible resources are controlled to perform a tiered response, resulting in a robust scheduling outcome.

[0006] In one embodiment, the step of constructing a dynamic value coefficient for energy storage based on the source-grid-load basic data, and performing day-ahead distributed bloat pre-scheduling based on the dynamic value coefficient for energy storage to obtain a robust safety envelope for energy storage status includes: An initial scenario set is generated based on the aforementioned source-network-load basic data; Based on the distribution distance, the initial scene set is reduced to a typical scene set; Construct a probabilistic fuzzy set based on the set of typical scenarios to obtain fuzzy set data; The ratio of peak net load to average net load in future periods is determined based on the source-grid-load basic data, and the dynamic value coefficient of energy storage is constructed based on the scaling factor and the ratio. Based on the energy storage dynamic value coefficient, a robust pre-scheduling is performed within the fuzzy set data to obtain the energy storage state robust safety envelope.

[0007] In one embodiment, the step of performing robust pre-scheduling based on the energy storage dynamic value coefficient within the fuzzy set data to obtain the robust safety envelope of the energy storage state includes: Within the fuzzy set data, the tiered call cost of each level of flexible resources is determined to obtain tiered cost data; Based on the tiered cost data and the energy storage dynamic value coefficient, a day-ahead pre-scheduling objective function is constructed to obtain the objective function data; Based on the objective function data, the minimum expected operating cost under the worst probability distribution is calculated to obtain the day-ahead scheduling plan, and the energy storage state robust safety envelope is generated based on the day-ahead scheduling plan.

[0008] In one embodiment, the step of performing intraday adaptive trajectory tracking based on the robust safety envelope of the energy storage state and real-time operating conditions to obtain real-time scheduling instructions includes: Obtain the actual net load observation vector within a preset time window period in the past; Calculate scene similarity based on the actual net load observation vector and the historical scene database; The scene similarity is normalized by introducing a normalized exponential function of sensitivity coefficient, and the real-time weights of each typical scene are updated to obtain weight update data. Based on the weighted update data and the robust safety envelope of the energy storage state, a dynamic reference trajectory is synthesized to obtain trajectory data; The trajectory data is subjected to rolling optimization tracking to obtain real-time scheduling instructions.

[0009] In one embodiment, the step of calculating scene similarity based on the actual net load observation vector and the historical scene database includes: The net load feature vectors of each typical scenario in the corresponding time period are extracted from the historical scenario database to obtain historical feature data; Calculate the geometric distance between the actual net load observation vector and the historical feature data; Based on the geometric distance, the morphological similarity between each typical scene and the real-time working condition is determined, and the scene similarity is determined.

[0010] In one embodiment, the step of performing rolling optimization tracking on the trajectory data to obtain real-time scheduling instructions includes: The deviation between the real-time energy storage status and the trajectory data is obtained to obtain trajectory deviation data; Based on the trajectory deviation data, predict the state evolution in the future prediction time domain to obtain predicted state data; Based on the predicted state data and the intraday tiered cost, an intraday rolling optimization objective function is constructed to obtain intraday objective function data; Based on the intraday objective function data, rolling optimization is performed in the prediction time domain to obtain real-time scheduling instructions.

[0011] In one embodiment, the step of controlling the flexibility resources to perform a tiered response according to the real-time scheduling instructions to obtain a toughened scheduling result includes: The output adjustment amount of the basic support resources is determined according to the real-time scheduling instructions to obtain basic adjustment data; Based on the real-time dispatch instructions, the dispatch output of the virtual power plant and the power reduction of interruptible loads are determined to obtain demand-side adjustment data; The forced load shedding amount is determined based on the basic adjustment data and the demand-side adjustment data. Based on the basic adjustment data, the demand-side adjustment data, and the forced load shedding, a tiered response is performed to obtain the resilience scheduling result.

[0012] Furthermore, to achieve the above objectives, this application also proposes a power grid active resilience dispatching device for ice storm scenarios, the device comprising: The data acquisition module is used to acquire basic data on the source, network, and load, as well as flexibility resource parameters, under ice storm scenarios. The pre-scheduling module is used to construct the energy storage dynamic value coefficient based on the source-grid-load basic data, and perform day-ahead distributed bar pre-scheduling based on the energy storage dynamic value coefficient to obtain the energy storage state robust safety envelope. The tracking module is used to perform intraday adaptive trajectory tracking based on the robust safety envelope of the energy storage state and the real-time operating conditions to obtain real-time scheduling instructions; The response module is used to control the flexibility resources to perform a tiered response according to the real-time scheduling instructions, so as to obtain the resilience scheduling result.

[0013] In addition, to achieve the above objectives, this application also proposes a power grid active resilience scheduling device for ice disaster scenarios. The device includes: a memory, a processor, and a power grid active resilience scheduling program for ice disaster scenarios stored in the memory and executable on the processor. The power grid active resilience scheduling program for ice disaster scenarios is configured to implement the steps of the power grid active resilience scheduling method for ice disaster scenarios as described above.

[0014] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, storing a power grid active resilience scheduling program for ice disaster scenarios. When the power grid active resilience scheduling program for ice disaster scenarios is executed by a processor, it implements the steps of the power grid active resilience scheduling method for ice disaster scenarios as described above.

[0015] In addition, to achieve the above objectives, the present invention also provides a computer program product, which includes a power grid active resilience scheduling program for ice disaster scenarios. When the power grid active resilience scheduling program for ice disaster scenarios is executed by a processor, it implements the steps of the power grid active resilience scheduling method for ice disaster scenarios as described above.

[0016] This application constructs a dynamic value coefficient for energy storage to guide day-ahead pre-scheduling by generating a robust safety envelope for the energy storage's state of charge, thereby incentivizing energy storage to store energy in advance before disasters escalate. Then, based on the safety envelope and real-time operating conditions, intraday adaptive trajectory tracking is performed, dynamically synthesizing and correcting the reference trajectory to obtain real-time dispatch instructions. Finally, flexible resources are controlled to provide a tiered response, prioritizing and orderly calling up resources at each level. This achieves synergy between day-ahead global forward planning and intraday dynamic correction, enhancing the grid's proactive resilience and supply reliability under ice storm scenarios. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the first embodiment of the active resilience-enhancing scheduling method for power grids in an ice storm scenario according to this application. Figure 2 This is an overall flowchart of the first embodiment of the active resilience-enhancing scheduling method for power grids in the ice storm scenario of this application; Figure 3 This is a flowchart illustrating the second embodiment of the active resilience-enhancing dispatching method for power grids in the context of ice storms according to this application. Figure 4 This is a flowchart illustrating the third embodiment of the active resilience-enhancing scheduling method for power grids in the context of ice storms presented in this application. Figure 5 This is a structural block diagram of the first embodiment of the active toughening dispatching device for power grids in the ice storm scenario of this application; Figure 6 This is a schematic diagram of the active toughening and dispatching equipment for power grids in the ice storm scenario described in this application.

[0020] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of this application and are not intended to limit this application.

[0022] It should be noted that the executing entity of the embodiments of this application can be a computing service device with data processing, network communication, and program execution functions, such as a personal computer or mobile phone, or an electronic device capable of realizing the above functions, such as the active resilience-enhancing dispatching device for power grids in the ice disaster scenario. The following uses the active resilience-enhancing dispatching device for power grids in the ice disaster scenario as an example to describe the following embodiments.

[0023] This application provides a method for active resilience-enhancing power grid scheduling in ice storm scenarios, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the active resilience-enhancing scheduling method for power grids in the context of an ice storm scenario, as described in this application.

[0024] Existing technologies primarily employ topology control based on pre-set ice melting and active disconnection to construct physical defenses, combined with flexible resource probabilistic scheduling mechanisms across multiple time scales to mitigate short-term power fluctuations in renewable energy sources. They generally utilize a combination of day-ahead static planning and intraday instantaneous rolling scheduling for disaster response. However, this approach suffers from significant drawbacks: poor perception of bidirectional source-load compression; existing models lack modeling of the coupled disaster-causing mechanism of extreme cold leading to a surge in heating load and icing causing a sharp drop in renewable energy output, making it difficult to capture the deep source-load mismatch risk during the long-term evolution of ice disasters; insufficient cross-period forward-looking scheduling capability; conventional rolling optimization focuses on transient balance, failing to fully exploit the cross-period energy shift potential of flexible resources such as energy storage, leading to premature depletion of the system's adjustment margin when disaster limits are reached; and poor intraday trajectory tracking adaptability; the static reference trajectory lacks a dynamic weight update mechanism based on real-time net load deviation, making the actual operating trajectory unable to flexibly adapt to rapidly changing extreme conditions, severely limiting the dynamic adjustment potential of flexible resources.

[0025] In this embodiment, the active resilience-enhancing scheduling method for power grids under ice storm scenarios includes the following steps: Step S10: Obtain basic source-grid-load data and flexibility resource parameters under the ice storm scenario; like Figure 2 As shown, Figure 2 The overall flowchart consists of five main modules: input layer, uncertainty set generation module for "source-network-load" scenario under ice disaster, day-ahead sub-Blu-ray bar pre-scheduling module, intraday adaptive trajectory tracking module, and output and evaluation module.

[0026] The input layer integrates extreme ice storm weather and source-grid-load forecast data, as well as the operation and cost parameters of multi-level flexible resources. The source-grid-load scenario uncertainty set generation module under ice storm conditions constructs a source-grid-load probability model and combines scenario sampling and reduction techniques to ultimately establish a fuzzy set based on Wasserstein distance. The day-ahead fuzzy bar pre-scheduling module aims to minimize the expected operating cost under the worst-case probability distribution. It prioritizes low-cost resources through tiered cost allocation and introduces dynamic value assessment to incentivize energy storage to proactively reserve energy before a disaster. The intraday adaptive trajectory tracking module employs a rolling optimization mechanism based on MPC. By dynamically synthesizing energy storage reference trajectories adapted to ice storm changes, it coordinates the scheduling of flexible resources under multiple constraints to minimize intraday operating costs and trajectory deviations. The output and evaluation module outputs the tiered response status and output curves of flexible resources, the actual evolution trajectory and safety envelope of energy storage SOC, and the total system operating cost and load shedding amount under ice storm scenarios.

[0027] It should be noted that the ice storm scenario refers to the operating condition of the power grid encountering extreme low temperatures and icing weather conditions, specifically a disaster environment where increased ice thickness on transmission lines and a sharp drop in renewable energy output occur simultaneously. This scenario characterizes the extreme weather threats faced by the power grid. Additionally, the source-grid-load basic data is a set of fundamental information describing the operating status of power sources, the power grid, and loads. This includes renewable energy output forecasts, grid topology parameters, and power load demand data, which support the construction of subsequent dispatch models. Furthermore, flexibility resource parameters describe the configuration information of the operating characteristics and adjustment capabilities of various adjustable resources. This includes parameters such as the capacity boundaries and cost coefficients of conventional units, energy storage, virtual power plants, and interruptible loads, which are used to define the resource constraints for dispatch optimization.

[0028] Understandably, obtaining basic data on the source, grid, and load, as well as flexible resource parameters in the context of ice storms, refers to collecting extreme ice storm weather forecasts, renewable energy output data, load demand data, and capacity and cost parameters of multi-level flexible resources to obtain the basic input data required for the scheduling model.

[0029] Through the above steps, unified collection of multi-source heterogeneous data in ice storm scenarios was achieved, providing a complete data foundation for day-ahead pre-scheduling and intraday tracking.

[0030] In the specific implementation, the basic data required by the model are collected and integrated, including: forecast data for extreme ice storms (such as ice thickness, wind speed, and ambient temperature); historical and forecast data for wind and solar power output from renewable energy sources; forecast data for electricity load demand; and capacity configuration, operating boundaries, and tiered response cost parameters for multi-level flexible resources (including conventional units, energy storage, virtual power plants, and interruptible loads). After preprocessing, these basic data serve as input variables for subsequent uncertainty scenario construction and multi-timescale collaborative optimization.

[0031] Step S20: Construct the dynamic value coefficient of energy storage based on the source-grid-load basic data, and perform day-ahead distributed bar pre-schedule based on the dynamic value coefficient of energy storage to obtain the robust safety envelope of energy storage status; It should be noted that the dynamic value coefficient of energy storage is a quantitative indicator characterizing the differentiated importance of the energy storage state of charge (SOC) at different time periods. It is a weighted parameter jointly determined by the scaling factor and the future net load peak-to-average ratio (PAPR), used to guide energy storage to proactively reserve energy before disasters intensify. Furthermore, day-ahead sub-Blule pre-scheduling refers to a scheduling method that considers the uncertainty of probability distribution at the day-ahead time scale and seeks the optimal solution under the worst-case distribution. It is a pre-scheduling model built based on sub-Blule optimization theory, used to generate scheduling plans with disaster resilience safety margins. Further, the robust safety envelope of the energy storage state of charge (SOC) describes the allowable operating range boundary formed after day-ahead optimization. It is the upper and lower limits of the SOC with safety margins generated by sub-Blule pre-scheduling, used to guide trajectory tracking during the intraday phase.

[0032] Understandably, constructing a dynamic value coefficient for energy storage based on source-grid-load data, and performing day-ahead pre-scheduling of the energy storage system based on the dynamic value coefficient to obtain the robust safety envelope of the energy storage state refers to using source-grid-load data to calculate the dynamic value coefficient of energy storage and substituting it into the sub-Bluer optimization model to solve, thereby obtaining the robust safety operating range of the energy storage state of charge.

[0033] Through the above steps, a quantitative assessment of the energy shift potential of energy storage across time periods and the day-ahead global optimal pre-scheduling are achieved, providing a reference boundary with a safety margin for the intraday stage.

[0034] Step S30: Perform intraday adaptive trajectory tracking based on the robust safety envelope of the energy storage status and real-time operating conditions to obtain real-time dispatch instructions; It should be noted that real-time operating conditions refer to the actual operating status of the power grid at the current moment, including information such as actual net load, actual renewable energy output, and real-time equipment status, reflecting the true operating conditions of the power grid. Additionally, intraday adaptive trajectory tracking refers to a scheduling method that dynamically adjusts and follows a reference trajectory based on real-time operating conditions on an intraday timescale. This involves using model predictive control and rolling optimization to bring the energy storage status close to the dynamic reference trajectory, eliminating deviations between day-ahead plans and real-time operation. Furthermore, real-time dispatch commands are immediate operational commands generated based on intraday optimization results for controlling flexibility resources. These commands include control signals encompassing unit output, energy storage charging and discharging, and load regulation, guiding the real-time response of power grid equipment.

[0035] Understandably, obtaining real-time dispatch instructions by performing intraday adaptive trajectory tracking based on the robust safety envelope of energy storage status and real-time operating conditions refers to synthesizing a dynamic reference trajectory based on real-time operating conditions and the day-ahead safety envelope, and then obtaining the optimal control strategy in the time domain through rolling optimization.

[0036] Through the above steps, effective coordination between daily planning and intraday operations is achieved, enabling real-time scheduling to absorb fluctuations without deviating from the global optimal strategy.

[0037] Step S40: Control the flexibility resources to perform a tiered response according to the real-time scheduling instructions to obtain the resilience scheduling result.

[0038] It should be noted that tiered response refers to a scheduling mechanism that prioritizes different levels of flexibility resources based on cost, prioritizing low-cost basic resources, then medium-cost demand-side resources, and finally considering high-cost load shedding measures. This mechanism aims to minimize operating costs while maintaining supply-demand balance. Additionally, the resilience-enhanced scheduling result is the grid operating status and resource output configuration obtained after active resilience-enhanced scheduling. This includes scheduling outputs containing information such as resource output curves at each level, energy storage status trajectories, and load shedding amounts. It characterizes the grid's resilience enhancement effect under ice storm scenarios.

[0039] Understandably, controlling flexible resources to respond in a tiered manner according to real-time dispatch instructions to obtain resilient dispatch results means sequentially calling basic support resources, virtual power plants, interruptible loads, and forced load shedding according to the priority order of real-time dispatch instructions, thereby obtaining resilient dispatch results.

[0040] Through the above steps, the orderly and coordinated use of flexible resources at multiple levels is achieved, which prevents power grid collapse while avoiding redundant scheduling costs.

[0041] In one feasible implementation, step S40 includes: Step S401: Determine the output adjustment amount of basic support resources according to the real-time scheduling instructions, and obtain basic adjustment data; It should be noted that basic support resources refer to traditional power generation resources that provide basic power supply and regulation capabilities, namely thermal power units, hydropower units, and conventional energy storage, which are power sources with stable output capabilities and are used to ensure the basic power balance of the power grid. Additionally, output adjustment refers to the amount of active power that basic support resources need to increase or decrease according to dispatch instructions; that is, the amount of power change that the units adjust upwards or downwards from their current output, used to respond to real-time changes in the supply-demand gap. Furthermore, basic adjustment data is structured information describing the output adjustment of basic support resources, i.e., a set of parameters including the adjustment amount of each unit and the adjusted output value, used to guide the actual regulation operations of basic resources.

[0042] Understandably, determining the output adjustment amount of basic support resources based on real-time dispatch instructions and obtaining basic adjustment data refers to analyzing the output adjustment requirements of thermal power units, hydropower units, and conventional energy storage in real-time dispatch instructions to obtain basic adjustment data.

[0043] Through the above steps, rapid response of basic support resources is achieved, providing primary power balance guarantee for the power grid.

[0044] Step S402: Determine the dispatch output of the virtual power plant and the power reduction of interruptible loads according to the real-time dispatch instructions to obtain demand-side adjustment data; It should be noted that a virtual power plant refers to a unified dispatch unit formed by aggregating various distributed energy sources and controllable loads. In other words, it is a virtualized resource that integrates distributed power sources, energy storage, and controllable loads, presenting itself as a single power plant, and is used to provide flexible demand-side regulation capabilities. Furthermore, dispatch output refers to the active power support provided by the virtual power plant to the grid according to dispatch instructions; that is, the net power output from the aggregated resources within the virtual power plant to the grid, used to supplement insufficient regulation of basic resources. Further, interruptible loads refer to load resources whose power supply can be interrupted according to contractual agreements during grid emergencies; that is, the portion of industrial large users or commercial loads that can be temporarily cut off, used to quickly reduce load demand when there is a severe imbalance between supply and demand. Additionally, power reduction refers to the amount of electricity consumed by interruptible loads reduced according to dispatch instructions; that is, the power difference between the current electricity consumption level and the minimum allowable level, used to alleviate the power supply pressure on the grid. Furthermore, demand-side adjustment data is a comprehensive set of information describing the dispatch output of virtual power plants and the power reduction of interruptible loads. It includes a set of parameters that include the amount of demand-side resource adjustment and the state after adjustment, and is used to guide the coordinated response of demand-side resources.

[0045] Understandably, determining the dispatch output of the virtual power plant and the power reduction of interruptible loads based on real-time dispatch instructions to obtain demand-side adjustment data refers to analyzing the power support requirements of the virtual power plant and the power reduction instructions of interruptible loads in the real-time dispatch instructions to obtain demand-side adjustment data.

[0046] Through the above steps, the flexible resources on the demand side are effectively activated, providing an important power balancing supplement when the basic resources are insufficient.

[0047] Step S403: Determine the forced load shedding amount based on the basic adjustment data and demand-side adjustment data; It should be noted that the forced load shedding amount refers to the load power that must be cut off when the supply and demand balance cannot be met even after the adjustment of basic resources and demand-side resources. It is a quantitative indicator of the ultimate load control measures taken by the power grid to ensure safe and stable operation, and it is used to prevent power grid collapse.

[0048] Understandably, determining the mandatory load shedding amount based on basic adjustment data and demand-side adjustment data means assessing the remaining supply-demand gap after adjusting basic support resources and demand-side resources, thereby determining the load power that must be shedding.

[0049] Through the above steps, the ultimate safeguard measures for severe supply and demand imbalances have been quantified, providing the last line of defense for the safe operation of the power grid.

[0050] Step S404: Based on the basic adjustment data, demand-side adjustment data and forced load shedding, perform a tiered response to obtain the resilience scheduling result.

[0051] Understandably, the resilient scheduling result is obtained by performing a tiered response based on basic adjustment data, demand-side adjustment data, and forced load shedding. This means that basic resource adjustment, demand-side resource allocation, and forced load shedding are executed sequentially according to cost priority to obtain the resilient scheduling result.

[0052] Through the above steps, multi-level resources are allocated in an orderly manner according to priority, which maximizes the satisfaction of technical scheduling objectives while avoiding redundant scheduling costs and enhances the proactive resilience of the power grid in ice storm scenarios.

[0053] In the actual implementation, the following results are output after the model calculation is completed: The stepped response status and output curves of multi-level flexible resources throughout the entire cycle of ice disaster evolution; The dynamic evolution trajectory of the actual SOC of the energy storage system and its dynamic curve with respect to the safety envelope; Total system operating cost and unavoidable load shedding losses under extreme ice disaster evolution scenarios; The above outputs enable a quantitative evaluation of multi-timescale coordination strategies in long-duration ice storm scenarios, helping to form a forward-looking and optimized scheduling scheme that avoids premature depletion of flexible resources, effectively resisting the "two-way squeeze of source and load" caused by ice storms, and minimizing the risk of system load reduction.

[0054] This embodiment achieves unified collection of multi-source heterogeneous information under extreme disaster conditions by acquiring basic source-grid-load data and flexibility resource parameters in the context of ice storm scenarios. By constructing a dynamic value coefficient for energy storage and performing day-ahead distributed pre-scheduling based on this coefficient, a robust safety envelope for energy storage status is obtained. This effectively guides energy storage to reserve energy in advance before the disaster intensifies, providing a reference boundary with a safety margin for intraday scheduling. By performing intraday adaptive trajectory tracking based on the robust safety envelope of energy storage status and real-time operating conditions, real-time scheduling commands are obtained, achieving synergy between day-ahead global planning and intraday dynamic correction. This ensures that real-time scheduling does not deviate from the globally optimal strategy while absorbing fluctuations. By controlling flexibility resources to perform tiered responses according to real-time scheduling commands, resilient scheduling results are obtained, enabling the orderly use of basic support resources, virtual power plants, interruptible loads, and forced load shedding. Therefore, a closed-loop synergy between day-ahead pre-scheduling and intraday tracking is formed in the long-duration evolution scenario of ice storms, improving the grid's proactive resilience capabilities and supply reliability.

[0055] Reference Figure 3 , Figure 3 This is a flowchart illustrating the second embodiment of the active resilience-enhancing scheduling method for power grids in the context of ice storms according to this application.

[0056] In one feasible implementation, step S20 may include: Step S201: Generate an initial scene set based on the source-network-load basic data; It should be noted that the initial scenario set refers to the original discrete scenario group obtained through the scenario generation method, which describes the uncertainty of source, grid and load. That is, it is a data set containing a large number of samples of renewable energy output and load demand combinations, which is used to characterize the probabilistic characteristics of source and load fluctuations under ice disaster scenarios.

[0057] Understandably, generating an initial scenario set based on source-grid-load basic data means using the uncertainty information in the source-grid-load basic data to build a probabilistic model and sample to generate discrete scenarios, thereby obtaining an initial scenario set.

[0058] Through the above steps, the source load uncertainty in the ice disaster scenario was discretized, providing a basis of original samples for subsequent scenario reduction.

[0059] Step S202: Reduce the initial scene set based on the distribution distance to obtain a typical scene set; It's important to note that distribution distance is a metric measuring the degree of difference between two probability distributions; it's a distance function used to quantify the deviation of probability characteristics between different scene sets, and it's used to select representative typical scenes. Scene reduction refers to the process of selecting a small number of representative scenes from a large initial scene set to reduce computational complexity. This involves retaining the key statistical features of the original distribution while reducing the number of scenes, thus improving model solving efficiency. Furthermore, the typical scene set refers to the simplified scene group obtained after scene reduction, which reflects the morphological characteristics of the original distribution. It's a dataset containing a finite number of representative source-load combination samples, used to support the subsequent construction of probabilistic fuzzy sets.

[0060] Understandably, reducing the initial scene set based on the distribution distance to obtain a typical scene set means calculating the distribution distance between each scene in the initial scene set and selecting and retaining representative scenes accordingly, thereby obtaining a typical scene set.

[0061] Through the above steps, we have achieved effective simplification of large-scale scenarios, reducing the computational burden while retaining the characteristics of long-tail risks.

[0062] Step S203: Construct a probabilistic fuzzy set based on the typical scenario set to obtain fuzzy set data; It should be noted that a probabilistic fuzzy set refers to a set of distributions centered on an empirical distribution and considering the range of uncertainty of the probability distribution. In other words, it is a fuzzy description containing multiple candidate probability distributions that the true distribution may fall into, used to characterize the robust boundary of source load uncertainty. Furthermore, fuzzy set data is a set of information describing the range and structural characteristics of the probabilistic fuzzy set; that is, it contains parameterized data including the empirical distribution, distribution radius, and candidate distribution boundaries, used to define the uncertainty search space for day-ahead pre-scheduling.

[0063] Understandably, constructing a probabilistic fuzzy set based on a set of typical scenarios to obtain fuzzy set data means using the empirical distribution and distribution distance radius of the set of typical scenarios to construct a fuzzy set containing multiple possible distributions, thereby obtaining fuzzy set data.

[0064] Through the above steps, robust modeling of source load uncertainty is achieved, providing a probabilistic candidate space including extreme scenarios for distributed robust pre-scheduling.

[0065] In the implementation, key influencing factors such as icing thickness are first considered. A "source-network-load" probabilistic model is established using a general generating function and adaptive kernel density estimation. A large-scale initial discrete scene is then generated using the Latin hypercube sampling (LHS) method. Subsequently, a scene reduction technique based on Wasserstein distance is employed to extract a finite number of typical scenes and construct an empirical probability distribution. The formula for calculating the Wasserstein distance is as follows:

[0066] In the formula, π is the joint probability distribution of ξ and ξE; ξ E Historical data on intermittent renewable energy output; the marginal probability distributions of ξ and ξE are P and PE, respectively; For ξ and ξ E The distance between them is expressed in any possible norm form. After defining the Wasserstein distance, a probabilistic fuzzy set P is constructed with the empirical distribution PE as the center and θ as the radius. This provides a long-tailed candidate space containing limit boundaries for day-ahead scheduling, and its expression is:

[0067] In the formula, θ is the Wasserstein radius, which characterizes the robustness of the model to the uncertainty of the probability distribution.

[0068] Step S204: Determine the ratio of peak net load to average net load in the future period based on the source-grid-load basic data, and construct the dynamic value coefficient of energy storage based on the scaling factor and the ratio; It should be noted that peak net load refers to the maximum net load demand of the power grid within a specific future period, i.e., the upper limit of the net demand power after subtracting renewable energy output from the total load in the forecast time domain. It is used to characterize the most severe supply-demand gap in the future period. Average net load, on the other hand, refers to the average net load demand of the power grid within a specific future period, i.e., the average net demand power after subtracting renewable energy output from the total load in the forecast time domain. It is used to characterize the average supply-demand level in the future period. Furthermore, the ratio is the value obtained by dividing peak net load by average net load, i.e., a relative indicator characterizing the severity of net load fluctuations in the future period, used to quantify the peak-valley difference in supply-demand imbalance. Additionally, the scaling factor is an adjustment parameter used to balance the differences in the magnitude of different optimization objectives. It is a weighting coefficient set appropriately to prevent the energy storage value guidance term from being submerged in the global objective function, and is used to ensure the effectiveness of dynamic value guidance for energy storage.

[0069] Understandably, determining the ratio of peak net load to average net load for future periods based on source-grid-load basic data, and constructing the dynamic value coefficient of energy storage based on the scaling factor and the ratio, refers to extracting net load forecast information for future periods from source-grid-load basic data, calculating the peak-to-average ratio, and then constructing the dynamic value coefficient of energy storage by combining it with the scaling factor.

[0070] Through the above steps, the forward-looking quantification of the value of energy storage across time periods is achieved, enabling energy storage state of charge to receive differentiated weight incentives at different time periods.

[0071] Step S205: Based on the dynamic value coefficient of energy storage, perform sub-robust pre-scheduling within the fuzzy set data to obtain the robust safety envelope of energy storage state.

[0072] Understandably, obtaining a robust safety envelope for energy storage status by performing robust pre-scheduling based on the dynamic value coefficient of energy storage within fuzzy set data means considering the worst-case probability distribution and minimizing the expected operating cost within the fuzzy set data, while using the dynamic value coefficient of energy storage to guide the advance reservation of energy storage, thereby obtaining a robust safety envelope for energy storage status.

[0073] Through the above steps, a globally optimal pre-scheduling that balances economy and robustness is achieved on the daytime scale, providing an energy storage operation boundary with a safety margin for the intraday phase.

[0074] In one feasible implementation, step S205 includes: Step S2051: Determine the tiered call cost of each level of flexibility resources within the fuzzy set data to obtain tiered cost data; It should be noted that tiered dispatch cost refers to the unit dispatch cost of each level of flexible resources, differentiated according to response priority. This means that the cost coefficients increase progressively from basic support resources, virtual power plants, interruptible loads, to forced load shedding, and are used to guide the model to prioritize the dispatch of low-cost resources. Furthermore, tiered cost data is structured information describing the tiered dispatch cost of each level of flexible resources. It includes a set of parameters containing the unit cost coefficients and priority ranking of each resource level, and is used to construct the cost component of the day-ahead pre-scheduling objective function.

[0075] Understandably, determining the tiered call costs of various levels of flexible resources within fuzzy set data to obtain tiered cost data means setting differentiated hierarchical cost parameters based on resource type and response characteristics within fuzzy set data to obtain tiered cost data.

[0076] Through the above steps, the differentiated quantification of flexible resource allocation costs was achieved, providing a hierarchical cost basis for the subsequent construction of the objective function.

[0077] Step S2052: Construct the day-ahead pre-scheduling objective function based on the tiered cost data and the dynamic value coefficient of energy storage, and obtain the objective function data; It should be noted that the day-ahead pre-scheduling objective function refers to the mathematical expression used to optimize scheduling decisions on the day-ahead time scale. It is a comprehensive performance index that includes the operating costs of various levels of flexibility resources, dynamic value assessment items for energy storage, and load shedding penalties. It is used to evaluate the economy and safety of different scheduling schemes. Furthermore, the objective function data is a set of information describing the specific composition and parameter configuration of the day-ahead pre-scheduling objective function. It includes structured data containing cost coefficients, value weights, and penalty factors, which guides the solution of the sub-Bruker optimization problem.

[0078] Understandably, constructing a day-ahead pre-scheduling objective function based on tiered cost data and energy storage dynamic value coefficients to obtain objective function data means integrating tiered cost data, energy storage dynamic value coefficients, and supply-demand balance constraints into an optimization objective expression, thereby obtaining objective function data.

[0079] Through the above steps, a mathematical expression of the multi-objective trade-offs in day-ahead scheduling is realized, enabling the synergistic optimization of economic efficiency and forward-looking energy storage reserves within a unified framework.

[0080] Step S2053: Based on the objective function data, solve for the minimum expected operating cost under the worst probability distribution to obtain the day-ahead scheduling plan, and generate the energy storage state robust safety envelope based on the day-ahead scheduling plan.

[0081] It should be noted that the worst-case probability distribution refers to the extreme probability distribution within the probability fuzzy set that maximizes the expected operating cost, i.e., the most unfavorable source-load uncertainty distribution for grid operation. It is used to test the robustness of dispatch strategies under extreme scenarios. Furthermore, the expected operating cost refers to the weighted average cost of dispatching various levels of flexible resources under the worst-case probability distribution; that is, the mathematical expectation of the operating cost of all possible scenarios after considering the probability distribution. It is used to evaluate the comprehensive economic efficiency of the dispatch scheme. Further, the day-ahead dispatch plan refers to the optimal resource output and energy storage charging / discharging arrangement obtained through optimization on the day-ahead time scale. It is a time-series decision scheme including unit output, energy storage status, and load regulation, used to guide real-time operation during the intraday phase.

[0082] Understandably, finding the minimum expected operating cost under the worst probability distribution based on the objective function data to obtain the day-ahead scheduling plan, and generating the robust safety envelope of the energy storage state based on the day-ahead scheduling plan, means finding the worst distribution that maximizes the expected operating cost within the probabilistic fuzzy set and minimizing that cost. Then, based on the optimal solution, the allowable operating range of the energy storage state of charge is extracted, thereby obtaining the robust safety envelope of the energy storage state.

[0083] Through the above steps, robust optimization solutions under extreme uncertainties are achieved, and the generated robust safety envelope of energy storage state provides dynamic boundary constraints that take into account both economy and safety for intraday scheduling.

[0084] In its implementation, a day-ahead pre-scheduling model was constructed: aiming to minimize the expected operating cost generated by calling up flexible resources at all levels for supply and demand balance under the worst-case probability distribution within the fuzzy set range. In the formula, f(x,ξ) represents the total operating cost of the system in scenario ξ, which consists of the scheduling cost of flexible resources at all levels and the dynamic value of energy storage. On the one hand, the system subdivides the scheduling costs at each level (i.e., FLevel0, FLevel1, FLevel2 in the formula) and utilizes the differences in the cost coefficients of each level to guide the model to prioritize calling up low-cost conventional resources and shifting loads, only considering the high cost of load shedding in extreme scenarios. On the other hand, considering the vulnerability of supply and demand balance under the evolution of ice storms, the model introduces a dynamic value assessment term for energy storage into the objective function. By assigning dynamic weights to the state of charge, energy storage systems are incentivized to proactively reserve capacity before disasters escalate, thus providing an adaptive safety margin for intraday periods. αt is the dynamic value coefficient of energy storage, determined by the scaling factor η and the peak and average net loads over the future period τ. The ratio of these factors determines the dynamic value assessment of the State of Charge (SOC) of energy storage, guiding the pre-positioning of flexibility resources before supply and demand shortages. Here, the scaling factor η plays a crucial role in balancing the order-of-magnitude differences between different optimization objectives. Since the system's basic operating costs and load shedding penalty costs are typically extremely high, introducing and appropriately setting η can effectively prevent numerical flooding of the energy storage value guidance term when solving the global objective function, ensuring that its incentive effect on pre-positioning energy storage is truly effective.

[0085]

[0086]

[0087]

[0088]

[0089]

[0090]

[0091]

[0092] In the formula, x represents the dispatch decision variable, which specifically includes the active power output PG,i,t of conventional units (thermal power units, hydropower units, wind power, and photovoltaic power), the charging and discharging power (Pch,i,t, Pdis,i,t) and state of charge (SOCt) of the energy storage system, the dispatch output PVPP,i,t of the virtual power plant, the interruptible load reduction power PIL,m,t, and the forced load shedding amount Pcut; ωn represents the load importance weight of node n; CL0, CL1, and CL2 are the corresponding graded unit cost coefficients.

[0093] After solving the problem, this module outputs the day-ahead optimal scheduling plan and generates a robust safety envelope of the energy storage SOC with disaster resilience margin, which serves as a reference benchmark for lower-level scheduling.

[0094] The day-ahead to intraday multi-timescale resilience enhancement model proposed in this paper needs to satisfy a series of physical operation and control constraints. These include: grid network topology and power flow constraints; output boundary constraints of multi-level flexible resources (thermal power, hydropower, energy storage, wind power, photovoltaic power, virtual power plants, and interruptible loads); and energy storage adaptive trajectory tracking constraints specific to the intraday stage.

[0095] To ensure the safe and stable operation of the power grid during ice storms, each node must maintain a balance between active and reactive power, and the power flow of lines and node voltages must not exceed limits.

[0096] Active and reactive power balance constraints at nodes:

[0097]

[0098] In the formula: , Let M and Mvar be the total active and reactive power output of node i during time period t, respectively. , Let represent the actual active and reactive load demands of node i during time period t, in MW and Mvar, respectively. , Let represent the active and reactive power flow from node i to the adjacent node j branch during time period t, in MW and Mvar, respectively. , Let represent the active and reactive power flows from adjacent node k into the branch of node i during time period t, in MW and Mvar, respectively. , Let i represent the set of adjacent nodes with node i as the head (power outflow) and the head (power inflow) respectively.

[0099] Relationship between voltage and power at both ends of the line:

[0100] In the formula: , For the scene Down Time-of-day branch Voltage at both ends of the node, kV; branch road The resistance, Ω; branch road The reactance, Ω; The rated voltage is kV.

[0101] Branch flow constraints:

[0102] In the formula, branch road The maximum active power limit that can be tolerated, in MW Mvar is the reactive power limit.

[0103] Node voltage constraints:

[0104] In the formula, , For nodes The upper and lower limits of the voltage, in kV.

[0105] Line capacity constraints:

[0106] In the formula, Pij,t is the line transmission power between node i and node j in time period t; This represents the maximum power capacity that line ij is allowed to transmit.

[0107] During the evolution of ice storms, various basic support resources and flexible adjustment resources need to respond to dispatch instructions within their respective physical limits.

[0108] Hydropower units generate electricity from runoff; the water flowing through the turbine converts its rotational kinetic energy into electrical energy. The characteristic function of a hydropower station's power generation per unit time relates to the flow rate and water head. for:

[0109] In the formula, ηi is the comprehensive output coefficient of the i-th hydropower unit; and Δt represents the power generation flow and head of the i-th hydropower unit at time t, respectively; Δt is the time interval.

[0110] In addition, the hydropower unit i at time t also satisfies the following constraints: Hydropower station power generation flow constraints:

[0111] Dynamic water balance constraints:

[0112]

[0113] Reservoir storage capacity constraints:

[0114] In the formula: , These are the upper and lower limits for power generation flow, respectively; and These represent inflow runoff and outflow, respectively, in m³ / s. For storage capacity; , These are the upper and lower limits of the reservoir capacity of the hydropower station, in m3.

[0115] Thermal power unit output constraints:

[0116] In the formula: The actual power of unit g during time period t. and These are the maximum and minimum permissible active power output limits for the unit, respectively.

[0117] The constraint on load shedding primarily limits the power reduction during any given period from exceeding its maximum responsive capacity. The specific formula is:

[0118] In the formula, PIL,m,t represents the power reduction of interruptible load m during time period t. This represents the maximum allowable power reduction limit for interruptible load m.

[0119] Virtual power plant support power upper and lower limit constraints:

[0120] Because virtual power plants aggregate various distributed resources, their output adjustments are often limited by ramp-up rates, as shown in the following formula:

[0121] In the formula, To provide power output for virtual power plant j during time period t; and These represent the upper and lower limits of the active power support that virtual power plant j can provide, respectively. and These are the upward and downward ramp rate limits for virtual power plant j, respectively. This is the scheduling time interval.

[0122] This embodiment reduces computational complexity by generating an initial scenario set and then reducing it, while retaining the long-tail risk characteristics of ice storms. By constructing a probabilistic fuzzy set, it provides a robust boundary for day-ahead pre-scheduling that includes extreme distribution candidates. By determining the ratio of peak net load to average net load in future periods and combining it with a scaling factor to construct a dynamic value coefficient for energy storage, it achieves forward-looking quantification of the value of energy storage across time periods. By determining the tiered call cost within the probabilistic fuzzy set and constructing a day-ahead pre-scheduling objective function, it guides the model to prioritize the use of low-cost resources and consider load shedding costs in extreme scenarios. By solving for the minimum expected operating cost under the worst-case probability distribution, it obtains the day-ahead scheduling plan, thereby generating a robust safety envelope for energy storage status. Thus, it achieves globally optimal pre-scheduling that balances economy and robustness on the day-ahead time scale. The generated robust safety envelope for energy storage status provides a dynamic operating boundary with safety margins for the intraday phase, enabling energy storage to proactively reserve energy before disaster intensifies, improving cross-time period adjustment capabilities and supply reliability in extreme scenarios.

[0123] This embodiment refers to Figure 4 , Figure 4 This is a flowchart illustrating the third embodiment of the active resilience-enhancing scheduling method for power grids in the context of ice storms described in this application.

[0124] In one feasible implementation, step S30 includes: Step S301: Obtain the actual net load observation vector within the past preset time window length; It should be noted that the actual net load observation vector refers to the time-series data set of the actual net load of the power grid within a preset time window period in the past. That is, it is the net demand power sequence obtained by subtracting the actual wind power output and actual photovoltaic power output from the actual total load, and it is used to characterize the real supply and demand status of the power grid in the near future. In addition, the preset time window length refers to the length of the past period used to observe the real-time operating conditions in intraday adaptive trajectory tracking, that is, the fixed time span for retrospectively collecting historical data, which is used to determine the data range for similarity calculation.

[0125] Understandably, obtaining the actual net load observation vector within a preset time window in the past refers to collecting data on the actual total load, actual wind power output, and actual photovoltaic power output within a fixed time period in the past and calculating the net load sequence to obtain the actual net load observation vector.

[0126] Through the above steps, the quantitative extraction of the recent real operating status of the power grid was achieved, providing a reliable real-time operating condition benchmark for subsequent scenario similarity calculation.

[0127] Step S302: Calculate scene similarity based on the actual net load observation vector and the historical scene database; Understandably, calculating scene similarity based on the actual net load observation vector and the historical scene database means comparing and analyzing the net load characteristics of each typical scene in the historical scene database to obtain scene similarity that characterizes the degree of similarity between the real-time operating conditions and the historical scene.

[0128] Through the above steps, the form matching between real-time operating conditions and typical current scenarios was achieved, providing a quantitative basis for adaptive weight updates.

[0129] In one feasible implementation, step S302 includes: Step S3021: Extract the net load feature vector of each typical scenario in the corresponding time period from the historical scenario database to obtain historical feature data; It should be noted that the historical scenario database refers to the collection of typical scenarios built and stored during the pre-scheduling phase of the daytime sub-segmentation, i.e., a database containing net load feature vectors of each typical scenario at different time periods, used to support scenario matching and trajectory synthesis during the intraday phase. Additionally, the net load feature vector refers to feature data describing the net load variation patterns of typical scenarios within corresponding time periods, i.e., a vector representation of the net load values ​​of a scenario at different times, used to characterize the typical morphological features of the scenario. Furthermore, historical feature data refers to the collection of net load feature vectors of each typical scenario extracted from the historical scenario database, i.e., historical scenario morphological information used for comparison with real-time operating conditions, used for subsequent geometric distance calculations.

[0130] Understandably, extracting the net load feature vectors of each typical scenario in the corresponding time period from the historical scenario database to obtain historical feature data means accessing the historical scenario database and reading the net load feature vectors of each typical scenario in the matching time period to obtain historical feature data.

[0131] Through the above steps, the characteristics of typical current scenarios can be quickly retrieved, providing a standardized data foundation for comparing the morphology of real-time operating conditions with historical scenarios.

[0132] Step S3022: Calculate the geometric distance between the actual net load observation vector and historical feature data; It should be noted that geometric distance refers to a numerical value that measures the difference between two vectors in space. It is the distance value obtained by taking the square root of the sum of the squares of the differences between the vector components. It is used to quantify the degree of morphological deviation between the actual net load observation vector and historical feature data.

[0133] Understandably, calculating the geometric distance between the actual net load observation vector and historical feature data means subtracting the corresponding components of the actual net load observation vector and historical feature data and calculating the Euclidean distance value to obtain the geometric distance.

[0134] Through the above steps, the numerical quantification of the differences between real-time operating conditions and historical scene forms was achieved, providing an accurate distance metric for similarity determination.

[0135] Step S3023: Determine the morphological similarity between each typical scene and the real-time working condition based on the geometric distance, and determine the scene similarity.

[0136] Understandably, determining the morphological similarity between typical scenarios and real-time operating conditions based on geometric distance means converting geometric distance into a similarity index that reflects the degree of morphological closeness, thereby obtaining the scene similarity.

[0137] The above steps achieve the transformation from distance measurement to similarity assessment, making scene matching results more intuitive and usable.

[0138] Step S303: Normalize the scene similarity by introducing a normalized exponential function of the sensitivity coefficient, update the real-time weights of each typical scene, and obtain the weight update data. It's important to note that the sensitivity coefficient is a parameter that adjusts the sensitivity of the normalized exponential function to similarity differences. A larger coefficient indicates a greater likelihood that the model will trust the weight adjustment factor of the most similar single scene, and it controls the concentration of scene weight distribution. The normalized exponential function, also known as the Softmax function, is a mathematical function that transforms any real-valued vector into a probability distribution. It's a probability mapping operation that uses exponential and normalization operations to sum the output values ​​to one, and it's used to convert geometric distance into probabilistic weights usable for scheduling. Furthermore, real-time weights refer to the confidence coefficients assigned to each typical scene at the current moment for synthesizing dynamic reference trajectories. These probabilistic weight values ​​reflect the degree of matching between the scene and real-time conditions, and they guide the weighted aggregation of the day-ahead envelope.

[0139] Understandably, normalizing scene similarity by introducing a normalized exponential function of sensitivity coefficient and updating the real-time weights of each typical scene to obtain weight update data means using a normalized exponential function combined with sensitivity coefficient to perform a probabilistic transformation of scene similarity, thereby obtaining real-time weight update data for each typical scene.

[0140] Through the above steps, an adaptive conversion from scene similarity to probability weights is achieved, enabling the model to automatically increase the confidence of the most similar scene when the working conditions change drastically.

[0141] Step S304: Synthesize the dynamic reference trajectory based on the weighted update data and the robust safety envelope of the energy storage state to obtain trajectory data; It should be noted that the dynamic reference trajectory refers to the reference curve adapted to the current operating conditions formed by weighted aggregation of the energy storage state envelopes of each typical scenario based on real-time weights. In other words, it is the energy storage state of charge target trajectory that integrates the day-ahead optimization information of all typical scenarios and is corrected in real time. It is used to guide the real-time tracking of energy storage during the intraday period.

[0142] Understandably, obtaining trajectory data by synthesizing dynamic reference trajectory data based on weighted updated data and robust safety envelope of energy storage status refers to using real-time weights to weight and fuse the robust safety envelope of energy storage status for each typical scenario, thereby obtaining dynamic reference trajectory data adapted to the current operating conditions.

[0143] Through the above steps, the adaptive fusion of the day-ahead safety envelope and real-time operating conditions is achieved, generating an energy storage tracking target that fits the current operating status.

[0144] Step S305: Perform rolling optimization tracking on the trajectory data to obtain real-time scheduling instructions.

[0145] Understandably, rolling optimization tracking of trajectory data to obtain real-time scheduling instructions refers to repeatedly optimizing and solving the dynamic reference trajectory in the prediction time domain and executing the first step of control to obtain real-time scheduling instructions.

[0146] Through the above steps, the energy storage status is closely followed by the dynamic reference trajectory, ensuring that intraday scheduling does not deviate from the global optimal strategy while smoothing out real-time fluctuations.

[0147] In one feasible implementation, step S305 includes: Step S3051: Obtain the deviation between the real-time energy storage status and the trajectory data to obtain trajectory deviation data; Understandably, obtaining the deviation between the real-time energy storage status and the trajectory data, and thus obtaining the trajectory deviation data, means collecting the actual state of charge of the energy storage at the current moment and calculating the difference between it and the corresponding value of the dynamic reference trajectory, thereby obtaining the trajectory deviation data.

[0148] Through the above steps, the deviation between the actual operating state of energy storage and the target trajectory is quantified in real time, providing deviation input for rolling optimization.

[0149] Step S3052: Predict the state evolution in the future prediction time domain based on the trajectory deviation data to obtain the predicted state data; Understandably, predicting the state evolution in the future prediction time domain based on trajectory deviation data and obtaining predicted state data means combining the current trajectory deviation with the real-time operating conditions of the power grid to extrapolate the evolution trend of energy storage and power grid status in the future period, thereby obtaining predicted state data.

[0150] Through the above steps, we can make an advance prediction of the power grid's operating status in future periods, providing a forward-looking basis for rolling optimization.

[0151] Step S3053: Construct an intraday rolling optimization objective function based on the predicted state data and intraday tiered cost to obtain intraday objective function data; It should be noted that intraday tiered cost refers to the differentiated unit dispatch cost set according to the response priority of each level of flexible resources on the intraday time scale. This is the cost coefficient that increases progressively for basic support resources, virtual power plants, interruptible loads, and forced load shedding, and it is used for the economic trade-offs in intraday rolling optimization. Furthermore, the intraday rolling optimization objective function is a mathematical expression used to optimize real-time dispatch decisions within the intraday forecast time domain. It is a comprehensive performance index that includes intraday tiered cost and trajectory deviation penalty terms, and it is used to evaluate the economy and tracking accuracy of different real-time dispatch schemes.

[0152] Understandably, constructing an intraday rolling optimization objective function based on predicted status data and intraday tiered costs to obtain intraday objective function data means integrating predicted status data, trajectory deviation penalty terms, and intraday tiered costs of resources at all levels into an optimization objective expression, thereby obtaining intraday objective function data.

[0153] Through the above steps, a mathematical expression of the multi-objective trade-off in intraday scheduling is realized, enabling real-time economic optimization and trajectory tracking accuracy to work together within a unified framework.

[0154] Step S3054: Based on the intraday objective function data, perform rolling optimization in the prediction time domain to obtain real-time scheduling instructions.

[0155] Understandably, obtaining real-time scheduling instructions by performing rolling optimization solutions based on intraday objective function data in the prediction time domain means solving for the optimal value of intraday objective function data and extracting the initial control strategy in the prediction time domain, thereby obtaining real-time scheduling instructions.

[0156] Through the above steps, rolling optimization solutions on the intraday time scale are achieved, enabling real-time scheduling commands to respond promptly to changes in operating conditions and approximate the dynamic reference trajectory.

[0157] In its implementation, this module employs a rolling optimization mechanism based on Model Predictive Control (MPC): to address the issue of real-time operating conditions deviating from typical daytime scenarios, leading to the failure of upper-level scheduling commands, this paper matches the current real-time operating conditions with a historical scenario library and dynamically adjusts the tracking target. By sensing the changing trend of the system's net load in real time, the confidence weights of the energy storage state envelope for each typical daytime scenario are dynamically adjusted, thereby synthesizing a dynamic reference trajectory that adapts to the current operating conditions and guiding energy storage resources to maintain effective tracking during real-time scheduling.

[0158] At real-time scheduling time t, the actual net load observation vector within the past Hwin period is defined as follows: The calculation formula is as follows:

[0159] In the formula: , , , They are time points The system's actual net load, actual total load, actual wind power output, and actual photovoltaic power output; This represents the current real-time scheduling moment. This refers to the length of the time window for real-time operating condition observation.

[0160] Simultaneously, from the typical scenario library constructed during the day-ahead scheduling phase, the net load feature vector of each scenario k in the corresponding time period is extracted. The morphological similarity dk(t) between the current real-time operating condition and each typical scenario from the day-ahead is quantified using Euclidean distance:

[0161] In the formula: N is the total number of typical scenarios; For scene indexing. For the past The actual net load observation vector within the time period; For the first Net load feature vectors for a typical scenario in the corresponding time period; For the first A typical scenario at time The net load value.

[0162] To transform geometric distance into probabilistic weights that can guide scheduling, a Softmax function with a sensitivity coefficient is introduced to normalize the similarity. The real-time weight ωk(t) update rules for each typical scenario at time t are as follows:

[0163] In the formula, β is the sensitivity adjustment coefficient. The larger the β is, the more the model tends to trust the single most similar scenario. This is the average distance between all scenes at the current moment, used for numerical normalization to prevent exponential explosion. Through this strategy, when real-time operating conditions change abruptly, the model can automatically identify and increase the weight of the closest scene, achieving adaptive perception of changes in operating conditions.

[0164] Based on the updated real-time weights ωk(t), the energy storage charge state envelopes optimized under various typical scenarios during the day-ahead scheduling phase are weighted and aggregated to generate a dynamic reference trajectory to guide future operation within the prediction time domain Hpred. :

[0165] This dynamic trajectory integrates day-ahead optimization information from all typical scenarios and undergoes weighted correction based on real-time matching results. In the lower-level MPC optimization model, real-time state is penalized... The deviation ensures that while absorbing real-time fluctuations, the system can return to the planned economic and safe range, achieving effective coordination between long and short time scales.

[0166] Objective function and constraints: Within each forecast window, the MPC model will solve for the real-time dispatch instructions that minimize intraday tiered costs and trajectory deviation penalties. Optimization variables include: output of basic conventional units; energy storage charging and discharging power; the amount of virtual power plants and interruptible loads called; and the amount of forced load shedding, etc.

[0167] Objective function: In the prediction time domain τ∈[t, t+Hpred-1], the objective function is established as follows:

[0168]

[0169] In the formula, , and These correspond to basic operating costs, flexible resource allocation costs, and forced load shedding penalty costs, respectively. To avoid redundancy, the mathematical calculation logic for these three cost components is completely consistent with the formula in the upper-level model. In the specific solution, simply replace the day-ahead planning variables in the original formula with the real-time scheduling decision variables in the current forecast time domain, and continue using the corresponding hierarchical cost coefficients.

[0170] Unlike the upper-level model, the lower-level model no longer includes a reward term to guide advance energy storage reserves, but instead introduces a dynamic trajectory tracking penalty term. To force real-time energy storage state Closely follow the dynamic reference trajectory generated by the upper layer to ensure that intraday scheduling does not deviate from the global optimal strategy while smoothing out deviations.

[0171] The constraints cover: active and reactive power balance at nodes; branch power flow and node voltage safety boundaries; operation and ramping constraints of hydro / thermal power units; capacity limits for tiered call of flexible resources; and deviation control constraints for dynamic trajectory tracking of energy storage. However, under the MPC rolling optimization architecture, the above constraints need to be re-established in the prediction time domain Hpred at each scheduling time t.

[0172] This embodiment achieves quantitative extraction of the recent real supply and demand status of the power grid by acquiring the actual net load observation vector within a preset time window. By calculating the geometric distance between the actual net load observation vector and the net load feature vectors of each typical scenario in the historical scenario database, the scenario similarity, which characterizes the degree of deviation between the real-time operating condition and the historical scenario, is obtained. By introducing a normalized exponential function of the sensitivity coefficient to normalize the scenario similarity and updating the real-time weights of each typical scenario, adaptive perception that automatically increases the confidence of the most similar scenario when the operating condition changes abruptly is achieved. Based on the real-time weights and the robust safety envelope of the energy storage status, a dynamic reference trajectory is synthesized to obtain an energy storage tracking target that fits the current operating status. By acquiring the deviation between the real-time energy storage status and the dynamic reference trajectory and predicting the state evolution in the future prediction time domain, an intraday rolling optimization objective function is constructed in combination with intraday tiered costs and solved by rolling optimization in the prediction time domain to obtain real-time dispatch instructions. This enables adaptive fusion of the day-ahead global safety envelope and intraday real-time operating conditions, allowing energy storage to closely follow the dynamic reference trajectory without deviating from the global optimal strategy while absorbing real-time fluctuations. This improves the intraday dynamic resilience and real-time supply reliability of the power grid in ice storm scenarios.

[0173] Reference Figure 5 , Figure 5 This is a structural block diagram of the first embodiment of the active resilience-enhancing dispatching device for power grids in the ice storm scenario of this application.

[0174] like Figure 5 As shown in the embodiments of this application, the active resilience-enhancing dispatching device for power grids in ice storm scenarios includes: Data acquisition module 10 is used to acquire basic source-grid-load data and flexibility resource parameters under ice disaster scenarios; The pre-scheduling module 20 is used to construct the energy storage dynamic value coefficient based on the source-grid-load basic data, and to perform day-ahead distributed bar pre-scheduling based on the energy storage dynamic value coefficient to obtain the energy storage state robust safety envelope. The tracking module 30 is used to perform intraday adaptive trajectory tracking based on the robust safety envelope of the energy storage status and real-time operating conditions to obtain real-time scheduling instructions; The response module 40 is used to control the flexible resources to perform a tiered response according to the real-time scheduling instructions, so as to obtain the toughened scheduling result.

[0175] Based on the first embodiment of the active resilience-enhancing dispatching device for power grids in the ice disaster scenario described in this application, a second embodiment of the active resilience-enhancing dispatching device for power grids in the ice disaster scenario is proposed.

[0176] In this embodiment, the pre-scheduling module 20 is also used to generate an initial scenario set based on the source-grid-load basic data; The initial scene set is reduced based on the distribution distance to obtain a typical scene set; Construct a probabilistic fuzzy set based on a set of typical scenarios to obtain fuzzy set data; Based on the source-grid-load basic data, determine the ratio of peak net load to average net load in the future period, and construct the dynamic value coefficient of energy storage based on the scaling factor and the ratio. Based on the dynamic value coefficient of energy storage, a robust pre-scheduling is performed within the fuzzy set data to obtain the robust safety envelope of energy storage status.

[0177] In this embodiment, the pre-scheduling module 20 is also used to determine the tiered call cost of each level of flexibility resources within the fuzzy set data, and obtain tiered cost data. Based on the tiered cost data and the dynamic value coefficient of energy storage, a day-ahead pre-scheduling objective function is constructed, and the objective function data is obtained; The minimum expected operating cost under the worst probability distribution is obtained by solving the objective function data, and the day-ahead scheduling plan is obtained. The robust safety envelope of energy storage state is generated based on the day-ahead scheduling plan.

[0178] In this embodiment, the tracking module 30 is also used to obtain the actual net load observation vector within a preset time window in the past; Calculate scene similarity based on the actual net load observation vector and the historical scene database; By introducing a normalized exponential function of sensitivity coefficient to normalize scene similarity, the real-time weights of each typical scene are updated to obtain weight update data. The trajectory data is obtained by synthesizing a dynamic reference trajectory based on the weighted update data and the robust safety envelope of the energy storage status. The trajectory data is subjected to rolling optimization tracking to obtain real-time scheduling instructions.

[0179] In this embodiment, the tracking module 30 is also used to extract the net load feature vector of each typical scene in the corresponding time period from the historical scene library to obtain historical feature data; Calculate the geometric distance between the actual net load observation vector and historical feature data; The morphological similarity between each typical scenario and the real-time operating condition is determined based on the geometric distance, thus determining the scenario similarity.

[0180] In this embodiment, the tracking module 30 is also used to obtain the deviation between the real-time energy storage status and the trajectory data, and to obtain trajectory deviation data; Based on trajectory deviation data, predict the future state evolution within the prediction time domain to obtain predicted state data; Based on the predicted state data and the intraday tiered cost, an intraday rolling optimization objective function is constructed, and the intraday objective function data is obtained; Based on intraday objective function data, rolling optimization is performed in the prediction time domain to obtain real-time scheduling instructions.

[0181] In this embodiment, the response module 40 is further configured to determine the output adjustment amount of the basic support resources according to the real-time scheduling instructions, and obtain basic adjustment data; Based on real-time dispatch instructions, the dispatch output of the virtual power plant and the power reduction of interruptible loads are determined to obtain demand-side adjustment data; The forced load shedding amount is determined based on the basic adjustment data and the demand-side adjustment data. Based on the basic adjustment data, demand-side adjustment data, and forced load shedding, a tiered response is performed to obtain the resilience scheduling result.

[0182] Other embodiments or specific implementations of the active resilience dispatching device for power grids in ice storm scenarios described in this application can be found in the above-mentioned method embodiments, and will not be repeated here.

[0183] This application provides a power grid active resilience scheduling device for ice disaster scenarios. The power grid active resilience scheduling device for ice disaster scenarios includes: at least one processor; and a memory communicatively connected to at least one processor; wherein the memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor to enable at least one processor to execute the power grid active resilience scheduling method for ice disaster scenarios in the above embodiment 1.

[0184] The following reference Figure 6This document illustrates a structural schematic diagram of a power grid active resilience dispatching device suitable for implementing the embodiments of this application in an ice storm scenario. The power grid active resilience dispatching device in the ice storm scenario of this application embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 6 The power grid active resilience dispatching device shown in the ice storm scenario is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments in this application.

[0185] like Figure 6 As shown, the active resilience dispatching device for the power grid in an ice storm scenario may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in the read-only memory 1002 or the program loaded from the storage device 1003 into the random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the active resilience dispatching device for the power grid in an ice storm scenario. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, a touch screen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the active resilience dispatching equipment for ice storm scenarios to communicate wirelessly or wiredly with other equipment to exchange data. While the figure shows active resilience dispatching equipment for ice storm scenarios with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0186] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0187] The active resilience-enhancing dispatching device for ice storm scenarios provided in this application, employing the active resilience-enhancing dispatching method for ice storm scenarios described in the above embodiments, can solve the technical problem of how to improve the active resilience capability of power grid dispatching in long-duration ice storm scenarios. Compared with the prior art, the beneficial effects of the active resilience-enhancing dispatching device for ice storm scenarios provided in this application are the same as those of the active resilience-enhancing dispatching method for ice storm scenarios provided in the above embodiments, and other technical features in this active resilience-enhancing dispatching device for ice storm scenarios are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0188] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0189] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0190] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the active resilience scheduling method for power grids in the ice disaster scenario described in the above embodiments.

[0191] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0192] The aforementioned computer-readable storage medium may be included in the active resilience-enhancing dispatching equipment for power grids in ice disaster scenarios; or it may exist independently and not be assembled into the active resilience-enhancing dispatching equipment for power grids in ice disaster scenarios.

[0193] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the active resilience-enhancing dispatching device for power grids in ice storm scenarios, the device can write computer program code for performing the operations of this application using one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++; and also conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (e.g., via the Internet using an Internet service provider).

[0194] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0195] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0196] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described active resilience-enhancing dispatching method for power grids in ice disaster scenarios. This addresses the technical problem of how to improve the active resilience capability of power grid dispatching in long-duration ice disaster scenarios. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the active resilience-enhancing dispatching method for power grids in ice disaster scenarios provided in the above embodiments, and will not be elaborated upon here.

[0197] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described active resilience-enhancing scheduling method for power grids in ice storm scenarios.

[0198] The computer program product provided in this application can solve the technical problem of how to improve the active resilience capability of power grid dispatching under long-term ice storm evolution scenarios. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the active resilience dispatching method for power grids under ice storm scenarios provided in the above embodiments, and will not be repeated here.

[0199] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the scope of protection of this application.

Claims

1. A method for proactively resilient power grid scheduling in ice storm scenarios, characterized in that, The method includes: Acquire basic data on source, network, and load, as well as flexibility resource parameters, under ice storm scenarios; Based on the source-grid-load basic data, a dynamic value coefficient for energy storage is constructed, and day-ahead distributed bar pre-scheduling is performed based on the dynamic value coefficient for energy storage to obtain a robust safety envelope for energy storage status. Based on the robust safety envelope of the energy storage state and the real-time operating conditions, intraday adaptive trajectory tracking is performed to obtain real-time scheduling instructions; Based on the real-time scheduling instructions, the flexible resources are controlled to perform a tiered response, resulting in a robust scheduling outcome.

2. The method as described in claim 1, characterized in that, The steps of constructing a dynamic value coefficient for energy storage based on the source-grid-load basic data, and performing day-ahead distributed bar pre-scheduling based on the dynamic value coefficient for energy storage to obtain a robust safety envelope for energy storage status include: An initial scenario set is generated based on the aforementioned source-network-load basic data; Based on the distribution distance, the initial scene set is reduced to a typical scene set; Construct a probabilistic fuzzy set based on the set of typical scenarios to obtain fuzzy set data; The ratio of peak net load to average net load in future periods is determined based on the source-grid-load basic data, and the dynamic value coefficient of energy storage is constructed based on the scaling factor and the ratio. Based on the energy storage dynamic value coefficient, a robust pre-scheduling is performed within the fuzzy set data to obtain the energy storage state robust safety envelope.

3. The method as described in claim 2, characterized in that, The step of performing robust pre-scheduling based on the energy storage dynamic value coefficient within the fuzzy set data to obtain the robust safety envelope of the energy storage state includes: Within the fuzzy set data, the tiered call cost of each level of flexible resources is determined to obtain tiered cost data; Based on the tiered cost data and the energy storage dynamic value coefficient, a day-ahead pre-scheduling objective function is constructed to obtain the objective function data; Based on the objective function data, the minimum expected operating cost under the worst probability distribution is calculated to obtain the day-ahead scheduling plan, and the energy storage state robust safety envelope is generated based on the day-ahead scheduling plan.

4. The method as described in claim 1, characterized in that, The step of performing intraday adaptive trajectory tracking based on the robust safety envelope of the energy storage state and real-time operating conditions to obtain real-time scheduling instructions includes: Obtain the actual net load observation vector within a preset time window period in the past; Calculate scene similarity based on the actual net load observation vector and the historical scene database; The scene similarity is normalized by introducing a normalized exponential function of sensitivity coefficient, and the real-time weights of each typical scene are updated to obtain weight update data. Based on the weighted update data and the robust safety envelope of the energy storage state, a dynamic reference trajectory is synthesized to obtain trajectory data; The trajectory data is subjected to rolling optimization tracking to obtain real-time scheduling instructions.

5. The method as described in claim 4, characterized in that, The step of calculating scene similarity based on the actual net load observation vector and the historical scene database includes: The net load feature vectors of each typical scenario in the corresponding time period are extracted from the historical scenario database to obtain historical feature data; Calculate the geometric distance between the actual net load observation vector and the historical feature data; Based on the geometric distance, the morphological similarity between each typical scene and the real-time working condition is determined, and the scene similarity is determined.

6. The method as described in claim 4, characterized in that, The step of performing rolling optimization tracking on the trajectory data to obtain real-time scheduling instructions includes: The deviation between the real-time energy storage status and the trajectory data is obtained to obtain trajectory deviation data; Based on the trajectory deviation data, predict the state evolution in the future prediction time domain to obtain predicted state data; Based on the predicted state data and the intraday tiered cost, an intraday rolling optimization objective function is constructed to obtain the intraday objective function data; Based on the intraday objective function data, rolling optimization is performed in the prediction time domain to obtain real-time scheduling instructions.

7. The method as described in claim 1, characterized in that, The step of controlling the flexibility resources to perform a tiered response according to the real-time scheduling instructions to obtain the resilience scheduling result includes: The output adjustment amount of the basic support resources is determined according to the real-time scheduling instructions to obtain basic adjustment data; Based on the real-time dispatch instructions, the dispatch output of the virtual power plant and the power reduction of interruptible loads are determined to obtain demand-side adjustment data; The forced load shedding amount is determined based on the basic adjustment data and the demand-side adjustment data. Based on the basic adjustment data, the demand-side adjustment data, and the forced load shedding, a tiered response is performed to obtain the resilience scheduling result.

8. A power grid active resilience-enhancing dispatching device for ice storm scenarios, characterized in that, The device includes: The data acquisition module is used to acquire basic data on the source, network, and load, as well as flexibility resource parameters, under ice storm scenarios. The pre-scheduling module is used to construct the energy storage dynamic value coefficient based on the source-grid-load basic data, and perform day-ahead distributed bar pre-scheduling based on the energy storage dynamic value coefficient to obtain the energy storage state robust safety envelope. The tracking module is used to perform intraday adaptive trajectory tracking based on the robust safety envelope of the energy storage state and the real-time operating conditions to obtain real-time scheduling instructions; The response module is used to control the flexibility resources to perform a tiered response according to the real-time scheduling instructions, so as to obtain the resilience scheduling result.

9. A power grid active resilience-enhancing dispatching device for ice storm scenarios, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the active resilience-enhancing dispatch method for power grids in ice storm scenarios as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the active resilience-enhancing scheduling method for the power grid in the ice disaster scenario as described in any one of claims 1 to 7.

Images