Power distribution network collaborative optimization operation method and device, equipment and medium

Abstract

The application discloses a power distribution network cooperative optimization operation method and device, equipment and medium, and relates to the technical field of power systems; including: acquiring power distribution network topology parameters and parameters of multi-element flexible resources; constructing a data center space-time flexible load, mobile energy storage operation and electric vehicle charging station charging and discharging scheduling model; constructing a data center-mobile energy storage cooperative optimization model with the minimum system comprehensive operation cost as the target under the normal operation scene of the power distribution network and solving to obtain a cooperative scheduling strategy; constructing a data center-electric vehicle charging station elastic promotion model with the minimum system comprehensive loss as the target under the post-disaster recovery scene of the power distribution network under extreme disturbance and solving to obtain a load recovery strategy; verifying the normal and post-disaster scenes to output the normal cooperative scheduling and post-disaster elastic recovery scheme. The application solves the problem that the existing power distribution network optimization method lacks unified cooperative scheduling of multi-element flexible resources, resulting in high normal operation cost and low post-disaster key load recovery level.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to methods, devices, equipment and media for the coordinated optimization of distribution network operation. Background Technology

[0002] With the large-scale integration of distributed wind power, photovoltaics, and other renewable energy sources, and the rapid development of new loads such as data centers and electric vehicles, the operation mode of the distribution network is transforming from the traditional unidirectional power supply to a complex system involving coordinated interaction between power sources, grids, loads, and storage. Against this backdrop, the distribution network faces challenges such as increased net load fluctuations, rising peak-hour power purchases from the upstream grid, increased risk of node voltage exceeding limits, and increased network losses. Simultaneously, extreme weather events or line faults further highlight the distribution network's insufficient capacity to guarantee critical loads during post-disaster recovery.

[0003] Currently, some research has explored the optimization of distribution network operation. One type is the traditional distribution network optimization method, which typically establishes power balance, voltage, and line capacity constraints based on distributed generation output, conventional loads, and power interaction with the upstream grid, aiming to minimize electricity purchase costs or network losses. This type of method often treats load as rigid demand and energy storage as a fixed node resource. Another type involves scheduling methods for data centers, mobile energy storage, or electric vehicles. For example, it achieves energy consumption management through the start-up, shutdown, and task migration of data center servers, or utilizes the charging, discharging, and location transfer of mobile energy storage to support distribution network operation. However, these methods often only optimize for a single resource or local scenario, lacking a collaborative scheduling mechanism that integrates diverse and flexible resources (such as data centers, mobile energy storage, and electric vehicle charging stations) into the unified distribution network operation framework. In addition, existing disaster recovery methods mainly rely on network reconstruction and load restoration by distributed generation or emergency power sources, failing to fully combine the computing power task migration of data centers with the charging and discharging support capabilities of electric vehicle charging stations to achieve complementary synergy between the demand and supply sides.

[0004] In summary, existing technologies suffer from the following main problems: First, they fail to adequately utilize the spatiotemporal adjustability of diverse and flexible resources. Data centers are often simplified as rigid loads, and energy storage scheduling is limited to fixed nodes, making it difficult to leverage capabilities such as task relocation, cross-site migration, and spatial transfer. This results in high peak-hour external power consumption, significant network losses, and limited renewable energy absorption capacity. Second, there is a disconnect between routine operation optimization and post-disaster recovery modeling. There is a lack of support for the coordinated scheduling of diverse resources in different scenarios within a unified technical framework. Especially after extreme disturbances, it is difficult to prioritize the recovery of critical loads through the coordination of task migration / reduction and charge / discharge scheduling, thus reducing the resilience and recovery efficiency of the distribution network. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for the coordinated optimization of distribution network operation, which addresses the problem that existing distribution network optimization methods lack unified and coordinated scheduling of diverse flexible resources, resulting in high normal operating costs and low recovery levels of critical loads after disasters.

[0006] In view of this, the first aspect of the present invention provides a method for coordinated optimization of distribution network operation, the method comprising:

[0007] Acquire basic operational data, including distribution network topology parameters and parameters of various flexible resources, including data centers, mobile energy storage and electric vehicle charging stations;

[0008] Based on the power distribution network topology parameters and the parameters of the diverse flexible resources, a spatiotemporal flexible load model for data centers, a mobile energy storage operation model, and a charging and discharging scheduling model for electric vehicle charging stations are constructed.

[0009] In the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the operation model of the mobile energy storage, a collaborative optimization model of data center-mobile energy storage with the goal of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy.

[0010] In the scenario of post-disaster recovery from extreme disturbances in the power distribution network, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a data center-electric vehicle charging station elastic boosting model with the goal of minimizing the overall system loss is constructed and solved to obtain the load recovery strategy.

[0011] Modeling and verification are performed for both normal operation scenarios and post-disaster recovery scenarios, and the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy are output.

[0012] Optionally, the step of constructing a data center spatiotemporal flexible load model, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model based on the distribution network topology parameters and the parameters of the diverse flexible resources includes:

[0013] Based on the parameters of the data center among the parameters of the aforementioned diverse flexible resources, a spatiotemporal flexible load model for the data center is established using a three-layer coupling structure of workload scheduling, computing resource allocation, and power.

[0014] Based on the parameters of mobile energy storage in the distribution network topology parameters and the parameters of the diverse flexible resources, a mobile energy storage operation model is established based on the transfer time, dwell node, charging time and discharging time of mobile energy storage between distribution system nodes.

[0015] Based on the parameters of electric vehicle charging stations in the parameters of the aforementioned diverse flexible resources, and based on the costs incurred by electric vehicle charging stations in purchasing electricity from the grid and the benefits gained by electric vehicle charging stations in participating in load shedding during emergency situations, a charging and discharging scheduling model for electric vehicle charging stations is established.

[0016] Optionally, under the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, a data center-mobile energy storage collaborative optimization model with the objective of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy, including:

[0017] Based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, a first objective function is established under the normal operation scenario of the distribution network with the goal of minimizing the overall system operating cost.

[0018] The first set of constraints is constructed, including power balance constraints, power exchange constraints, voltage constraints at distribution network nodes, data center constraints, and mobile energy storage constraints.

[0019] The collaborative optimization model of data center-mobile energy storage, which consists of the first objective function and the first constraint, is solved to obtain the collaborative scheduling strategy.

[0020] Optionally, the expression for the first objective function is:

[0021] ;

[0022] In the formula, The cost of electricity purchased from the upstream power grid. Costs incurred due to line losses during power distribution network transmission. The charging, discharging, operation, and dispatch costs of mobile energy storage systems, The cost of electricity consumption required for data center operation.

[0023] Optionally, in the scenario of post-disaster recovery from extreme disturbances in the power distribution network, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a data center-electric vehicle charging station elastic boosting model with the objective of minimizing the overall system loss is constructed and solved to obtain a load recovery strategy, including:

[0024] Based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a second objective function is established under the scenario of post-disaster recovery of the distribution network under extreme disturbances, with the goal of minimizing the overall system loss.

[0025] Construct second constraints, including electric vehicle constraints, data center constraints, and power distribution network constraints;

[0026] The load recovery strategy is obtained by solving the data center-electric vehicle charging station elasticity improvement model composed of the second objective function and the second constraint.

[0027] Optionally, the expression for the second objective function is:

[0028] ;

[0029] In the formula, Let n be the load shedding penalty coefficient. For nodes At any moment Load demand; For nodes At any moment The power supply load; A function representing the operating cost of electric vehicle charging stations; This is a function for the operating cost of internet data centers.

[0030] Optionally, the step of modeling and validating normal operation and post-disaster recovery scenarios respectively, and outputting the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy, includes:

[0031] Based on a pre-built improved IEEE 69-node distribution network, a distribution network simulation model under normal operation scenarios and a distribution network fault recovery simulation model under extreme disturbance post-disaster recovery scenarios are established.

[0032] In the power distribution network simulation model under the normal operation scenario, the cooperative scheduling strategy is executed, and the normal cooperative scheduling scheme is obtained through simulation calculation;

[0033] In the power distribution network fault recovery simulation model under the post-disaster recovery scenario, the load recovery strategy is executed, and a post-disaster resilient recovery scheme is obtained through simulation calculation.

[0034] A second aspect of the present invention provides a distribution network collaborative optimization operation device, the device comprising:

[0035] The acquisition unit is used to acquire basic operational data, including distribution network topology parameters and parameters of various flexible resources, wherein the various flexible resources include data centers, mobile energy storage and electric vehicle charging stations;

[0036] The first construction unit is used to construct a data center spatiotemporal flexible load model, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model based on the distribution network topology parameters and the parameters of the diverse flexible resources.

[0037] The second construction unit is used to construct and solve a data center-mobile energy storage collaborative optimization model with the goal of minimizing the overall system operating cost, based on the data center spatiotemporal flexible load model and the mobile energy storage operation model, under the normal operation scenario of the distribution network, so as to obtain a collaborative scheduling strategy.

[0038] The third construction unit is used to construct and solve a data center-electric vehicle charging station elastic boosting model with the goal of minimizing the overall system loss, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, in the scenario of post-disaster recovery of the distribution network under extreme disturbances; and obtain the load recovery strategy.

[0039] The output unit is used to model and verify the normal operation scenario and the post-disaster recovery scenario respectively, and output the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy.

[0040] A third aspect of the present invention provides a distribution network collaborative optimization operation device, the device comprising a processor and a memory:

[0041] The memory is used to store program code and transmit the program code to the processor;

[0042] The processor is used to execute the steps of the distribution network collaborative optimization operation method as described in the first aspect above, according to the instructions in the program code.

[0043] A fourth aspect of the present invention provides a computer-readable storage medium for storing program code for executing the power distribution network collaborative optimization operation method described in the first aspect above.

[0044] As can be seen from the above technical solutions, the present invention has the following advantages:

[0045] This invention provides a method for coordinated optimization of power distribution network operation. First, it acquires the topology parameters of the power distribution network, as well as the parameters of data centers, mobile energy storage, and electric vehicle charging stations, laying a data foundation for unified coordinated scheduling. Then, it constructs a spatiotemporal flexible load model for the data center, an operation model for mobile energy storage, and a charging / discharging scheduling model for electric vehicle charging stations. This transforms the data center from a rigid load into a spatiotemporally flexible resource with task relocation and cross-station migration capabilities; transforms mobile energy storage from a fixed node resource into a flexible adjustment entity capable of optimized charging / discharging and location transfer; and transforms electric vehicle charging stations into flexible support nodes that can provide load reduction. This method integrates the data center, mobile energy storage, and electric vehicle... By incorporating diverse and flexible resources such as charging stations into the unified optimization framework of the distribution network, the spatiotemporal adjustability of the load side and the energy side support capabilities jointly participate in the optimization of distribution network operation. This solves the problem of insufficient utilization of the spatiotemporal adjustability characteristics of diverse resources and the inability to coordinate and schedule them in a unified manner in existing technologies. Then, under the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the operation model of mobile energy storage, a data center-mobile energy storage collaborative optimization model with the goal of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy. The spatiotemporal distribution of the load is reshaped by the task time shifting and cross-site migration of the data center, while the net load fluctuation is smoothed by the charging, discharging and location transfer of mobile energy storage. This approach aims to reduce network losses and, while meeting constraints such as power balance and node voltage, decrease peak-hour external power demand, improve load curve shape, reduce overall system operating costs and network losses, and enhance the controllability of the distribution network, thus solving the problem of high normal operating costs. Subsequently, in the scenario of post-disaster recovery from extreme disturbances in the distribution network, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a data center-electric vehicle charging station elastic boosting model with the goal of minimizing overall system losses is constructed and solved to obtain a load recovery strategy. This strategy releases power margin during critical periods by migrating and reducing computing tasks in the data center, and integrates this with the power supply of the electric vehicle charging station. The charging and discharging support capabilities create a complementary synergy between the demand and supply sides. Under constraints such as islanded power balance, fault isolation, and load importance levels, priority is given to restoring primary loads. This achieves dynamic matching of supply and demand and flexible and rational allocation of resources during the recovery period, reducing load shedding losses, improving the recovery level of total load and primary loads, enhancing the continuity and efficiency of the post-disaster recovery process, and solving the problem of low recovery levels of critical loads after disasters. Finally, modeling and verification are performed for both normal operation scenarios and post-disaster recovery scenarios, outputting normal collaborative scheduling schemes and post-disaster flexible recovery schemes. This simultaneously improves the economy of normal operation and the flexibility of post-disaster recovery within a unified technical framework. Therefore, this invention solves the problems of high normal operation costs and low recovery levels of critical loads after disasters caused by the lack of unified collaborative scheduling of diverse and flexible resources in existing technologies. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 A flowchart illustrating a method for collaborative optimization of power distribution network operation provided in an embodiment of the present invention;

[0048] Figure 2 This is a schematic diagram of a power distribution network collaborative optimization operation device provided in an embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0050] Terminology Explanation:

[0051]

[0052] Please see Figure 1 The present invention provides a method for coordinated optimization of distribution network operation, comprising:

[0053] Step 101: Obtain basic operational data, including distribution network topology parameters and parameters of various flexible resources, wherein the various flexible resources include data centers, mobile energy storage and electric vehicle charging stations;

[0054] It should be noted that the first step is to identify the main operating entities involved in the system and their interactions, including the distribution network, the upstream power grid, distributed wind turbines, distributed photovoltaic units, data centers, mobile energy storage, and electric vehicle charging stations. The distribution network topology parameters include network topology structure, node and line parameters, node basic load data, voltage operating limits, and upstream power grid interaction power constraints, which are used for subsequent power flow calculations and constraint modeling.

[0055] The parameters for diverse and flexible resources are as follows: parameters for data centers (task arrival volume, delay window, cross-site migration relationship, server activation scale and power parameters), used to characterize the spatiotemporal adjustability of computing load; parameters for mobile energy storage (initial resident node, initial state of charge, charging and discharging efficiency, maximum charging and discharging power and inter-node transfer time), used to achieve joint optimization of charging and discharging and location transfer; and parameters for electric vehicle charging stations (arrival time of electric vehicles in the station, departure time, target charging amount, maximum charging and discharging power and adjustable load boundary), used to support emergency power supply after disasters.

[0056] In addition to obtaining the aforementioned distribution network topology parameters and parameters for diverse flexible resources, the specific implementation also requires obtaining distributed wind power and photovoltaic output data and nodal loads to construct constraints, as detailed in the following embodiment. Distributed wind power and photovoltaic output data refers to the predicted values ​​of available active and reactive power output for each time period, embedded with power balance constraints to improve the absorption capacity of new energy sources. Nodal loads refer to the basic electricity demand for each time period, classified according to importance level (such as primary load, industrial load, etc.), and are used for power balance and load shedding loss calculations.

[0057] Furthermore, in specific implementation, the aforementioned basic data also needs to be discretized by time period, initialized and verified by parameter according to a unified scheduling cycle, to form an input dataset for unified modeling and collaborative optimization of diverse flexible resources, providing a data foundation for the subsequent construction of a data center spatiotemporal flexible load model, a mobile energy storage operation model and an electric vehicle charging station charging and discharging scheduling model.

[0058] Step 102: Based on the distribution network topology parameters and the parameters of the diverse flexible resources, construct a data center spatiotemporal flexible load model, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model.

[0059] It should be noted that this step is based on the basic operational data obtained in step 101, namely the parameters of the distribution network topology and the parameters of the multiple flexible resources, to construct a unified mathematical model of the distribution network's multiple flexible resources; the distribution network's multiple flexible resources include data centers, mobile energy storage and electric vehicle charging stations, and the unified mathematical model of the distribution network's multiple flexible resources covers the spatiotemporal flexible load model of data centers, the operation model of mobile energy storage and the charging and discharging scheduling model of electric vehicle charging stations.

[0060] It is understandable that the spatiotemporal flexible load characteristics of data centers include task migration characteristics in the time dimension and cross-site migration characteristics in the spatial dimension. Specifically, the task migration characteristic in the time dimension refers to the ability of some computing tasks to be moved forward or backward between different time periods, while meeting business deadlines, latency tolerance windows, and service quality constraints, thereby achieving load shaving and valley filling. The cross-site migration characteristic in the spatial dimension refers to the ability of computing tasks to migrate and redistribute between different data centers under the condition of multiple data centers operating collaboratively, enabling the data center load to be spatially reconfigured based on the power balance of distribution network nodes, line flow, and node voltage operating status.

[0061] Based on the above characteristics, the data center is modeled as a schedulable flexible load, and its flexible adjustment capability is embedded into the distribution network optimization model through task time relocation constraints, cross-site migration constraints, task flow conservation constraints, and service quality constraints. Based on the resource operation mechanism and scheduling characteristics, a spatiotemporal flexible load model for the data center, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model are constructed respectively, as detailed in the corresponding embodiments below.

[0062] Step 103: Under the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, construct a data center-mobile energy storage collaborative optimization model with the goal of minimizing the overall system operating cost and solve it to obtain the collaborative scheduling strategy;

[0063] It should be noted that in this step, when considering the normal operation scenario of the distribution network, a data center-mobile energy storage collaborative optimization model is constructed with the goal of minimizing the overall system operating cost by coupling the characteristics of the data center spatiotemporal flexible load model and the mobile energy storage operation model. In the solution process, the distribution network safety operation constraints and resource operation boundary constraints are combined to obtain the initial collaborative scheduling strategy under the normal scenario.

[0064] Specifically, in one example, the normal operation scenario of the distribution network is modeled using an improved IEEE 69-node distribution network system, which integrates distributed wind turbines, distributed photovoltaic (PV) generators, a data center, and mobile energy storage. Wind and PV generators represent the output of distributed renewable energy, the data center participates in distribution network scheduling as a flexible load with time-shifting and spatial migration capabilities, and mobile energy storage participates in node-side power support as a flexible resource with charging, discharging, and location-shifting capabilities. By integrating these resources into the structure, data center task scheduling, mobile energy storage energy scheduling, and distribution network power flow constraints can be coupled into the same optimization problem. This verifies the improvement effect of the data center-mobile energy storage collaborative optimization model on peak purchased power, system operating costs, network losses, and node voltage operating status under normal operation scenarios. The specific process of step 103 is detailed in the corresponding embodiment below.

[0065] It should be further noted that mobile energy storage in the normal operation scenario of the distribution network can be replaced or expanded into mobile energy storage vehicles, towable energy storage devices, and combinations of fixed and mobile energy storage, as long as the state of charge, power boundary and location / node support constraints are met.

[0066] Step 104: In the scenario of post-disaster recovery from extreme disturbances in the power distribution network, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, construct and solve the elastic boosting model of the data center-electric vehicle charging station with the goal of minimizing the overall system loss, and obtain the load recovery strategy.

[0067] It should be noted that in this step, when considering the post-disaster load recovery scenario under extreme disturbances in the distribution network (i.e., under the post-disaster recovery scenario of extreme disturbances in the distribution network), by coupling the characteristics of the data center spatiotemporal flexible load model and the electric vehicle charging station charging and discharging scheduling model, a data center-electric vehicle charging station elastic enhancement model with the goal of minimizing the overall system loss is constructed and solved, and the load recovery strategy of the distribution network is output.

[0068] For example, after an extreme disturbance or main grid failure occurs in a distribution network, damaged lines are disconnected, and the distribution network may be divided into several islands. Available power resources within these islands include distributed generation and emergency power sources. During post-disaster recovery, the distribution network needs to allocate power and restore loads under constraints such as power flow, voltage, line capacity, fault isolation, and repair schedule. On the grid side, network reconfiguration and distributed generation (DG) dispatch improve power accessibility and allocate limited power supply capacity. On the demand side, distributed generation (IDC) shifts power demand from islands with large power gaps or early periods to islands with more sufficient power margins or later periods through task migration / reduction, releasing power margins in critical windows. On the supply side, distributed generation systems (EVCS) reallocate power in time and space through charging, discharging, and cross-regional movement, thereby providing compensatory support for critical loads within the islands. The core of their synergy lies in the fact that the IDC provides demand-side flexibility through "peak shaving and congestion relief," making it easier for EVCS and DG support to reach the target load under network constraints. Meanwhile, the EVCS provides supply-side capabilities for "critical period energy replenishment and cross-domain support," ensuring more stable power supply for IDC task transfers. Through this complementary synergy between supply-side support and demand-side peak shaving, the model can improve load recovery levels and reduce load shedding losses while meeting operational constraints, achieving an overall improvement in system resilience. The specific process of this step is illustrated in the corresponding implementation example below.

[0069] It should be further noted that, in the scenario of post-disaster recovery of the distribution network under extreme disturbances, electric vehicle charging stations can be replaced or expanded into charging and battery swapping stations with aggregated charging and discharging capabilities, vehicle-to-grid interaction aggregates, or other adjustable electric transportation resources; the model should still include access time period, target power, charging and discharging power, and energy state constraints.

[0070] Step 105: Model and verify the normal operation scenario and the post-disaster recovery scenario respectively, and output the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy.

[0071] It should be noted that the "cooperative scheduling strategy" and "load recovery strategy" obtained through the optimization model are respectively placed in two pre-built distribution network simulation models for simulation operation verification to test their feasibility and effectiveness in different scenarios, and finally output specific solutions that can be used for actual scheduling. For example, for the normal operation scenario, a simulation model including wind power, photovoltaic, data center and mobile energy storage can be built based on the improved IEEE 69-node system. The data such as the task migration path of the data center in each time period, the start and stop status of the server, and the stationing node and charging and discharging power of the mobile energy storage given in the cooperative scheduling strategy are input into the model. The time-series power flow simulation calculation is used to verify whether the constraints such as node voltage not exceeding the limit and line not being overloaded are met. At the same time, the power purchase cost and network loss are calculated. If the verification is successful, a day-ahead scheduling plan table containing the operation instructions of each resource every 15 minutes is output as the normal cooperative scheduling scheme. For post-disaster recovery scenarios, a fault location (e.g., line 7-8 disconnected) is set up in the same test system and islanded. Data such as the load shedding ratio of each node, the data center task reduction, and the charging and discharging power of electric vehicle charging stations, given in the load recovery strategy, are input into the fault recovery simulation model. Dynamic simulation is used to verify whether constraints such as power balance within the island and priority power supply for primary loads are met. If the verification is successful, a post-disaster elastic recovery plan including recovery timing, load shedding allocation table, and charging station charging and discharging plan is output. In this way, through simulation verification and plan output of two specific scenarios, it is ensured that the proposed strategy is not only theoretically optimal but also realistically feasible in the simulation environment. See the corresponding implementation examples below for details.

[0072] In one embodiment, step 102 includes:

[0073] Step 1021: Based on the parameters of the data center in the parameters of the diverse flexible resources, a spatiotemporal flexible load model of the data center is established by adopting a three-layer coupling structure of workload scheduling, computing resource allocation and power.

[0074] Specifically, data centers are modeled as a type of schedulable flexible load, and a three-layer coupled model of workload scheduling, computing resource allocation, and power optimization is established. At the workload scheduling level, allocation and reception variables of workloads between front-end servers and data centers are introduced to characterize cross-site task inflow / outflow, and the actual business scale completed in the current period is expressed by the relationship between processing volume and discard / transfer volume, ensuring that all business flows satisfy non-negativity and conservation logic.

[0075] At the workload scheduling level, the specific modeling process is as follows: Let... For Internet mobile data centers (i.e., data centers, IDCs) at nodes ,time Below Total workload of the type; For IDC at the node ,time The following is allocated to the front-end server. Type is The workload; This refers to the set of the number of front-end servers; For time Remove IDC from node Transferred to target node d A type of work task represents a work task transferred from one node to another; D is the set of network nodes; For IDC at the node ,time Below Type of workload transfer; For IDC at the node ,time The processed For a given workload, the specific modeling of task interactions between data centers (and thus the modeling at the workload scheduling level) is as follows:

[0076] (2)

[0077] Understandably, workload scheduling determines the types and quantities of computing tasks processed by each data center at different times, as well as the temporal and spatial distribution of these tasks. This includes local processing of tasks, cross-data center migration, and time-shifting within latency tolerance windows. Computing resource allocation and power consumption constitute the physical mapping of the actual power load of the data center. Computing resource allocation determines the number, frequency, and computing power allocation of activated servers based on workload scheduling results, thus determining the power of the data center's IT equipment. Power consumption is calculated by adding the power consumption of infrastructure such as cooling and lighting to the IT equipment power, forming the total power drawn by the data center from the power distribution network. The core logic of this three-layer coupling is as follows: workload scheduling, as the upper decision-making layer, changes the computing resource requirements of each data center at different times through time-shifting and cross-site migration of tasks; computing resource allocation, as the middle layer, translates scheduling decisions into server start / stop status and processing capacity; and power consumption, as the execution layer, quantifies resource allocation into actual power consumption and incorporates it into the power balance of the power distribution network. Through this layered coupling structure, the flexible adjustment capability of the data center can be described by a unified mathematical relationship between power, tasks, and resources, thereby embedding the spatiotemporal adjustable characteristics of computing load into the power distribution network collaborative optimization model.

[0078] Step 1022: Based on the parameters of mobile energy storage in the distribution network topology parameters and the parameters of the multiple flexible resources, establish a mobile energy storage operation model based on the transfer time, dwell node, charging time and discharging time of mobile energy storage between distribution system nodes;

[0079] Specifically, the charging and discharging capacity of mobile energy storage is not only limited by its maximum power output and battery capacity, but also needs to consider the demand of the power grid. When discharging, mobile energy storage (MES) should prioritize serving nodes with lower voltage or higher loads. By dispatching energy storage to where it is most needed, peak loads on the power grid can be reduced and voltage regulation achieved. The specific modeling process for the mobile energy storage operation model is as follows: Let... For MES in time From node Transfer to target node Time; For MES at the node The initial time; For MES at the node Charging time; For MES slave nodes Transfer to target node The movement of time; For MES at the node The discharge time; For MES at the node The end time. The duration of the MES movement between two specified nodes within the power distribution system is modeled as follows (i.e., the mobile energy storage operation model):

[0080] (3)

[0081] It should be noted that establishing a mobile energy storage operation model based on the transfer time, dwell node, charging time, and discharging time between nodes in the power distribution system essentially couples the location and energy state of the mobile energy storage into a unified spatiotemporal decision variable. The transfer time determines the time step required for the mobile energy storage to reach the target node from the current node, directly affecting its dispatchability in different time periods. The dwell node identifies the location of the mobile energy storage in a specific time period, determining the node where its charging and discharging power is injected into the power distribution network. The charging and discharging times respectively characterize the time intervals during which the mobile energy storage exchanges energy with the grid, and together with the state of charge evolution, constitute the energy feasible region. These elements are not independent but are interconnected through time continuity constraints and location uniqueness constraints: mobile energy storage can only be located at one node at a time, and charging or discharging operations can only be performed while residing at a certain node; neither charging nor discharging is allowed during the transfer process. Based on this logic, the modeling formula for the mobility duration given in step 1022 describes the temporal relationship between the mobility time required for the MES to move from node i to target node j and the initial time, charging time, discharging time, and end time. This transforms the transfer time, dwell node, charging time, and discharging time into analytically expressible temporal constraints, thereby embedding the temporal scheduling capability of mobile energy storage into the distribution network collaborative optimization model. It is understandable that this formula is crucial to the mobile energy storage operation model, ensuring that the spatial transfer and temporal scheduling of mobile energy storage remain consistent in the optimization solution, enabling subsequent power balance constraints, state of charge constraints, and charging / discharging power constraints to be modeled based on the correct temporal position.

[0082] Step 1023: Based on the parameters of electric vehicle charging stations in the parameters of the diversified flexible resources, and based on the costs incurred by electric vehicle charging stations in purchasing electricity from the grid and the benefits obtained by electric vehicle charging stations in participating in load reduction during emergency situations, establish a charging and discharging scheduling model for electric vehicle charging stations.

[0083] Specifically, the scheduling of electric vehicle charging stations needs to consider the charging and discharging behavior of electric vehicles. The cost function of an electric vehicle charging station consists of two main parts: the cost incurred by the charging station in purchasing electricity from the grid, and the revenue gained by the charging station from participating in load shedding during emergency situations.

[0084] The specific process of constructing the charging and discharging scheduling model for electric vehicle charging stations is as follows: Let... To indicate at time The economic cost coefficient is used to measure the cost of load dispatch and reflects the electricity price in the electricity market. To indicate charging station The reward coefficient is used to reward the load reduction or flexibility provided by the charging station, reflecting the compensation for the charging station's participation in load management; The time interval. Electric vehicle cost function. (That is, the electric vehicle charging station charging and discharging scheduling model) can be expressed as:

[0085] (4)

[0086] In the formula, Let be the rated charging power of charging station c at time t; Let be the actual charging power of charging station c at time t.

[0087] Understandably, based on the parameters of electric vehicle charging stations (including the arrival and departure times of electric vehicles within the station, target charging volume, maximum charging and discharging power, and adjustable load boundaries), and considering the costs incurred by charging stations in purchasing electricity from the grid and the benefits gained from participating in load shedding during emergencies, a charging and discharging scheduling model for electric vehicle charging stations is established. The key is to treat the charging station as a flexible adjustment node that can both absorb power from the grid (charging) and reduce power absorption or even reverse discharge (load shedding) during emergencies. Electricity purchase costs drive charging stations to increase electricity purchases during off-peak hours and reduce purchases during peak hours, reflecting economic objectives. Meanwhile, load shedding benefits incentivize charging stations to proactively reduce charging power or even discharge to the grid during scenarios of power shortages in the distribution network or post-disaster recovery, in order to support critical loads. These two parts are quantified through the economic cost coefficient (reflecting electricity market prices) and the incentive coefficient (reflecting compensation for charging stations' participation in load management) in the cost function, forming the electric vehicle cost function expression given in step 1023. This expression couples the economic operation of charging stations with the grid dispatch requirements: during normal operation, the optimization model tends to reduce electricity purchase costs; during post-disaster recovery, adjusting the reward coefficient can guide charging stations to prioritize providing load reduction services.

[0088] In one embodiment, step 103 includes:

[0089] Step 1031: Based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, establish a first objective function under the normal operation scenario of the distribution network with the goal of minimizing the overall system operating cost;

[0090] Specifically, a first objective function is established under the normal operation scenario of the distribution network, with the goal of minimizing the overall system operating cost. The overall system operating cost of this invention mainly includes the cost of electricity purchased from the upstream power grid. Costs incurred due to line losses during power distribution network transmission The charging and discharging operation, operation and dispatch costs of mobile energy storage systems Electricity consumption costs required for data center operation The first objective function is expressed as:

[0091] (5)

[0092] Understandably, based on the aforementioned spatiotemporal flexible load model for data centers and the aforementioned mobile energy storage operation model, a first objective function is established under the normal operation scenario of the distribution network, with the goal of minimizing the overall system operating cost. Essentially, this unifies and quantitatively couples the economic operation objective of the distribution network with the scheduling decisions of diverse flexible resources. Specifically, the four costs in the first objective function correspond to the economic consumption of different stages in the operation of the distribution network: the electricity purchase cost from the upstream grid reflects the direct cost of the distribution network obtaining power from the main grid, significantly affected by peak-hour power purchases; line loss cost reflects energy loss during power transmission, closely related to power flow distribution and load curve shape; the charging and discharging operation, operation, and scheduling costs of the mobile energy storage system cover energy storage charging and discharging efficiency losses, battery depreciation, and additional costs incurred due to location relocation; and the power consumption cost of data center operation is directly related to workload scheduling and server start / stop status. Minimizing the sum of the four costs mentioned above is the objective of the optimization model. This allows the model to reshape the spatiotemporal distribution of load through task time shifting and cross-site migration within the data center, and to smooth net load fluctuations and reduce network losses through the charging, discharging, and relocation of mobile energy storage. This synergistically reduces peak-hour electricity demand and network losses, ultimately minimizing the overall system operating cost. Specifically, the electricity consumption cost required for data center operation is directly determined by the three-layer coupling relationship between workload scheduling, computing resource allocation, and power in the data center's spatiotemporal flexible load model, reflecting the actual electricity expenditure incurred by the data center due to computing tasks. The charging, discharging, operation, and scheduling costs of the mobile energy storage system originate from the state-of-charge evolution, charging / discharging power boundaries, and node location transfer constraints in the mobile energy storage operation model. These costs encompass energy losses during charging and discharging, battery aging, and additional expenses incurred during spatial relocation.

[0093] Therefore, the first objective function expression given in step 1031 is the mathematical embodiment of the aforementioned modeling idea, namely, "to minimize the overall operating cost and couple two flexible resources, data center and mobile energy storage", which provides guidance for solving the data center-mobile energy storage collaborative scheduling strategy under normal scenarios.

[0094] Step 1032: Construct the first set of constraints, including power balance constraints, power exchange constraints, voltage constraints at distribution network nodes, data center constraints, and mobile energy storage constraints.

[0095] Specifically, by mapping the operation strategies of IDC (data center) and MES (mobile energy storage) to the equivalent power injection / consumption of corresponding nodes, and embedding power balance, power exchange limit and voltage constraints of distribution network, power balance constraints, power exchange constraints, voltage constraints of distribution network nodes, data center constraints and mobile energy storage constraints are established to obtain the first constraint condition.

[0096] in:

[0097] a) Distribution network power balance constraints: To ensure the conservation of active power in the distribution network at all times, wind power and photovoltaic output, and energy storage discharge at nodes constitute active power injection, while energy storage charging, data center power consumption, and conventional loads constitute active power consumption. The net injection should equal the sum of the active power flow from that node to adjacent nodes. Among these, For nodes time Active power of wind turbine generators; For nodes time Active power of photovoltaic units; and For nodes time The active and reactive power of the load; For time line The active power flow. The power balance constraint of the distribution network can be expressed as:

[0098] (6)

[0099] In the formula, The charging power of the mobile energy storage system (MES) to the distribution network node b where it resides during time period t; The discharge power of the mobile energy storage system (MES) to the distribution network node b where it resides during time period t; The active power consumed by the Internet data center at time t to the distribution network node b where it resides.

[0100] b) Distribution network power exchange constraints: A maximum allowable value is set for the active power exchanged from the upstream power grid to the distribution network to ensure that the optimization results do not draw power from the upstream power grid without restriction at any time. Among these, and They are time Active and reactive power provided by the upstream power grid; and They are time The maximum allowable values ​​of active and reactive power provided by the upstream power grid, and the power exchange constraints of the distribution network are expressed as follows:

[0101] (7)

[0102] c) Distribution network node voltage constraints: To ensure the voltage quality and equipment safety of the distribution network, upper and lower operating limits are set for the voltage amplitude of each node, requiring that the voltage of each node must be within the allowable range at any given time. For example, For nodes time The voltage amplitude, and For nodes The lower limit of voltage; For nodes The upper limit of voltage, specifically constrained as follows:

[0103] (8)

[0104] d) Data Center Constraints: At the computing resource level, this paper further characterizes the available computing power of a data center using the number of active servers and gives an upper limit to processing capacity, i.e., the processing volume does not exceed the computing resource capacity determined by the server scale. time Front-end server allocation Type of workload; This represents the maximum number of servers that are currently in use. This represents the server's processing capacity coefficient. For latency-sensitive workloads, the acceptable latency period is; This is the initial time point; This is a time threshold; For the node ,time The capacity transfer ratio coefficient. The relevant constraints on IDC as a flexible load can be expressed as:

[0105] (9)

[0106] In the formula, Allocate the completed workload of type y to node b, time period t, and front-end server.

[0107] (10)

[0108] In the formula, This refers to the workload at node b and time period t, where the front-end server x receives y-type data and processes it locally. Let y be the workload of type y that migrates from node b to target node d during time period t; This represents the number of servers that are active in the data center at node b and time period t. For node b and time period t, the task time shift amount is y-shaped.

[0109] e) Mobile Energy Storage Constraints: MES can be deployed on any node in the power distribution system, but MES cannot be located in two or more locations simultaneously. The binary decision variable represents the MES in time. Is it at the node? Its constraints are expressed as:

[0110] (11)

[0111] Based on this, we assume For MES at the node time Maximum charging power; For MES at the node time Given the maximum discharge power, the power constraint for MES charging or discharging is:

[0112] (11)

[0113] Step 1033: Solve the data center-mobile energy storage collaborative optimization model composed of the first objective function and the first constraint to obtain the collaborative scheduling strategy.

[0114] It should be noted that the data center-mobile energy storage collaborative optimization model, consisting of the first objective function and the first constraint, is a mixed-integer nonlinear programming problem. It is usually solved by calling commercial solvers such as Gurobi and CPLEX after linearization. Under the premise of satisfying the first constraint, the model minimizes the overall system operating cost. The collaborative scheduling strategy obtained by the solution can generate the day-ahead scheduling plan under the normal operation of the distribution network, clarifying the task processing volume, cross-site migration path and server start-up and shutdown status of the data center in each time period, as well as the stationing nodes, charging and discharging power and state of charge trajectory of the mobile energy storage. Thus, by relocating tasks and scheduling energy storage, peak-hour electricity purchase costs are reduced, network losses are reduced, voltage quality is improved and the capacity for renewable energy consumption is enhanced.

[0115] In one embodiment, step 104 includes:

[0116] Step 1041: Based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, establish a second objective function for the post-disaster recovery scenario of extreme disturbances in the power distribution network with the goal of minimizing the overall system loss;

[0117] Specifically, this embodiment takes minimizing the overall system loss, i.e., minimizing the sum of the distribution network post-disaster load shedding loss cost, data center task control and operation cost, and electric vehicle charging station power scheduling cost, as its core objective. A second objective function is established for the distribution network post-disaster recovery scenario under extreme disturbances, expressed as:

[0118] (12)

[0119] In the formula, Let n be the load shedding penalty coefficient. For nodes At any moment Load demand; For nodes At any moment The power supply load; A function representing the operating cost of electric vehicle charging stations; This is a function for the operating cost of internet data centers.

[0120] Understandably, based on the aforementioned spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a second objective function is established under the scenario of post-disaster recovery from extreme disturbances in the distribution network, with the goal of minimizing the overall system loss. Its core lies in incorporating three key costs in the post-disaster recovery process—load shedding loss, data center task control cost, and electric vehicle charging station power scheduling cost—into the optimization objective, thereby achieving the optimal match between recovery resources and load demand while satisfying the second constraint condition in step 1042. Specifically, the load shedding loss term... Directly reflects the economic losses of unpowered loads, driving the model to prioritize critical users such as primary loads; data center task control and operation costs. The task migration / reduction decisions derived from the spatiotemporal flexible load model of data centers reflect the cost of releasing power margin through the spatiotemporal relocation of computing load; power scheduling costs of electric vehicle charging stations. The cost of electricity purchase and the revenue from load reduction in the electric vehicle charging station's charging and discharging scheduling model represent the economic behavior of charging stations participating in emergency support. Minimizing the sum of these three factors is used as the second objective function. This allows the optimization model to coordinate the decision-making of data center task migration / reduction strategies and electric vehicle charging station charging and discharging plans under limited power supply capacity after a disaster. It proactively shifts power demand from power-scarce islands or time periods to power-sufficient islands or time periods, thereby reducing total load shedding losses and improving critical load recovery levels. Therefore, the expression for the second objective function given in step 1041 is the mathematical embodiment of the aforementioned modeling idea of ​​"minimizing overall system loss while coupling two flexible resources: data centers and electric vehicle charging stations," providing a clear optimization guide for solving load recovery strategies in post-disaster recovery scenarios.

[0121] Step 1042: Construct the second set of constraints, including electric vehicle constraints, data center constraints, and power distribution network constraints;

[0122] The constraints are as follows;

[0123] 1. Electric vehicle constraints: These include electric vehicle charging task and energy state constraints, electric vehicle capacity constraints, and electric vehicle charging power constraints.

[0124] a) Electric vehicle charging task and state of energy constraints: The target state of charge for each electric vehicle when leaving a charging station must be greater than or equal to its target charge amount. For electric vehicles The target charging volume is for electric vehicles. The expected charging energy at the end of the entire charging cycle; For electric vehicles The amount of charge at the charging station upon arrival; For electric vehicles At any moment The charging power; For charging efficiency; For electric vehicles The moment of departure; For electric vehicles The arrival time. Specifically expressed as:

[0125] (13)

[0126] b) Electric vehicle charging tasks and state of energy constraints: Electric vehicle capacity constraints. Constraints are placed on the capacity of the electric vehicle battery to ensure that the charging amount is always within the battery's capacity range. For electric vehicles At any moment The charging energy; , Electric vehicles At any moment The minimum and maximum charging energy are specifically expressed as:

[0127] (14)

[0128] c) Electric vehicle charging power constraints: Data center power consumption The maximum power consumption of the computing server must not be exceeded. Specifically, it can be expressed as:

[0129] (15)

[0130] 2. Data center constraints;

[0131] To ensure that the power consumption of data center computing services is controlled within maximum computing capacity, the task... In data center Minimum computed power and maximum calculated power The following formula should be satisfied:

[0132] (16)

[0133] in, Let d be the IT computing power of task d in data center i and time t.

[0134] Meanwhile, task d is in the data center time Computing server power consumption It should not exceed the task In data center Maximum computing server power consumption , expressed as:

[0135] (17)

[0136] For an IDC to operate smoothly and maintain service continuity under dynamic load changes, it needs sufficient capacity to handle sudden surges in workload and computing capacity demands. Assuming... For the task In data center time The enabled status, For data centers Computing capacity, For the task In data center The power consumption coefficient, For the task In data center time Task utilization should meet the following requirements:

[0137] (18)

[0138] in, This represents the activation status of task d at time t in data center i; The power consumption of task d in the largest IT server in data center i; For the task In data center Time period The actual power consumption of IT servers; For the task In data center Time period Status of IT equipment.

[0139] Construct constraints for the data center cooling system, assuming... , Data Center time Minimum and maximum power of the refrigeration system It is a constant; This is a binary variable representing a data center. At any moment Is the refrigeration system working? , Let the minimum and maximum internal temperatures be respectively. This constraint can be expressed as follows:

[0140] (19)

[0141] in, Let be the internal temperature of data center i at time t.

[0142] 3. Distribution network constraints: The physical operation constraints of the distribution network are simultaneously incorporated into the model solution framework from seven aspects: power balance, line power, power supply ratio, node voltage, maximum power supply constraint, line power flow direction constraint, and fault isolation constraint.

[0143] a) Power balance constraint, expressed as:

[0144] (20)

[0145] in, Let be the rated active power of data center i at time t. Let be the actual active power of data center i at time t.

[0146] (twenty one)

[0147] in, , They are nodes At the moment Total active power supply and total reactive power supply; , Wind turbine generator sets At any moment Active power and reactive power; , Photovoltaic generator sets At any moment Active power and reactive power; This is a binary variable representing whether line l is at time [time]. Enabled (1 indicates enabled, 0 indicates disabled); , These are the times for line l at time l. The active power flow and reactive power flow.

[0148] b) Line power constraint, expressed as:

[0149] (twenty two)

[0150] (twenty three)

[0151] (twenty four)

[0152] in, For nodes To the node The power between; For nodes To the node The power between; The maximum apparent power of line l; This is a binary variable representing line l at time [time]. The status of line l is 1, which indicates that line l is enabled, and 0 indicates that line l is disabled. It is a constant; For nodes At any moment The voltage amplitude; For nodes At any moment The voltage amplitude; Let be the resistance of line l; The reactance of line l; , These are the times for line l at time l. The active power and reactive power.

[0153] c) Power supply ratio constraint, the expression is:

[0154] (25)

[0155] in: is the load dispatching ratio coefficient, representing the proportion of power supplied to node n at time t.

[0156] d) Node voltage constraints, expressed as:

[0157] (26)

[0158] in, , They are nodes The minimum and maximum voltage amplitudes.

[0159] e) Maximum power supply constraint, expressed as:

[0160] (27)

[0161] (28)

[0162] in, , These are the maximum active power and reactive power supplied, respectively.

[0163] f) Line power flow direction constraint, expressed as:

[0164] (29)

[0165] (30)

[0166] g) Fault isolation constraints, expressed as:

[0167] (31)

[0168] (32)

[0169] in, is the load dispatch ratio coefficient, representing the proportion of power supplied to node m at time t.

[0170] Step 1043: Solve the data center-electric vehicle charging station elasticity improvement model composed of the second objective function and the second constraint to obtain the load recovery strategy.

[0171] It should be noted that the data center-electric vehicle charging station resilience improvement model, consisting of the second objective function and the second constraint, is a mixed-integer nonlinear programming problem. It is usually solved by linearization and then using commercial solvers such as Gurobi and CPLEX. Under the premise of satisfying the second constraint, such as island power balance, line capacity, node voltage, fault isolation, electric vehicle charging and discharging timing, data center task migration / reduction, and load power supply ratio, the model minimizes the overall system loss (i.e., the sum of load shedding loss, data center task control cost, and electric vehicle charging station power scheduling cost). The load recovery strategy obtained from the solution can generate a dynamic recovery plan for the post-disaster recovery scenario of the distribution network under extreme disturbances. It clarifies the load shedding ratio of each node in each time period, the task migration path and task reduction amount of the data center, and the charging and discharging power and state of charge trajectory of the electric vehicle charging station. In this way, by coordinating the spatiotemporal relocation of data center computing load and the energy support of electric vehicle charging station, the power supply of key users such as primary loads is prioritized, the overall load recovery rate and primary load recovery rate are improved, the economic loss of load shedding is reduced, and the post-disaster resilience and recovery efficiency of the distribution network are enhanced.

[0172] In one embodiment, step 105 includes:

[0173] Based on a pre-built improved IEEE 69-node distribution network, a distribution network simulation model under normal operation scenarios and a distribution network fault recovery simulation model under extreme disturbance post-disaster recovery scenarios are established. In the distribution network simulation model under normal operation scenarios, the cooperative scheduling strategy is executed, and a normal cooperative scheduling scheme is obtained through simulation calculation. In the distribution network fault recovery simulation model under disaster recovery scenarios, the load recovery strategy is executed, and a post-disaster elastic recovery scheme is obtained through simulation calculation.

[0174] Specifically, based on a pre-built improved IEEE 69-node distribution network (i.e., distribution network test system), simulation models of the distribution network under normal operation scenarios and distribution network fault recovery scenarios under extreme disturbance and post-disaster recovery scenarios are established. The distribution network test system is constructed as follows: on the basis of a standard IEEE 69-node distribution network, two wind farms, two photovoltaic power plants, four data centers, and four electric vehicle charging stations (EVCS) are connected, and the loads are divided into primary loads, industrial loads, commercial loads, and residential loads according to their importance. At the same time, multiple flexible resources such as the upstream grid and mobile energy storage are set up. For normal operation scenarios, the data center-mobile energy storage collaborative scheduling strategy obtained in step 1033 (including the data center's task processing volume, cross-site migration path, server start / stop status, and the mobile energy storage's stationary nodes, charging / discharging power, and state of charge trajectory) is used as the input to the simulation model. After executing the collaborative scheduling strategy, the distribution network time-series power flow simulation calculation is performed to evaluate the feasibility of the scheduling strategy under constraints such as power balance, node voltage, and line capacity. Peak-hour electricity purchase cost, network loss, and renewable energy absorption rate are also statistically analyzed. Finally, the normal collaborative scheduling scheme (i.e., the day-ahead scheduling plan and resource operation curve) is output.

[0175] For post-disaster recovery scenarios under extreme disturbances, the main grid, distributed wind turbines, distributed photovoltaic units, data centers, electric vehicle charging stations, critical loads, and fault locations are marked in the improved IEEE 69-node distribution network simulation model, and fault isolation and islanding logic are set up. The data center-electric vehicle charging station elastic boosting strategy obtained in step 1043 (including the load shedding ratio of each node in each time period, the task migration path and task reduction of the data center, and the charging and discharging power and state of charge trajectory of the electric vehicle charging station) is used as the input of the post-disaster simulation model. After executing the load recovery strategy, fault recovery timing simulation calculation is performed. Under the premise of satisfying constraints such as islanding power balance, line capacity, node voltage, fault isolation, electric vehicle charging and discharging timing, data center task migration / reduction, and load power supply ratio, the load recovery process in each time period is dynamically simulated, prioritizing the power supply of primary loads, and statistically analyzing indicators such as total load recovery rate, primary load recovery rate, and economic loss of load shedding. Finally, a post-disaster elastic recovery plan (including load recovery timing plan, load shedding allocation table, data center task adjustment instructions, and charging station charging and discharging plan) is output.

[0176] Through the above simulation verification, the improvement effect of the proposed collaborative optimization operation method on the system economy under normal operation can be quantitatively evaluated, as well as its role in enhancing the critical load guarantee capability and system resilience in post-disaster recovery scenarios. It should be noted that the improved IEEE 69-node distribution network and resource configuration used in this example are only for illustrating the modeling and verification methods of this invention and do not constitute a limitation on the application scenarios and node configurations of this invention.

[0177] It should be further noted that the distribution network testing system is not limited to the improved IEEE 69-node distribution network. Those skilled in the art can also replace it with an actual park distribution network, urban distribution network or other standard node system; when replacing, the corresponding topology, lines, node load and resource access parameters must be provided.

[0178] In summary, the present invention provides a method for coordinated optimization of distribution network operation. First, it acquires the topology parameters of the distribution network, as well as the parameters of data centers, mobile energy storage, and electric vehicle charging stations, laying a data foundation for unified coordinated scheduling. Then, it constructs a spatiotemporal flexible load model for the data center, an operation model for mobile energy storage, and a charging / discharging scheduling model for electric vehicle charging stations. This transforms the data center from a rigid load into a spatiotemporally flexible resource with task relocation and cross-station migration capabilities; transforms mobile energy storage from a fixed node resource into a flexible adjustment entity capable of optimized charging / discharging and location transfer; and transforms electric vehicle charging stations into flexible support nodes that can provide load reduction. This method integrates the data center, mobile energy storage, and electric vehicle charging stations. By incorporating diverse and flexible resources such as electric vehicle charging stations into the unified optimization framework of the distribution network, the spatiotemporal adjustability of the load side and the energy side support capabilities jointly participate in the optimization of distribution network operation. This solves the problem of insufficient utilization of the spatiotemporal adjustability characteristics of diverse resources and the inability to coordinate and schedule them in a unified manner in existing technologies. Next, under the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the operation model of mobile energy storage, a data center-mobile energy storage collaborative optimization model with the objective of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy. This strategy reshapes the spatiotemporal distribution of load through task time shifting and cross-site migration of the data center, while smoothing the net load through the charging, discharging, and location transfer of mobile energy storage. By mitigating fluctuations and reducing network losses, and thereby reducing peak-hour external power demand, improving load curve shape, lowering overall system operating costs and network losses, and enhancing the controllability of distribution network operation while meeting constraints such as power balance and node voltage, the problem of high normal operating costs is solved. Subsequently, in the scenario of distribution network post-disaster recovery from extreme disturbances, based on the spatiotemporal flexible load model of data centers and the charging and discharging scheduling model of electric vehicle charging stations, a data center-electric vehicle charging station elastic boosting model with the goal of minimizing overall system losses is constructed and solved to obtain a load recovery strategy. This strategy releases power margin during critical periods by migrating and reducing computing tasks in the data center, and integrates with electric vehicle charging stations. The charging and discharging support capabilities create a complementary synergy between the demand and supply sides. Under constraints such as islanded power balance, fault isolation, and load importance levels, priority is given to ensuring the restoration of primary loads. This achieves dynamic matching of supply and demand and flexible and rational allocation of resources during the recovery period, reducing load shedding losses, improving the recovery level of total load and primary loads, enhancing the continuity and efficiency of the post-disaster recovery process, and solving the problem of low recovery levels of critical loads after disasters. Finally, modeling and verification are performed for both normal operation scenarios and post-disaster recovery scenarios, outputting normal collaborative scheduling schemes and post-disaster flexible recovery schemes. This simultaneously improves the economy of normal operation and the flexibility of post-disaster recovery within a unified technical framework. Therefore, this invention solves the problems of high normal operation costs and low recovery levels of critical loads after disasters caused by the lack of unified collaborative scheduling of diverse and flexible resources in existing technologies.

[0179] The above is a method for coordinated optimization of power distribution network operation provided in the embodiments of the present invention. The following is a device for coordinated optimization of power distribution network operation provided in the embodiments of the present invention.

[0180] Please see Figure 2 The present invention provides a distribution network collaborative optimization operation device, comprising:

[0181] The acquisition unit 201 is used to acquire basic operational data, including distribution network topology parameters and parameters of diverse flexible resources, distributed wind power and photovoltaic output data, and node loads; the diverse flexible resources include data centers, mobile energy storage, and electric vehicle charging stations;

[0182] The first construction unit 202 is used to construct a data center spatiotemporal flexible load model, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model based on the distribution network topology parameters and the parameters of the diverse flexible resources.

[0183] The second construction unit 203 is used to construct a data center-mobile energy storage collaborative optimization model with the goal of minimizing the overall system operating cost, based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model under the normal operation scenario of the distribution network. The model is solved using the normal operation constraints of the distribution network constructed based on the distributed wind power and photovoltaic power output data and the node load, as well as the operation constraints of the data center and mobile energy storage, as the first constraint conditions, to obtain the collaborative scheduling strategy.

[0184] The third construction unit 204 is used to construct a data center-electric vehicle charging station elastic enhancement model with the goal of minimizing the overall system loss, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, in the scenario of post-disaster recovery of the distribution network under extreme disturbances. The load recovery strategy is obtained by solving the post-disaster operation constraints of the distribution network constructed based on the distributed wind power and photovoltaic power output data and the node load, as well as the operation constraints of the data center and the electric vehicle charging station, as the second constraint conditions.

[0185] The output unit 205 is used to model and verify the normal operation and post-disaster recovery scenarios respectively, and output the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy.

[0186] In one embodiment, the first building unit 202 is specifically used for:

[0187] Based on the parameters of the data center among the parameters of the aforementioned diverse flexible resources, a spatiotemporal flexible load model for the data center is established using a three-layer coupling structure of workload scheduling, computing resource allocation, and power.

[0188] Based on the parameters of mobile energy storage in the distribution network topology parameters and the parameters of the diverse flexible resources, a mobile energy storage operation model is established based on the transfer time, dwell node, charging time and discharging time of mobile energy storage between distribution system nodes.

[0189] Based on the parameters of electric vehicle charging stations in the parameters of the aforementioned diverse flexible resources, and based on the costs incurred by electric vehicle charging stations in purchasing electricity from the grid and the benefits gained by electric vehicle charging stations in participating in load shedding during emergency situations, a charging and discharging scheduling model for electric vehicle charging stations is established.

[0190] In one embodiment, the second building unit 203 is specifically used for:

[0191] Based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, a first objective function is established under the normal operation scenario of the distribution network with the goal of minimizing the overall system operating cost.

[0192] The first set of constraints is constructed, including power balance constraints, power exchange constraints, voltage constraints at distribution network nodes, data center constraints, and mobile energy storage constraints.

[0193] The collaborative optimization model of data center-mobile energy storage, which consists of the first objective function and the first constraint, is solved to obtain the collaborative scheduling strategy.

[0194] The expression for the first objective function is:

[0195] ;

[0196] in, The cost of electricity purchased from the upstream power grid. Costs incurred due to line losses during power distribution network transmission. The charging, discharging, operation, and scheduling costs of mobile energy storage systems, The cost of electricity consumption required for data center operation.

[0197] In one embodiment, the third building unit 204 is specifically used for:

[0198] Based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a second objective function is established under the scenario of post-disaster recovery of the distribution network under extreme disturbances, with the goal of minimizing the overall system loss.

[0199] Construct second constraints, including electric vehicle constraints, data center constraints, and power distribution network constraints;

[0200] The load recovery strategy is obtained by solving the data center-electric vehicle charging station elasticity improvement model composed of the second objective function and the second constraint.

[0201] The expression for the second objective function is:

[0202] ;

[0203] in, Let n be the load shedding penalty coefficient. For nodes At any moment Load demand; For nodes At any moment The power supply load; A function representing the operating cost of electric vehicle charging stations; This is a function for the operating cost of internet data centers.

[0204] In one embodiment, the output unit 205 is specifically used for:

[0205] Based on a pre-built improved IEEE 69-node distribution network, a distribution network simulation model under normal operation scenarios and a distribution network fault recovery simulation model under extreme disturbance post-disaster recovery scenarios are established.

[0206] In the power distribution network simulation model under the normal operation scenario, the cooperative scheduling strategy is executed, and the normal cooperative scheduling scheme is obtained through simulation calculation;

[0207] In the power distribution network fault recovery simulation model under the post-disaster recovery scenario, the load recovery strategy is executed, and a post-disaster resilient recovery scheme is obtained through simulation calculation.

[0208] In summary, the distribution network collaborative optimization operation device provided by this invention first acquires the distribution network topology parameters and parameters of data centers, mobile energy storage, and electric vehicle charging stations, laying a data foundation for unified collaborative scheduling. Then, it constructs a spatiotemporal flexible load model for data centers, a mobile energy storage operation model, and a charging / discharging scheduling model for electric vehicle charging stations, respectively. This transforms data centers from rigid loads into spatiotemporally flexible resources with task relocation and cross-station migration capabilities; mobile energy storage from fixed node resources into a flexible adjustment entity capable of optimized charging / discharging and location transfer; and electric vehicle charging stations into flexible support nodes that can provide load reduction. This allows for the coordinated operation of data centers, mobile energy storage, and electric vehicle charging stations. By incorporating diverse and flexible resources such as electric vehicle charging stations into the unified optimization framework of the distribution network, the spatiotemporal adjustability of the load side and the energy side support capabilities jointly participate in the optimization of distribution network operation. This solves the problem of insufficient utilization of the spatiotemporal adjustability characteristics of diverse resources and the inability to coordinate and schedule them in a unified manner in existing technologies. Next, under the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the operation model of mobile energy storage, a data center-mobile energy storage collaborative optimization model with the objective of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy. This strategy reshapes the spatiotemporal distribution of load through task time shifting and cross-site migration of the data center, while smoothing the net load through the charging, discharging, and location transfer of mobile energy storage. By mitigating fluctuations and reducing network losses, and thereby reducing peak-hour external power demand, improving load curve shape, lowering overall system operating costs and network losses, and enhancing the controllability of distribution network operation while meeting constraints such as power balance and node voltage, the problem of high normal operating costs is solved. Subsequently, in the scenario of distribution network post-disaster recovery from extreme disturbances, based on the spatiotemporal flexible load model of data centers and the charging and discharging scheduling model of electric vehicle charging stations, a data center-electric vehicle charging station elastic boosting model with the goal of minimizing overall system losses is constructed and solved to obtain a load recovery strategy. This strategy releases power margin during critical periods by migrating and reducing computing tasks in the data center, and integrates with electric vehicle charging stations. The charging and discharging support capabilities create a complementary synergy between the demand and supply sides. Under constraints such as islanded power balance, fault isolation, and load importance levels, priority is given to ensuring the restoration of primary loads. This achieves dynamic matching of supply and demand and flexible and rational allocation of resources during the recovery period, reducing load shedding losses, improving the recovery level of total load and primary loads, enhancing the continuity and efficiency of the post-disaster recovery process, and solving the problem of low recovery levels of critical loads after disasters. Finally, modeling and verification are performed for both normal operation scenarios and post-disaster recovery scenarios, outputting normal collaborative scheduling schemes and post-disaster flexible recovery schemes. This simultaneously improves the economy of normal operation and the flexibility of post-disaster recovery within a unified technical framework. Therefore, this invention solves the problems of high normal operation costs and low recovery levels of critical loads after disasters caused by the lack of unified collaborative scheduling of diverse and flexible resources in existing technologies.

[0209] Furthermore, this embodiment of the invention also provides a distribution network collaborative optimization operation device, the device including a processor and a memory:

[0210] The memory is used to store program code and transmit the program code to the processor;

[0211] The processor is used to execute the steps of the power distribution network collaborative optimization operation method as described in the above method embodiments, according to the instructions in the program code.

[0212] Furthermore, this embodiment of the invention also provides a computer-readable storage medium for storing program code, which is used to execute the power distribution network collaborative optimization operation method described in the above method embodiments.

[0213] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0214] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0215] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0216] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0217] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0218] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for coordinated optimization of distribution network operation, characterized in that, include: Acquire basic operational data, including distribution network topology parameters and parameters of various flexible resources, including data centers, mobile energy storage and electric vehicle charging stations; Based on the power distribution network topology parameters and the parameters of the diverse flexible resources, a spatiotemporal flexible load model for data centers, a mobile energy storage operation model, and a charging and discharging scheduling model for electric vehicle charging stations are constructed. In the normal operation scenario of the distribution network, based on the spatiotemporal flexible load model of the data center and the operation model of the mobile energy storage, a collaborative optimization model of data center-mobile energy storage with the goal of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy. In the scenario of post-disaster recovery from extreme disturbances in the power distribution network, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a data center-electric vehicle charging station elastic boosting model with the goal of minimizing the overall system loss is constructed and solved to obtain the load recovery strategy. Modeling and verification are performed for both normal operation scenarios and post-disaster recovery scenarios, and the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy are output.

2. The method for coordinated optimization of distribution network operation according to claim 1, characterized in that, The step of constructing a data center spatiotemporal flexible load model, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model based on the distribution network topology parameters and the parameters of the diverse flexible resources includes: Based on the parameters of the data center among the parameters of the aforementioned diverse flexible resources, a spatiotemporal flexible load model for the data center is established using a three-layer coupling structure of workload scheduling, computing resource allocation, and power. Based on the parameters of mobile energy storage in the distribution network topology parameters and the parameters of the diverse flexible resources, a mobile energy storage operation model is established based on the transfer time, dwell node, charging time and discharging time of mobile energy storage between distribution system nodes. Based on the parameters of electric vehicle charging stations in the parameters of the aforementioned diverse flexible resources, and based on the costs incurred by electric vehicle charging stations in purchasing electricity from the grid and the benefits gained by electric vehicle charging stations in participating in load shedding during emergency situations, a charging and discharging scheduling model for electric vehicle charging stations is established.

3. The method for coordinated optimization of distribution network operation according to claim 1, characterized in that, In the scenario of normal operation of the distribution network, based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, a data center-mobile energy storage collaborative optimization model with the objective of minimizing the overall system operating cost is constructed and solved to obtain a collaborative scheduling strategy, including: Based on the spatiotemporal flexible load model of the data center and the mobile energy storage operation model, a first objective function is established under the normal operation scenario of the distribution network with the goal of minimizing the overall system operating cost. The first set of constraints is constructed, including power balance constraints, power exchange constraints, voltage constraints at distribution network nodes, data center constraints, and mobile energy storage constraints. The collaborative optimization model of data center-mobile energy storage, which consists of the first objective function and the first constraint, is solved to obtain the collaborative scheduling strategy.

4. The method for coordinated optimization of distribution network operation according to claim 3, characterized in that, The expression for the first objective function is: ； In the formula, The cost of electricity purchased from the upstream power grid. Costs incurred due to line losses during power distribution network transmission. The charging, discharging, operation, and scheduling costs of mobile energy storage systems, The cost of electricity consumption required for data center operation.

5. The method for coordinated optimization of distribution network operation according to claim 1, characterized in that, In the scenario of post-disaster recovery from extreme disturbances in the power distribution network, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a data center-electric vehicle charging station elastic boosting model with the objective of minimizing the overall system loss is constructed and solved to obtain a load recovery strategy, including: Based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, a second objective function is established under the scenario of post-disaster recovery of the distribution network under extreme disturbances, with the goal of minimizing the overall system loss. Construct second constraints, including electric vehicle constraints, data center constraints, and power distribution network constraints; The load recovery strategy is obtained by solving the data center-electric vehicle charging station elasticity improvement model composed of the second objective function and the second constraint.

6. The method for coordinated optimization of distribution network operation according to claim 5, characterized in that, The expression for the second objective function is: ； In the formula, Let n be the load shedding penalty coefficient. For nodes At any moment Load demand; For nodes At any moment The power supply load; A function representing the operating cost of electric vehicle charging stations; This is a function for the operating cost of internet data centers.

7. The method for coordinated optimization of distribution network operation according to claim 1, characterized in that, The process involves modeling and validating normal operation and post-disaster recovery scenarios, respectively, and outputting a normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and a post-disaster elastic recovery scheme corresponding to the load recovery strategy, including: Based on a pre-built improved IEEE 69-node distribution network, a distribution network simulation model under normal operation scenarios and a distribution network fault recovery simulation model under extreme disturbance post-disaster recovery scenarios are established. In the power distribution network simulation model under the normal operation scenario, the cooperative scheduling strategy is executed, and the normal cooperative scheduling scheme is obtained through simulation calculation; In the power distribution network fault recovery simulation model under the post-disaster recovery scenario, the load recovery strategy is executed, and a post-disaster resilient recovery scheme is obtained through simulation calculation.

8. A distribution network collaborative optimization operation device, characterized in that, include: The acquisition unit is used to acquire basic operational data, including distribution network topology parameters and parameters of various flexible resources, wherein the various flexible resources include data centers, mobile energy storage and electric vehicle charging stations; The first construction unit is used to construct a data center spatiotemporal flexible load model, a mobile energy storage operation model, and an electric vehicle charging station charging and discharging scheduling model based on the distribution network topology parameters and the parameters of the diverse flexible resources. The second construction unit is used to construct and solve a data center-mobile energy storage collaborative optimization model with the goal of minimizing the overall system operating cost, based on the data center spatiotemporal flexible load model and the mobile energy storage operation model, under the normal operation scenario of the distribution network, so as to obtain a collaborative scheduling strategy. The third construction unit is used to construct and solve a data center-electric vehicle charging station elastic boosting model with the goal of minimizing the overall system loss, based on the spatiotemporal flexible load model of the data center and the charging and discharging scheduling model of the electric vehicle charging station, in the scenario of post-disaster recovery of the distribution network under extreme disturbances; and obtain the load recovery strategy. The output unit is used to model and verify the normal operation scenario and the post-disaster recovery scenario respectively, and output the normal collaborative scheduling scheme corresponding to the collaborative scheduling strategy and the post-disaster elastic recovery scheme corresponding to the load recovery strategy.

9. A distribution network collaborative optimization operation device, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the power distribution network collaborative optimization operation method according to any one of claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code for executing the distribution network collaborative optimization operation method according to any one of claims 1-7.

Images