Method and device for coordinated planning of data centers and energy storage based on node carbon intensity

Abstract

This application provides a method and apparatus for collaborative planning of data centers and energy storage based on nodal carbon intensity, belonging to the field of energy dispatch. The method includes: determining the calculation method for node net load and branch power flow based on the tree topology of the distribution network, and defining a recursive calculation rule for node carbon intensity; updating the net load of each node in the distribution network according to planning disturbances, and synchronously updating the branch power flow of the distribution network to obtain branch power flow data after planning disturbances; substituting the branch power flow data into the recursive calculation rule for node carbon intensity, replacing only the power flow input parameters, and recalculating the carbon intensity distribution of each node in the distribution network; constructing a carbon-oriented two-layer collaborative planning framework for data centers and shared energy storage, using the carbon intensity distribution of each node in the distribution network as the core basis for carbon constraints and optimization, and through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity determination, and time-sharing scheduling optimization, performing carbon convergence judgment and cost accounting on the planning scheme, and outputting the optimal planning scheme.

Description

Technical Field

[0001] This application relates to the field of energy dispatching technology, and in particular to a method and apparatus for coordinated planning of data centers and energy storage based on nodal carbon intensity. Background Technology

[0002] With the comprehensive advancement of the "dual carbon" goals, data centers, as a new type of high-energy-consuming load, directly impact the carbon emission pattern of the power distribution network through their site selection, layout, and operation mode, thus having a profound influence on the achievement of regional carbon reduction targets. Meanwhile, shared energy storage, as a key means of flexibly adjusting the load of the power distribution network and enhancing the absorption capacity of renewable energy, can effectively mitigate the high-energy-consumption fluctuations of data centers and optimize the carbon emission distribution of the power distribution network. Therefore, conducting coordinated planning for data centers and shared energy storage has significant theoretical value and engineering practical implications.

[0003] Currently, research on carbon-oriented collaborative planning of data centers and energy storage has been carried out in related technologies. The main idea is to build a node carbon intensity model based on carbon emission flow theory, quantify the dynamic impact of various factors on node carbon intensity, and build a two-layer collaborative planning framework. The upper layer is used for data center site selection and screening, and the lower layer is used for energy storage site selection, capacity determination and time-sharing scheduling optimization. Preliminary collaboration is achieved through iterative feedback between the upper and lower layers, and the solution is completed by using an algorithm that combines enumeration traversal and commercial solvers. However, existing planning methods still have many shortcomings and are difficult to meet the actual engineering needs and optimization requirements under the "dual carbon" goal: First, site planning and operation scheduling are disconnected, and the "site selection first, verification later" model is often adopted. There is a lack of linkage solution process, resulting in poor feasibility and low solution efficiency. At the same time, the site selection process does not fully consider the actual engineering constraints and is out of touch with the actual engineering situation. Second, carbon costs are not effectively integrated into the optimization goal. Traditional scheduling is mostly economically driven and does not incorporate carbon orientation into the core optimization dimension. It is impossible to dynamically adjust the energy storage operation strategy according to carbon price and carbon intensity, resulting in a disconnect between economic efficiency and carbon reduction goals. Furthermore, the impact of dynamic carbon potential on the planning scheme is not considered. Third, energy storage modeling is too simplified, and a single total energy storage or fixed capacity setting is often adopted. The constraints of different nodes, the linkage between power and energy, and the investment, operation cost and efficiency characteristics of energy storage throughout its entire life cycle are not fully considered.

[0004] Therefore, there is an urgent need for a method to link planning and scheduling, and to coordinate carbon orientation and economic efficiency, so as to improve the feasibility, economic efficiency and emission reduction effect of planning schemes. Summary of the Invention

[0005] In view of this, this application provides a data center and energy storage collaborative planning method and device based on nodal carbon intensity, which can realize the linkage between planning and scheduling, the synergy between carbon orientation and economics, and improve the feasibility, economics and emission reduction effect of planning schemes.

[0006] Specifically, this application is implemented through the following technical solution:

[0007] The first aspect of this application provides a data center and energy storage collaborative planning method based on nodal carbon intensity, the method comprising:

[0008] A dynamic carbon emission flow model is constructed for the distribution network. The calculation methods for node net load and branch power flow are determined based on the tree topology of the distribution network. At the same time, the recursive calculation rules for node carbon intensity are defined.

[0009] A carbon intensity feedback model for data centers and shared energy storage is constructed. Based on the planning disturbances caused by data center site selection, shared energy storage site selection and capacity determination, and charging and discharging scheduling, the net load of each node in the distribution network is first updated, and then the power flow of the distribution network branches is updated based on the changes in the net load of the nodes, so as to obtain the branch power flow data after the planning disturbance.

[0010] Substitute the branch power flow data into the node carbon intensity recursive calculation rule, replace only the power flow input parameters, and recalculate the carbon intensity distribution of each node in the distribution network.

[0011] A carbon-oriented dual-layer collaborative planning framework for data centers and shared energy storage is constructed. The carbon intensity distribution of each node in the distribution network is used as the core basis for carbon constraints and optimization. Through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity determination and time-sharing scheduling optimization, carbon convergence judgment and cost accounting are performed on the planning scheme, and the optimal planning scheme is output.

[0012] A second aspect of this application provides a data center and energy storage collaborative planning device based on nodal carbon intensity, the device comprising a construction module, an update module, a calculation module and an output module;

[0013] The construction module is used to build a dynamic carbon emission flow model for the distribution network, determine the calculation method of node net load and branch power flow based on the tree topology of the distribution network, and define the recursive calculation rules of node carbon intensity.

[0014] The update module is used to construct a carbon intensity feedback model for data centers and shared energy storage. Based on the planning disturbances caused by the location selection of data centers, the location and capacity determination of shared energy storage, and the charging and discharging scheduling, it first updates the net load of each node in the distribution network, and then updates the power flow of the distribution network branches based on the changes in the net load of the nodes, so as to obtain the branch power flow data after the planning disturbance.

[0015] The calculation module is used to substitute the branch power flow data into the node carbon intensity recursive calculation rule, and recalculate the carbon intensity distribution of each node in the distribution network by only replacing the power flow input parameters.

[0016] The output module is used to construct a carbon-oriented dual-layer collaborative planning framework for data centers and shared energy storage. It takes the carbon intensity distribution of each node in the distribution network as the core basis for carbon constraints and optimization. Through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity setting, and time-sharing scheduling optimization, it performs carbon convergence judgment and cost accounting on the planning scheme and outputs the optimal planning scheme.

[0017] The data center and energy storage collaborative planning method and device based on nodal carbon intensity provided in this application first construct a dynamic carbon emission flow model. Based on a tree topology, it standardizes the calculation of node net load and branch power flow, and defines the recursive rules for node carbon intensity. This achieves a precise, efficient, and physically reasonable quantitative expression of carbon intensity at the distribution network node level, providing a unified and reliable carbon sensing foundation for subsequent planning. Then, through a carbon intensity feedback model, it updates node net load and branch power flow in real time according to planning disturbances from data centers and energy storage. Carbon intensity can be recalculated simply by replacing the power flow parameters, forming a dynamic mapping between planning behavior and changes in carbon distribution. The relationship ensures that carbon intensity responds in real time to power changes and that calculations are simple and stable. Finally, a carbon-oriented two-layer collaborative planning framework is constructed, with node carbon intensity as the core constraint and optimization basis. Carbon convergence and cost optimization are achieved through iterative feedback of upper-layer site selection and lower-layer energy storage configuration and scheduling. The overall process is progressive and logically complete, from basic carbon flow modeling and dynamic carbon feedback to two-layer collaborative optimization. It not only achieves accurate characterization and dynamic transmission of carbon emission characteristics, but also achieves collaborative optimization and decoupled solution of data center and energy storage planning. While improving computational efficiency, it ensures that the planning scheme meets carbon emission constraints, grid security and economic optimization. Attached Figure Description

[0018] Figure 1 A flowchart of the data center and energy storage collaborative planning method based on nodal carbon intensity provided in Embodiment 1 of this application;

[0019] Figure 2 This is a schematic diagram of the structure of the data center and energy storage collaborative planning device based on nodal carbon intensity provided in Embodiment 2 of this application. Detailed Implementation

[0020] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0021] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0022] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0023] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0024] Figure 1 This is a flowchart illustrating the data center and energy storage collaborative planning method based on nodal carbon intensity provided in Embodiment 1 of this application. Please refer to... Figure 1 The method provided in this embodiment may include:

[0025] S101. Construct a dynamic carbon emission flow model for the distribution network, determine the calculation method of node net load and branch power flow based on the tree topology of the distribution network, and define the recursive calculation rules for node carbon intensity.

[0026] Specifically, a distribution network is a power distribution network composed of distribution substations, distribution lines, distribution transformers, various load nodes, and distributed power sources. It is used to step down the voltage of the high-voltage transmission system and distribute it to various user nodes. In this application, the distribution network specifically refers to a distribution network with a radial tree topology, which has a clearly defined root node (power source end), branches, and leaf nodes (load ends). Power flows from the root node to each load node along the branches, and it is the physical carrier for the access and operation of data centers and shared energy storage.

[0027] Dynamic carbon emission flow models can reflect the spatiotemporal distribution and transmission patterns of carbon emissions during power transmission in power distribution networks in real time. Based on dynamic operating conditions such as load changes, renewable energy output, data center access, and energy storage charging and discharging behavior, this model can quantify the carbon emission contributions and transmission relationships of each branch and node, distinguishing it from static carbon emission statistics and dynamically reflecting the carbon emission transfer patterns at different times and under different operating scenarios.

[0028] Node net load refers to the net power that a load node in a distribution network actually needs to draw from the upstream grid after deducting the active power consumed by the node itself and the output of distributed renewable energy generation connected locally. Branch flow refers to the magnitude and direction of active power transmitted on branches between two adjacent nodes in a distribution network, reflecting the actual transmission status of electrical energy in the distribution network.

[0029] In specific implementation, the calculation method for determining the node net load based on the tree topology of the distribution network includes: taking the actual power load of each node in the distribution network as the basis, deducting the output of renewable energy accessed locally by the node to obtain the initial net load of the node; performing non-negative processing on the initial net load to form the final node net load; the calculation dimension of the node net load covers all typical days of distribution network operation and the entire time period of each typical day.

[0030] Specifically, the actual electricity load of each node in the distribution network on each typical day and at each time period is obtained; for each node, the local renewable energy generation output is deducted from its actual electricity load to obtain the initial net load; the initial net load is processed to be non-negative, and values ​​less than zero are set to zero to form the final node net load used for calculation; the node net load of all nodes in the distribution network on all typical days and at all times of each typical day is calculated in the above manner.

[0031] For example, in one embodiment, the node net load can be expressed as:

[0032] ;

[0033] in, The net load of node b under typical day s and time period t; The actual electricity load of node b under typical day s and time period t; Provide renewable energy output for node b under typical day s and time period t.

[0034] Optionally, the calculation method for determining branch power flow based on the tree topology of the distribution network includes: for each branch of the distribution network, determining all subtree nodes covered downstream of the target branch, summing the net load of all subtree nodes to obtain the power flow value of the target branch; wherein, the power flow calculation adopts a bottom-up single traversal method from the leaf node to the root node of the distribution network, and the power flow calculation dimension covers all typical days of distribution network operation and the entire time period of each typical day.

[0035] Specifically, for each branch in the distribution network, all subtree nodes downstream of the branch are determined; the net load of all subtree nodes downstream of the branch is summed to obtain the power flow value of the corresponding branch; the power flow calculation of each branch is completed sequentially from the leaf node to the root node in a bottom-up single traversal order; the power flow values ​​of all branches are calculated in the above manner for all typical days and for all time periods of each typical day.

[0036] For example, in one embodiment, the branch power flow can be represented as:

[0037] ;

[0038] ;

[0039] in, The baseline power flow value for branch l under typical day s and time period t; The baseline net load for node j under typical day s and time period t; Let j be the set of all child nodes of node j; The baseline power flow value for the branch connecting node j and child node k under typical day s and time period t; The set of all nodes within the downstream subtree of branch l; This represents the baseline net load of node n within the subtree under typical days s and time periods t.

[0040] Optionally, a recursive calculation rule for node carbon intensity is defined, including: for any target node to be calculated in the distribution network, determining the upstream power supply node of the target node and the upstream adjacent branch connecting the upstream power supply node and the target node, using the carbon intensity of the upstream power supply node as the calculation benchmark for the carbon intensity of the target node; extracting the benchmark power flow value of the upstream adjacent branch, taking the non-negative step of the benchmark power flow value, and multiplying it by the carbon intensity of the upstream power supply node to obtain the positive power contribution value of the upstream adjacent branch; extracting the output of local renewable energy connected to the target node, taking the non-negative step of the local renewable energy output, and multiplying it by the equivalent zero-carbon weight of the renewable energy to obtain the target... The zero-carbon compensation contribution value of the node is calculated; the positive power contribution value is added to the zero-carbon compensation contribution value to obtain the numerator value for calculating the carbon intensity of the target node; the reference power flow value is non-negative, and a preset protection term is superimposed to obtain the denominator value for calculating the carbon intensity of the target node; the numerator value is divided by the denominator value to obtain the carbon intensity of the target node in the corresponding time period and scenario. The carbon intensity calculation of all nodes is completed in a bottom-up recursive order, starting from the leaf nodes of the distribution network, until the root node; in particular, reverse power flow scenarios where the reference power flow value of the upstream adjacent branch is negative, and abnormal calculation scenarios where the reference power flow value is close to zero are handled separately.

[0041] Specifically, for all typical days of distribution network operation and the entire time period of each typical day, for any target node to be calculated in the distribution network, based on the tree topology of the distribution network, the direct upstream power supply node of the target node and the upstream adjacent branch connecting the upstream power supply node and the target node are located. The carbon intensity of the upstream power supply node in the corresponding time period and scenario is used as the calculation benchmark for the carbon intensity of the target node. The benchmark power flow value of the upstream adjacent branch in the corresponding time period and scenario is extracted, and the benchmark power flow value is processed by non-negation (setting values ​​less than zero to zero). Then, the processed power flow value is multiplied by the carbon intensity of the upstream power supply node to obtain the positive power contribution value of the upstream adjacent branch to the target node. The output of local renewable energy connected to the target node in the corresponding time period and scenario is extracted, and the output is processed by non-negation (setting values ​​less than zero to zero). Then, the processed output value is multiplied by the equivalent zero-carbon weight of renewable energy to obtain the zero-carbon compensation contribution value of the target node. The positive power contribution value of the upstream adjacent branch is added to the zero-carbon compensation contribution value of the target node to obtain the numerator value of the carbon intensity calculation of the target node. The reference power flow value of the upstream adjacent branch is processed to be non-negative, and then a preset protection term is added to obtain the denominator value for the carbon intensity calculation of the target node, avoiding calculation anomalies caused by a denominator of zero. The numerator value of the carbon intensity calculation of the target node is divided by the corresponding denominator value to obtain the carbon intensity of the target node in this time period and scenario. Following a bottom-up recursive order, starting from the leaf nodes of the distribution network, the above operations are performed on all nodes in the distribution network in sequence to complete the carbon intensity calculation of each node in the corresponding time period and scenario, up to the root node of the distribution network. For reverse power flow scenarios where the reference power flow value of the upstream adjacent branch is negative, and for calculation anomaly scenarios where the reference power flow value approaches zero, separate correction processing is performed to ensure the stability and accuracy of carbon intensity calculation. Following the above steps, the carbon intensity calculation of all nodes in the distribution network is completed in sequence for all typical days and for each typical day throughout the entire time period, forming a complete distribution network node carbon intensity distribution.

[0042] For example, in one embodiment, the nodal carbon intensity can be expressed as:

[0043] ;

[0044] in, The carbon intensity of target node b under typical day s and time period t; The baseline power flow values ​​for branch (a, b) under typical day s and time period t; Carbon intensity of upstream power supply node a under typical day s and time period t; To provide power output for the local renewable energy source connected to target node b under typical day s and time period t; Equivalent zero-carbon weight for local renewable energy output at target node b; The actual electricity load of target node b under typical day s and time period t; This is a preset protection item.

[0045] The method provided in this embodiment uses the carbon intensity of the upstream power supply node as a benchmark, combines the non-negative power flow processing of the upstream branch with the weighted calculation of zero-carbon compensation of local renewable energy, introduces protection terms to avoid denominator anomalies, and calculates the node carbon intensity from bottom to top recursively. On the one hand, the recursive calculation based on tree topology closely matches the power and carbon emission transmission patterns of radial distribution networks. It can efficiently complete the carbon intensity calculation of all nodes with only a single traversal, without the need for complex power flow iteration, thus significantly improving computational efficiency. On the other hand, the non-negative processing of branch power flow and renewable energy output effectively eliminates the interference of reverse power on carbon intensity calculation, accurately quantifies the carbon transmission contribution of the upstream grid and the emission reduction compensation effect of local zero-carbon power sources, and truly reflects the spatiotemporal propagation and dilution patterns of carbon emissions in the distribution network. At the same time, the setting of the denominator superposition protection term completely avoids the calculation anomalies in scenarios where the power flow approaches zero or reverse power flow, ensuring the robustness and stability of carbon intensity calculation. The resulting dynamic node carbon intensity can accurately reflect the carbon emission levels of each node under different time periods and operating scenarios, providing accurate, reliable, and dynamic node-level carbon potential basis for subsequent carbon-oriented data center and shared energy storage dual-layer collaborative planning. This achieves deep coupling between planning and carbon emission reduction targets, improving the economy and emission reduction effect of the planning scheme.

[0046] S102. Construct a carbon intensity feedback model for data centers and shared energy storage. Based on the planning disturbances caused by the location selection of data centers, the location and capacity determination of shared energy storage, and the charging and discharging scheduling, first update the net load of each node in the distribution network, and then update the power flow of the distribution network branches based on the changes in the net load of the nodes, so as to obtain the power flow data of the branches after the planning disturbance.

[0047] Specifically, the carbon intensity feedback model refers to the calculation process used to dynamically update the net load and branch power flow of the distribution network nodes, and then recalculate the node carbon intensity distribution, after introducing planning disturbance factors such as data centers and shared energy storage. Data centers, as a new type of high-energy-consuming and controllable load within the distribution network, are used to support various computing, storage, and network services. Their core function is to consume electrical energy to maintain equipment operation. In this application, the location and operating scale of the data center directly determine its load injection amount at each node of the distribution network, and are a key disturbance source affecting the distribution network's net load and power flow distribution. Shared energy storage, as a flexible adjustment resource in the distribution network, is used to share the energy storage needs of multiple regions. Its core function is to determine the configuration location and capacity through site selection and capacity allocation, and to achieve peak shaving and valley filling of the distribution network load and optimize power transmission through time-sharing charging and discharging scheduling, thereby adjusting the node net load and branch power flow and helping to improve carbon emission reduction.

[0048] It should be noted that the introduction of data centers and shared energy storage will directly generate new electricity loads at the corresponding nodes, altering their power consumption levels. The charging and discharging behavior of shared energy storage will adjust the node power state in real time, increasing the node's net load during charging and decreasing it during discharging. The node net load is the foundation of power balance in the distribution network, and its changes will directly alter the transmission and distribution of electricity among the network's branches, leading to a redistribution of power flow. To accurately reflect the true power operation status of the distribution network after the planned disturbance and provide precise input parameters for subsequent carbon intensity recalculation, it is necessary to first update the net load of each node in the distribution network, and then update the branch power flow based on the changing patterns of the node net load, obtaining the branch power flow data after the planned disturbance. This data will support accurate feedback on the dynamic carbon intensity distribution.

[0049] In practice, based on the planning disturbances caused by data center site selection, shared energy storage site selection and capacity determination, and charging and discharging scheduling, the net load of each node in the distribution network is first updated. This includes: for any target node in the distribution network, extracting the original net load of the target node as the initial base value for net load update; based on the data center site selection decision, if the target node is selected as a data center construction node, the corresponding data center operating load is superimposed on the original net load; if the node is not selected, the original net load remains unchanged to obtain the current net load of the target node; from the current net load of the target node, the output of local renewable energy connected to the target node is deducted to obtain the intermediate net load of the target node; based on the shared energy storage site selection and capacity determination and charging and discharging scheduling decision, for the shared energy storage connected to the target node, the difference between the charging power and the discharging power is superimposed on the intermediate net load; wherein, the charging power increases the node load, and the discharging power decreases the node load; the net load value after superimposing the energy storage influence is processed to be non-negative to obtain the final net load of the target node after update.

[0050] Specifically, the original net load of the target node calculated in the above embodiments under the corresponding time period and scenario is extracted, and this original net load is used as the initial base value for this net load update; based on the site combination scheme output by the upper-layer data center site selection planning, it is determined whether the target node is selected as a data center construction node. If the target node is a selected construction node, the data center operating load under the corresponding time period and scenario is superimposed on its original net load; if the target node is not selected as a construction node, its original net load value remains unchanged to obtain the current net load of the target node; the local renewable energy output connected to the target node under the corresponding time period and scenario is extracted, and the local renewable energy output is subtracted from the current net load of the target node to obtain the intermediate net load of the target node; based on the lower-layer shared energy storage site selection, capacity setting, and charging / discharging... The scheduling decision determines whether the target node is connected to shared energy storage. If the target node is connected to shared energy storage, the charging power and discharging power of the shared energy storage in the corresponding time period and scenario are obtained, the difference between the charging power and discharging power is calculated, and this difference is added to the intermediate net load of the target node. A positive value for charging power increases the node load, and a negative value for discharging power decreases the node load. If the target node is not connected to shared energy storage, the intermediate net load value remains unchanged. The net load value after adding the impact of shared energy storage is processed to be non-negative, and values ​​less than zero are set to zero to obtain the final net load of the target node after updating in the corresponding time period and scenario. Following the above steps, the net load of all nodes in the distribution network is updated in sequence for all typical days and for all time periods of each typical day, forming the net load distribution of distribution network nodes after the planned disturbance.

[0051] For example, in one embodiment, the updated net load of each node in the distribution network is represented as follows:

[0052] ;

[0053] in, The updated net load of node b under typical day s and time period t; For decision variable vectors; The original net load of node b under typical day s and time period t; For data center operating load; It provides power to the local renewable energy source connected to node b; Charging power for shared energy storage; For the discharge power of shared energy storage; This is the shared energy storage access indicator coefficient; it is 1 if access is granted, and 0 otherwise.

[0054] The method provided in this implementation updates the node net load based on the original net load, superimposed with the data center operating load, deducted from the local renewable energy output, superimposed with the difference in charging and discharging power of shared energy storage, and finally performs non-negative processing. On the one hand, it fully integrates planning disturbance factors such as data center site selection, shared energy storage site selection and capacity setting, and charging and discharging scheduling into the net load calculation, accurately quantifying the dynamic impact of high-energy-consuming data center load injection and flexible adjustment of shared energy storage on the node power state, and truly reflecting the actual power demand of each node in the distribution network after planning disturbances. On the other hand, the non-negative processing of the final net load effectively eliminates the interference of reverse power flow on subsequent power flow and carbon intensity calculations, ensuring the physical rationality of the calculation. At the same time, this update method strictly covers all typical days and all time periods, and can fully characterize the spatiotemporal power distribution characteristics of the distribution network. It provides accurate and reliable input parameters for subsequent branch power flow updates and node carbon intensity recalculation, supports the dynamic closed-loop operation of the carbon intensity feedback model, and lays a solid foundation for the accurate optimization of two-layer collaborative planning.

[0055] Optionally, the power flow of distribution network branches is updated based on changes in node net load, including: for any target branch in the distribution network, determining the baseline power flow value of the target branch when there is no planned disturbance; determining the change in node net load caused by the planned disturbance based on the updated node net load; determining the access node and subtree range of each shared energy storage station according to the site selection and capacity determination decision of shared energy storage, and generating a branch-energy storage subtree association identifier; the identifier is used to distinguish whether the target branch belongs to the shared energy storage subtree; for the target branch that belongs to the shared energy storage subtree, calculating the sum of net load changes in the downstream subtree range of the target branch based on the changes in node net load, and obtaining the power flow correction amount of the corresponding branch; superimposing the baseline power flow value of the target branch with the power flow correction amount to obtain the updated branch power flow after the planned disturbance; wherein, for the target branch that does not belong to the shared energy storage subtree, the baseline power flow value remains unchanged.

[0056] Specifically, for any target branch in the distribution network, the baseline power flow value of the target branch under the corresponding time period, scenario, and without planned disturbance is extracted, and this baseline power flow value is used as the initial base value for this power flow update. Based on the node net load data before and after the planned disturbance, for each node in the distribution network, the net load change (updated net load minus original net load) under the corresponding time period and scenario is calculated to obtain a complete set of node net load changes. According to the lower-level shared energy storage site selection and capacity determination decision, the access nodes of each shared energy storage station are determined. Based on the tree topology of the distribution network, the downstream subtree range corresponding to each energy storage access node is delineated, and a branch-energy storage subtree association identifier is generated to distinguish whether the target branch belongs to the subtree range of a certain shared energy storage. For each target branch, based on the branch-energy storage subtree association identifier... The process involves determining whether the target branch belongs to the shared energy storage subtree: If the target branch belongs to the shared energy storage subtree, all nodes in the downstream subtree of the branch are located, and the net load change of all nodes in the subtree under the corresponding time period and scenario is summed to obtain the power flow correction of the target branch; the baseline power flow value of the target branch is superimposed with the power flow correction to obtain the updated branch power flow of the target branch after the planned disturbance; if the target branch does not belong to the shared energy storage subtree, its baseline power flow value remains unchanged, and no power flow correction is performed; following a bottom-up single traversal order, the power flow update calculation of all branches is completed sequentially from the leaf node of the distribution network to the root node; in the above manner, the power flow update of all branches of the distribution network is completed sequentially for all typical days and all time periods of each typical day to obtain the complete branch power flow data after the planned disturbance.

[0057] For example, in one embodiment, the updated power flow of the distribution network branches is represented as follows:

[0058] ;

[0059] in, The updated branch power flow value for branch l under typical day s and time period t; For decision variable vectors; The baseline power flow value of branch line l under typical day s and time period t without planned disturbance; To perform a summation operation on all shared energy storage stations j within the distribution network, A collection of shared energy storage stations; This serves as the identifier for the branch-energy storage subtree association. The charging power of the shared energy storage station j under typical day s and time period t; The discharge power of the shared energy storage station j under typical day s and time period t.

[0060] The method provided in this embodiment, based on the original baseline power flow, combined with the net load change of nodes, and using the branch-energy storage subtree association identifier, only corrects and updates the affected branches. On the one hand, it accurately captures the physical impact of planning disturbances such as data center site selection and load injection, shared energy storage site selection and capacity setting, and charging and discharging scheduling on the power distribution of the distribution network. It calculates the power flow correction only for branches located within the downstream subtree of shared energy storage, avoiding repeated traversal of all branches in the distribution network and significantly improving the computational efficiency of power flow updates. On the other hand, it fully reflects the power change law of the shared energy storage regulation effect transmitted downstream along the distribution network topology, ensuring that the updated branch power flow can truly and accurately reflect the actual power transmission state of the distribution network after the planning disturbance. This provides high-quality and high-reliability input data for the subsequent accurate recalculation of node carbon intensity, realizes the accurate mapping between planning disturbances and carbon emission feedback, and effectively supports the stable operation of the carbon intensity feedback model and the high-precision solution of the two-layer collaborative planning.

[0061] S103. Substitute the branch power flow data into the node carbon intensity recursive calculation rule, replace only the power flow input parameters, and recalculate the carbon intensity distribution of each node in the distribution network.

[0062] Specifically, the power flow data of each branch of the distribution network updated after the planning disturbance is extracted, and this power flow data is used as a new input parameter for the recursive calculation of node carbon intensity. The recursive calculation rule for node carbon intensity defined in S101 is followed, except that the power flow input parameter in the rule is replaced with the updated branch power flow value after the current planning disturbance, instead of the baseline power flow value without planning disturbance. For any target node to be calculated in the distribution network, the recursive calculation rule for node carbon intensity is continued to be executed, and the carbon intensity calculation of all nodes in the distribution network is completed sequentially for all typical days and for the entire time period of each typical day, so as to obtain the carbon intensity distribution of each node in the distribution network after the planning disturbance. In this embodiment, it will not be described in detail.

[0063] S104. Construct a carbon-oriented dual-layer collaborative planning framework for data centers and shared energy storage. Take the carbon intensity distribution of each node in the distribution network as the core basis for carbon constraints and optimization. Through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity determination and time-sharing scheduling optimization, carbon convergence judgment and cost accounting are performed on the planning scheme, and the optimal planning scheme is output.

[0064] In specific implementation, the upper-level data center site selection planning includes: determining the set of candidate data center nodes within the distribution network, setting the number of data center sites, installed capacity, and power supply security constraints; generating multiple data center site combination schemes as decision schemes to be optimized; inputting each site combination scheme into the lower-level shared energy storage planning, and judging whether the current data center site combination meets the preset carbon budget constraints based on the carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme; selecting feasible data center site combinations that meet both carbon budget constraints and site construction constraints with the optimization objective of minimizing the annual comprehensive cost; and outputting the feasible data center site combinations to the lower level as input conditions for the lower-level shared energy storage site selection and capacity determination.

[0065] Specifically, based on the distribution network topology, relevant standards such as GB50174, and actual engineering construction requirements, pre-site selection constraint verification is performed on each node within the distribution network to determine the set of candidate data center nodes that meet the engineering construction conditions. Constraints such as the number of data center sites, the installed capacity of a single site, and the power supply security of the distribution network are set. An enumeration traversal method is used to generate all candidate node combinations that meet the constraints of the number of sites, installed capacity, and power supply security, forming multiple data center site combination schemes as decision-making schemes to be optimized. For each data center site combination scheme, it is sequentially input into the lower-level shared energy storage planning module, calling the carbon intensity feedback model. Based on the planning disturbance corresponding to the site combination, the net load update and branch power flow update of each node in the distribution network are sequentially completed to obtain the branch power flow data after the planning disturbance. The updated branch power flow data is then substituted into the node carbon intensity recursive calculation. The rules involve replacing only the power flow input parameters and recalculating the carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme. Based on the node carbon intensity distribution corresponding to the current planning scheme, combined with the power load of each node and the operating load of the data center, the total carbon emissions of the distribution network under the full typical day and all time period are calculated to determine whether the current data center site combination meets the preset carbon budget constraints. With the optimization objective of minimizing the annual comprehensive cost, an annual cost calculation model including data center investment cost, operating cost, and carbon cost is constructed to evaluate the economic efficiency and emission reduction effect of all site combination schemes that meet the constraints. Feasible data center site combinations that meet both carbon budget constraints and engineering constraints such as the number of sites, installed capacity, and power supply security are selected. The selected feasible data center site combinations are output to the lower layer as input conditions for the lower layer's shared energy storage site selection, capacity determination, and charging and discharging scheduling.

[0066] Optionally, the lower-level shared energy storage site selection and capacity determination includes: receiving feasible data center site combinations output from the upper-level data center site selection plan; incorporating the node load changes corresponding to the feasible data center site combinations into the constraints of the shared energy storage plan; determining the candidate access node set for shared energy storage within the distribution network; and defining the selection range of candidate nodes based on the distribution network topology, node load distribution, and carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme; setting the capacity upper limit, investment cost threshold, charging and discharging power constraints, and state of charge constraints for shared energy storage as the basic constraints for site selection and capacity determination; and addressing the upper-level data storage requirements. For each feasible data center site combination output by the layer, a shared energy storage site selection and capacity optimization model is constructed by combining the carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme. With the goal of optimizing the investment cost, operation and maintenance cost and carbon penalty cost of shared energy storage, while satisfying the node carbon intensity constraints, distribution network power flow security constraints and energy storage's own operation constraints, candidate access nodes and corresponding capacities for shared energy storage that are suitable for the current data center site combination are selected. The selected shared energy storage site selection and capacity optimization scheme is fed back to the upper-level planning, and the carbon intensity distribution of each node in the distribution network is updated again through the carbon intensity feedback model.

[0067] Specifically, the system receives feasible data center site combinations from the upper-level data center site selection plan, extracts the data center operating load of each node corresponding to the site combination under all typical days and all time periods, and incorporates the load changes of the corresponding nodes into the constraints of the shared energy storage plan. Based on the tree topology of the distribution network, the load distribution after the original load of each node and the new load of the data center are superimposed, and the carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme, the system verifies the access conditions of each node in the distribution network, determines the candidate access node set for shared energy storage, and defines the screening range of candidate nodes. The system sets the single-site capacity upper limit, total investment cost threshold, upper and lower limits of charging and discharging power constraints, and all-time state of charge constraints for shared energy storage as the basic constraints for shared energy storage site selection and capacity determination. For each feasible data center site combination output from the upper level, the system extracts the carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme and constructs a shared energy storage site selection and capacity determination optimization model. The system aims to minimize the investment cost, operation and maintenance cost, and carbon penalty cost of shared energy storage. To optimize the objectives and simultaneously satisfy constraints on node carbon intensity, distribution network power flow security, energy storage charging and discharging power, state of charge, capacity ceiling, and investment cost threshold, the optimization model is solved to select candidate access nodes and corresponding capacities for shared energy storage that are suitable for the current data center site combinations, forming a shared energy storage site selection, capacity determination, and charging / discharging scheduling scheme. The selected shared energy storage site selection, capacity determination, and charging / discharging scheduling scheme is fed back to the upper-level planning layer. The carbon intensity feedback model is invoked, and based on the planning disturbances caused by the shared energy storage site selection, capacity determination, and charging / discharging scheduling, the net load of each node in the distribution network is updated, and the power flow of the distribution network branches is updated based on the changes in node net load, resulting in branch power flow data after the planning disturbance. The updated branch power flow data is substituted into the node carbon intensity recursive calculation rules, replacing only the power flow input parameters, to recalculate the updated carbon intensity distribution of each node in the distribution network. Following this method, the shared energy storage site selection and capacity determination planning suitable for all feasible data center site combinations is completed, forming a complete set of shared energy storage planning schemes. The updated carbon intensity distribution is synchronously fed back to both the upper and lower layers. The upper layer, based on the fed-back carbon intensity distribution, determines whether the current data center site combination meets the carbon constraint requirements. If not, the site combination is adjusted and re-output. The lower layer, synchronously based on the fed-back carbon intensity distribution, optimizes the site selection, capacity determination, and charging / discharging scheduling plan for shared energy storage. The above steps are repeated to achieve bidirectional iterative feedback between the upper-layer data center site selection and the lower-layer shared energy storage planning, until both the site combination output by the upper layer and the energy storage planning scheme output by the lower layer can make the node carbon intensity distribution meet the carbon convergence requirements, and the planning schemes of the upper and lower layers reach a state of coordinated adaptation. The iteration is then terminated. Based on the coordinated adaptation planning scheme of the upper and lower layers, the annual comprehensive cost is calculated, and the optimal coordinated planning scheme is finally determined.

[0068] Optionally, carbon convergence judgment is performed on the planning scheme, including: verifying whether the carbon intensity of all nodes in the distribution network meets the preset carbon budget threshold on all typical days and all time periods, and whether the fluctuation range of carbon intensity is within the preset tolerance range.

[0069] The method provided in this embodiment has two aspects. First, the dynamic carbon emission flow model is based on the tree topology of the distribution network, standardizes the calculation of node net load and branch power flow, and defines the node carbon intensity calculation rules by using bottom-up recursion, non-negative processing, and adding protection terms to the denominator. This can accurately depict the spatiotemporal transmission law of carbon emissions, avoid calculation anomalies and reverse power flow interference, and achieve efficient, stable, and physically reasonable quantification of node carbon intensity, providing a reliable carbon perception basis for subsequent planning. Second, under planning disturbances, the node net load is updated first, and then the branch power flow is updated. By superimposing data center load, deducting renewable energy output, including energy storage charging and discharging, and performing non-negative processing, the method accurately reflects the impact of load and regulation resources on the load. The impact of node power is addressed by using branch-energy storage subtree association identifiers to correct only affected branches, significantly improving power flow update efficiency and ensuring that power calculations accurately reflect the grid operating status, providing precise input for carbon intensity recalculation. Thirdly, the two-layer collaborative planning framework uses iterative feedback from upper-layer data center site selection and lower-layer shared energy storage site selection, capacity determination, and scheduling. With node carbon intensity as the core constraint and optimization basis, it takes into account carbon budget, grid security, and optimal annual comprehensive cost, achieving linkage between planning and scheduling, and synergistic unity between carbon emission reduction and economic efficiency. This effectively solves problems such as the separation of planning and scheduling, lack of deep integration of carbon costs, and simplistic modeling in traditional methods, significantly improving the feasibility, economic efficiency, and emission reduction effect of the planning scheme.

[0070] Corresponding to the aforementioned embodiment of a data center and energy storage collaborative planning method based on node carbon intensity, this application also provides an embodiment of a data center and energy storage collaborative planning device based on node carbon intensity.

[0071] Figure 2 This is a schematic diagram of the data center and energy storage collaborative planning device based on nodal carbon intensity provided in Embodiment 2 of this application. Please refer to... Figure 2 The apparatus provided in this embodiment includes a construction module 210, an update module 220, a calculation module 230, and an output module 240.

[0072] The construction module 210 is used to construct a dynamic carbon emission flow model for the distribution network, determine the calculation method of node net load and branch power flow based on the tree topology of the distribution network, and define the recursive calculation rules of node carbon intensity.

[0073] The update module 220 is used to construct a carbon intensity feedback model for data centers and shared energy storage. Based on the planning disturbances formed by data center site selection, shared energy storage site selection and capacity setting, and charging and discharging scheduling, it first updates the net load of each node in the distribution network, and then updates the power flow of the distribution network branches based on the changes in the net load of the nodes, so as to obtain the branch power flow data after the planning disturbance.

[0074] The calculation module 230 is used to substitute the branch power flow data into the node carbon intensity recursive calculation rule, and only replace the power flow input parameters to recalculate the carbon intensity distribution of each node in the distribution network.

[0075] The output module 240 is used to construct a carbon-oriented dual-layer collaborative planning framework for data centers and shared energy storage. It takes the carbon intensity distribution of each node in the distribution network as the core basis for carbon constraints and optimization. Through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity setting and time-sharing scheduling optimization, it performs carbon convergence judgment and cost accounting on the planning scheme and outputs the optimal planning scheme.

[0076] The apparatus of this embodiment can be used to perform... Figure 1 The steps of the method embodiment shown are similar in principle and process, and will not be repeated here.

[0077] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0078] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0079] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A data center and energy storage collaborative planning method based on nodal carbon intensity, characterized in that, The method includes: A dynamic carbon emission flow model is constructed for the distribution network. The calculation methods for node net load and branch power flow are determined based on the tree topology of the distribution network. At the same time, the recursive calculation rules for node carbon intensity are defined. A carbon intensity feedback model for data centers and shared energy storage is constructed. Based on the planning disturbances caused by data center site selection, shared energy storage site selection and capacity determination, and charging and discharging scheduling, the net load of each node in the distribution network is first updated, and then the power flow of the distribution network branches is updated based on the changes in the net load of the nodes, so as to obtain the branch power flow data after the planning disturbance. Substitute the branch power flow data into the node carbon intensity recursive calculation rule, replace only the power flow input parameters, and recalculate the carbon intensity distribution of each node in the distribution network. A carbon-oriented dual-layer collaborative planning framework for data centers and shared energy storage is constructed. The carbon intensity distribution of each node in the distribution network is used as the core basis for carbon constraints and optimization. Through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity determination and time-sharing scheduling optimization, carbon convergence judgment and cost accounting are performed on the planning scheme, and the optimal planning scheme is output.

2. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, The calculation method for determining the net load of nodes based on the tree topology of the distribution network includes: Based on the actual power load of each node in the distribution network, the initial net load of the node is obtained by deducting the output of renewable energy connected locally to the node. The initial net load is processed to be non-negative, forming the final node net load; the calculation dimension of the node net load covers all typical days of distribution network operation and the entire time period of each typical day.

3. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, The calculation methods for determining branch power flow based on the tree topology of the distribution network include: For each branch of the distribution network, all subtree nodes covered downstream of the target branch are determined, and the net load of all subtree nodes is summed to obtain the power flow value of the target branch. The power flow calculation adopts a bottom-up single traversal method from the leaf node to the root node of the distribution network, and the power flow calculation dimension covers all typical days of distribution network operation and the entire time period of each typical day.

4. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, Define the recursive calculation rules for nodal carbon intensity, including: For any target node to be calculated in the distribution network, the upstream power supply node of the target node and the upstream adjacent branch connecting the upstream power supply node and the target node are determined, and the carbon intensity of the upstream power supply node is used as the calculation benchmark for the carbon intensity of the target node. The reference power flow value of the upstream adjacent branch is extracted, and after the reference power flow value is processed to be non-negative, it is multiplied with the carbon intensity of the upstream power supply node to obtain the positive power contribution value of the upstream adjacent branch. Extract the local renewable energy output connected to the target node, process the local renewable energy output to be non-negative, and multiply it by the equivalent zero-carbon weight of the renewable energy to obtain the zero-carbon compensation contribution value of the target node. The positive power contribution value is added to the zero carbon compensation contribution value to obtain the molecular value of the carbon intensity calculation of the target node; The reference power flow value is processed to be non-negative, and a preset protection term is added to obtain the denominator value for the carbon intensity calculation of the target node. Divide the numerator value by the denominator value to obtain the carbon intensity of the target node in the corresponding time period and scenario. Following a bottom-up recursive order, start from the leaf node of the distribution network and calculate the carbon intensity of all nodes in sequence until the root node. In particular, reverse power flow scenarios where the reference power flow value of the upstream adjacent branch is negative, and abnormal calculation scenarios where the reference power flow value is close to zero are handled separately.

5. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, Based on the planning disturbances caused by data center site selection, shared energy storage site selection and capacity determination, and charging and discharging scheduling, the net load of each node in the distribution network is first updated, including: For any target node in the distribution network, the original net load of the target node is extracted as the initial base value for net load updating; Based on the data center site selection decision, if the target node is selected as the data center site construction node, the corresponding data center operating load is added to the original net load; if no node is selected, the original net load remains unchanged, and the current net load of the target node is obtained. The intermediate net load of the target node is obtained by subtracting the local renewable energy output connected to the target node from the current net load of the target node. Based on the site selection, capacity determination, and charging / discharging scheduling decisions for shared energy storage, the difference between the charging power and the discharging power is added to the intermediate net load for shared energy storage connected to the target node; wherein, the charging power increases the node load, and the discharging power decreases the node load. The net load value after superimposed energy storage effect is processed to be non-negative, and the final net load after the target node is updated is obtained.

6. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, Then, update the power flow of the distribution network branches based on changes in node net load, including: For any target branch in the distribution network, determine the baseline power flow value of the target branch without planned disturbance; Based on the updated node net load, determine the change in node net load caused by the planning disturbance. Based on the site selection and capacity determination decision of shared energy storage, the access node and subtree range of each shared energy storage station are determined, and a branch-energy storage subtree association identifier is generated; the identifier is used to distinguish whether the target branch belongs to the shared energy storage subtree. For a target branch belonging to a shared energy storage subtree, based on the net load change of the node, the total net load change within the downstream subtree of the target branch is calculated to obtain the power flow correction amount of the corresponding branch. The baseline power flow value of the target branch is superimposed with the power flow correction to obtain the updated branch power flow after the planning disturbance; among them, for the target branch that does not belong to the shared energy storage subtree, the baseline power flow value is kept unchanged.

7. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, Upper-level data center site selection planning includes: Determine the set of candidate nodes for data centers within the power distribution network, and set the number of data center construction sites, installed capacity, and power supply security constraints. Generate multiple data center site combination schemes as decision-making options to be optimized; Each site combination scheme is input into the lower-level shared energy storage plan, and based on the carbon intensity distribution of each node of the distribution network corresponding to the current planning scheme, it is determined whether the current data center site combination meets the preset carbon budget constraint. With the goal of minimizing the annual overall cost, feasible combinations of data center sites that meet both carbon budget constraints and site construction constraints are selected. The feasible site combinations of the data center are output to the lower layer as input conditions for the site selection and capacity determination of the lower layer's shared energy storage.

8. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, Lower-level shared energy storage site selection and capacity determination include: Receive feasible site combinations of data centers from the upper-level data center site selection planning output, and incorporate the node load changes corresponding to the feasible site combinations of data centers into the constraints of the shared energy storage planning. Determine the set of candidate access nodes for shared energy storage within the distribution network, and define the screening range of candidate nodes by combining the distribution network topology, node load distribution, and carbon intensity distribution of each node in the distribution network corresponding to the current planning scheme. Set the capacity limit, investment cost threshold, charging and discharging power constraints, and state of charge constraints for shared energy storage as the basic constraints for site selection and capacity determination; For each feasible site combination of data centers output from the upper layer, and in combination with the carbon intensity distribution of each node of the distribution network corresponding to the current planning scheme, a shared energy storage site selection and capacity optimization model is constructed. With the goal of optimizing the investment cost, operation and maintenance cost and carbon penalty cost of shared energy storage, while also meeting the constraints of node carbon intensity, power flow security of distribution network and the operation constraints of energy storage itself, candidate access nodes and corresponding capacities for shared energy storage that are suitable for the current combination of data center sites are selected. The selected shared energy storage site selection and capacity determination schemes are fed back to the upper-level planning, and the carbon intensity distribution of each node in the distribution network is updated again through the carbon intensity feedback model.

9. The data center and energy storage collaborative planning method based on nodal carbon intensity according to claim 1, characterized in that, The carbon convergence assessment of the planning scheme includes: Verify that the carbon intensity of all nodes in the distribution network meets the preset carbon budget threshold on all typical days and at all times, and that the fluctuation range of carbon intensity is within the preset tolerance range.

10. A data center and energy storage collaborative planning device based on nodal carbon intensity, characterized in that, The device includes a construction module, an update module, a calculation module, and an output module; The construction module is used to build a dynamic carbon emission flow model for the distribution network, determine the calculation method of node net load and branch power flow based on the tree topology of the distribution network, and define the recursive calculation rules of node carbon intensity. The update module is used to construct a carbon intensity feedback model for data centers and shared energy storage. Based on the planning disturbances caused by the location selection of data centers, the location and capacity determination of shared energy storage, and the charging and discharging scheduling, it first updates the net load of each node in the distribution network, and then updates the power flow of the distribution network branches based on the changes in the net load of the nodes, so as to obtain the branch power flow data after the planning disturbance. The calculation module is used to substitute the branch power flow data into the node carbon intensity recursive calculation rule, and recalculate the carbon intensity distribution of each node in the distribution network by only replacing the power flow input parameters. The output module is used to construct a carbon-oriented dual-layer collaborative planning framework for data centers and shared energy storage. It takes the carbon intensity distribution of each node in the distribution network as the core basis for carbon constraints and optimization. Through iterative feedback of upper-layer data center site selection planning and lower-layer shared energy storage site selection, capacity setting, and time-sharing scheduling optimization, it performs carbon convergence judgment and cost accounting on the planning scheme and outputs the optimal planning scheme.

Images