A distribution network source and load cooperative scheduling method and system for a green electricity direct connection scene

Abstract

The application relates to the technical field of power distribution network operation control, and discloses a power distribution network source and load collaborative scheduling method and system for a green power direct connection scene. Green power output, load and energy storage operation data of a plurality of grid-connected green power direct connection projects are acquired, public power grid support power, support time period and support path are determined, and a direct connection support occupied segment is constructed; according to the occupied relationship of each segment to the shared support margin of power distribution equipment, conflict equipment-time period units are identified and a direct connection support conflict cluster is formed; the conflict cluster is subjected to connection compression and interaction attribution, the support reduction responsibility is determined in combination with the project regulation capacity boundary; the residual responsibility power and the residual regulation capacity are used to adjust the energy storage and flexible load power, and the support conflict is rechecked. The scheme improves the rationality and safety of multi-project support conflict identification, responsibility distribution and scheduling execution.

Description

Technical Field

[0001] This invention relates to the field of distribution network operation and control technology, and in particular to a distribution network source-load coordinated scheduling method and system for green electricity direct connection scenarios. Background Technology

[0002] With the widespread integration of new energy sources, energy storage, and flexible loads into distribution networks, existing smart dispatching technologies typically rely on power forecasting, power flow calculations, and optimization algorithms to coordinate the regulation of source-load-storage resources. However, when multiple green power direct-connection projects simultaneously require support from the public grid, the support paths and time periods of different projects may overlap, resulting in overlapping occupancy of shared support margins for distribution equipment. Existing smart dispatching technologies usually focus on overall network power balance and capacity constraint optimization, making it difficult to accurately identify support conflicts jointly formed by multiple projects, and also difficult to quantify the actual contribution of each project to the conflict exceeding limits, leading to a mismatch between the allocation of support reduction responsibilities and the actual margin occupancy. Therefore, how to accurately identify support conflicts among multiple projects and reasonably determine the regulation responsibilities of each project remains a problem that distribution network coordinated dispatching needs to solve in green power direct-connection scenarios. Summary of the Invention

[0003] To address the difficulty of accurately identifying support conflicts arising from multiple direct-connection green energy projects and rationally allocating regulation responsibilities using existing technologies, this invention provides a distribution network source-load coordinated scheduling method and system for direct-connection green energy scenarios. By combining the public grid support power, support time period, and support path of each grid-connected direct-connection green energy project, direct-connection support occupancy segments are formed. Direct-connection support conflict clusters are identified based on the occupancy relationship of each segment on the shared support margin of distribution equipment. Connectivity compression and interactive attribution are performed on the occupancy segments within the conflict clusters. The support reduction responsibility of each project is determined by combining the project's regulation capacity boundary, thereby coordinating and adjusting the energy storage charging and discharging power and flexible load power. Iterative correction is performed through conflict review, ensuring that the regulation responsibility matches the actual crowding-out relationship of each project on the distribution equipment's support margin. This improves the rationality and feasibility of multi-project source-load coordinated scheduling while meeting the safety operation constraints of distribution equipment.

[0004] This invention provides a distribution network source-load coordinated scheduling method for green electricity direct connection scenarios, the method comprising the following steps:

[0005] Step S1: Obtain the green power output forecast curves, load planning curves, energy storage operation plans, grid connection capacity, and project regulation capacity boundaries of multiple grid-connected green power direct connection projects in the target distribution network, as well as the topology, rated transmission capacity, base operating power, and safety reserve capacity of the target distribution network; determine the initial autonomous dispatch curve of each grid-connected green power direct connection project based on the green power output forecast curves, load planning curves, and energy storage operation plans of each project.

[0006] Step S2: Based on each initial autonomous dispatch curve, determine the public grid support power that each grid-connected green energy direct connection project needs to input from the public grid during each dispatch period. Determine the continuous dispatch periods where the public grid support power exceeds the preset support activation threshold as support periods. Based on the grid connection point and topology of each grid-connected green energy direct connection project in the target distribution network, determine the support path through which the public grid support power passes, and combine the corresponding grid-connected green energy direct connection project, support period, support path, and public grid support power into a direct connection support occupancy segment.

[0007] Step S3: For each power distribution device in the target power distribution network, determine the shared support margin of the device based on the rated transmission capacity, the base operating power and the safety reserved capacity of the device during the corresponding scheduling period; map the public grid support power corresponding to each direct connection support segment to the power distribution device in its support path to obtain the support margin occupancy of each direct connection support segment on the power distribution device.

[0008] Step S4: Based on the overlapping support time periods and support paths of each direct-connection support segment, determine the associated direct-connection support segments; for the distribution equipment jointly occupied by the associated direct-connection support segments, summarize the support margin occupancy of the distribution equipment in each associated direct-connection support segment according to the scheduling time period to obtain the total equipment time period support occupancy; when the total equipment time period support occupancy exceeds the equipment shared support margin of the distribution equipment in the corresponding scheduling time period, combine the distribution equipment and the scheduling time period into a conflicting equipment-time period unit, and determine the excess portion as the conflict over-limit; determine the direct-connection support segments that jointly participate in the same conflicting equipment-time period unit as directly associated, and form a direct-connection support conflict cluster with the directly or indirectly associated direct-connection support segments and the conflicting equipment-time period units they participate in;

[0009] Step S5: For each direct-connection support conflict cluster, a support occupancy connectivity compression and boundary compensation interactive attribution mechanism is adopted to compress the direct-connection support occupancy segments within the direct-connection support conflict cluster to obtain the support occupancy proxy nodes; based on the changes in the graph constraint alliance crowding value before and after each support occupancy proxy node joins individually and together with associated proxy nodes, the margin crowding contribution is determined; based on the project adjustment capacity boundary, boundary compensation is performed on the margin crowding contribution to obtain the support reduction responsibility of each grid-connected green electricity direct-connection project;

[0010] Step S6: Determine the adjustment state transition period based on the support reduction responsibility, energy storage operation plan, load plan curve, and project adjustment capacity boundary of each grid-connected green electricity direct connection project; perform adaptive original-dual scheduling of responsibility margin based on the remaining responsibility power and remaining adjustment capacity of each project during the adjustment state transition period, and update the responsibility execution rate and the projection boundaries of the original and dual variables; adjust the energy storage charging and discharging power and flexible load power of each grid-connected green electricity direct connection project based on the updated responsibility execution rate and projection boundaries to obtain the source-load coordinated scheduling curve; update the direct connection support occupied segment based on the source-load coordinated scheduling curve, and redetermine the direct connection support conflict cluster; when there is still a direct connection support conflict cluster, return to step S5, update the support reduction responsibility of each grid-connected green electricity direct connection project based on the redetermined direct connection support conflict cluster, and execute step S6 again until the support margin occupied by each distribution equipment does not exceed the shared support margin of the corresponding equipment; issue energy storage control instructions and flexible load control instructions to the corresponding grid-connected green electricity direct connection project based on the adjusted source-load coordinated scheduling curve.

[0011] This invention also provides a distribution network source-load coordinated scheduling system for green electricity direct connection scenarios, the system comprising:

[0012] The autonomous scheduling determination module is used to acquire the green power output prediction curves, load planning curves, energy storage operation plans, grid connection capacity and project regulation capacity boundaries of multiple grid-connected green power direct connection projects in the target distribution network, as well as the topology and power distribution equipment operation data of the target distribution network, and to determine the initial autonomous scheduling curves of each grid-connected green power direct connection project.

[0013] The support occupancy construction module is used to determine the public grid support power and support time period based on each initial autonomous scheduling curve, determine the support path based on the grid connection point and topology of each grid-connected green electricity direct connection project, and construct the direct connection support occupancy segment.

[0014] The support margin determination module is used to determine the shared support margin of each power distribution device based on the rated transmission capacity, base operating power and safety reserve capacity of each device, and to determine the support margin occupancy of each direct-connection support segment for the corresponding power distribution device.

[0015] The support conflict identification module is used to construct direct support conflict clusters and determine the number of conflicts exceeding the limit based on the overlapping relationship of support time periods, the overlapping relationship of support paths, and the comparison results of support margin occupancy and device shared support margin of each direct support segment.

[0016] The adjustment responsibility determination module is used to perform connectivity compression on the direct support occupied segments within each direct support conflict cluster to obtain the support occupied proxy nodes; based on the changes in the graph constraint alliance crowding value before and after each support occupied proxy node joins individually and together with associated proxy nodes, the margin crowding contribution is determined, and boundary compensation is performed according to the project adjustment capacity boundary to obtain the support reduction responsibility amount.

[0017] The scheduling execution module is used to determine the adjustment state transition period and divide the adjustment state segment according to the support reduction responsibility amount. Within each adjustment state segment, it performs adaptive original-dual scheduling of responsibility margin, adjusts the energy storage charging and discharging power and flexible load power, and obtains the source-load coordinated scheduling curve. Based on the source-load coordinated scheduling curve, it redetermines the direct connection support conflict cluster, and triggers the adjustment responsibility determination module to update the support reduction responsibility amount when there is still a direct connection support conflict cluster, until the support margin occupancy of each power distribution equipment does not exceed the corresponding equipment shared support margin, and issues control commands to the corresponding grid-connected green electricity direct connection project.

[0018] By adopting the above solution, the beneficial effects achieved by the present invention are as follows:

[0019] This invention associates the public grid support power of each grid-connected green electricity direct-connection project with the support time period and support path, constructs a direct-connection support occupancy segment that can characterize the relationship between the project, time period, path, and support power, and maps each occupancy segment to the distribution equipment in its support path. This enables equipment-level and time period-level description of the shared support margin of distribution equipment occupied by multiple green electricity direct-connection projects. By comparing the total support occupancy of equipment time periods with the shared support margin of equipment, conflicting equipment-time period units are identified and direct-connection support conflict clusters are formed. This improves the conflict identification capability in the case of overlapping support paths and support time periods of multiple projects, and solves the problem that it is difficult to locate specific conflicting equipment, conflicting time periods, and participating projects based solely on the overall grid power balance results.

[0020] This invention compresses the direct support occupancy segments within the direct support conflict cluster to form support occupancy proxy nodes that retain support paths, margin occupancy, and project affiliation. It determines the margin occupancy contribution based on the change in the occupancy value of the graph-constrained alliance when a proxy node participates alone or jointly with associated proxy nodes, thus distinguishing between the independent occupancy effect of a single project and the collaborative occupancy effect of multiple projects. Furthermore, it performs boundary compensation based on the project's adjustment capacity boundary, allocating the responsibility exceeding the project's adjustment capacity to other projects that can reduce the remaining conflict excess. This improves the matching degree between the support reduction responsibility and the actual support path, margin occupancy relationship, and project adjustment capacity, solving the problem that averaging responsibility according to support power ratios can easily lead to responsibility imbalance or unexecutable adjustment instructions.

[0021] This invention implements adaptive original-dual scheduling of energy storage charging and discharging power and flexible load power based on the support reduction responsibility, remaining responsibility power, and remaining regulation capacity of each grid-connected green power direct connection project. This realizes the executable conversion of support reduction responsibility into specific energy storage control and flexible load control quantities. By dynamically updating the responsibility execution rate and projection boundary through adjustment state transition identification, and re-identifying direct connection support conflict clusters after forming the source-load coordinated scheduling curve, a closed-loop processing process of responsibility determination, coordinated regulation, and conflict review is formed. This enhances the adaptability of the scheduling scheme to changes in energy storage status, changes in flexible load regulation capacity, and changes in distribution network operation boundaries. While ensuring that the support margin occupancy of distribution equipment does not exceed the equipment shared support margin, it improves the safety, rationality, and feasibility of coordinated scheduling of green power direct connection projects. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a distribution network source-load coordinated scheduling method for green electricity direct connection scenarios provided by the present invention.

[0023] Figure 2 This is a flowchart illustrating the interactive attribution process for supporting occupancy connectivity compression and boundary compensation proposed in Embodiment 5 of the present invention. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0025] Example 1, as Figure 1 As shown, this invention provides a distribution network source-load coordinated scheduling method for green electricity direct connection scenarios, which includes the following steps:

[0026] Step S1: Obtain the green power output prediction curves, load planning curves, energy storage operation plans, grid connection capacity, and project regulation capacity boundaries of multiple grid-connected green power direct connection projects in the target distribution network, as well as the topology, rated transmission capacity, base operating power, and safety reserve capacity of the target distribution network; determine the initial autonomous dispatch curve of each grid-connected green power direct connection project based on the green power output prediction curves, load planning curves, and energy storage operation plans of each project; the project regulation capacity boundary refers to the allowable power range, energy range, and duration range when adjusting the power at the grid connection point of the grid-connected green power direct connection project under the conditions of meeting energy storage operation constraints, flexible load operation constraints, and project grid connection constraints;

[0027] Step S2: Based on each initial autonomous dispatch curve, determine the public grid support power that each grid-connected green energy direct connection project needs to input from the public grid during each dispatch period. Determine the continuous dispatch periods where the public grid support power exceeds the preset support activation threshold as support periods. Based on the grid connection point and topology of each grid-connected green energy direct connection project in the target distribution network, determine the support path through which the public grid support power passes, and combine the corresponding grid-connected green energy direct connection project, support period, support path, and public grid support power into a direct connection support occupancy segment.

[0028] Step S3: For each power distribution device in the target power distribution network, determine the shared support margin of the device based on the rated transmission capacity, the base operating power and the safety reserved capacity of the device during the corresponding scheduling period; map the public grid support power corresponding to each direct connection support segment to the power distribution device in its support path to obtain the support margin occupancy of each direct connection support segment on the power distribution device.

[0029] Step S4: Based on the overlapping support time periods and support paths of each direct-connection support segment, determine the associated direct-connection support segments; for the distribution equipment jointly occupied by the associated direct-connection support segments, summarize the support margin occupancy of the distribution equipment in each associated direct-connection support segment according to the scheduling time period to obtain the total equipment time period support occupancy; when the total equipment time period support occupancy exceeds the equipment shared support margin of the distribution equipment in the corresponding scheduling time period, combine the distribution equipment and the scheduling time period into a conflicting equipment-time period unit, and determine the excess portion as the conflict over-limit; determine the direct-connection support segments that jointly participate in the same conflicting equipment-time period unit as directly associated, and form a direct-connection support conflict cluster with the directly or indirectly associated direct-connection support segments and the conflicting equipment-time period units they participate in;

[0030] Step S5: For each direct-connection support conflict cluster, a support occupancy connectivity compression and boundary compensation interactive attribution mechanism is adopted to perform connectivity compression on the direct-connection support occupancy segments within the direct-connection support conflict cluster to obtain support occupancy proxy nodes; based on the changes in graph constraint alliance crowding value before and after each support occupancy proxy node joins individually and jointly with associated proxy nodes, the margin crowding contribution is determined; based on the project's adjustment capacity boundary, boundary compensation is performed on the margin crowding contribution to obtain the support reduction responsibility of each grid-connected green electricity direct-connection project; the support reduction responsibility refers to the public grid support power that should be reduced during the corresponding scheduling period, allocated to the corresponding grid-connected green electricity direct-connection project to eliminate the conflict excess of conflict equipment-time period units;

[0031] Step S6: Determine the adjustment state transition period based on the support reduction responsibility, energy storage operation plan, load plan curve, and project adjustment capacity boundary of each grid-connected green electricity direct connection project; perform adaptive original-dual scheduling of responsibility margin based on the remaining responsibility power and remaining adjustment capacity of each project during the adjustment state transition period, and update the responsibility execution rate and the projection boundaries of the original and dual variables; adjust the energy storage charging and discharging power and flexible load power of each grid-connected green electricity direct connection project based on the updated responsibility execution rate and projection boundaries to obtain the source-load coordinated scheduling curve; update the direct connection support occupied segment based on the source-load coordinated scheduling curve, and redetermine the direct connection support conflict cluster; when there is still a direct connection support conflict cluster, return to step S5, update the support reduction responsibility of each grid-connected green electricity direct connection project based on the redetermined direct connection support conflict cluster, and execute step S6 again until the support margin occupied by each distribution equipment does not exceed the shared support margin of the corresponding equipment; issue energy storage control instructions and flexible load control instructions to the corresponding grid-connected green electricity direct connection project based on the adjusted source-load coordinated scheduling curve.

[0032] In one specific implementation, taking a 10kV distribution network containing three grid-connected green electricity direct-connection projects A, B and C as an example, a dispatch period is set every 15 minutes; Project A, Project B and Project C are connected to nodes NA, NB and NC respectively. The common grid support path of the three projects all passes through distribution transformer T1. The support paths of Project A and Project B also pass through distribution line L2, and the support path of Project C passes through distribution line L3.

[0033] The green electricity output forecast curves, load planning curves, and energy storage operation plans for three projects are obtained. Taking 10:00 to 10:45 as an example, the public grid support power for Project A during the three scheduling periods is 0.70MW, 0.72MW, and 0.68MW, respectively; for Project B, it is 0.60MW, 0.58MW, and 0.62MW; and for Project C, it is 0.40MW, 0.42MW, and 0.38MW. The preset support activation threshold is 0.20MW. Since the above public grid support power all exceed the support activation threshold, 10:00 to 10:45 is determined as the support period for the three projects, and direct connection support occupancy segments corresponding to Project A, Project B, and Project C are generated respectively.

[0034] The rated transmission capacity of distribution transformer T1 is 5.00MW, the base operating power during the corresponding dispatch period is 3.30MW, and the safety reserve capacity is 0.30MW. Therefore, its shared support margin is 1.40MW. The rated transmission capacity of distribution line L2 is 3.00MW, the base operating power is 1.60MW, and the safety reserve capacity is 0.30MW. Therefore, its shared support margin is 1.10MW. Mapping the public grid support power of each project to the corresponding support path, with the support power transmission coefficient set to 1, the support margin occupancy of distribution transformer T1 for projects A, B, and C is equal to their public grid support power, and the support margin occupancy of distribution line L2 for projects A and B is also equal to their public grid support power.

[0035] During the three scheduling periods of 10:00-10:15, 10:15-10:30, and 10:30-10:45, the total equipment time-period support occupancy of distribution transformer T1 is 1.70MW, 1.72MW, and 1.68MW, respectively, with corresponding conflict over-limits of 0.30MW, 0.32MW, and 0.28MW, respectively; the total equipment time-period support occupancy of distribution line L2 is 1.30MW, with corresponding conflict over-limits of 0.20MW. Therefore, distribution transformer T1 and distribution line L2 are combined with the aforementioned scheduling periods to form conflicting equipment-time-period units. Since the three directly connected support occupancy segments jointly participate in the conflicting equipment-time-period unit corresponding to distribution transformer T1, and projects A and B also jointly participate in the conflicting equipment-time-period unit corresponding to distribution line L2, the three directly connected support occupancy segments and the conflicting equipment-time-period units they participate in form a direct connected support conflict cluster.

[0036] For the direct support conflict cluster, the direct support occupancy segments of Project A, Project B, and Project C are converted into corresponding support occupancy proxy nodes. A compressed association graph is constructed based on the support margin occupancy relationship of the proxy nodes on distribution transformer T1 and distribution line L2. By calculating the change in graph constraint alliance crowding value before and after each proxy node joins individually and jointly with associated proxy nodes, the margin crowding contribution of Project A, Project B, and Project C is obtained. Further, combining the support paths and project adjustment capacity boundaries of the three projects, the support reduction responsibility amounts corresponding to the three scheduling periods are determined: 0.11MW, 0.12MW, and 0.11MW for Project A; 0.09MW, 0.10MW, and 0.09MW for Project B; and 0.10MW, 0.10MW, and 0.08MW for Project C.

[0037] Based on the energy storage operation plan, load plan curve, and project regulation capacity boundary of each project, 10:15 is determined as the regulation state transition period, and the remaining responsible power and remaining regulation capacity of each project before and after the transition are calculated. After implementing the responsibility margin adaptive original-dual scheduling, during the scheduling period from 10:15 to 10:30, the energy storage discharge power of Project A is increased by 0.08MW, and the flexible load power is reduced or transferred by 0.04MW; the energy storage discharge power of Project B is increased by 0.06MW, and the flexible load power is reduced or transferred by 0.04MW; the energy storage discharge power of Project C is increased by 0.07MW, and the flexible load power is reduced or transferred by 0.03MW, thereby completing the support reduction responsibility of 0.12MW, 0.10MW, and 0.10MW respectively.

[0038] After the adjustments were completed, the public grid support power for Projects A, B, and C during the 10:15 to 10:30 dispatch period decreased to 0.60MW, 0.48MW, and 0.32MW, respectively, and the total equipment time-period support occupancy of distribution transformer T1 decreased to 1.40MW; the sum of the support margin occupancy of Projects A and B on distribution line L2 decreased to 1.08MW. After adjusting the remaining two dispatch periods in the same way, the total equipment time-period support occupancy of distribution transformer T1 did not exceed 1.40MW, and the total equipment time-period support occupancy of distribution line L2 did not exceed 1.10MW, and there were no direct-connection support conflict clusters after the re-determination. Based on the adjusted source-load coordinated dispatch curve, corresponding energy storage charging and discharging power control commands and flexible load power control commands were issued to Projects A, B, and C, respectively.

[0039] Example 2: This example is based on Example 1 and further explains the process of determining the public power grid support power and support period in step S2.

[0040] For each scheduling period in each initial autonomous scheduling curve, the power direction of the corresponding initial exchange power is determined; when the power direction is the public grid supplying power to the corresponding grid-connected green electricity direct connection project, the non-negative part of the corresponding initial exchange power is determined as the public grid support power;

[0041] Screening of public grid support periods where the support power exceeds the support activation threshold, merging time-continuous support periods to obtain candidate support periods;

[0042] Candidate support periods with a duration not less than the support duration threshold are identified as support periods, and the start and end times of the support periods, as well as the public grid support power corresponding to each scheduling period, are recorded.

[0043] Example 3: This example is based on Example 2 and further explains the process of constructing the direct connection support occupied segment in step S2.

[0044] For each support period, the public grid support power of the corresponding grid-connected green electricity direct connection project is added as incremental load to the grid connection point of the grid-connected green electricity direct connection project, and incremental power flow calculation is performed while keeping the power injection of other nodes in the target distribution network unchanged.

[0045] Determine the transmission power of each power distribution device before and after the superimposed incremental load, and determine the non-negative difference between the two as the incremental transmission power formed by the public grid support power on the corresponding power distribution device.

[0046] Distribution equipment whose incremental transmission power exceeds the path identification threshold is identified as supporting associated equipment, and the support path for the corresponding public power grid support power is determined according to the electrical connection relationship between each supporting associated equipment and the grid connection point.

[0047] The corresponding grid-connected green electricity direct connection project, support period, support path, and incremental transmission power on each support-related device are combined to obtain the direct connection support occupied segment.

[0048] Example 4 further explains the process of determining the device shared support margin and support margin occupancy in step S3, based on Example 1. Specifically, the direct connection support occupancy segment is generated using the method described in Example 3.

[0049] Obtain the current limit and transmission power limit of each power distribution device, as well as the voltage limit of each node in the target power distribution network; for each power distribution device and each scheduling period, subtract the absolute value of the device's baseline operating power and the safety reserve capacity from the rated transmission capacity of the corresponding device, and determine the non-negative part of the resulting difference as the initial support margin;

[0050] For each power distribution equipment and each scheduling period, determine the direct support occupied segment of the power distribution equipment through which the support path passes; taking the target power distribution network benchmark operating state of the corresponding scheduling period as the starting state, according to the proportion of public grid support power of each direct support occupied segment, synchronously increase the incremental load of the grid connection point corresponding to each direct support occupied segment, and perform power flow calculation after each increase until the current or transmission power of the power distribution equipment reaches the corresponding limit, or the voltage of any node included in the support path reaches the corresponding voltage limit;

[0051] The increase in transmission power of the power distribution equipment relative to the equipment's baseline operating power when any limit is reached is determined as the safety support limit, and the smaller value between the initial support margin and the safety support limit is determined as the equipment-shared support margin of the power distribution equipment in the corresponding scheduling period.

[0052] For each direct-connection support segment, calculate the transmission power increment caused by the unit public grid support power increment at the corresponding grid connection point on each distribution equipment in its support path, and obtain the support power transfer coefficient of the direct-connection support segment to each distribution equipment; the product of the public grid support power of the direct-connection support segment and the corresponding support power transfer coefficient is determined as the support margin occupancy of the direct-connection support segment to the corresponding distribution equipment.

[0053] Example 5, as Figure 2 As shown, this embodiment is based on embodiment four and further explains the process of determining the margin crowding contribution and support reduction responsibility in step S5.

[0054] Step S51: For each direct connection support conflict cluster, determine the conflicting device-time period units contained therein as device-time period coordinates; generate a support occupancy vector based on the support margin occupancy of each direct connection support occupancy segment at each device-time period coordinate, and generate an over-limit response vector based on the change in the conflict over-limit amount of each device-time period coordinate before and after removing the corresponding direct connection support occupancy segment; combine the support occupancy vector and the over-limit response vector into a support occupancy response fingerprint, and establish association edges between the occupancy nodes of each direct connection support occupancy segment that jointly participates in the conflicting device-time period unit, to obtain a support occupancy association graph;

[0055] Step S52: Extract the occupancy nodes belonging to the same grid-connected green electricity direct connection project from the support occupancy association graph to form a project occupancy subgraph; for adjacent occupancy nodes in the project occupancy subgraph, calculate the difference between the corresponding support occupancy response fingerprints, and the deviation between the over-limit response vector obtained by removing adjacent occupancy nodes and the sum of the over-limit response vectors of each node; sequentially merge adjacent occupancy nodes whose differences do not exceed the corresponding thresholds and whose graph remains connected after merging to obtain a connected compressed node group; wherein, the project occupancy subgraph refers to a graph structure that takes the direct connection support occupancy segment corresponding to the same grid-connected green electricity direct connection project as the occupancy node, and is formed by the association edges connecting the occupancy nodes in the support occupancy association graph; the occupancy nodes in the project occupancy subgraph inherit the support occupancy response fingerprint, support time period and support path of the corresponding direct connection support occupancy segment, and the association edge indicates that the corresponding direct connection support occupancy segment has overlapping support time periods and jointly occupies at least one power distribution equipment.

[0056] Step S53: Convert each connected compressed node group into a support occupancy proxy node, and establish a compressed mapping relationship between the support occupancy proxy node and its contained direct-connected support occupancy segments and the grid-connected green electricity direct-connection projects; add the support margin occupancy of each direct-connected support occupancy segment in the same device-time period coordinate to obtain the physical occupancy vector of the support occupancy proxy node; map the associated edges connecting different connected compressed node groups to the connection edges between proxy nodes to obtain a compressed association graph;

[0057] Step S54: For the proxy node alliance composed of supporting occupancy proxy nodes, superimpose the physical occupancy vectors of each proxy node in the alliance, and summarize the portion exceeding the shared support margin of the corresponding device to obtain the initial alliance crowding value; take each connected component formed by the proxy node alliance in the compressed association graph as an effective alliance, and add the initial alliance crowding values ​​of each effective alliance to obtain the graph-constrained alliance crowding value; calculate the change in the graph-constrained alliance crowding value before and after each supporting occupancy proxy node joins different effective alliances to obtain the counterfactual marginal crowding amount;

[0058] The graph-restricted alliance crowding-out value is represented as follows:

[0059] ;

[0060] in, Indicates a proxy node alliance In compressing the correlation diagram Under constraints, graph-based alliances limit value crowding out; This refers to the alliance of proxy nodes that support the current crowding-out value calculation; This represents a compressed relational graph consisting of each supporting proxy node and the connecting edges between the proxy nodes; Indicates a proxy node alliance In compressing the correlation diagram The set of connected components formed in the middle; Indicates one of the valid alliances; Indicates the number of the power distribution equipment. Indicates the number of the scheduling period. This indicates support for occupying proxy nodes; Indicates proxy node In power distribution equipment and scheduling period The corresponding support margin occupancy on the device-time coordinate; This indicates the corresponding device sharing support margin; Indicates the length of the scheduling period; ;

[0061] Step S55: Based on the number of nodes in the effective alliance and the shared device-time occupancy intensity between the proxy nodes and the effective alliance, the counterfactual marginal crowding amount is weighted to obtain the basic marginal attribution amount of each supporting occupancy proxy node; for two supporting occupancy proxy nodes with related edges, the value change generated by their joint joining of the effective alliance is subtracted from the value change generated by their individual joining to obtain the collaborative crowding correction amount, which is then allocated according to the support margin occupancy amount of the two in the shared device-time coordinate; the basic marginal attribution amount and the collaborative crowding correction amount are combined and contribution conservation correction is performed so that the sum of the attribution amounts of each proxy node is equal to the graph-constrained alliance crowding value when all proxy nodes participate; according to the compression mapping relationship, the corrected proxy node attribution amounts are summarized to the corresponding grid-connected green electricity direct connection project to obtain the margin crowding contribution amount of each grid-connected green electricity direct connection project;

[0062] Step S56: Generate an initial support reduction responsibility curve based on the margin crowding contribution of each grid-connected green electricity direct connection project and its support margin occupancy ratio in different time periods; truncate the initial support reduction responsibility curve according to the project regulation capacity boundary to obtain the support reduction responsibility curve within the boundary, and determine the difference before and after truncation as the amount of responsibility to be compensated; determine the reduction relationship between the power reduction of the public grid support and the conflict excess of each conflicting equipment-time period unit according to the support path of each grid-connected green electricity direct connection project, and allocate the amount of responsibility to be compensated to other grid-connected green electricity direct connection projects that can reduce the remaining conflict excess according to the reduction relationship, remaining regulation capacity and margin crowding contribution, until the conflict excess is eliminated or each project reaches the project regulation capacity boundary, and obtain the support reduction responsibility; when there is still a conflict excess, generate a distribution network restriction identifier according to the remaining conflict excess.

[0063] Example 6: This example is based on Example 5 and further explains the process of obtaining the source-load coordinated scheduling curve in step S6.

[0064] Step S61: For each scheduling period within this scheduling cycle, determine the energy storage charging and discharging power margin based on the energy storage operation plan and project regulation capacity boundary of each grid-connected green electricity direct connection project; determine the flexible load power regulation margin based on the load plan curve and project regulation capacity boundary; normalize the support reduction responsibility, energy storage charging and discharging power margin, and flexible load power regulation margin according to the corresponding project regulation capacity boundary, and arrange them according to the grid-connected green electricity direct connection projects and scheduling periods to obtain a regulation state vector sequence; for each regulation state vector, use extended rows and columns to record the first-order terms of each state component, use the main diagonal elements to record the second-order terms of each state component, and use off-diagonal elements to record the product terms between different state components to obtain a regulation state characterization matrix sequence;

[0065] Step S62: Set candidate time windows and candidate segmentation periods with different time scales in the adjusted state representation matrix sequence, calculate the difference between the average representation matrix before and after each candidate segmentation period, and obtain the state difference matrix of each candidate segmentation period; perform singular value decomposition on each state difference matrix to obtain a low-rank state difference matrix; determine the transition score based on the Frobenius norm of each low-rank state difference matrix, and determine the candidate segmentation period with the transition score exceeding the preset transition threshold as the candidate transition period;

[0066] The state difference matrix is ​​represented as follows:

[0067] ;

[0068] in, Indicates candidate time window Intra-candidate segmentation time period The corresponding state difference matrix; Indicates the starting time period of the candidate time window; Indicates the end of the candidate time window; Indicates the candidate segmentation time period; Indicates the scheduling period; Indicates the scheduling period The corresponding regulation state representation matrix; This represents the average representation matrix prior to the candidate segmentation period; This represents the average representation matrix after the candidate segmentation period; the preceding square root term is used to balance the difference in scale caused by the different number of scheduling periods before and after the candidate segmentation period.

[0069] Perform singular value decomposition on the state difference matrix:

[0070] ;

[0071] in, The left singular vector matrix representing the state difference matrix; This represents a singular value matrix in which the singular values ​​are arranged in descending order. The right singular vector matrix representing the state difference matrix; superscript This indicates the matrix transpose.

[0072] Retain the first r singular values ​​and their corresponding left and right singular vectors to reconstruct the matrix, obtaining the low-rank state difference matrix:

[0073] ;

[0074] in, Indicates that the rank is pre-reserved; Indicates from the previous A matrix composed of left singular vectors; Indicates from the previous A diagonal matrix composed of singular values; Indicates from the previous A matrix composed of right singular vectors; This represents the low-rank state difference matrix after retaining the main state change components.

[0075] The corresponding transition score is expressed as:

[0076] ;

[0077] in, Indicates candidate segmentation time period The corresponding transition score; Represents the Frobenius norm; Indicates the index of the singular value to be retained; Indicates the first A set of singular values ​​are retained, and the transition score is used to characterize the overall change in the main regulatory state components before and after the candidate segmentation period.

[0078] Step S63: Divide the scheduling periods within the candidate time window of each candidate transition period into direction identification group and transition positioning group according to the period number; redetermine the low-rank state difference matrix corresponding to each candidate segmentation period based on the direction identification group; divide the low-rank state difference matrix with the largest Frobenius norm by its Frobenius norm to obtain the transition direction matrix; calculate the matrix inner product of each adjustment state characterization matrix sequence and the transition direction matrix in the transition positioning group to obtain the transition projection value; and determine the candidate segmentation period with the smallest sum of squared deviations of the transition projection values ​​before and after the candidate segmentation period as the adjustment state transition period; redefine the adjustment state segment according to each adjustment state transition period.

[0079] Step S64: For each grid-connected green electricity direct-connection project and each regulation status section, the support reduction responsibility amount corresponding to each scheduling period that has not yet been scheduled is accumulated according to the scheduling period length to obtain the remaining responsibility electricity, and the responsibility execution rate is determined according to the remaining responsibility electricity and the corresponding remaining scheduling duration; the remaining regulation capacity of energy storage and the remaining regulation capacity of flexible load are determined according to the energy storage operation plan, load plan curve and project regulation capacity boundary; the responsibility execution rate is allocated according to the proportion of the remaining regulation capacity of energy storage and the remaining regulation capacity of flexible load to the total regulation capacity of the two; the responsibility execution rate allocated to energy storage is converted into the candidate energy storage power adjustment amount to increase the energy storage discharge power or reduce the energy storage charging power, and the responsibility execution rate allocated to flexible load is converted into the candidate flexible load power adjustment amount to reduce or transfer the load power; the candidate energy storage power adjustment amount and the candidate flexible load power adjustment amount are set as original variables, the Lagrange multipliers used to constrain the original variables to complete the responsibility execution rate are set as dual variables, and the projection boundaries of the original variables and dual variables are determined;

[0080] Step S65: Project the original variables onto the corresponding projection boundaries to obtain the executable energy storage power adjustment amount and the executable flexible load power adjustment amount, and determine the support reduction response power based on the executable power adjustment amount; update the original variables and dual variables based on the deviation between the support reduction response power and the responsibility execution rate, and take the projection residual corresponding to the original variable with the largest remaining regulation capacity as the dual stability term until the deviation does not exceed the responsibility deviation threshold; when entering the next regulation state segment, redetermine the responsibility execution rate and projection boundaries, and connect the adjustment results of each regulation state segment to obtain the source-load coordinated scheduling curve.

[0081] Example 7: This example provides a distribution network source-load coordinated scheduling system for green electricity direct connection scenarios. The system is used to execute the methods in any of the preceding examples, and the system includes:

[0082] The autonomous scheduling determination module is used to acquire the green power output prediction curves, load planning curves, energy storage operation plans, grid connection capacity and project regulation capacity boundaries of multiple grid-connected green power direct connection projects in the target distribution network, as well as the topology and power distribution equipment operation data of the target distribution network, and to determine the initial autonomous scheduling curves of each grid-connected green power direct connection project.

[0083] The support occupancy construction module is used to determine the public grid support power and support time period based on each initial autonomous scheduling curve, determine the support path based on the grid connection point and topology of each grid-connected green electricity direct connection project, and construct the direct connection support occupancy segment.

[0084] The support margin determination module is used to determine the shared support margin of each power distribution device based on the rated transmission capacity, base operating power and safety reserve capacity of each device, and to determine the support margin occupancy of each direct-connection support segment for the corresponding power distribution device.

[0085] The support conflict identification module is used to construct direct support conflict clusters and determine the number of conflicts exceeding the limit based on the overlapping relationship of support time periods, the overlapping relationship of support paths, and the comparison results of support margin occupancy and device shared support margin of each direct support segment.

[0086] The adjustment responsibility determination module is used to perform connectivity compression on the direct support occupied segments within each direct support conflict cluster to obtain the support occupied proxy nodes; based on the changes in the graph constraint alliance crowding value before and after each support occupied proxy node joins individually and together with associated proxy nodes, the margin crowding contribution is determined, and boundary compensation is performed according to the project adjustment capacity boundary to obtain the support reduction responsibility amount.

[0087] The scheduling execution module is used to determine the adjustment state transition period and divide the adjustment state segment according to the support reduction responsibility amount. Within each adjustment state segment, it performs adaptive original-dual scheduling of responsibility margin, adjusts the energy storage charging and discharging power and flexible load power, and obtains the source-load coordinated scheduling curve. Based on the source-load coordinated scheduling curve, it redetermines the direct connection support conflict cluster, and triggers the adjustment responsibility determination module to update the support reduction responsibility amount when there is still a direct connection support conflict cluster, until the support margin occupancy of each power distribution equipment does not exceed the corresponding equipment shared support margin, and issues control commands to the corresponding grid-connected green electricity direct connection project.

[0088] The present invention and its embodiments have been described above. This description is not restrictive. The accompanying drawings are only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this description and designs a similar structure and embodiment without departing from the spirit of the present invention, such design should fall within the protection scope of the present invention.

Claims

1. A distribution network source-load coordinated scheduling method for green electricity direct connection scenarios, characterized in that, The method includes the following steps: Step S1: Obtain the green power output forecast curves, load planning curves, energy storage operation plans, grid connection capacity, and project regulation capacity boundaries of multiple grid-connected green power direct connection projects in the target distribution network, and determine the initial autonomous dispatch curve; Step S2: Based on each initial autonomous scheduling curve, construct the direct connection support occupied segment; Step S3: For each power distribution device in the target power distribution network, determine the shared support margin of the device; map the public grid support power corresponding to each directly connected support segment to the power distribution device in its support path to obtain the support margin occupancy. Step S4: Based on the overlapping support time periods and support paths of each direct support segment, determine the associated direct support segments; summarize the support margin usage of each associated direct support segment, and combine it with the device shared support margin to construct a direct support conflict cluster; Step S5: For each direct support conflict cluster, adopt the support occupancy connectivity compression and boundary compensation interactive attribution mechanism to perform connectivity compression on the direct support occupancy segments within the direct support conflict cluster to determine the margin crowding contribution; according to the project adjustment capacity boundary, perform boundary compensation on the margin crowding contribution to obtain the support reduction responsibility. Step S6: Based on the support reduction responsibility, determine the adjustment state transition period, execute the responsibility margin adaptive original-dual scheduling, obtain the source-load cooperative scheduling curve, and issue instructions.

2. The distribution network source-load coordinated scheduling method for green electricity direct connection scenarios according to claim 1, characterized in that, Step S2 specifically involves: determining the public grid support power required from the public grid for each grid-connected green electricity direct connection project during each scheduling period based on the initial autonomous scheduling curves; defining the continuous scheduling period where the public grid support power exceeds the preset support activation threshold as the support period; determining the support path through which the public grid support power passes based on the grid connection point and topology of each grid-connected green electricity direct connection project in the target distribution network; and combining the grid-connected green electricity direct connection project, support period, support path, and public grid support power into a direct connection support occupancy segment.

3. The distribution network source-load coordinated scheduling method for green electricity direct connection scenarios according to claim 2, characterized in that, The process of determining the contribution of margin crowding out includes the following steps: Step E1: For each direct connection support conflict cluster, with each direct connection support occupied segment as the occupied node, establish association edges between the occupied nodes of the jointly participating conflicting devices-time units to obtain the support occupancy association graph. Step E2: Extract occupancy nodes from the support occupancy association graph to form a project occupancy subgraph; for adjacent occupancy nodes in the project occupancy subgraph, calculate the difference between the corresponding support occupancy response fingerprints to obtain a connected compressed node group; Step E3: Convert each connected compressed node group into a support occupancy proxy node, and establish a compressed mapping relationship between the support occupancy proxy node and its contained direct-connected support occupancy segments and the grid-connected green electricity direct-connection projects; add the support margin occupancy of each direct-connected support occupancy segment in the same device-time period coordinate to obtain the physical occupancy vector; map the associated edges connecting different connected compressed node groups to the connection edges between proxy nodes to obtain the compressed association graph; Step E4: For the proxy node alliance composed of supporting occupancy proxy nodes, superimpose the physical occupancy vector of each proxy node in the alliance, take each connected component formed by the proxy node alliance in the compressed association graph as an effective alliance, compute the graph to restrict the change of alliance occupancy value, and obtain the counterfactual marginal occupancy amount. Step E5: Weight the counterfactual marginal crowding amount to obtain the basic marginal attribution amount of each supporting proxy node, and perform contribution conservation correction; according to the compression mapping relationship, summarize the corrected proxy node attribution amount to the corresponding grid-connected green electricity direct connection project to obtain the margin crowding contribution amount.

4. The distribution network source-load coordinated scheduling method for green electricity direct connection scenarios according to claim 3, characterized in that, Step E4 specifically involves: for the proxy node alliance composed of supporting proxy nodes, superimposing the physical occupancy vectors of each proxy node in the alliance, and summing up the portion exceeding the shared support margin of the corresponding device to obtain the initial alliance crowding value. Each connected component formed by the agent node alliance in the compressed association graph is taken as a valid alliance, and the initial alliance crowding value of each valid alliance is added together to obtain the graph-bound alliance crowding value; the change in the graph-bound alliance crowding value before and after each supporting occupying agent node joins different valid alliances is calculated to obtain the counterfactual marginal crowding amount.

5. The distribution network source-load coordinated scheduling method for green electricity direct connection scenarios according to claim 1, characterized in that, The process of obtaining the source-load coordinated scheduling curve includes the following steps: Step S61: Determine the energy storage charging and discharging power margin and the flexible load power regulation margin, and combine them with the support reduction responsibility to obtain the regulation state characterization matrix sequence; Step S62: Set candidate time windows and candidate segmentation periods with different time scales in the adjusted state representation matrix sequence, calculate the difference between the average representation matrix before and after each candidate segmentation period, obtain the state difference matrix and perform singular value decomposition to obtain the low-rank state difference matrix; obtain the transition score based on the Frobenius norm of each low-rank state difference matrix, and determine the candidate transition period. Step S63: Determine the adjustment state transition time period based on the candidate transition time period, and divide the adjustment state segment; Step S64: For each grid-connected green electricity direct connection project and each regulation status section, the support reduction responsibility amount corresponding to each scheduling period that has not yet been scheduled is accumulated according to the scheduling period length to obtain the remaining responsibility electricity and determine the responsibility execution rate; set as the original variable and dual variable, and determine the projection boundary; Step S65: Project the original variables to the corresponding projection boundary to determine the support reduction response power; update the original variables and dual variables according to the deviation between the support reduction response power and the responsibility execution rate to obtain the source-load cooperative scheduling curve.

6. The distribution network source-load coordinated scheduling method for green electricity direct connection scenarios according to claim 5, characterized in that, Step S63 specifically involves: dividing the scheduling periods within the candidate time window of each candidate transition period into a direction identification group and a transition positioning group according to the period number; re-determining the low-rank state difference matrix corresponding to each candidate segmentation period based on the direction identification group; dividing the low-rank state difference matrix with the largest Frobenius norm by its Frobenius norm to obtain the transition direction matrix; calculating the matrix inner product of each adjustment state characterization matrix sequence in the transition positioning group with the transition direction matrix to obtain the transition projection value; and determining the candidate segmentation period with the smallest sum of squared deviations of the transition projection values ​​before and after the candidate segmentation period as the adjustment state transition period.

7. The distribution network source-load coordinated scheduling method for green electricity direct connection scenarios according to claim 1, characterized in that, The instructions in step S6 include energy storage control instructions and flexible load control instructions.

Images