CN-122026526-A - Power grid dispatching method and system based on aggregation unit

CN122026526ACN 122026526 ACN122026526 ACN 122026526ACN-122026526-A

Abstract

The invention discloses a power grid dispatching method and a system based on an aggregation unit, and belongs to the field of power system dispatching, wherein the method comprises the steps of acquiring dynamic characteristic vectors of distributed resources in a power grid, aggregating the dynamic characteristic vectors into an aggregation matching unit by adopting a clustering algorithm, generating and screening candidate combinations based on a rapid greedy multi-starting-point selection algorithm to obtain final candidate combinations, establishing a dispatching optimization model according to the final candidate combinations, solving to obtain optimal power instructions of the aggregation matching units, and further realizing accurate dispatching of the power grid; therefore, by implementing the method and the device, the problems of static solidification of the aggregation model, difficulty in real-time adaptation of the dynamic characteristics of resources, high scheduling cost and low efficiency in the prior art can be solved.

Inventors

YE JICHAO
LIU BIN
JI QINGFENG
XIA TONG
JI AOYING
JIANG CHANG
CHENG XIANG
ZHANG HANBING
CHEN YUEJUN
LU WU
HUANG HUI
XU YONGHAI
WU XIAOGANG
LIU FANGZHOU
LIU LINPING

Assignees

国网浙江省电力有限公司丽水供电公司

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. An aggregation unit-based power grid dispatching method is characterized by comprising the following steps: Acquiring dynamic characteristic vectors corresponding to distributed resources in the power grid; clustering each distributed resource by adopting a clustering algorithm according to the dynamic characteristic vector corresponding to each distributed resource to obtain each aggregation matching unit; Generating candidate combinations by adopting a rapid greedy multi-starting-point selection algorithm according to the aggregation matching units, and screening the candidate combinations to obtain final candidate combinations, wherein each candidate combination comprises a plurality of aggregation matching units; And establishing a dispatching optimization model according to the final candidate combination, solving the dispatching optimization model to obtain optimal power instructions corresponding to all the aggregation matching units in the final candidate combination, and dispatching the power grid according to all the optimal power instructions.
2. The power grid dispatching method based on the aggregation unit as claimed in claim 1, wherein the obtaining the dynamic characteristic vector corresponding to each distributed resource in the power grid specifically comprises: and acquiring operation data corresponding to each distributed resource in the power grid, and acquiring dynamic characteristic vectors corresponding to each distributed resource according to the operation data corresponding to each distributed resource.
3. The power grid dispatching method based on the aggregation units as claimed in claim 1, wherein the clustering algorithm is adopted to cluster each distributed resource according to the dynamic characteristic vector corresponding to each distributed resource, so as to obtain each aggregation matching unit, specifically: Processing each distributed resource by adopting a resource similarity formula and a clustering algorithm according to the dynamic characteristic vector corresponding to each distributed resource to obtain a clustering set, wherein the clustering set comprises a plurality of aggregation units; According to the clustering set, obtaining an aggregation unit characteristic vector corresponding to each aggregation unit in the clustering set, and matching each aggregation unit in the clustering set with the aggregation unit characteristic vector corresponding to each aggregation unit to obtain each aggregation matching unit.
4. The power grid dispatching method based on the aggregation units as claimed in claim 1, wherein the generating each candidate combination by adopting a fast greedy multi-start selection algorithm according to each aggregation matching unit comprises the following steps: Randomly selecting a plurality of aggregation matching units from the aggregation matching units as initial units; Traversing each initial unit, constructing an initialization candidate combination according to the initial unit traversed currently, and generating candidate combinations corresponding to the initial unit traversed currently by adopting a quick greedy multi-starting-point selection algorithm according to the initial unit traversed currently and a residual unit set traversed currently, wherein the residual unit set traversed currently is obtained according to each candidate combination and the initial unit traversed currently; and taking the candidate combination corresponding to each initial unit as each candidate combination.
5. The power grid dispatching method based on the aggregation units as claimed in claim 4, wherein the generating the candidate combination corresponding to the initial unit of the current traversal by adopting a fast greedy multi-start selection algorithm according to the initial unit of the current traversal and the residual unit set of the current traversal comprises the following steps: And according to the currently traversed initial unit and the currently traversed residual unit set, iteratively updating the initialization candidate combination by adopting a quick greedy multi-start selection algorithm until the initialization candidate combination of the current iteration meets a first preset condition, and outputting the initialization candidate combination of the current iteration as a candidate combination corresponding to the currently traversed initial unit.
6. The power grid dispatching method based on the aggregation unit as claimed in claim 1, wherein the screening of each candidate combination is performed to obtain a final candidate combination, specifically: calculating by adopting an evaluation function value calculation method according to each aggregation matching unit in each candidate combination to obtain an evaluation function value corresponding to each candidate combination; And sorting the candidate combinations according to the evaluation function values corresponding to the candidate combinations to obtain sorting results, and determining the final candidate combination according to the sorting results.
7. The power grid dispatching method based on the aggregation unit according to claim 1, wherein the dispatching optimization model is built according to the final candidate combination, specifically: according to the final candidate combination, taking the minimized power grid dispatching cost as an optimization target, and establishing an objective function; establishing scheduling variable constraint and power grid mapping constraint according to the final candidate combination; And establishing the dispatching optimization model according to the objective function, the dispatching variable constraint and the power grid mapping constraint.
8. The power grid dispatching system based on the aggregation unit is characterized by comprising a data acquisition module, a first processing module, a second processing module and a dispatching module; the data acquisition module is used for acquiring dynamic characteristic vectors corresponding to distributed resources in the power grid; The first processing module is used for clustering the distributed resources by adopting a clustering algorithm according to the dynamic characteristic vectors corresponding to the distributed resources to obtain aggregation matching units; The second processing module is used for generating candidate combinations by adopting a quick greedy multi-starting-point selection algorithm according to the aggregation matching units and screening the candidate combinations to obtain final candidate combinations, wherein each candidate combination comprises a plurality of aggregation matching units; And the scheduling module is used for establishing a scheduling optimization model according to the final candidate combination, solving the scheduling optimization model to obtain optimal power instructions corresponding to all the aggregation matching units in the final candidate combination, and further scheduling the power grid according to all the optimal power instructions.
9. The power grid dispatching system based on the aggregation unit as claimed in claim 8, wherein the obtaining the dynamic characteristic vector corresponding to each distributed resource in the power grid specifically comprises: and acquiring operation data corresponding to each distributed resource in the power grid, and acquiring dynamic characteristic vectors corresponding to each distributed resource according to the operation data corresponding to each distributed resource.
10. The power grid dispatching system based on the aggregation units as claimed in claim 8, wherein the clustering algorithm is adopted to cluster each distributed resource according to the dynamic characteristic vector corresponding to each distributed resource, so as to obtain each aggregation matching unit, specifically: Processing each distributed resource by adopting a resource similarity formula and a clustering algorithm according to the dynamic characteristic vector corresponding to each distributed resource to obtain a clustering set, wherein the clustering set comprises a plurality of aggregation units; According to the clustering set, obtaining an aggregation unit characteristic vector corresponding to each aggregation unit in the clustering set, and matching each aggregation unit in the clustering set with the aggregation unit characteristic vector corresponding to each aggregation unit to obtain each aggregation matching unit.

Description

Power grid dispatching method and system based on aggregation unit Technical Field The invention relates to the field of power systems, in particular to a power grid dispatching method and system based on an aggregation unit. Background In a modern power system, with the rapid access of distributed photovoltaic, wind power, energy storage and flexible load, the system has the characteristics of double-side high randomization, dispersion and diversification of source load. In the prior art, a method of aggregation is generally adopted to aggregate distributed resources with different geographical dispersion and capacity into one or a plurality of virtual entities with larger specifications and easier management, such as a virtual power plant or a virtual energy storage system. The method aims at enabling an unordered distributed resource group to be equivalent to a small number of controllable aggregation units similar to a traditional power plant, and then carrying out unified interaction and scheduling with a power grid main station. However, such existing aggregation models have fundamental limitations. Since they are mostly built based on offline historical statistics or preset typical operating scenarios, they are static and fixed in nature. The method can not accurately reflect the difference of dynamic response characteristics of units in a resource group in real time and the evolution of the synergistic potential of the units under different time scales, so that the problem that the precision of an aggregation model is seriously not matched with the real-time scheduling requirement exists. Disclosure of Invention The invention provides a power grid dispatching method and system based on an aggregation unit, which can solve the problem that the dispatching cost is reduced while the dispatching efficiency of a power grid is difficult to ensure in the prior art. In a first aspect, an embodiment of the present invention provides a power grid dispatching method based on an aggregation unit, including: Acquiring dynamic characteristic vectors corresponding to distributed resources in the power grid; clustering each distributed resource by adopting a clustering algorithm according to the dynamic characteristic vector corresponding to each distributed resource to obtain each aggregation matching unit; Generating candidate combinations by adopting a rapid greedy multi-starting-point selection algorithm according to the aggregation matching units, and screening the candidate combinations to obtain final candidate combinations, wherein each candidate combination comprises a plurality of aggregation matching units; And establishing a dispatching optimization model according to the final candidate combination, solving the dispatching optimization model to obtain optimal power instructions corresponding to all the aggregation matching units in the final candidate combination, and dispatching the power grid according to all the optimal power instructions. The method and the device can accurately capture dynamic response characteristics of distributed resources in a power grid in real time by acquiring dynamic characteristic vectors of the distributed resources, break the limitation that a traditional static aggregation model depends on offline data, cluster the distributed resources by combining a clustering algorithm to obtain an aggregation matching unit, enable the distributed resources with different geographical dispersion and characteristics to realize scientific and reasonable classification and aggregation, remarkably improve aggregation precision, generate and screen candidate combinations by means of a quick greedy multi-starting-point selection algorithm, efficiently mine synergistic potential of the aggregation matching unit, screen final candidate combinations with more optimal value, finally establish a scheduling optimization model based on the final candidate combinations and solve optimal power instructions, realize optimal management and control of power grid scheduling cost on the basis of meeting scheduling variable constraint and power grid mapping constraint, improve scientificity and execution efficiency of scheduling decisions, effectively balance power grid scheduling efficiency and cost, and guarantee reliability and flexibility of power grid operation. As a preferable example of the first aspect, the obtaining a dynamic characteristic vector corresponding to each distributed resource in the power grid specifically includes: and acquiring operation data corresponding to each distributed resource in the power grid, and acquiring dynamic characteristic vectors corresponding to each distributed resource according to the operation data corresponding to each distributed resource. In the preferred example, the dynamic characteristic vector is generated by acquiring the operation data corresponding to each distributed resource of the power grid, so that the dynamic characteristic vector can be di