CN-122026387-A - Multi-objective optimization method and system for power grid bearing capacity by considering source-load interaction

CN122026387ACN 122026387 ACN122026387 ACN 122026387ACN-122026387-A

Abstract

The invention discloses a multi-objective optimization method and system for power grid bearing capacity considering source-load interaction, and relates to the technical field of power system and power distribution network optimization control. The method comprises the steps of obtaining power grid structure parameters, node net injection prediction baseline data, source load controllable resource information and historical prediction error samples, constructing a resource flexibility envelope and node aggregation control quantity feasible region, determining node voltage calibration parameters and branch current calibration parameters, establishing voltage opportunity constraint, branch current opportunity constraint and budget summarization constraint, further solving a bearing capacity multi-objective optimization model, obtaining node bearing capacity parameters and node aggregation control quantity, decomposing and forming a resource level control quantity sequence, and executing adjustment. According to the invention, finer optimization analysis can be performed on the power grid bearing capacity under the condition of considering source-load interaction and operation uncertainty, and safety constraint satisfaction and resource execution feasibility are considered.

Inventors

LI XUE
YANG BO
DING YIHAN
SONG JINZHAO
Kuang Hongjiang
ZHANG LIJIE
YANG GUANG
LIU HAOTIAN
LIU YAN
SUN XIAOXI

Assignees

国网吉林省电力有限公司白城供电公司

Dates

Publication Date: 20260512
Application Date: 20260407

Claims (9)

1. The power grid bearing capacity multi-objective optimization method considering source load interaction is characterized by comprising the following steps of: Acquiring power grid operation data, wherein the power grid operation data comprises power grid structure parameters, node net injection prediction baseline data, source load controllable resource information and historical prediction error samples; establishing a resource flexibility envelope of each source load controllable resource in the optimized time domain based on the source load controllable resource information, establishing a cross-period coupling constraint on the source load controllable resource with time sequence coupling, and aggregating the resource flexibility envelope into a node aggregation flexibility envelope according to the mapping relation between each source load controllable resource and the node, thereby defining a node aggregation control quantity feasible domain; Determining node voltage calibration parameters and branch current calibration parameters based on historical prediction error samples and grid structure parameters, setting probability budget parameters, and establishing voltage opportunity constraint, branch current opportunity constraint and budget summarization constraint according to the node voltage calibration parameters, the branch current calibration parameters and the probability budget parameters; the method comprises the steps of taking node bearing capacity parameters, node aggregation control quantity and probability budget parameters as decision variables, determining node net injection quantity based on node net injection prediction baseline data and node aggregation control quantity, constructing and solving a bearing capacity multi-objective optimization model under the node aggregation control quantity feasible region, voltage opportunity constraint, branch current opportunity constraint and budget summarization constraint based on the limiting net injection quantity formed by a smooth limiting injection function of the node bearing capacity parameters and the node net injection quantity, and determining node bearing capacity parameter optimization values and node aggregation control quantity optimization values of each discrete period of each node, wherein the node bearing capacity parameter optimization values are used for representing the bearing net injection upper limit of the corresponding node in the corresponding discrete period; and according to the node aggregation control quantity optimization value, solving a resource-level control quantity sequence meeting node aggregation consistency constraint, resource flexibility envelope and cross-period coupling constraint in an optimization time domain, and sending the resource-level control quantity sequence to a corresponding source load controllable resource to execute adjustment.
2. The multi-objective optimization method for the power grid bearing capacity considering source load interaction according to claim 1 is characterized in that the source load controllable resources comprise source side controllable resources and load side controllable resources, wherein constraint parameters of the source side controllable resources at least comprise distributed power output upper and lower limits, climbing constraint, energy storage charging and discharging power constraint and energy storage state of charge constraint in the source load controllable resource information, and constraint parameters of the load side controllable resources at least comprise adjustable load adjustment upper and lower limits, accumulated adjustment quantity constraint, charging facility power constraint and energy demand constraint; The cross-period coupling constraint at least comprises an energy storage state of charge balance constraint, an adjustable load accumulation adjustment constraint and a charging facility period energy balance constraint.
3. The method for optimizing power grid bearing capacity with source load interaction considered as claimed in claim 1, wherein aggregating the resource flexibility envelope into a node aggregation flexibility envelope according to the mapping relation between each source load controllable resource and the node, thereby defining a node aggregation control amount feasible region, comprises: Is provided with A limited set of directions in the control space is aggregated for the nodes, Is a node In discrete time periods Is used for controlling the quantity of node aggregation, For mapping to nodes Node feasible set obtained by common induction of resource flexibility envelope and cross-period coupling constraint of each source load controllable resource Discrete time period And direction vector Calculating a support value, satisfying: ; Constructing an external approximation polyhedron of a node aggregation control quantity feasible domain based on the support value, wherein the external approximation polyhedron meets the following conditions: ; In the formula, For the limited set of directions The first of (3) The number of direction vectors is chosen to be the same, Representing a finite set of directions The total number of directions in (a); Representing nodes In discrete time periods Along the first edge Support values for the individual direction vectors; Representing a transpose; Representing nodes In discrete time periods Is a polygon approximated outside the node aggregation control quantity feasible region; the outer approximation polyhedron As a node In discrete time periods Is a node aggregation control quantity feasible region.
4. The method for multi-objective optimization of power grid bearing capacity with source load interaction considered as claimed in claim 1, wherein determining the node voltage calibration parameter and the branch current calibration parameter based on the historical prediction error sample and the power grid structure parameter comprises: Calculating an active power calibration amplitude and a reactive power calibration amplitude for an active prediction error and a reactive prediction error respectively, and performing robust calibration on the active power calibration amplitude and the reactive power calibration amplitude based on a coverage statistic and a discrete degree statistic, wherein for any power type In discrete time periods Power calibration amplitude of (2) The method meets the following conditions: ; In the formula, Which represents the active power of the electric motor, Representing reactive power; representing power type At the position of Is a prediction error sample sequence of (2); Representation of The operator of the sub-position is used for the sub-position, Representing a quantile level; Representing a median absolute deviation operator; Representing the coverage factor; representing robust coefficients; Each node is in discrete time period Active power calibration amplitude of (a) constitutes an active calibration amplitude vector Each node is in discrete time period Reactive power calibration amplitude of (2) to form reactive calibration amplitude vector ; The method comprises the steps of establishing a sensitivity matrix from active power disturbance and reactive power disturbance to node voltage and branch current by adopting a power distribution network linearization power flow model, and calibrating power amplitude based on the sensitivity matrix And in the iterative solving process, weighting and updating the linearization operating point according to a preset damping factor, and applying trust domain constraint to the node aggregation control quantity between adjacent iterations.
5. The method for multi-objective optimization of power grid bearing capacity with consideration of source load interaction according to claim 1, wherein the probability budget parameters comprise node voltage probability budget parameters And branch current probability budget parameters ; Recording device Is a node In discrete time periods Is used for the calibration of the node voltage parameter, Is a branch circuit In discrete time periods The branch current calibration parameter of (2) then the voltage tightening amount And the amount of current tightening The following respectively satisfy: ; ; In the formula, Representing the coverage factor; An inverse cumulative distribution function operator representing a standard normal distribution, an ; Based on the voltage tightening amount And the amount of current tightening Establishing a voltage opportunity constraint, expressed as: ; The branch current opportunity constraint is expressed as: ; budget summary constraints are expressed as: ; In the formula, Representing nodes In discrete time periods Is used to predict the node voltage for the linearization, 、 Respectively nodes Upper and lower allowable voltage limits of (2); Representing branches In discrete time periods Is used for the linear prediction of the branch current, Representing branches Upper allowable current limit of (2); representing discrete time periods A voltage budget upper limit for (a); representing discrete time periods Is set to the current budget upper limit.
6. The method for multi-objective optimization of power grid bearing capacity with source load interaction considered as claimed in claim 5, wherein the net injection quantity of the nodes is expressed as: ; The net injection amount for clipping is expressed as: ; In the formula, Representing nodes In discrete time periods Is used to determine the net injection quantity of the node, Representing nodes In discrete time periods Is an active net injection predicted baseline value of (1), Representing nodes In discrete time periods The node aggregation control amount of the active component; Representing nodes In discrete time periods Is used to limit the net injection amount of (c) in the injection, Representing nodes In discrete time periods Is defined by the node bearing capacity parameter of (a), Representing the smoothing coefficient; determining a node clipping ratio according to the difference value between the node net injection quantity and the clipping net injection quantity, wherein the node clipping ratio is expressed as follows: ; In the formula, Representing nodes In discrete time periods Is used for limiting the proportion of the node transmission, Representing a preset positive lower limit of the node bearing capacity parameter; the node limit transmission proportion is used for representing the relative limit transmission degree of the node generated by the bearing capacity limitation in the corresponding discrete time period, and writing the relative limit transmission degree into an objective function or constraint condition of the bearing capacity multi-objective optimization model.
7. The power grid bearing capacity multi-objective optimization method considering source load interaction according to claim 6, wherein the objective function of the bearing capacity multi-objective optimization model comprises a node bearing capacity lifting term, a node limit transmission inhibition term and a budget distribution matching term, and the expression is: ; In the formula, Representing an objective function; 、、 Objective function weight coefficients respectively representing a node bearing capacity lifting item, a node limited transmission inhibiting item and a budget distribution matching item, and are all larger than 0; Representing nodes Node importance coefficients of (a); representing discrete time periods Budget distribution matching terms of (c) and satisfy ; Wherein, the ; ; Representing nodes In discrete time periods Is used to determine the actual voltage budget distribution duty cycle of (a), Representing nodes In discrete time periods A reference voltage budget distribution duty cycle of (a); Representing branches In discrete time periods Is used to determine the actual current budget distribution duty cycle of (a), Representing branches In discrete time periods Is set, the reference current budget distribution duty cycle of (a); as a total number of nodes, 、 Respectively any node in the node set Node voltage probability budget parameters, node voltage calibration parameters; For the total number of branches, 、 Respectively any branch in the branch set A branch current probability budget parameter, a branch current calibration parameter; is a regularization constant, and 。
8. The method for optimizing power grid bearing capacity with consideration of source load interaction according to claim 1, wherein the solving the resource-level control quantity sequence meeting node aggregation consistency constraint, resource flexibility envelope and cross-period coupling constraint in the optimizing time domain comprises the following steps: Establishing a resource-level control quantity sequence decomposition model aiming at source load controllable resources mapped to the same node, wherein the resource-level control quantity sequence decomposition model reduces the deviation of a resource-level control quantity sequence of each source load controllable resource relative to a corresponding reference control quantity in an optimization time domain and suppresses jump of a resource-level control quantity of an adjacent discrete period on the premise of meeting node aggregation consistency constraint, resource flexibility envelope corresponding to each source load controllable resource and cross-period coupling constraint; And converting the solved resource-level control quantity sequence into an active power set value, a reactive power set value, a charge-discharge power set value, a load adjustment quantity or a charge power set value corresponding to the source load controllable resource, and sending the active power set value, the reactive power set value, the charge-discharge power set value, the load adjustment quantity or the charge power set value to the corresponding source load controllable resource to execute adjustment.
9. A power grid bearing capacity multi-objective optimization system considering source load interaction, which is used for realizing the power grid bearing capacity multi-objective optimization method considering source load interaction according to any one of claims 1-8, and is characterized in that the system comprises: the data acquisition module is used for acquiring power grid operation data, including power grid structure parameters, node net injection prediction baseline data, source load controllable resource information and historical prediction error samples; The resource modeling module is used for constructing a resource flexibility envelope of each source load controllable resource in the optimization time domain based on the source load controllable resource information, and establishing cross-period coupling constraint on the source load controllable resource with time sequence coupling; the node aggregation module is used for aggregating the resource flexibility envelope into a node aggregation flexibility envelope according to the mapping relation between each source load controllable resource and the node, and limiting a node aggregation control quantity feasible region according to the node aggregation flexibility envelope; The calibration and constraint construction module is used for determining node voltage calibration parameters and branch current calibration parameters based on the historical prediction error samples and the power grid structure parameters, setting probability budget parameters and establishing voltage opportunity constraints, branch current opportunity constraints and budget summarization constraints; The bearing capacity optimizing module is used for taking the node bearing capacity parameter, the node aggregation control quantity and the probability budget parameter as decision variables, determining the node net injection quantity based on the node net injection prediction baseline data and the node aggregation control quantity, constructing and solving a bearing capacity multi-objective optimizing model based on the limiting net injection quantity formed by the node bearing capacity parameter and the node net injection quantity through a smooth limiting injection function, and determining node bearing capacity parameter optimizing values and node aggregation control quantity optimizing values of each node in discrete time periods; And the resource decomposition and execution module is used for solving a resource-level control quantity sequence meeting the node aggregation consistency constraint, the resource flexibility envelope and the cross-period coupling constraint in an optimization time domain according to the node aggregation control quantity optimization value, and sending the resource-level control quantity sequence to a corresponding source load controllable resource to execute adjustment.

Description

Multi-objective optimization method and system for power grid bearing capacity by considering source-load interaction Technical Field The invention relates to the technical field of optimal control of a power system and a power distribution network, in particular to a multi-objective power grid bearing capacity optimization method and system considering source-load interaction. Background With the access of resources such as a distributed power supply, an energy storage device, an adjustable load, an electric vehicle charging facility and the like to a power distribution network, the operation of the power distribution network is gradually changed from the traditional unidirectional power supply to a mode of bidirectional interaction of source and load and dynamic change of operation state. In this context, the grid load capacity is no longer dependent on the plant rated capacity alone, but is also closely related to node injection variation, line operating conditions, load fluctuations, and adjustable resource responsiveness. In the prior art, the power grid bearing capacity evaluation method mostly adopts a static analysis or deterministic boundary check mode, and accessible capacity is generally estimated based on constraints such as node voltage, branch current, transformer capacity and the like. The method is simple to realize, but the cooperative adjustment capability of multi-class source-load resources is not considered sufficiently, and particularly under the condition that energy storage energy constraint, load time sequence constraint or charging demand constraint and the like exist, the actual adjustable space is difficult to accurately reflect. In addition, although the existing part of methods consider prediction errors or operation uncertainties, the existing part of methods are usually processed by adopting uniform margin or fixed safety coefficient, so that the differences of risk levels of different nodes, different branches and different times are difficult to embody, the bearing capacity assessment is easily deviated from conservation, or the safety boundary control is not fine enough. In addition, in the prior art, the problem of insufficient connection between an optimization result and an equipment execution process exists, and even if a node-level adjustment result is obtained, an effective mechanism for further implementing the node-level adjustment result into various source load resource actual control sequences is lacking, so that the engineering application effect of a scheme is influenced. Therefore, it is necessary to provide a power grid bearing capacity optimization method capable of taking the source-load interaction characteristics, uncertainty influence, power grid safety constraint and resource execution feasibility into consideration. Disclosure of Invention Aiming at the problems that the bearing capacity of a power distribution network is difficult to finely evaluate, the risk constraint is difficult to uniformly model and the optimization result is difficult to realize to a resource execution layer under the conditions of source load interaction and uncertainty, the invention provides a multi-target optimization method and system for the bearing capacity of the power distribution network, which consider the source load interaction. The invention realizes the aim through the following technical scheme: a power grid bearing capacity multi-objective optimization method considering source load interaction comprises the following steps: Acquiring power grid operation data, wherein the power grid operation data comprises power grid structure parameters, node net injection prediction baseline data, source load controllable resource information and historical prediction error samples; establishing a resource flexibility envelope of each source load controllable resource in the optimized time domain based on the source load controllable resource information, establishing a cross-period coupling constraint on the source load controllable resource with time sequence coupling, and aggregating the resource flexibility envelope into a node aggregation flexibility envelope according to the mapping relation between each source load controllable resource and the node, thereby defining a node aggregation control quantity feasible domain; Determining node voltage calibration parameters and branch current calibration parameters based on historical prediction error samples and grid structure parameters, setting probability budget parameters, and establishing voltage opportunity constraint, branch current opportunity constraint and budget summarization constraint according to the node voltage calibration parameters, the branch current calibration parameters and the probability budget parameters; the method comprises the steps of taking node bearing capacity parameters, node aggregation control quantity and probability budget parameters as decision variables, determining node net injection