CN-122000879-A - Dynamic adjustment method for limit power of power grid section based on boundary condition correction model

CN122000879ACN 122000879 ACN122000879 ACN 122000879ACN-122000879-A

Abstract

The invention relates to the technical field of safe operation and scheduling of an electric power system, and discloses a dynamic adjustment method for power grid section quota power based on a boundary condition correction model, which comprises the steps of collecting and processing multi-source operation data to form a standardized operation boundary condition time sequence; obtaining limit correction value true value by obtaining limit transmissible power of a target transmission section at historical moment or in an off-line simulation scene, training a supervised learning time sequence prediction model to obtain a boundary condition correction model, training an unsupervised learning model based on historical normal operation data to establish a reference model of a normal operation domain, multiplying a limit correction value predicted value output by the boundary condition correction model by a dynamic safety coefficient at a corresponding moment to obtain a final correction value after safety adjustment, and adding the final correction value and static reference limit power to obtain dynamic section limit power. The invention introduces unsupervised learning to construct a normal operation domain, and realizes automatic identification of extreme and unseen working conditions.

Inventors

LI XIAOPENG
WU JUN
Chen Keao
DENG ZHISEN
ZHAO JING
ZHANG GUOFANG
YANG XIAOLEI
LI JINLONG

Assignees

国网四川省电力公司电力科学研究院

Dates

Publication Date: 20260508
Application Date: 20260126

Claims (10)

1. A dynamic adjustment method for the limit power of a power grid section based on a boundary condition correction model is characterized by comprising the following steps: aiming at a target power transmission section, collecting and processing multi-source operation data influencing the power transmission capacity of the target power transmission section to form a standardized operation boundary condition time sequence, wherein each moment t in the operation boundary condition time sequence corresponds to a characteristic vector b (t) representing the operation state at the moment; Obtaining the limit transmissible power of the target transmission section at historical time or in an off-line simulation scene, and calculating the limit transmissible power and the preset static reference limit power Obtaining the true value of the limit correction quantity at the corresponding moment or scene From the characteristic vector b (t) and the limit correction amount Forming a training sample set; The real-time and historical operation boundary condition time sequence is input into the boundary condition correction model when in online application, and the limit correction quantity prediction sequence of the future time period is obtained by prediction; training an unsupervised learning model based on historical normal operation data to establish a reference model of a normal operation domain, inputting a current or predicted operation boundary condition time sequence into the reference model, and calculating the degree of deviation of the operation boundary condition time sequence from the normal operation domain to obtain an abnormality index According to the abnormality index Calculating a dynamic safety factor Wherein the safety factor is Value of (2) with degree of abnormality Is monotonically decreasing with increasing; Multiplying the predicted value of the limit correction quantity output by the boundary condition correction model by the dynamic safety coefficient at the corresponding moment Obtaining a final correction after safety adjustment, and combining the final correction with the static reference limit power And adding to obtain the dynamic section quota power for power grid safety monitoring and scheduling.
2. The dynamic adjustment method for the power grid section limit power based on the boundary condition correction model according to claim 1, wherein the specific process of constructing the time sequence of the operation boundary condition is as follows: normalizing or standardizing the continuous operation data; Periodically encoding the periodic time data through sine and cosine function conversion; processing discrete state data by adopting binarization, single-heat coding or embedded vector coding; and splicing the processed dimensional features into the unified feature vector b (t).
3. The dynamic adjustment method for the limit power of the power grid section based on the boundary condition correction model as set forth in claim 1, wherein the specific process of obtaining the limit transmissible power of the target power transmission section in the historical moment or the off-line simulation scene is as follows: Extracting a transmissibility evaluation value of the target section, which is obtained through safe constraint unit combination, safe constraint economic dispatch or total transmissibility evaluation under the condition that a preset N-1 safe and stable constraint condition is met at a corresponding historical moment, from historical operation records of a power grid dispatching automation system and a safe check system, and taking the transmissibility evaluation value as a limit transmissibility of the historical moment; and/or the number of the groups of groups, And on an offline simulation platform, carrying out power flow calculation and safety and stability analysis aiming at a set operation boundary condition scene, and solving the maximum power which can be safely transmitted by the target section on the premise of meeting the safety and stability guide rule of the power system and the thermal stability constraint of equipment, wherein the maximum power is used as the limit transmissible power in the simulation scene.
4. The dynamic adjustment method of grid section limit power based on boundary condition correction model as set forth in claim 1, wherein the boundary condition correction model adopts a deep neural network model with a codec structure, the input of the deep neural network model is an operation boundary condition characteristic sequence in a historical time window, the output is a limit correction prediction sequence in a future time window, and the deep neural network model uses the static reference limit power Residual modeling is performed for the reference.
5. The method of claim 1, wherein the boundary condition correction model is configured to output a plurality of probability quantile values of the quota correction for the future period simultaneously, the probability quantile comprising a median value and at least one conservative estimate value greater than the median, and wherein the training loss function of the boundary condition correction model combines a mean square error loss for the median value and a quantile regression loss for each quantile.
6. The method for dynamically adjusting the power grid section limit power based on the boundary condition correction model according to claim 4 or 5, wherein a weight is allocated to each training sample when the boundary condition correction model is trained, and the weight is inversely related to a safety margin of an operation condition corresponding to the training sample, so that the smaller the safety margin, the greater the contribution of the training sample to the training loss function in the training process.
7. The method for dynamically adjusting power grid section limit power based on boundary condition correction model as set forth in claim 1, wherein said unsupervised learning model is a model for constructing said reference model using a self-encoder or a variational self-encoder, said anomaly index And measuring by calculating the error norm between the output vector reconstructed by the reference model and the original input vector of the current operation boundary condition feature vector.
8. A method for dynamically adjusting power grid section limit based on a boundary condition correction model according to claim 1 or 7, wherein the safety factor is as follows Is calculated as follows: , wherein, Is the lowest safety coefficient preset, when the abnormality degree index is the same as the abnormality degree index When the preset threshold is exceeded, a rollback mechanism is started to ensure the safety coefficient Is arranged as Or cut off the final correction to a preset conservative upper limit.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the dynamic adjustment method of grid section limit power based on a boundary condition correction model as claimed in any one of claims 1 to 8 when the program is executed by the processor.
10. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, implements a dynamic adjustment method for grid section limit power based on a boundary condition correction model as claimed in any one of claims 1 to 8.

Description

Dynamic adjustment method for limit power of power grid section based on boundary condition correction model Technical Field The invention relates to the technical field of safe operation and scheduling of power systems, in particular to a dynamic adjustment method for power grid section quota power based on a boundary condition correction model. Background The stability limit of the power transmission section of the power grid is a key constraint condition for guaranteeing the safe and stable operation of the power system. The current method mainly prepares a static or stepped section power limit according to the 'safety and stability guidelines of the power system' and related stability limit regulations through offline power flow calculation and stability analysis under a typical operation mode and a preset accident set. The quota is applied to a dispatch monitoring system for monitoring the section flow in real time and determining if the cross-sectional flow is out of limit. However, with the wide access of large-scale intermittent new energy and the increasingly complex operation modes of the power grid, the actual transmissibility of the power transmission section is significantly affected by multi-dimensional operation boundary conditions such as load distribution, new energy output, equipment maintenance planning, meteorological environment and the like. The traditional static quota system and the electronic quota management method based on the rule base are difficult to accurately reflect the continuous change of the boundary conditions in real time. In actual operation, the limit adjustment is usually carried out only through manual experience or semi-automatic mode, so that response is lagged, work is heavy, and two adverse situations are easily caused, namely, the limit is too conservative under most conventional working conditions, the actual power transmission capacity of the power grid is not fully utilized, the resource utilization efficiency and the new energy consumption level are reduced, and the judgment of safety margin is not fine enough under extreme or complex boundary working conditions, so that the operation risk can be hidden. Therefore, a new method capable of automatically sensing the operation boundary condition and realizing intelligent, dynamic and safe adjustment of the section limit power is needed to be researched, so that the transmission potential is fully excavated on the premise of ensuring the safety of the power grid. Disclosure of Invention In order to solve the problems in the prior art, the invention aims to provide a dynamic adjustment method, equipment and storage medium for power grid section limit power based on a boundary condition correction model. The invention dynamically calculates the limit correction quantity through the data driving model on the basis of being completely compatible with the existing static limit system, and introduces an unsupervised safety adjustment mechanism, so as to realize prospective, self-adaptive and risk-controllable dynamic fine adjustment of the section limit, thereby improving the utilization efficiency of the power transmission capacity of the power grid and the safety operation level. The invention is realized by the following technical scheme: in a first aspect, the present invention provides a dynamic adjustment method for power grid section limit power based on a boundary condition correction model, including: S1, aiming at a target power transmission section, acquiring and processing multi-source operation data influencing the power transmission capacity of the target power transmission section to form a standardized operation boundary condition time sequence, wherein each moment t in the operation boundary condition time sequence corresponds to a characteristic vector b (t) representing the operation state at the moment; S2, obtaining the limit transmissible power of the target transmission section in a historical moment or an off-line simulation scene, and calculating the limit transmissible power and a preset static reference limit power Obtaining the true value of the limit correction quantity at the corresponding moment or sceneFrom the characteristic vector b (t) and the limit correction amountForming a training sample set; S3, taking the operation boundary condition time sequence in the history period as input, taking the quota correction truth sequence of the same moment and the future period as an output target, training a supervised learning time sequence prediction model (boundary condition correction model) to obtain a boundary condition correction model; S4, training an unsupervised learning model based on historical normal operation data to establish a reference model of a normal operation domain, inputting a current or predicted operation boundary condition time sequence into the reference model, and calculating the degree of deviation of the operation boundary condition time sequence from the normal