Transient Voltage Stability Labeling With Semi-Supervised Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data-driven technologies for assessing transient voltage stability in power grids face challenges in ensuring the accuracy and reliability of labeled samples, primarily due to the subjective nature of voltage threshold and time window settings.
Innovation Solution
A method and device for labeling transient voltage stability samples in a power grid based on semi-supervised learning, which involves obtaining transient voltage time series trajectories, preliminary labeling, constructing a voltage stability sample set, and using semi-supervised clustering and classification learning to refine the labeling, ensuring interactive verification and repeated iteration until all samples pass verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If engineering criterion based on fixed voltage threshold and time threshold is used, then the monitoring process is simple and easy to operate, but the adaptability to different scenarios and power grids is insufficient
Solution Approach 1:
The patent transforms fixed voltage threshold and time threshold parameters into dynamic thresholds learned from data. The machine learning model automatically adapts threshold values to different power grid scenarios, replacing subjective engineering settings with objective data-driven parameters that improve adaptability while maintaining operational simplicity.
Solution Approach 2:
The patent replaces the mechanical engineering criterion approach (fixed thresholds) with a data-driven machine learning system. This substitution enables the system to automatically adapt to different scenarios through learning from historical data, eliminating the need for manual threshold adjustment while maintaining ease of operation.
2Ease of operation
If engineering criterion based on fixed voltage threshold and time threshold is used, then the monitoring process is simple, but the accuracy and reliability of transient voltage stability assessment result is difficult to guarantee
Solution Approach 1:
The patent replaces the mechanical fixed-threshold criterion with an intelligent machine learning-based assessment system. This substitution maintains operational simplicity while significantly improving measurement precision by learning optimal decision boundaries from extensive training data, thereby guaranteeing higher assessment accuracy and reliability.
Solution Approach 2:
The machine learning model performs self-learning and self-optimization from historical power grid data, automatically improving its assessment accuracy without requiring manual intervention for parameter tuning. This self-service capability ensures high measurement precision while maintaining ease of operation.
3Ease of manufacture
If engineering criterion is used to label samples during sample generation, then the labeling process is straightforward, but the reliability of labeled samples is insufficient
Solution Approach 1:
The patent replaces the mechanical engineering criterion labeling approach with a machine learning-based labeling system. This substitution maintains straightforward labeling process while significantly improving sample labeling reliability by using learned patterns from extensive training data to generate more accurate and trustworthy labels for transient voltage stability samples.
Data Source
AI summary
A method and device for labeling transient voltage stability samples in a power grid based on semi-supervised learning are provided. The method includes: S1: obtaining a transient voltage time series trajectory V formed for each load bus in the power grid under N transient operating scenarios; S2: preliminarily labeling the stability status of each transient operating scenario with a voltage time series dataset V, and integrating the labeling result Yi into a class label dataset Y; S3: constructing a voltage stability sample set S={(Vi, Yi)|1≤i≤N}, dividing S into sample subsets Su and Sk; S4: labeling samples in Su by using a semi-supervised clustering learning method and a semi-supervised classification learning method to obtain result datasets Yu1 and Yu2 respectively; S5: performing interactive verification on Yu1 and Yu2, and updating Su and Sk; and S6: performing repeated iteration on the S4 and the S5.