Special transformer user load feature extraction method and device based on multi-dimensional time sequence data, equipment and medium

By preprocessing and extracting features from multidimensional time series data, the problem of insufficient fusion of multi-source heterogeneous data in traditional methods is solved, and high-precision extraction of user load features of dedicated transformers and improvement of sample quality are achieved.

CN122451285APending Publication Date: 2026-07-24CHINA SOUTHERN POWER GRID DIGITAL GRID GRP CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202610487199.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-14
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Traditional methods for extracting load characteristics from dedicated transformer users do not take into account the fusion and feature construction of multi-source heterogeneous data, and ignore the important influence of external environmental data on load characteristics, resulting in high data noise and single feature dimensions, which affects the accuracy of data processing.

Method used

By preprocessing the collected multidimensional time series data, spatial and temporal repair values ​​of the data points to be repaired are obtained, and fusion is performed based on fusion weights to construct a multidimensional time series dataset. Time series feature sets are extracted, target feature sets are screened, and a spatiotemporal feature sample library of special transformer user load is constructed.

Benefits of technology

It improves the accuracy of repairing abnormal and missing data, comprehensively quantifies load characteristics, improves data processing accuracy and sample quality, and ensures the structure and conciseness of the feature library.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to a special transformer user load feature extraction method and device based on multi-dimensional time sequence data, equipment and a medium. The method comprises the following steps: preprocessing a collected original data set to obtain a reference data set, obtaining a to-be-repaired data point in the reference data set, respectively obtaining a spatial repair value and a time repair value of the to-be-repaired data point, and fusing the spatial repair value and the time repair value based on a fusion weight to obtain a data repair value, updating a data value corresponding to the to-be-repaired data into a data repair value to obtain a repair data set, performing time sequence alignment on the load repair data set and the environment repair data set to obtain a multi-dimensional time sequence data set, extracting a time sequence feature set of the multi-dimensional time sequence data set, screening the time sequence feature set to obtain a target feature set, and storing the target feature set in a database; and constructing a special transformer user load spatio-temporal feature sample library based on the target feature set. By adopting the method, the data processing precision can be improved.
Need to check novelty before this filing date? Find Prior Art