State detection data characteristic parameter extraction and normalized transition method

Abstract

The invention discloses a state detection data characteristic parameter extraction and normalized transition method. The state detection data characteristic parameter extraction and standardization transition method comprises steps of S1 studying a parsing technology for various types of tests according to semi-structured data, adopting an infrared thermography method and a partial discharging map method to extract key contents from the text and storing key contents in a relation database, S2, studying a normalized multi-medium device mixed storage technology and performing normalized storage on an extracted characteristic quantity and key data by targeting an unstructured bulk mass data and on the basis of big data storage and map detection efficiency optimization technology of a computer framework, and S3, studying the analysis and diagnosis technology of the state detection data and performs personalized display on the basis of the normalized model and the characteristic quantity of various kinds of detection data. The state detection data characteristic parameter extraction and normalized transition method is simple in characteristic parameter processing and extraction, accurate, complete and short in time consuming, can explore a connection between the partial discharge characteristic vector and an insulated aged state, and provides the basis to the partial discharge signal.

Description

technical field [0001] The invention relates to the technical field of state detection of electric equipment, in particular to a method for extracting characteristic parameters of state detection data and normalizing conversion. Background technique [0002] Due to the large amount of power equipment map and video data, and limited server resources, in order to make full use of limited resources to improve processing performance, research distributed flow computing technology in big data, realize efficient allocation and processing of power equipment state change detection tasks, and task scheduling Optimization means that different tasks that process different business data will be run in parallel to maximize the use of system computing resources, shorten the overall data analysis time, and improve the performance of parallel task processing. However, the processing of a large amount of unstructured data (including text, images, videos, etc.) is complex and time-consuming. ...

AI Technical Summary

Problems solved by technology

However, the processing of a large amount of unstructured data (including text, images, videos, etc.) is complex and time-consuming

[0003] Moreover, the collected original substation infrared images contain a lot of random noise and clutter interference, which will affect the quality of subsequent image segmentation, thereby affecting the extraction of target features.

Moreover, the extracted data of the existing state detection data feature parameters is incomplete, the extraction of the target hotspot area is not accurate, and the data parameters converted by using the existing parameter conversion method are not standardized.