10kV feeder line fault prediction method based on CNN and LightGBM
A fault prediction and feeder technology, applied in the field of intelligent distribution network, can solve problems such as poor objectivity and real-time performance, insufficient rapidity, flexibility and predictability, and complex distribution network structure, and achieve good rapidity and timeliness. sexual effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0054] see Figure 1 to Figure 8 , a 10kV feeder fault prediction method based on CNN and LightGBM, mainly includes the following steps:
[0055] 1) Obtain the original data of the distribution network and preprocess the original data of the distribution network.
[0056] Further, the distribution network raw data includes meteorological data and distribution network information data. The original data of the distribution network is divided into three categories based on the feeder faults, which respectively represent x under different types of feeder faults i year i Raw data of the feeder. The feeder faults include faults caused by operating factors, faults caused by equipment failures and faults caused by weather factors. The feeders described in this embodiment are all 10kV feeders.
[0057] Further, the steps of preprocessing the raw data of the distribution network are as follows:
[0058] 1.1) Complement the original data of the distribution network by using the te...
Embodiment 2
[0102] A 10kV feeder fault prediction method based on CNN and LightGBM mainly includes the following steps:
[0103] 1) Obtain the original data of the distribution network and preprocess the original data of the distribution network.
[0104] 2) Extract features from the raw data of the distribution network, and construct a feature set f{f 1 , f 2 … f 15 , f 16 , L}; L is the label, indicating whether the feeder is faulty or not. Among them, the element f 1 , element f 2 , element f 3 , element f 4 , element f 5 , element f 6 is an inherent attribute. element f 7 , element f 8 , element f 9 , element f 10 , element f 11 , element f 12 , element f 13 characteristics for statistical analysis. element f 14 , element f 15 , element f 16 Deep Temporal Features Extracted for Convolutional Neural Networks (CNN).
[0105] 3) Using the LightGBM algorithm to establish a distribution network fault prediction model.
[0106] 4) Input the real-time data of the 10kV ...
Embodiment 3
[0108] A 10kV feeder fault prediction method based on CNN and LightGBM, the main steps are shown in Example 2, wherein the distribution network original data set feature parameters include inherent attribute features, statistical analysis features and deep time series features.
[0109] The inherent attribute features include line properties f 1 , line length f 2 , Substation f 3 , line commissioning time f 4 , the number of line equipment f 5 and line equipment manufacturers f 6 .
[0110] The statistical analysis features include time statistical features, meteorological statistical features and fault autocorrelation statistical features, wherein the time statistical features include summer characterization parameters f 7 and the weekend representation parameter f 8 . Meteorological statistical features include the maximum value of temperature in a day f 9 , the minimum temperature f in a day 10 , the maximum value of humidity f 11 and the maximum wind speed f 12 ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com