Browse Prior Art Database

Method and System for Device Fault Pre-Warning By Correlation Analysis and Filters

IP.com Disclosure Number: IPCOM000247187D
Publication Date: 2016-Aug-15
Document File: 5 page(s) / 114K

Publishing Venue

The IP.com Prior Art Database

Abstract

Security and reliability is very important to power grid. Conventionally, equipment fault warning rely mainly on the basis of scoring system, which is very rough. This invention is based on all information the utility could collect, calculate the correlation of influence factors and the fault. Then build a trigger and fileter system, which trigger an event when some conditions are satisfied and filter the events, to finally get important events.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 59% of the total text.

Page 01 of 5

Method and System for Device Fault Pre-Warning By Correlation Analysis and Filters


1. Background (a) Importance

Security and reliability is very important to power grid.

More data available for utilities. There are many on-line monitoring system. CBM (Condition Based Maintenance) is more and morepopular in the industry. Information the utility could collect, including the environment and weather, device itself, and so on.

Benefit: To compare the status of the device in a time-line and between peer devices, recognize the fault sign, to warning the utility. Help to decide the best maintain opportunity. To implement the whole life-cycle management. Help to plan the resources of maintenance and save costs. To better emergency response. Predict the probability to prevent the fault event.

(b) Concept

   Fault in power grid is caused by device failure, environmental and bad weather (like thunder, rain, and typhoon), and external damage (people, animal, construction, vehicles).
(c) Traditional method

   To access the status of device, use a score system. It's very rough, and too dependent on the experience, without many quantityanalysis. Weather data and other data is not really used to predict the fault.

Main Idea

Based on all information the utility could collect, calculate the correlation of influence factor and the fault. We can triggeran event when another event happened had has significant correlation. Build an event filter system, to finally get important events.

Advantage and benefit

By event mechanism, utility can flexible define the time and devices concerned.

More acurrate than scoring system.

Events can be evaluated separately, which means a distribution computing could be used to improve efficiency. Device fault pre-warning can be noticed in time.

Steps
Step1: Analyse the influencing factors and classification.
Step 1.1: Define influencing factors and classification

Analyze all characteristics which are related to the fault of devices


Find all influencing factors of devices
Divide the influencing factors into two categories: Events and Attributes.

Step 1.2: Discretization the oroginal continuous values

Discretize all factors based on apriori knowledge or classification methods.

1



Page 02 of 5


(Y) Years of Age


(L) Load Rate

discrete

value range

discrete

value range

y1

Y<=2

r1

L<=50%

y2

2<Y<=12

r2

50%<L<=75%

y3

Y>12

r3

75%<L<=100%

r4

L>1...