Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

A Method for Health Condition Evaluation of Wind Turbines

IP.com Disclosure Number: IPCOM000244653D
Publication Date: 2016-Jan-05
Document File: 6 page(s) / 465K

Publishing Venue

The IP.com Prior Art Database

Related People

Yan Pei: AUTHOR [+3]

Abstract

The invention is a flexible wind turbine condition monitoring method using any data available. The WTCM method quantifies the wind turbine condition as feature vector using the correlation relationship between the available data. Normal feature vector dataset in which all the data is the feature vector of the wind turbine when the condition of the turbine is normal is generated firstly. Condition of a wind turbine is considered as abnormal if the characteristic of its feature is different from those in the normal feature vector data, and vice versa.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 35% of the total text.

Page 01 of 6

Method for Health Condition Evaluation of Wind Turbines

Abstract

The invention is a flexible wind turbine condition monitoring method using any data available. The WTCM method quantifies the wind turbine condition as feature vector using the correlation relationship between the available data. Normal feature vector dataset in which all the data is the feature vector of the wind turbine when the condition of the turbine is normal is generated firstly. Condition of a wind turbine is considered as abnormal if the characteristic of its feature is different from those in the normal feature vector data, and vice versa.

Background:

Wind power has become the most popular renewable energy promising to replace traditional pollutive thermal power generation because of its rich resources, mature technology, and zero emission. Global wind turbine installation had reached 318GW by end of 2013. With rapid installation increase of wind farms, expensive O&M (operation and maintenance) cost and downtime electricity sale loss develop to be more and more pressing issues. Taking a 2MW wind turbine as example, as evaluated by master thesis of KTH, about 248.4kUSD annual cost arises, incl. 242.7kUSD O&M cost and
5.7kUSD electricity sale loss.

Statement of the Problem:

Therefore, this invention aims to solve such a problem:

Wind turbines are complex electro-mechanical systems and are always located in places where the environment is very harsh, with rapidly changing temperature, air pressure and alternating load operation conditions. Thus, wind turbines are subjected to different kind of failures. These failures may cause great loss for wind farm owners and may pose danger for the grid. Furthermore, wind farms are often built in rural areas where the traffic is inconvenience, such as mountainous area or at sea. The inconvenience of traffic increase the O&M (operation & maintenance) cost. To reduce the O&M cost, WTCM (wind turbine condition monitoring) system was developed to help wind farm owners to realize condition based O&M.

On most wind turbines, many sensors are mounted to detect different signals related to the condition of the wind turbine. How to use these signals to determine the condition of a wind turbine is the problem our invention deals with.

Core idea:

WT is a complex system, and the signals detected by different sensors mounted on WT are related to each other. The correlation relationship between the detected signals is related to the operation condition of the WT. The change of WT condition would change the correlation relationship between the detected signals. Our invention uses the correlation relationship to determine whether a WT is normal or abnormal.

We reserve all rights in this document and in the information contained therein. Reproduction, use or disclosure to third parties without express authority is strictly forbidden. Ó 2014 ABB Ltd.


Page 02 of 6

Our invention comprises the following steps:


- Collect the available signals...