Location:Home > Engineering science > Power Engineering and Engineering Thermophysics > Thermal Power Engineering > Research on Wind Turbine Blade Defect Diagnose Based on Computer Vision

Research on Wind Turbine Blade Defect Diagnose Based on Computer Vision

Downloads: []
Tutor: ZhangYanPing
School: Huazhong University of Science and Technology
Course: Thermal Power Engineering
Keywords: wind turbine,fault diagnosis,fault monitoring,blade,software system
CLC: TM315
Type: Master's thesis
Year:  2013
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

With the advancement of technology, the demand for energy is increasing whiletraditional fossil fuel prices continue to rise and the pollution of the environment isincreasingly being perceived by people, resulting in wind energy utilization increasinglysubject to the attention of the whole society as a clean and renewable energy. The installedcapacity of the wind turbine is into a rising trend in recent years. With the increase in theinstalled capacity of wind machine, failure problems have become increasingly prominent,and gotten the constant attention of operators and researchers. The increase of wind turbineunit capacity and the environment in which the wind turbine blade which the main part ofthe wind turbine work makes the failures rising incidence, and the supervision of the windturbine blade failures and determine the fault species has a very important practicalsignificance.This article researched from the practical point of view into the determination of typesof wind turbine blade failures and fault location. Through the study of the Fault locationmethod of wind turbine blades, this paper proposes a novel method of removingbackground of the wind blade, and uses positioning ribbons and positioning points on bladezoning to extract fault characteristics and the growth of the fault. Study on the light andshooting angle of the wind turbine blades extract, this page got the applicable range of lightand shooting angle. When research in the wind turbine blade failures image featureextraction methods, an improved algorithm of Manifold learning was proposed£¬and thethree methods which are double-tree wavelet¡¢Manifold learning and improved algorithm ofManifold learning were used to extract information from blade images, then the extracteddata was used to train SVM to get the most useful one.Wind turbine fault monitoring system is designed on the basis of the wind vane faultmonitoring system which as a detailed description of the installation of the camera¡¢thesystem hardware¡¢software constitution. This system can improve the safety and reliabilityof the wind turbine blade state, ensure the operation stable of the wind turbine.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top