Inspection based on convolutional neural networks of defects in wind turbine blades

ES3078200T3Undetermined Publication Date: 2026-09-09SIEMENS GAMESA RENEWABLE ENERGY AS
View PDF -1 Cites 0 Cited by

Patent Information

Application Number
ES2019752931T
Authority / Receiving Office
ES · ES
Patent Type
Patents
Current Assignee / Owner
Priority Date
2018-08-03
Filing Date
2019-07-23
Publication Date
2026-09-09
Estimated Expiration
2039-07-23

AI Technical Summary

Technical Problem

Manual inspection of wind turbine blades for defects is time-consuming and not cost-efficient, with poor detection accuracy.

Method used

A method using convolutional neural networks (CNNs) for automated blade defect detection and localization, involving two stages: global image segmentation and localized refinement, utilizing trained models for blade outline and defect classification.

Benefits of technology

Enables rapid, accurate, and cost-effective pixel-level blade defect determination without requiring skilled annotators, reducing annotation costs and improving detection quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000009_0000
    Figure 00000009_0000
  • Figure 00000009_0001
    Figure 00000009_0001
  • Figure 00000009_0002
    Figure 00000009_0002
Patent Text Reader

Abstract

A computer system (CS) automatically executes a computerized method for determining blade defects. In step S1), an interface (IF) of the computer system (CS) receives an image (O1) of a wind turbine containing at least a portion of one or more blades. The image has a specified number of pixels in height and width. Step S2) consists of two consecutive steps, S2a) and S2b), executed by a processing unit (PU) of the computer system (CS). In step S2a), the image (O1) is analyzed to determine the blade outline. In step S2b), a modified image (AI) is created from the analyzed image (O1), containing only the blade information. Finally, step S3) consists of the processing unit (PU) analyzing the modified image (AI) to determine a blade defect (BD) and / or a blade defect type (BDT).As a result, the processing unit (PU) generates blade defects (BD) and / or blade defect types (BDT).
Need to check novelty before this filing date? Find Prior Art