Wind Turbine Component Health Monitoring via Residual Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current wind turbine monitoring systems require manual interpretation of threshold value exceedances, leading to inefficiencies in interpreting large-scale monitoring data and varying interpretations across different components, which can result in suboptimal maintenance planning and increased costs.

Innovation Solution

A system with an independent data processing environment that receives and processes data from wind turbines using component-specific monitoring algorithms to automatically establish a health value for components, incorporating both direct and indirect monitoring data, as well as failure rate inputs, to provide a uniform and actionable health status.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If manual interpretation of threshold value exceedances is used, then system complexity is reduced, but productivity and consistency of maintenance planning deteriorate

Engineering Contradiction:
Improvesystem complexityVSAvoidmaintenance planning efficiency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The monitoring system automatically calculates health values and generates maintenance recommendations without requiring manual interpretation. The system serves itself by processing monitoring data through algorithms that assess component health status and generate actionable insights, eliminating the need for human analysts to manually interpret threshold exceedances

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system transforms raw monitoring data into meaningful health values by applying mathematical models and algorithms. Instead of simply monitoring threshold exceedances, the system calculates composite health parameters that provide a unified assessment of component status, enabling automated decision-making

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If manual interpretation of monitoring data is used, then device complexity is reduced, but measurement precision and consistency of health assessment deteriorate

Engineering Contradiction:
Improveinterpretation system complexityVSAvoidhealth assessment consistency
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system automatically and consistently applies the same health assessment algorithms to all components, eliminating variability introduced by different human interpreters. The standardized automated process ensures that identical monitoring data always produces identical health assessments, improving measurement precision and consistency

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system transforms diverse monitoring parameters into a unified health value scale, enabling consistent comparison across different component types. By converting various sensor readings into standardized health assessments, the system achieves precision and uniformity in evaluating component condition

Inventive Principle:
Principle #35Parameter changes

3Reliability

If component-specific monitoring algorithms are implemented, then reliability of health assessment is improved, but device complexity increases

Engineering Contradiction:
Improvehealth assessment accuracyVSAvoidmonitoring system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The monitoring system is divided into component-specific modules, each with dedicated algorithms tailored to the unique characteristics of individual component types. This segmentation allows each algorithm to be optimized for its specific component while maintaining overall system manageability through modular architecture

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Despite being component-specific, the algorithms follow a universal framework and methodology for health assessment. The standardized approach to calculating health values across different components provides reliability and consistency, while the modular design keeps complexity manageable through reuse of common computational patterns

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3638900B1Independent monitoring system for a wind turbine
Publication Date: 2023.06.07 K B ELECTRONICS INC
  • EP3638900B1 patent drawingFigure 1
  • EP3638900B1 patent drawingFigure 2
  • EP3638900B1 patent drawingFigure 3

AI summary

The invention relates to a system for monitoring of wind turbine components comprising an independent data processing environment adapted to: receive a first category of data input related to operation of the wind turbine, process the received data input by one or more component specific monitoring algorithms adapted to establish an estimated component value related to a component to be monitored based on received first category data input having at least indirectly impact on the component, wherein the component specific monitoring algorithm is adapted to establish a component residual as the difference between the estimated component value and received first category of data input of the component to be monitored, and wherein the component specific monitoring algorithm furthermore is adapted to establish a component specific health value of the component to be monitored based on the established residual and put the health value at disposal for data processors outside the environment.