Rotorcraft Rotor Defect Detection Using Neural Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing rotorcraft rotor defects and adjustment errors are incomplete and imperfect, often failing to detect real defects and generating false indications, requiring lengthy expert intervention and being inefficient in reducing inspection costs.
Innovation Solution
The use of an Artificial Neural Network (ANN) with supervised competitive learning (SCLN) to analyze vibration data, identifying defective blades and adjustment errors by distinguishing between modulus and phase data, and employing a competitive layer of neurons with iterative weight modification algorithms like LVQ and SOM for accurate classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional automatic analysis tools are used to diagnose rotor defects, then analysis speed is improved, but detection reliability deteriorates due to incomplete detection and false positives
Solution Approach 1:
The diagnosis system is segmented into multiple specialized neural networks: a first SCLN for detecting presence/absence of defects and adjustment errors, and a second SCLN for locating specific defective components. This segmentation allows each network to specialize in specific diagnostic tasks, improving both speed and reliability compared to single traditional analysis tools.
Solution Approach 2:
The patent introduces vibration signature analysis as an intermediary step between raw accelerometer data and defect diagnosis. The system transforms time-domain vibration signals into frequency-domain signatures, which serve as intermediate representations that enhance detection accuracy while maintaining automated processing speed.
2Reliability
If expert intervention is used to analyze vibration data, then detection reliability is improved, but analysis time increases significantly
Solution Approach 1:
The neural network system performs self-service by automatically analyzing vibration signatures and generating diagnostic results without requiring expert intervention. The SCLNs autonomously process vibration data, detect defects, identify adjustment errors, and locate problematic components, eliminating the time-consuming manual analysis while maintaining high reliability through the competitive learning mechanism.
Solution Approach 2:
The system incorporates feedback mechanisms where the neural networks continuously learn from vibration pattern recognition results. The competitive learning process provides feedback that refines the detection and localization capabilities, enabling the system to achieve expert-level reliability through automated iterative improvement rather than manual expert review.
Data Source
AI summary
The invention relates to a method of detecting and identifying a defect or an adjustment error of a rotorcraft rotor using an artificial neural network (ANN), the rotor having a plurality of blades and a plurality of adjustment members associated with each blade; the network (ANN) is a supervised competitive learning network (SSON, SCLN, SSOM) having an input to which vibration spectral data measured on the rotorcraft is applied, the network outputting data representative of which rotor blade presents a defect or an adjustment error or data representative of no defect, and where appropriate data representative of the type of defect that has been detected.