Rotorcraft Structural Fault Detection via Virtual Load Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The impracticality of installing specialized sensors on vehicles to detect and isolate every possible structural fault condition due to high costs, weight, and maintenance challenges, as well as the limitations of conventional sensors for load measurement.
Innovation Solution
A structural diagnostic system that utilizes virtual monitoring of vehicle responses, leveraging real-time sensor data and empirical models to detect and isolate faults by compensating for normal variations in operating states, allowing a few load sensors to generate multiple sensed features sensitive to various faults.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If specialized sensors are installed to detect and isolate every structural fault condition, then fault detection capability is improved, but device complexity, cost, and weight increase
Solution Approach 1:
The patent applies universality by using a small number of multi-functional load sensors that can detect multiple types of structural faults through virtual monitoring. Instead of installing specialized sensors for each fault type, the system uses general-purpose load sensors combined with empirical models to detect various fault conditions including cracks, delamination, and structural degradation across multiple locations and modes.
Solution Approach 2:
The patent employs copying by creating virtual models (empirical models) that replicate the expected behavior of healthy structural components. These virtual models generate predicted load signatures that are compared against actual sensor measurements, allowing fault detection without physical inspection or specialized sensors. The virtual monitoring system copies normal operational patterns to identify deviations indicating faults.
2Device complexity
If virtual monitoring with few sensors is used, then device complexity is reduced, but measurement precision may deteriorate
Solution Approach 1:
The patent introduces an intermediary computational layer (empirical models and signal processing algorithms) that transforms limited sensor measurements into comprehensive fault detection capabilities. The empirical models act as intermediaries between the physical sensors and the fault detection logic, enabling precise fault identification by comparing predicted versus actual load responses through mathematical relationships rather than direct physical measurement at every potential fault location.
3Ease of manufacture
If conventional load sensors are used, then ease of manufacture is improved, but the ability to detect multiple fault types deteriorates
Solution Approach 1:
The patent applies parameter changes by utilizing multiple features extracted from load sensor signals (magnitude, frequency, phase, temporal patterns) to detect different fault types with the same physical sensor. By analyzing various parameters of the load response and comparing them against empirical models, the system achieves versatile fault detection capability using conventional sensors, transforming limited physical measurements into comprehensive diagnostic information.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system, method, and computer program for detecting and isolating structural faults is provided. To detect and isolate structural faults, a plurality of sensed features that corresponds to sensor data from the monitoring sub-system and a plurality of estimated features that corresponds to the plurality of sensed features can be generated. Further, the plurality of sensed features and the plurality of estimated features can be compared to produce a plurality of residuals. From the plurality of residuals, at least one structural fault within the vehicle are determined and isolated for further use of maintenance decision support.