Component Health Monitoring Using Dynamic Peak Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional health monitoring systems face challenges in accurately determining the health state of components due to noisy signals and varying degradation trends, leading to issues with setting proper thresholds for detection, resulting in either premature or missed notifications.
Innovation Solution
A health monitoring system that uses a sensor, such as an accelerometer, to generate status indicators, which are compared to a dynamic threshold calculated as the mean value of historical data plus a configurable standard deviation, with peak counts exceeding a pre-determined threshold triggering health notifications and maintenance actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional threshold-based monitoring is used, then the system is simple to operate, but it leads to premature or missed detections due to noisy signals and varying degradation trends
Solution Approach 1:
The patent segments the continuous degradation signal into discrete degradation events by identifying peaks that exceed the dynamic threshold. Each peak represents a distinct degradation event, allowing the system to count individual events rather than analyzing continuous noisy signals. This segmentation approach improves detection accuracy while maintaining manageable system complexity.
Solution Approach 2:
The patent implements a dynamic threshold that adapts to varying degradation trends and noise levels. The threshold is calculated as mean + n*standard_deviation, where both mean and standard deviation are updated based on historical data. This dynamic adaptation allows the system to maintain high detection accuracy across different operating conditions without requiring complex manual tuning.
2Reliability
If a fixed threshold is used for health monitoring, then the analysis method is simple, but it results in too-early detections when set low or missed detections when set high
Solution Approach 1:
The system uses feedback from historical status indicator data to continuously update the dynamic threshold. By calculating the mean and standard deviation from historical data and adjusting the threshold accordingly, the system automatically adapts to changing conditions. This feedback mechanism improves detection reliability while maintaining ease of operation, as the system self-adjusts without requiring complex manual intervention.
Solution Approach 2:
The patent changes the threshold parameter from a fixed value to a dynamic value that varies with operating conditions. The threshold is expressed as mean + n*standard_deviation, where n is a configurable multiplier. This parameter transformation allows the threshold to adapt to varying noise levels and degradation rates, improving reliability while keeping the configuration simple through the single parameter n.
3Adaptability or versatility
If trend analysis of smoothed features is used, then the monitoring approach is conventional and simple, but it fails to accurately detect health states for components with non-monotonic degradation patterns
Solution Approach 1:
Instead of analyzing the continuous trend of smoothed features, the patent inverts the approach by detecting discrete peaks that deviate from the trend. By identifying local maxima that exceed the dynamic threshold, the system can detect degradation events regardless of whether the overall trend is monotonic or non-monotonic. This inversion enables the system to handle diverse component degradation patterns while maintaining measurement precision.
Solution Approach 2:
The patent transitions from analyzing the one-dimensional continuous trend to detecting zero-crossing events in the derivative or peak events above a threshold. By changing the dimension of analysis from continuous magnitude to discrete event counting, the system becomes adaptable to various degradation patterns while preserving detection accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A health monitoring system (100) includes at least one sensor (102). A health monitoring unit (HMU) (104) is operatively connected to the at least one sensor to receive data therefrom and to generate at least one status indicator (SI) based on the data from the at least one sensor. A computing device (107) is operatively connected to the HMU to receive data therefrom and is configured to generate a health notification when a quantity of the SI that are greater than or equal to a pre-determined SI threshold exceed a pre-determined peak count threshold. A method for monitoring and determining the health of a component includes collecting data during a sample cycle from at least one sensor to generate a numerical SI for a given component with a computing device. The method includes identifying if the numerical SI qualifies as a peak SI, which is greater or equal to a pre-determined SI threshold.