Piecewise Linear Regression for Real-Time Equipment Life Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for monitoring and predicting the end-of-life of electronic equipment, such as those in aerospace, struggle with real-time diagnostics and prognostics due to high computational resource requirements, making it difficult to embed processing units on-board and providing timely alerts for maintenance.
Innovation Solution
A method that embeds software for monitoring and prognostics within the system to be monitored, using piecewise linear regression based on environmental measurements, allowing for real-time calculation of damage and life expectancy, with the ability to create new pieces based on linear correlation coefficients and trigger alarms before critical thresholds are exceeded.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a processing unit is used to analyze recorded values and provide system diagnosis, then diagnostic accuracy is improved, but the computing and memory resource requirements become too high to embed the processing unit in the monitored system
Solution Approach 1:
The patent segments the monitoring system into two parts: a simple embedded monitoring device that collects environmental data and stores it in memory, and a separate processing unit that performs the complex diagnostic analysis. This segmentation allows the embedded device to remain lightweight while still enabling accurate diagnostics through the external processing unit.
Solution Approach 2:
The patent introduces an intermediary approach where environmental data is collected and stored in a standardized format by the embedded monitoring device, then transferred to an external processing unit for analysis. This intermediary data storage and transfer mechanism decouples the resource-constrained embedded device from the computationally intensive diagnostic functions.
2Loss of time
If data is transferred to the processing unit only after a complete profile is obtained (e.g., after 350 flight hours), then data transfer frequency is reduced, but the time lag before diagnosis results are available becomes unacceptably long
Solution Approach 1:
The patent implements dynamic data transfer where the monitoring device can transfer data to the processing unit at multiple points: when a complete profile is obtained, but also earlier when intermediate results are available. This dynamic approach allows for more frequent updates without requiring transfer of complete profiles every time, balancing timeliness with transfer efficiency.
Solution Approach 2:
The patent enables preliminary diagnostic analysis to be performed on intermediate data before a complete profile is obtained. The processing unit can provide preliminary diagnosis results based on partial data, allowing maintenance planning to begin earlier while the complete profile is still being collected.
3Reliability
If the processing unit analyzes all recorded values to extract remaining lifespan information, then diagnostic completeness is improved, but the time required to obtain results extends to several days
Solution Approach 1:
The patent applies partial action by allowing the processing unit to analyze a subset of the recorded data to provide preliminary or approximate diagnostic results quickly. This partial analysis can be performed on key parameters or representative samples, providing timely insights without requiring complete analysis of all environmental data points.
Solution Approach 2:
The patent segments the diagnostic analysis into multiple stages: initial quick analysis of critical parameters, intermediate analysis as more data becomes available, and final comprehensive analysis when the complete profile is obtained. This staged approach provides progressively more accurate results over time, balancing speed and completeness.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
According to the invention, to determine an operating forecast for a system, measurements are taken of an environmental value in an environment surrounding the system. Then these measurements are processed in a processing unit in order to determine how much of the system's lifetime has been consumed. Said consumed lifetime is based on a history of the system in the environment. This is used to calculate a forecast of a good operating lifetime. According to the invention, this forecast determination is made by adaptive linear regression. Thus, all forecast calculations are simplified and can be performed by the system itself and in real-time.