Parallel Equipment Anomaly Prediction Using Correlation Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current predictive maintenance methods for industrial equipment are not sufficiently anticipatory, leading to unnecessary replacements and high false alarm rates, as they lack sensitivity and reliability in detecting impending failures in rotating machinery operating in parallel.
Innovation Solution
A method that records and processes measurements of operating parameters over time for redundant equipment, using coefficients of determination and linear regression to detect malfunctions by establishing thresholds and evaluating noise levels, thereby reducing false declarations and improving predictive accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current predictive maintenance methods are used to detect equipment failures, then maintenance can be scheduled in advance, but the sensitivity and reliability are insufficient leading to high false alarm rates and unnecessary replacements
Solution Approach 1:
The patent segments the fault detection process into multiple independent analysis stages: individual parameter monitoring, correlation analysis between parameters, and hierarchical decision-making levels. This segmentation allows each stage to focus on specific aspects of equipment health, reducing false alarms while maintaining sensitivity to actual faults.
Solution Approach 2:
The patent introduces correlation coefficients as intermediary metrics that mediate between raw operating parameters and failure predictions. By analyzing the correlation between parameters from different equipment operating in parallel, the system identifies subtle patterns that indicate developing faults without triggering false alarms from normal variations.
2Reliability
If equipment is monitored using traditional methods, then operational continuity is maintained, but faulty equipment may malfunction long before detection and the installation does not function optimally
Solution Approach 1:
The patent implements preliminary action by continuously monitoring correlation patterns between parameters of parallel equipment and establishing baseline correlations during normal operation. This allows the system to detect deviations from normal correlation patterns before they indicate actual failures, enabling proactive maintenance scheduling.
Solution Approach 2:
The system employs feedback mechanisms where correlation analysis results continuously inform the monitoring process. When correlation deviations are detected, the system adjusts monitoring sensitivity and triggers diagnostic procedures, creating a feedback loop that accelerates failure detection while maintaining operational reliability.
3Reliability
If preventive maintenance is performed according to fixed schedules, then equipment protection is ensured, but equipment functioning normally may be replaced unnecessarily
Solution Approach 1:
The patent replaces static fixed-schedule maintenance with dynamic condition-based maintenance. The system continuously adapts monitoring thresholds and maintenance schedules based on real-time correlation analysis of equipment parameters, allowing maintenance to be performed only when actual degradation patterns indicate necessity, thereby avoiding unnecessary replacements of healthy equipment.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A method for predicting an anomaly in the operation of devices, comprising the steps of taking a first device and a second device operating in parallel and in substantially the same way, each device comprising an operating parameter varying with time in a similar way between the first and the second devices; of collecting and of recording measurements of the parameter for the first and the second devices; of processing the collected measurements so as to detect a possible malfunctioning of one of the devices by establishing a determination coefficient between the measurements of the parameter for the first device and the measurements of the same parameter for the second device; of indicating the malfunctioning of one of the devices if the determination coefficient is below a threshold; and of indicating an absence of malfunctioning of one of the devices and/or of adjusting the threshold according to the coefficient if the latter is equal to or above the threshold.