Smart Sensor System Multivariate Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current condition-based maintenance approaches for industrial assets rely on univariate techniques that fail to effectively detect anomalies in complex systems due to their inability to consider interactions and relationships between multiple sensor measurements, leading to delayed detection of equipment degradation and increased operational costs.
Innovation Solution
The implementation of a smart sensor system that utilizes advanced condition monitoring methods, including the Hotelling's T2 statistic, Rank Permutation Transformation, and Likelihood Ratio Test, to analyze time sequential values and identify changes in asset system operating conditions, providing alerts for deviations from normal operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If univariate techniques are used to monitor individual sensor measurements, then the monitoring system is simple to implement, but the system fails to detect anomalies in complex systems where interactions between multiple sensors are critical
Solution Approach 1:
The patent combines multiple sensor measurements into a unified multivariate monitoring framework. By merging individual sensor data streams and analyzing them collectively using statistical methods (Hotelling's T2 statistic, Rank Permutation Transformation, Likelihood Ratio Test), the system captures interactions between sensors while maintaining a coherent monitoring structure that improves anomaly detection without excessive complexity
Solution Approach 2:
The monitoring system is designed to perform multiple functions: it simultaneously detects anomalies, identifies faulty components, and provides diagnostic information. The multivariate approach serves as a universal framework that handles various types of equipment degradation and fault conditions across different industrial assets, making the system adaptable to complex monitoring needs
2Measurement precision
If multivariate techniques are used to analyze interactions between sensor measurements, then anomaly detection accuracy improves, but the complexity of the monitoring system increases
Solution Approach 1:
The patent transforms sensor measurements through mathematical transformations (Rank Permutation Transformation, Hotelling's T2 statistic calculation, Likelihood Ratio Test) that change the parameter representation of the data. These transformations convert complex multivariate relationships into statistically interpretable forms that improve detection accuracy while managing computational complexity through established statistical frameworks
Solution Approach 2:
The patent introduces statistical intermediaries (test statistics, transformation functions) that mediate between raw sensor measurements and anomaly detection decisions. These intermediary mathematical constructs simplify the analysis of complex sensor interactions by providing structured ways to evaluate multivariate relationships without requiring direct analysis of all possible sensor combinations
3Reliability
If equipment degradation is detected early using advanced monitoring, then maintenance costs are reduced and operational availability increases, but the requirement for sophisticated analysis methods increases system complexity
Solution Approach 1:
The patent implements preliminary action by establishing baseline statistical models of normal equipment operation before degradation occurs. The system pre-configures monitoring thresholds and analysis methods (Hotelling's T2, Rank Permutation Transformation, Likelihood Ratio Test) that enable early detection of deviations from normal operation, allowing maintenance to be scheduled before failures occur, thus improving reliability while managing complexity through proactive setup
Solution Approach 2:
The monitoring system incorporates feedback mechanisms that continuously compare current sensor readings against established baselines and provide alerts when anomalies are detected. This feedback loop enables real-time adjustment of maintenance schedules and operational decisions based on actual equipment condition, improving operational availability while using systematic feedback rules to manage analysis complexity
Data Source
AI summary
A method for advanced condition monitoring of an asset system includes monitoring a variable of an asset system using the at least one sensor of a smart sensor system; determining whether the asset system has departed from normal operation; and identifying the variable of the asset system indicating the departure from normal operation. In another method, the time sequential values of the monitored variable is analyzed by using a Rank Permutation Transformation test, a Hotelling's T2 statistic test, and a Likelihood Ratio Test; and a change of an operating condition of the asset system is determined using the analyzed values. An alert is provided if necessary. A smart sensor system includes an on-board processing unit for performing the method of the invention.