Context-Aware Anomaly Detection for Technical Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data-based anomaly detection methods in technical systems often produce inaccurate results, leading to potential system shutdowns or failure to recognize anomalies, which can result in hazardous behavior.
Innovation Solution
A computer-implemented method that uses environment and system state vectors, combined with anomaly models to assess probabilities and conditional probabilities, enabling more accurate anomaly detection by considering environmental and system conditions, and allowing for self-improvement of the anomaly detection model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If data-based anomaly detection methods are used, then automated anomaly detection is enabled, but accuracy of anomaly detection deteriorates leading to false positives and negatives
Solution Approach 1:
The anomaly detection system is segmented into two independent models: an environment anomaly model that processes environment state vectors separately from a system anomaly model that processes system state vectors. This segmentation allows each model to specialize in its domain, improving overall detection accuracy while maintaining automation. The environment model assesses whether environmental conditions are anomalous, and the system model assesses whether system conditions are anomalous given the environment, creating a hierarchical detection structure that reduces false positives and negatives.
2Measurement precision
If environment state vector and system state vector are detected and processed through separate anomaly models, then accuracy of anomaly detection is improved, but device complexity increases
Solution Approach 1:
The environment state vector serves as an intermediary between the environment and the system anomaly model. The system anomaly model takes both the environment state vector and system state vector as inputs, using the environment vector as contextual information to better assess system anomalies. This intermediary approach allows the system to consider environmental context when detecting system anomalies, improving accuracy without requiring a completely complex rearchitecture of the detection system.
Data Source
AI summary
A computer-implemented method for detecting an anomaly in a technical system. The method includes detecting an environment state vector and a system state vector, the environment state vector including at least one first value which characterizes a physical environment condition or a physical operating condition of the technical system, and the system state vector including at least one second value which characterizes a physical condition of the technical system; ascertaining, using an environment anomaly model, an environment value which characterizes a probability or a probability density value with which the environment state vector occurs; ascertaining, using a system anomaly model, a system value which characterizes a conditional probability or a conditional probability density value with which the system state vector occurs if the environment state vector occurs; signaling the presence of an anomaly or signaling the absence of an anomaly based on the environment value and/or the system value.