CPS Anomaly Detection Using Auxiliary Variables and Diagnostic Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing anomaly detection systems in cyber-physical systems (CPS) primarily rely on raw data and fail to utilize synthetic data derived from primary variables, which often carry more useful information about the state of equipment, leading to inefficiencies in anomaly detection.
Innovation Solution
The method involves generating diagnostic rules to calculate auxiliary CPS variables based on primary variables, using methods such as smoothing, polynomial approximations, machine learning models, and statistical transformations, to enhance anomaly detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If anomaly detection systems rely only on raw primary CPS variables, then the system complexity remains low, but the anomaly detection accuracy is insufficient
Solution Approach 1:
The patent introduces auxiliary CPS variables as intermediary elements that mediate between primary variables and anomaly detection. These auxiliary variables (such as smoothed values, derivatives, statistical characteristics) act as intermediate representations that enhance the information content available for anomaly detection without directly increasing system hardware complexity.
Solution Approach 2:
The system performs preliminary processing of primary CPS variables to generate auxiliary variables before anomaly detection occurs. Operations such as smoothing, differentiation, and statistical analysis are conducted in advance to prepare enhanced data representations, allowing the anomaly detection algorithm to work with pre-processed, information-rich variables.
2Reliability
If external monitoring systems are deployed at all nodes, then anomaly detection coverage is complete, but the cost and complexity of servicing increases significantly
Solution Approach 1:
The patent creates a universal anomaly detection methodology that can be applied across different CPS nodes and types. The approach of generating auxiliary variables from primary variables is universally applicable to various equipment and process types, allowing a single standardized system design to serve multiple functions and locations without requiring node-specific customization.
Solution Approach 2:
The system segments the anomaly detection function into two parts: (1) generation of auxiliary variables from primary variables, and (2) anomaly detection using the auxiliary variables. This segmentation allows the complex processing to be localized at each node while the overall system maintains comprehensive coverage through standardized implementation.
3Device complexity
If EPS is used for anomaly detection, then the system is simple and integrated with ACS TP, but decision making is sluggish and includes human factor
Solution Approach 1:
The patent replaces human-based decision-making in EPS with automated anomaly detection algorithms that process auxiliary CPS variables. The system uses computational methods (statistical analysis, pattern recognition) to automatically detect anomalies based on deviations in auxiliary variables, eliminating human response delays and subjectivity while maintaining system simplicity.
4Reliability
If redundancy for all MMI is implemented, then faultless operation is ensured, but the cost becomes extremely high and technically infeasible
Solution Approach 1:
The patent uses auxiliary variables as intermediary representations that aggregate information from multiple primary variables and MMI readings. By processing and combining data through auxiliary variables (such as moving averages, statistical characteristics), the system achieves robust anomaly detection that is tolerant to individual MMI failures without requiring redundant hardware for each sensor.
Data Source
AI summary
A method for determination of anomalies in a cyber-physical system (CPS) includes generating one or more diagnostic rules configured to calculate at least one auxiliary CPS variable. One or more values of the at least one auxiliary CPS variable are calculated for a predefined output interval of time based on collected values of a group of primary CPS variables for a predefined input interval of time based on the generated diagnostic rule. An anomaly is determined based on the collected values of the group of primary CPS variables and the one or more calculated values of the at least one auxiliary CPS variable.