Cyber-Physical System Anomaly Detection Using Auxiliary Variables

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing anomaly detection systems in cyber-physical systems (CPS) primarily focus on critical data and do not effectively utilize auxiliary variables derived from primary CPS variables, which can carry more informative synthetic data, leading to incomplete anomaly detection due to noise in raw data and lack of consideration for derivatives and integrated transformations.

Innovation Solution

The method generates diagnostic rules to calculate auxiliary CPS variables from primary variables, using techniques such as smoothing, trend analysis, and machine learning models to enhance anomaly detection by considering both primary and auxiliary variables, thereby improving the accuracy of anomaly identification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If anomaly detection systems only use primary critical data from ACS TP, then the system complexity remains low, but the measurement precision and reliability of anomaly detection deteriorates due to noise in raw data and lack of informative synthetic data

Engineering Contradiction:
Improveanomaly detection accuracyVSAvoiddata processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces auxiliary CPS variables as intermediary elements that mediate between primary CPS variables and anomaly detection. These auxiliary variables are calculated using diagnostic rules that process primary variables through smoothing, trend analysis, and other transformations. The intermediary auxiliary variables filter noise while preserving anomaly information, thereby improving detection accuracy without directly increasing system complexity

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system performs preliminary processing of primary CPS variables by calculating auxiliary variables in advance before anomaly detection. Diagnostic rules pre-process the raw data through smoothing, window analysis, and trend calculations, preparing refined data that reduces noise and enhances anomaly visibility. This preliminary action prevents the need for complex real-time processing during anomaly detection

Inventive Principle:
Principle #10Preliminary action

2Reliability

If external monitoring systems are deployed at all critical nodes, then the coverage and reliability of monitoring improves, but the cost and device complexity increases significantly

Engineering Contradiction:
Improvemonitoring reliabilityVSAvoidsystem deployment complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates a universal anomaly detection system that integrates with the existing ACS TP and can monitor the entire enterprise through a single centralized platform. Instead of deploying separate external monitoring systems at each critical node, the unified system processes data from all CPS variables enterprise-wide, providing comprehensive monitoring coverage with a single multi-functional system that reduces overall deployment complexity

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent merges the anomaly detection functionality with the existing automated control system for technological processes (ACS TP). By combining monitoring and control functions into a unified system that shares data infrastructure and processing resources, the patent achieves enterprise-wide monitoring coverage without the need for separate external monitoring systems at each node, thereby reducing deployment complexity while maintaining reliability

Inventive Principle:
Principle #5Merging (Combining)

3Ease of operation

If emergency protection systems are used with simple architecture, then the ease of operation and response time improves, but the measurement precision deteriorates because EPS assumes faultless monitoring instruments which is not always true in practice

Engineering Contradiction:
Improvesystem operation simplicityVSAvoidanomaly detection accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The system performs preliminary processing of monitoring data by calculating auxiliary CPS variables that are smoothed and filtered versions of primary variables. This pre-processing prepares refined data that is more robust to instrument failures and noise, allowing the simple EPS architecture to operate with higher measurement precision without requiring complex redundant instrumentation

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces auxiliary variables as intermediary data elements that mediate between potentially faulty primary monitoring data and the emergency protection system decisions. These intermediary variables are calculated using diagnostic rules that account for instrument reliability issues, providing more accurate input to the EPS while maintaining its simple architecture and ease of operation

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4246888A1System and method for determination of anomalies in a cyber-physical system
Publication Date: 2023.09.20 AO KASPERSKY LAB
  • EP4246888A1 patent drawingFigure 1a
  • EP4246888A1 patent drawingFigure 1b
  • EP4246888A1 patent drawingFigure 1c

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.