Alarm Correlation Mapping for Industrial Alarm Flood Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial plants face challenges in identifying the root cause of alarm floods, which can lead to rapid propagation of alarms and increased operational risks due to the complexity of integrated systems, resulting in potential accidents and inefficiencies in managing abnormal process situations.
Innovation Solution
A method for monitoring industrial plants involves identifying abnormal episodes through alarm logs, extracting associated events, computing an alarm correlation map to describe correlations between these events, and classifying the episodes using the correlation map, thereby identifying the root cause of alarm floods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional alarm monitoring methods are used in complex industrial plants, then alarms can be displayed to operators, but alarm floods propagate quickly and become difficult to manage
Solution Approach 1:
The patent segments alarm floods into discrete abnormal episodes by identifying start and end times based on alarm rate thresholds. Each episode is treated as a separate analytical unit, allowing operators to manage individual episodes rather than overwhelming continuous alarm streams. This segmentation transforms the complex problem of alarm flood management into manageable discrete events.
Solution Approach 2:
The patent introduces an intermediary analysis layer that processes alarm data between the alarm generation system and the operator interface. This intermediary computes alarm correlation maps and generates abnormal event sequences, acting as a mediator that translates raw alarm data into meaningful patterns that operators can understand and act upon, reducing the complexity of direct alarm management.
2Loss of information
If alarm correlation analysis is performed on extracted abnormal events, then root cause identification is enabled, but computational requirements increase
Solution Approach 1:
The patent performs preliminary actions by pre-computing alarm correlation maps from extracted abnormal events before full analysis is needed. These correlation maps are stored and can be quickly referenced during actual alarm episodes, avoiding the need to perform computationally intensive correlation analysis in real-time. This preliminary computation reduces both information loss and computational energy requirements during critical operations.
Solution Approach 2:
The patent applies partial action by focusing computational resources on analyzing only the most relevant abnormal events within each episode rather than processing all alarm data equally. The system identifies key abnormal events that contribute most to root cause identification and performs detailed correlation analysis only on these selected events, reducing overall computational energy while maintaining effective root cause identification.
3Measurement precision
If multiple abnormal events are extracted from alarm logs, then comprehensive episode analysis is achieved, but analysis time increases
Solution Approach 1:
The patent changes parameters by transforming multiple abnormal events into a standardized alarm correlation map representation. This parameter transformation allows different types of abnormal events to be compared and analyzed using consistent metrics, enabling comprehensive analysis of multiple events without proportionally increasing analysis time. The standardized representation facilitates faster processing while maintaining measurement precision.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for monitoring an industrial plant is provided. The method includes: identifying an abnormal episode based on an alarm log (50); extracting abnormal events from the alarm log (50), the abnormal events being associated with the abnormal episode; computing from the extracted abnormal events an alarm correlation map describing correlations between the extracted abnormal events; and classifying the abnormal episode using the alarm correlation map.