Composite Regression Model for Process Deviation Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current process control systems in plants often operate in abnormal states due to undetected issues, leading to suboptimal performance and significant costs, as existing diagnostic tools primarily react to problems after they occur, rather than preventing them.

Innovation Solution

A method and system that utilize a composite model with multiple regression models to predict and prevent abnormal operations by collecting and analyzing data from process variables, generating a new model to replace existing ones when deviations are detected, and using a deviation detector to indicate significant deviations from predicted behavior.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional diagnostic tools are used to detect process abnormalities, then problems can be identified after they occur, but the system cannot prevent abnormal operations and incurs significant downtime and equipment damage costs

Engineering Contradiction:

Improveprocess operation reliabilityVSAvoiddowntime

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by continuously collecting process data and updating regression models to establish baseline normal operation patterns before abnormalities occur. The deviation detector proactively compares real-time data against these models to identify potential issues before they escalate into actual abnormalities, enabling preventive rather than reactive maintenance and eliminating downtime associated with unexpected failures

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback by continuously monitoring process variables, comparing them against predicted values from regression models, and generating alerts when deviations exceed thresholds. This closed-loop feedback mechanism enables real-time detection of abnormal trends and allows operators to take corrective actions before equipment damage occurs, significantly improving reliability and reducing downtime

Inventive Principle:

Principle #23Feedback

2Measurement precision

If multiple regression models are used to cover different operating regions, then prediction accuracy across varying conditions is improved, but model complexity and computational requirements increase

Engineering Contradiction:

Improveprediction accuracyVSAvoidmodel complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system segments the process operating space into multiple distinct operating regions, with each region having its own specialized regression model. This segmentation allows each model to be optimized for specific operating conditions, improving prediction accuracy within each region while maintaining manageable model complexity through localized rather than global modeling approaches

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system dynamically selects and switches between different regression models based on current operating conditions. The composite model structure allows automatic transition between regional models as process variables change, maintaining high prediction accuracy across varying conditions while avoiding the need for a single overly complex global model

Inventive Principle:

Principle #15Dynamics

3Reliability

If a composite model with multiple regional regression models is implemented, then detection of abnormal operations across different operating conditions is improved, but the difficulty of detecting and measuring deviations increases

Engineering Contradiction:

Improveabnormal operation detectionVSAvoiddeviation detection difficulty

Core Design Contradiction:

ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The system introduces an intermediary composite model structure that bridges multiple regional regression models and provides a unified interface for deviation detection. This intermediary layer handles the complexity of switching between regional models and presents a consistent deviation measurement to the detection algorithm, simplifying the overall detection process while maintaining high reliability across different operating conditions

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS8032341B2Modeling a process using a composite model comprising a plurality of regression models

Publication Date: 2011.10.04 FISHER ROSEMOUNT SYST INC

AI summary

A system for detecting abnormal operation of at least a portion of a process plant includes a composite model for modeling at least the portion of the process plant. The model may be configurable to include multiple regression models corresponding to multiple different operating regions of the portion of the process plant. A new model may be generated from two or more of the regression models, and the composite model may be revised to replace the two or more regression models with the new model. The system may also include a deviation detector configured to determine if the actual operation of the portion of the process plant deviates significantly from the operation predicted by the composite model. If there is a significant deviation, this may indicate an abnormal operation.