Model-Based Controller Variability Detection at Constraint Limits

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Model-based industrial process controllers experience a decline in performance over time due to factors like inaccurate models, misconfiguration, and operator actions, leading to increased variability and control giveaway, which reduces their effectiveness by up to fifty percent.

Innovation Solution

An apparatus and method that identify impacts and causes of variability or control giveaway by analyzing data from industrial process controllers, calculating standard deviations, and determining control giveaway values to generate graphical displays showing the effects on model-based controller performance, allowing for early detection and reduction of performance losses.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If model-based controllers are used to control industrial processes, then controller performance and precision are improved, but performance declines over time due to model inaccuracy, misconfiguration, and operator actions

Engineering Contradiction:
Improvecontroller performanceVSAvoidperformance stability over time
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The system performs preliminary actions by continuously monitoring controlled variables and calculating control giveaway values before significant performance degradation occurs. This enables early detection of model inaccuracy, misconfiguration, or operator actions that may lead to performance decline, allowing corrective measures to be taken proactively rather than reactively.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by continuously evaluating control giveaway values and comparing them against thresholds or historical data. This feedback loop enables the system to detect performance degradation trends and trigger alerts or corrective actions, maintaining reliability over time despite the natural decline in model-based controller performance.

Inventive Principle:
Principle #23Feedback

2Loss of time

If control giveaway values are calculated frequently to detect performance decline, then early detection capability is improved, but computational resources and processing time are consumed

Engineering Contradiction:
Improvedetection timeVSAvoidcomputational resource consumption
Core Design Contradiction:
Loss of timeVSUse of energy by moving object

Solution Approach 1:

The system applies partial action by calculating control giveaway values selectively based on identified periods when controlled variables are near their limits, rather than continuously computing for all time periods. This approach provides sufficient detection capability while reducing unnecessary computational overhead and energy consumption.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system changes parameters by adapting the frequency and intensity of control giveaway calculations based on process conditions. When controlled variables are close to limits, the system increases monitoring intensity; when they are well within acceptable ranges, monitoring intensity is reduced, optimizing the balance between detection speed and resource consumption.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3639098B1Apparatus and method for identifying impacts and causes of variability or control giveaway on model-based controller performance
Publication Date: 2022.10.19 HONEYWELL INTERNATIONAL INC
  • EP3639098B1 patent drawingFigure 1~2
  • EP3639098B1 patent drawingFigure 3
  • EP3639098B1 patent drawingFigure 4

AI summary

A method includes obtaining (504) data identifying values of one or more controlled variables associated with an industrial process controller (106). The method also includes identifying (506) periods when at least one of the one or more controlled variables has been moved to an associated limit by the controller. The method further includes, for each identified period, (i) identifying (510) a standard deviation of predicted values for the associated controlled variable and (ii) determining (512) a control giveaway value for the associated controlled variable based on the standard deviation. The control giveaway value is associated with an offset between the associated controlled variable's average value and the associated limit. In addition, the method includes identifying (516) variances in the one or more controlled variables using the control giveaway values and generating (520) a graphical display (1100, 1200, 1300) identifying one or more impacts or causes for at least some of the variances.