MPC Operator Support for Irregular Industrial Control Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Model predictive controllers (MPC) in industrial process control systems often require complex configuration and can malfunction during upset conditions, leading to operator uncertainty and loss of confidence, as they are configured as complex 'black box' technologies that do not react as expected in real-life scenarios.
Innovation Solution
A method and apparatus that utilize function blocks to obtain measurements of controlled and manipulated variables, detect irregular operating conditions, analyze case-specific conditions, and remove irregularities, providing operator support through heuristic approaches to address irregularities such as steady state value violations, oscillations, and excessive movements, thereby enhancing operator decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If MPC controllers are configured as complex black box technology, then adaptability and control capability are improved, but operator confidence and reliability deteriorate due to unexpected behavior during upset conditions
Solution Approach 1:
The patent introduces an intermediary system that sits between the MPC controller and the operator. This intermediary provides real-time monitoring, diagnostics, and explanation of controller behavior, translating the complex black box operations into understandable information for operators. This mediator restores operator confidence without reducing the adaptability of the MPC controller.
Solution Approach 2:
The patent implements comprehensive feedback mechanisms that provide operators with real-time information about controller status, performance, and decision-making processes. By feeding back explanatory information about why the controller is taking certain actions, especially during upset conditions, operators can understand and trust the system behavior, maintaining both adaptability and reliability.
2Reliability
If MPC applications are disabled or switched off during upset conditions, then operator confidence is maintained, but control capability and productivity are lost
Solution Approach 1:
The patent enables the MPC controller to self-diagnose and self-explain its behavior during upset conditions. By providing built-in monitoring and explanation capabilities, the controller serves itself by maintaining operator confidence through transparency, eliminating the need to shut down the controller while preserving productivity.
Solution Approach 2:
The intermediary system continuously monitors and explains controller behavior during upset conditions, maintaining operator confidence without requiring controller shutdown. This allows the MPC application to remain active and productive while operators trust the system through the intermediary's real-time explanations.
3Reliability
If detailed monitoring and analysis of irregular operating conditions is implemented, then operator support and reliability are improved, but system complexity and computational requirements increase
Solution Approach 1:
The patent segments the monitoring and analysis functions into modular components that can be independently implemented and managed. By dividing the complex monitoring task into separate functional blocks (detection, analysis, explanation generation), the system achieves detailed monitoring capability while keeping overall system complexity manageable through modular architecture.
Data Source
Figure 1
Figure 2
Figure 3~5
AI summary
A method includes obtaining (1002) measurements associated with a plurality of controlled variables (CVs) and a plurality of manipulated variables (MVs). The method also includes detecting (1004) an irregular operating condition (300) of a CV or an MV. The method also includes analyzing (1006) case specific operating conditions (400, 500, 600, 700, 800, 900) based on the irregular operating condition. The method further includes removing (1008) the irregular operating condition based on analyzing the case specific operating conditions.