AI Multivariable Control for Threshold-Stable Industrial Facilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current industrial control systems, particularly in complex facilities like nuclear power plants, rely on one-to-one relationships between sensors and actuators, which are inadequate in managing multi-variant interactions and maintaining optimal monitored variables during changes in uncontrolled variables, leading to inefficiencies and potential threshold violations.
Innovation Solution
Implementing a multi-variant control system utilizing artificial intelligence and machine learning to identify and optimize relationships between controlled and uncontrolled variables, allowing for more intelligent control strategies that minimize variations in monitored variables while achieving desired targets without violating operational thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a conventional one-to-one control system is used between sensors and actuators, then the system structure is simple, but the system cannot effectively manage multi-variant interactions and maintain optimal monitored variables during changes in uncontrolled variables
Solution Approach 1:
The patent implements a multi-variant control system where a single control framework can manage multiple controlled variables and their interactions simultaneously. The system uses AI/ML models to handle various types of relationships (direct and indirect) between controlled and monitored variables, making the control system universally applicable to complex multi-variant scenarios rather than requiring separate one-to-one control loops for each variable.
Solution Approach 2:
The patent introduces AI and machine learning models as intermediary components between sensors and actuators. These intermediary models analyze the behavior of distributed control systems, identify relationships between variables, and determine optimization strategies, thereby enabling sophisticated multi-variant control without requiring direct complex wiring or control logic between each sensor-actuator pair.
2Reliability
If a multi-variant control system utilizing AI and ML is implemented, then the system can effectively optimize controlled variables and maintain monitored variables within thresholds, but the device complexity increases
Solution Approach 1:
The patent implements a feedback mechanism where the AI/ML models continuously monitor the behavior of the distributed control system and the relationships between controlled and monitored variables. Based on this feedback, the system dynamically adjusts control strategies to maintain monitored variables within predetermined thresholds, thereby improving reliability through data-driven adaptive control.
Solution Approach 2:
The control system utilizes machine learning models that automatically learn and identify relationships between variables without requiring manual programming or intervention. The system self-optimizes by training models on historical data and autonomously determining control variable optimizations, reducing the need for complex manual control configuration while maintaining high reliability.
3Productivity
If conventional control systems are used, then the device complexity is low, but the productivity and operational efficiency are reduced due to inability to optimize controlled variables effectively
Solution Approach 1:
The patent employs machine learning models that are trained in advance on historical operational data to learn the behavior of the distributed control system and identify relationships between controlled and monitored variables. This preliminary training enables the system to make informed real-time control decisions that optimize productivity, rather than relying on simple reactive control loops.
Solution Approach 2:
The system dynamically changes control parameters by optimizing controlled variables based on identified relationships and current operational conditions. The AI/ML models determine optimal setpoints and adjustment strategies for controlled variables, enabling the system to adapt parameters in real-time to maximize productivity and operational efficiency while maintaining reliability.
Data Source
AI summary
The present disclosure is directed to a multi-variant control system for an industrial facility. In one form of a method, a processor monitors a behavior of a distributed control system with respect to at least a set of uncontrolled variables, a set of controlled variables, and a set of monitored variables; trains a model based on the monitored behavior utilizing at least one of artificial intelligence or machine learning; and identifies, based on the model, a subset of controlled variables for optimization. Further, the processor monitors a behavior of the industrial facility without the use of the distributed control system; optimizes, based on the monitored behavior, the subset of controlled variables to obtain a target result and restrict variation of values of the set of monitored values; and executes a control system for the industrial facility based on the optimization of the subset of controlled variables.