Industrial Process Control System with Expert System Override
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional control systems for industrial processes struggle to react appropriately to extreme and sudden changes in operating conditions, often moving outside their designed operating range, leading to instability and prolonged recovery times.
Innovation Solution
A control system that integrates an expert system based on fuzzy logic or machine learning algorithms, such as neuronal networks or decision trees, to select manipulated variables or output values, overriding conventional controllers to stabilize processes by triggering specific actions or sequences of actions, thereby maintaining operations within the designed range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a conventional controller is used for controlling an industrial process, then the controller operates properly within its designed operating range, but it cannot handle extreme changes of conditions that move it outside its defined operating range
Solution Approach 1:
An expert system acts as an intermediary between the process and the conventional controller. The expert system monitors process variables and detects when extreme conditions occur, then overrides the conventional controller's output with pre-determined manipulated variables designed to handle these specific extreme situations, allowing the system to adapt to conditions outside the conventional controller's designed range while maintaining reliability through the expert system's specialized handling
Solution Approach 2:
The control system dynamically switches between conventional controller operation and expert system override based on real-time process conditions. The selector module enables this dynamic transition, allowing the system to adapt its control strategy according to whether conditions are within or outside the conventional controller's operating range
2Adaptability or versatility
If the conventional controller moves outside its defined operating range, then it cannot provide appropriate control outputs, but switching to override control requires additional system complexity
Solution Approach 1:
The control system is segmented into distinct functional modules: the conventional controller for normal operation, the expert system for extreme condition handling, and a selector module for switching between them. This segmentation allows each component to be optimized for its specific function while working together as an integrated system, managing complexity through modular architecture
Solution Approach 2:
The expert system contains pre-programmed knowledge and rules for handling extreme conditions, allowing it to autonomously determine appropriate control actions without requiring complex real-time calculations or additional sensors. The system serves itself by having the expert system's database of extreme condition responses automatically activated when needed
Data Source
Figure 1~2
Figure 3
AI summary
For controlling an industrial process (12), a conventional controller (33), particularly a controller based on Model Predictive Control, is configured to calculate manipulated variables (uC) for controlling the process (12), based on process variables (y). The controller (33) is combined with an expert system (31), which determines output values (u1, u2, uN) for controlling the process (12), and a selector (32), which selects either the manipulated variables (uC) or the output values (u1, u2, uN) for controlling the process (12), based on operating conditions of the process. Using an expert system (31) to interrupt a conventional controller (33) from controlling an industrial process (12) has the advantage that a much broader range of operating conditions and emergency situations can be handled than with the conventional controller (33), and a wide range of predefined actions can be triggered for returning operations to a designed operating range of the conventional controller (33).