Embedded Multivariable Predictive Controller Simulator
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional simulators for multivariable predictive controllers in industrial process control systems fall short in accurately simulating real-time industrial processes, particularly in matching simulated and actual process models, which hinders validation and design optimization.
Innovation Solution
A multivariable predictive controller simulator that runs on an embedded platform, allowing seamless integration of offline process simulations with embedded controllers, enabling validation and memory optimization recommendations, and facilitating switching between simulation and real-time control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional simulators are used for multivariable predictive controllers, then simulation capability is provided, but accuracy in matching simulated and actual process models deteriorates
Solution Approach 1:
The patent creates a virtual copy of the embedded controller that replicates its exact control logic, algorithms, and parameters. This virtual controller runs in the simulator and produces identical control outputs to the real controller for given process conditions, enabling accurate validation without physical hardware.
Solution Approach 2:
The controller design, including control algorithms, tuning parameters, and logic, is fully developed and tested in the offline simulator before being deployed to the actual embedded controller. This preliminary validation ensures the controller will perform correctly when deployed, catching issues early in the design phase.
2Ease of manufacture
If offline simulation is implemented for controller validation, then design validation capability is improved, but integration with embedded controller execution deteriorates
Solution Approach 1:
The simulator is designed to serve multiple functions: offline validation of controller logic, performance prediction, operator training, and commissioning support. The same software platform can operate in different modes (simulation vs. real-time monitoring) depending on needs.
Solution Approach 2:
The patent introduces an intermediary layer that connects the offline simulator with the real embedded controller through standardized communication protocols. This intermediary enables data exchange and synchronization while maintaining independence of each system.
3Speed
If embedded platform computational resources are utilized, then execution speed is improved, but memory constraints worsen
Solution Approach 1:
The controller implements selective computation where not all control algorithms run at full complexity continuously. Less critical computations are reduced or skipped based on current process conditions, while critical control functions maintain full accuracy. This partial computation approach reduces memory and processing demands.
Solution Approach 2:
The controller dynamically adjusts operational parameters such as control cycle frequency, prediction horizon length, and model complexity based on available computational resources and current process urgency. During high-load conditions, parameters are adjusted to reduce computational burden while maintaining essential control functionality.
Data Source
AI summary
A method includes retrieving process variable values from a multivariable predictive controller. The multivariable predictive controller is executed on an embedded platform and is configured to control an industrial process. The method also includes simulating how the multivariable predictive controller would attempt to control the industrial process based on the process variable values. The method further includes transmitting simulated process variable values to the multivariable predictive controller.