MIMO MPC Function Block Segmentation for Industrial Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial process control and automation systems face challenges in efficiently implementing advanced control algorithms, particularly in providing optimal process control and robustness due to limitations in existing regulatory control methods like PID, which lack the capability to handle multiple inputs and outputs effectively.
Innovation Solution
The implementation of a MIMO MPC algorithm divided into three distinct function blocks: an input function block for data collection, an algorithm function block for executing the MPC algorithm, and an output function block for managing control signals, utilizing a shared memory space for data exchange and executing these blocks in foreground and background tasks to optimize processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If PID regulatory control methods are used, then the system is simple to implement, but the capability to handle multiple inputs and outputs effectively is limited
Solution Approach 1:
The control system is segmented into distinct functional blocks: input function blocks for data collection and preprocessing, algorithm function blocks for executing MIMO MPC calculations, and output function blocks for managing control signals. This segmentation enables handling of multiple inputs and outputs while maintaining organizational simplicity.
Solution Approach 2:
The function blocks are designed with universal interfaces and standardized data structures that can handle various types of inputs and outputs. The MIMO MPC algorithm itself provides multi-functionality by simultaneously processing multiple process variables and control actions across different process units.
2Speed
If intensive MIMO MPC calculations are performed in foreground tasks, then real-time control responsiveness is improved, but system robustness and availability deteriorate due to computational load
Solution Approach 1:
The computational tasks are segmented between foreground and background execution contexts. Foreground tasks handle real-time control responsiveness with simplified calculations, while background tasks perform intensive MIMO MPC computations. This segmentation ensures that real-time responsiveness is maintained while robustness is preserved through proper task scheduling.
Solution Approach 2:
Input data preprocessing is performed in advance in foreground tasks, preparing data structures for subsequent background calculations. This preliminary action reduces the computational burden during critical real-time control cycles and ensures that intensive calculations have optimized input data available when executed.
3Adaptability or versatility
If a unified control algorithm structure is used, then implementation is straightforward, but extensibility for alternative control algorithms is limited
Solution Approach 1:
The control system is organized into modular function blocks that can be independently configured and executed. Different algorithm function blocks can be instantiated to implement various control algorithms (MIMO MPC, PID, model-based controllers), and these blocks can be selectively activated based on process requirements without restructuring the entire system.
Solution Approach 2:
The function block architecture provides a universal framework that can accommodate multiple control algorithms through standardized interfaces and configuration parameters. The same basic function block structure supports different algorithm implementations, enabling extensibility while maintaining structural consistency across the control system.
Data Source
AI summary
A method includes executing a first function block configured to (i) receive multiple input data values associated with an industrial process control and automation system, (ii) pre-process the input data values, and (iii) store the pre-processed input data values in a shared memory. The method also includes executing a second function block configured to (i) receive references to the pre-processed input data values in the shared memory, (ii) invoke execution of one or more core functions to implement a process control algorithm, and (iii) store output data values in the shared memory. The method further includes executing a third function block configured to (i) receive references to the output data values in the shared memory, (ii) post-process the output data values, and (iii) output multiple output signals containing the post-processed output data values.