Multivariable Predictive Controller Tunable Trade-off Factor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multivariable Predictive Control (MPC) performance degrades over time due to process changes, requiring frequent model recalibration, which is resource-intensive and invasive, causing disruptions in process operations.
Innovation Solution
A tunable trade-off between optimal process operation and perturbation is achieved through a method that uses a user-specified trade-off factor to select between economic optimization and constraint control objective functions, allowing for low-amplitude step testing to maintain model accuracy without significant process disruption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional model recalibration is performed to maintain MPC control performance, then model accuracy is improved, but process operation is disrupted and resource consumption increases
Solution Approach 1:
The system performs automated closed-loop step testing where the MPC controller tests itself by introducing perturbations to its own manipulated variables and measuring the resulting controlled variable responses. This self-diagnostic capability enables automatic model recalibration without external intervention or process disruption, resolving the contradiction between maintaining model accuracy and ensuring continuous operation.
Solution Approach 2:
The system implements periodic closed-loop step testing at predetermined intervals to automatically update the dynamic process model. By scheduling these tests during normal operation rather than performing manual recalibration, the system maintains model accuracy while minimizing disruption to continuous process operation.
2Measurement precision
If closed-loop step testing is performed to identify process models, then model accuracy is improved, but optimal process operation is compromised due to variable perturbation
Solution Approach 1:
The system introduces small-amplitude perturbations to manipulated variables during closed-loop step testing, using just enough disturbance to generate identifiable process responses without significantly deviating from optimal operating conditions. This partial action approach maintains model identification accuracy while minimizing impact on process optimality.
Solution Approach 2:
The system uses the MPC controller's feedback mechanism to measure controlled variable responses to perturbations and automatically update the process model. The feedback loop enables the system to learn from the perturbations it introduces, improving model accuracy without requiring large deviations from normal operation that would compromise optimality.
Data Source
Figure 1~2B
Figure 3
Figure 4
AI summary
An integrated multivariable predictive controller (MPC) and tester is disclosed. The invention system provides optimal control and step testing of a multivariable dynamic process using a small amplitude step for model identification purposes, without moving too far from optimal control targets. A tunable parameter specifies the trade-off between optimal process operation and minimum movement of process variables, establishing a middle ground between running a MPC on the Minimum Cost setting and the Minimum Move setting. Exploiting this middle ground, embodiments carry out low amplitude step testing near the optimal steady state solution, such that the data is suitable for modeling purposes. The new system decides when the MPC should run in optimization mode and when it can run in constrained step testing mode. The invention system determines when and how big the superimposed step testing signals can be, such that the temporary optimization give-away is constrained to an acceptable range.