Automated Tuning Weights for Multivariable Model Predictive Controllers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for tuning multivariable model predictive controllers (MPCs) are cumbersome and time-consuming, often requiring trial and error, and do not ensure robust stability or optimal performance due to model uncertainties and lack of automated procedures that consider both performance and robustness requirements.
Innovation Solution
An automated method for tuning MPCs that calculates optimal tuning weights based on performance and robustness requirements, using closed-loop transfer functions to ensure stability, performance, and robustness, while allowing for user-defined preferences in measurement and actuator usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional trial and error tuning procedure is used for multivariable MPC, then tuning parameters can be adjusted, but the tuning process becomes very time-consuming and cumbersome
Solution Approach 1:
The system performs self-tuning by automatically calculating optimal tuning parameters based on process model and performance specifications, eliminating the need for manual trial-and-error adjustment by operators
Solution Approach 2:
The tuning parameters are pre-calculated offline based on process identification and performance requirements, so that the controller is ready for immediate deployment without time-consuming on-site tuning
2Reliability
If traditional tuning procedure is used, then some tuning parameters can be set, but robust stability cannot be ensured due to model uncertainties
Solution Approach 1:
The method incorporates feedback from process identification results and performance specification analysis to automatically adjust tuning parameters, ensuring robust stability margins are maintained despite model uncertainties
Solution Approach 2:
The system automatically adjusts tuning parameters based on calculated performance and robustness specifications, transforming the complex robust stability problem into a parameter optimization task that can be solved systematically
3Manufacturing precision
If manual tuning is performed without automated procedures, then some control performance can be achieved, but optimal performance and robustness cannot be simultaneously ensured
Solution Approach 1:
The manual mechanical tuning process is replaced with an automated computational system that uses process models and optimization algorithms to calculate optimal tuning parameters, achieving both optimal performance and robustness simultaneously
4Stability of the object's composition
If ill-conditioned process model is used with typical MPC design, then controller can be implemented, but actuators become saturated and system becomes unstable
Solution Approach 1:
The system performs preliminary analysis of the process model conditioning before controller design, and pre-calculates appropriate tuning parameters that prevent actuator saturation and ensure stability for ill-conditioned systems
Data Source
AI summary
A fast and reliable technique for tuning multivariable model predictive controllers (MPCs) that accounts for performance and robustness is provided. Specifically, the technique automatically yields tuning weights for the MPC based on performance and robustness requirements. The tuning weights are parameters of closed-loop transfer functions which are directly linked to performance and robustness requirements. Automatically searching the tuning parameters in their proper ranges assures that the controller is optimal and robust. This technique will deliver the traditional requirements of stability, performance and robustness, while at the same time enabling users to design their closed-loop behavior in terms of the physical domain. The method permits the user to favor one measurement over another, or to use one actuator more than another.