SS-MPC: A user-friendly software based on single shooting optimization to solve Model Predictive Control problems
Saeid Bayat, James T. Allison
Abstract
To optimize active engineering systems, closed-loop controllers are crucial due to model discrepancies and uncertain disturbances. However, simple controllers like PID have limitations. Model Predictive Control (MPC) addresses these issues by optimizing system behavior, ensuring constraint compliance, and delivering exceptional performance. This paper presents a user-friendly Matlab program for MPC design. The program allows interactive parameter adjustments, result visualization, and optional combination of MPC with a Kalman filter. This software aims to facilitate broader adoption of advanced control methods. This paper includes an example with an analytical solution and compares that with the MPC results. • A User-friendly software is developed to solve Model Predictive Control problems. • The Single Shooting method is used to convert the problem to a nonlinear program. • A Kalman filter can be activated by a flag. • The user can define a wide range of parameter values using a Graphical User Interface. • Several examples with closed-form solutions are used to demonstrate and investigate the software’s capabilities.