LMIs in Control/Click here to continue/Fundamentals of Matrix and LMIs/D-stability Settling time poles

LMI for Settling Time Poles

The following LMI allows for the verification that poles of a system will fall within a settling time constraint. This can also be used to place poles for settling time when the system matrix includes a controller, such as in the form A+BK.

The System
We consider the following system:

$$ \begin{align} \dot x(t)&=Ax \end{align}$$

or the matrix $$ A \in \mathbb{R}^{n\times n} $$, which is the state matrix.

The Data
The data required is the matrix A and the settling time $$t_s$$ you wish to verify.

The Optimization Problem
To begin, the constraint of the pole locations is as follows: $${(z+z^{*}) \over 2}+{4.6 \over t_s}{\leq}0$$, where z is a complex pole of A. We define $$2 Re(z){\leq}-\alpha$$. The goal of the optimization is to find a valid P > 0 such that the following LMI is satisfied.

The LMI: LMI for Settling Time Poles
The LMI problem is to find a matrix P > 0 satisfying:

$$\begin{align} AP+(AP)^T +\alpha P&<0\\ \end{align}$$

Conclusion:
If the LMI is found to be feasible, then the pole locations of A, represented as z, will meet the settling time specification of $${(z+z^{*}) \over 2}+{4.6 \over t_s}{\leq}0$$, and the poles of A satisfy the previously defined constraint.

Implementation
A link to Matlab codes for this problem in the Github repository:

https://github.com/maxwellpeterson99/MAE509Code

Related LMIs
- D-stabilization

- D-stability Controller

- D-stability Observer

Return to Main Page
-Matrix and LMI Properties and Tools