ctrbf
Compute controllability staircase form.
📝Syntax
[Abar, Bbar, Cbar, T, k] = ctrbf(A, B, C)
[Abar, Bbar, Cbar, T, k] = ctrbf(A, B, C, tol)
📥Input Arguments
Parameter Description
A State matrix: Nx-by-Nx matrix
B Input-to-state matrix: Nx-by-Nu matrix
C Output-to-state matrix: Ny-by-Nx matrix
tol scalar real (tolerance).
📤Output Arguments
Parameter Description
Abar Observability staircase state matrix.
Bbar Observability staircase input matrix.
Cbar Observability staircase output matrix.
T Similarity transform matrix.
k Vector: number of observable states.
📄Description

ctrbf(A, B, C) decomposes the given state-space system, defined by matrices A, B, and C, into the controllability staircase form.

This results in transformed matrices Abar, Bbar, and Cbar, along with a similarity transformation matrix T and a vector k.

The length of vector k is equal to the order of the system represented by A, and each entry in k denotes the number of controllable states factored out at each step of the transformation matrix computation.

The non-zero elements in k indicate the number of iterations required for T calculation, and the sum of k corresponds to the number of states in Ac, the controllable portion of Abar.

💡Examples
A = [-1.5  -0.5; 1     0];
B = [0.5; 0];
C = [0   1];
[Abar, Bbar, Cbar, T, k] = ctrbf(A, B, C)
🔗See Also
ctrb
🕔Version History
Version Description
1.0.0 initial version
Edit this page on GitHub