aresolutions.m

KALMAN FILTERING FOR FUZZY DYNAMIC SYSTEMS

function [pi1, pi2, P1, P2] = ARESolutions(Sw, Sv)

% function [pi1, pi2, P1, P2] = ARESolutions(Sw, Sv)
%
% Compute the Algebraic Ricatti Equation (ARE) solutions
% for the optimal controller and for the Kalman filter.
% INPUTS
%   Sw = process noise covariance matrix.
%   Sv = measurement noise covariance matrix.
% OUTPUTS
%   pi1 = optimal control ARE solution for the first local T-S model.
%   pi2 = optimal control ARE solution for the second local T-S model.
%   P1 = Kalman filter ARE solution for the first local T-S model.
%   P2 = Kalman filter ARE solution for the second local T-S model.

[A1, A2, B1, B2, h1, h2] = FuzzyModel([0 0 0]);

% Optimal control ARE solutions.
pi1 = dare(A1,B1,eye(3),eye(1),zeros(size(B1)),eye(3));
pi2 = dare(A2,B2,eye(3),eye(1),zeros(size(B2)),eye(3));

% Kalman filter ARE solutions.
P1 = dare(A1', eye(3), Sw, Sv, zeros(3,3), eye(3));
P2 = dare(A2', eye(3), Sw, Sv, zeros(3,3), eye(3));