cola_rr_lin.m

国外经典书籍MULTIVARIABLE FEEDBACK CONTROL-多变量反馈控制 的源码

function [f,g]=cola_lin_rr(X,u)
%
% cola_lin_rr - This function designed for use with 'cola_linearize.m' to
%        create a linear model of a L/D-V/B-distillation column.
%
%     x - states (liquid composition and liquid hold up)
%     u - inputs and disturbances (L/D, V/B, feedrate and feed
%         composition)
%     g - outputs (distillate and bottoms compostition)
%
%  NOTE: Everything here is row vectors rather than columnn vectors


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

NT=41;
% Splitting the states
x=X(1:NT);                          % Liquid composition
M=X(NT+1:2*NT);                     % Liquid hold up

% Inputs and disturbances
RT=u(1);  % L/D
RB=u(2);  % V/B
F= u(3);  % Feedrate
zF=u(4);  % Feed composition
qF=1.0;   % Feed liquid fraction 
          % Use qF=u(5) if qF is to be included in linear model

% P-Controllers for control of reboiler and condenser hold up.
KcB=10;  KcD=10;         % controller gains
MDs=0.5; MBs=0.5;        % Nominal holdups - these are rather small  
Ds=0.5; Bs=0.5;          % Nominal flows
MB=X(NT+1);  MD=X(2*NT); % Actual reboiler and condenser holdup
D=Ds+(MD-MDs)*KcD;       % Reflux
B=Bs+(MB-MBs)*KcB;       % Bottoms flow     

% Resulting reflux and boilup
LT = RT*D;
VB = RB*B;

% Store all inputs and disturbances
U(1)=LT; U(2)=VB; U(3)=D; U(4)=B; U(5)=F; U(6)=zF; U(7)=qF;

% This variable is not used
t=0;

xprime=colamod(t,X',U');

% Output
f=xprime';
g=[x(NT),x(1)];