太阳能电池的模型预测控制设计

%Book Title: Practical Design and Application of Linear MPC %Chapter: 8 %Section: 4 %Authors: Nassim Khaled and Bibin Pattel %Last Modified: 1/16/2018 close all %% System Identification of PV Cell %% Simulate the model sim('Chapter_8_Section_4_Model') %% Create time vector Time=0:1:length(Lambda_UV)-1; %% Change plotting settings set(0,'DefaultLineLineWidth',1) set(0,'DefaultAxesFontWeight','bold') set(0,'DefaultAxesFontName','Arial') set(0,'DefaultAxesFontSize',12) %% Grouping simulation data and neglecting first few seconds(Nstart*stp_sz) Nstart =2000; %Clip the data starting from Nstart as the first few seconds of the simulation should be discarded Measured_Outputs=[ Tm_f(Nstart:end)]; %Measured outputs Manipipulated_Variables=[ Lambda_UV(Nstart:end) ];% Pout_previous ]; %Manipulated variables %% Obtaining initial conditions at step time=Nstart Measured_Outputs_Nstart=Measured_Outputs(1,:); %Capturing the measured outputs at step time= Nstart Manipipulated_Variables_Nstart=Manipipulated_Variables(1,:); %Capturing the manipulated variables at step time= Nstart %% Forcing response to start from zero initial conditions Measured_Outputs_zero_initial_conditions=Measured_Outputs(1:end,:)-repmat(Measured_Outputs_Nstart,length(Measured_Outputs(1:end,:)),1); %Subtracting initial conditions for measured outputs to obtain zero response at step time= Nstart Manipipulated_Variables_zero_initial_conditions=Manipipulated_Variables(1:end,:)-repmat(Manipipulated_Variables_Nstart,length(Manipipulated_Variables(1:end,:)),1); %Subtracting initial conditions for manipulated variables to obtain zero actuation at step time= Nstart %% Prepare date for system identification stp_sz=10 data=iddata(Measured_Outputs_zero_initial_conditions(1:stp_sz:end,:),Manipipulated_Variables_zero_initial_conditions(1:stp_sz:end,:),stp_sz); %data is packaged for system identification using iddata %% Generate a preliminary 3rd order system that fits the data sys1=n4sid(data,3,'Form','canonical','DisturbanceModel','None','InputDelay',[0 ]','InitialState','zero'); %n4sid generates a preliminary system in the canonical form %% Generate a more refined system sys2 = pem(data,sys1,'InitialState','zero') %pem generates a more refined system that fits the data better %% Define the options for comparing the various identified systems opt = compareOptions('InitialCondition','Z'); %% Plot the comparison between the simulation data and the identified system sys2 figure compare(data,sys2,opt); %% Extract A, B, C and D matrices from sys2 A=sys2.A B=sys2.B C=sys2.C D=sys2.D %% Use the ss command to represent the results from sys2 in a state space format PV_Linear_Model = ss(A,B,C,D,stp_sz) save Chapter_8_Section_4_Linear_Model.mat PV_Linear_Model