function iduidemo(arg) %IDUIDEMO Runs the ident demo. % ARG is the number of the frame to be shown. % L. Ljung 4-4-94 % Copyright 1986-2004 The MathWorks, Inc. % $Revision: 1.14.2.2 $ $Date: 2004/04/10 23:19:34 $ set(0,'showhiddenhandles','on') Xsum = findobj(get(0,'children'),'flat','tag','sitb16'); set(0,'showhiddenhandles','off') number = 1; if isempty(Xsum) number = -1; ident; set(0,'showhiddenhandles','on') Xsum = findobj(get(0,'children'),'flat','tag','sitb16'); set(0,'showhiddenhandles','off') end XID = get(Xsum,'Userdata'); if arg==0 figname=idlaytab('figname',31); if ~figflag(figname) layout butwh=[mStdButtonWidth mStdButtonHeight]; butw=butwh(1);buth=butwh(2); ftb=mFrameToText/2; % Frame to button bb = 4; % between buttons etf = mEdgeToFrame; EdWH=[4*butw 11*mLineHeight+2*bb]; FigW = EdWH(1)+2*(etf+ftb); fig=figure('Units','Points','NumberTitle','off','Name',figname,'color',... get(0,'defaultUIcontrolbackgroundcolor'),'tag','sitb31',... 'HandleVisibility','callback','Resize','on',... 'Integerhandle','off','MenuBar','none',... 'DockControls','off',... 'vis','off','Defaulttextfontsize',10); pos = iduilay1(FigW,3); uicontrol(fig,'Units','Points','pos',pos(1,:),'style','frame')%... %,'backgroundcolor', [0.5,0.5,0.5]); XID.demo(2)=uicontrol(fig,'Units','Points','pos',pos(2,:),'style','push',... 'string','Next >>','callback','iduidemo(1);','userdata',1); XID.demo(3)=uicontrol(fig,'Units','Points','pos',pos(3,:),'style','push',... 'string','Prev <<','callback','iduidemo(2);','userdata',1,... 'enable','off'); XID.demo(4)=uicontrol(fig,'Units','Points','pos',pos(4,:),'style','push',... 'string','Close','callback','close(gcf)'); lev2=pos(1,2)+pos(1,4); uicontrol(fig,'Units','Points','pos',[etf lev2+etf [EdWH+[2*ftb 2*bb]]],... 'style','frame','backgroundcolor',[0.5,0.5,0.5]); XID.demo(1)=uicontrol(fig,'Units','Points','pos',[etf+ftb lev2+etf+bb EdWH],... 'style','text','Backgroundcolor',[1 1 1],'Foregroundcolor',[0,0,0],... 'max',20,'Horizontalalignment','left','Fontsize',10); % 'edit','max',2,'fontsize',13,'Horizontalalignment','left'); FigWH=[FigW lev2+2*etf+EdWH(2)+2*bb]; ScreenPos = get(0,'ScreenSize'); FigPos = [(ScreenPos(3)-FigWH(1)) FigWH(2)/4 FigWH]; set(fig,'pos',FigPos); set(fig,'Units','Pixels'); new_figpos = get(fig,'pos'); FigPos = [(ScreenPos(3)-new_figpos(3)) new_figpos(4)/4 new_figpos(3:4)]; set(fig,'pos',FigPos); set(get(fig,'children'),'units','norm') try if length(XID.layout)>30, if XID.layout(31,3) eval('set(fig,''pos'',XID.layout(31,1:4))','') end, end catch end figure(fig) set(fig,'vis','on') drawnow end arg=1; %if isempty(iduiwok(16)),number=-1;else number=1;end else number=get(XID.demo(1+arg),'userdata'); end if number==-1 text=str2mat(' The Graphical User Interface for the System Identification',... ' Toolbox (ident) is now being created. It is normally invoked',... ' by typing ident in the MATLAB command window.',... '');%,... %' Be patient, it may take a while to finish.'); set(XID.demo(1),'string',text) set(XID.demo(2),'userdata',1); set(XID.demo(3),'userdata',1,'enable','off');drawnow ident elseif number==1 set(XID.demo(3),'enable','off'); text=str2mat(... ' To follow the demo, please select menu items and click on buttons',... ' as prompted. You are of course free to explore other options during',... ' the demo, but the text assumes that you carry out all instructions.',... '',... ' Note that you can resize this figure (and most other ident figures)', ... ' if necessary. Try to resize now if you wish.', ... '',... ' 1) First type load dryer2 at the MATLAB command',... ' line. This loads the data variables u2 and y2 into the workspace.', ... ' This data is from an actual hairdryer. The input (u2) is the heating',... ' power and the output (y2) is the temperature of the outflow air.'); set(XID.demo(2),'userdata',2); set(XID.demo(3),'userdata',1); elseif number==2 set(XID.demo(3),'enable','on'); text=str2mat(... ' 2) In the ident window, press the popup menu Data and select',... ' Import. This will open a dialog box.', ... '',... ' 3) Go to the dialog box and enter u2 into the editable text box beside',... ' Input: and y2 into the editable text box beside Output:.',... '',... ' 4) Enter a Sampling interval: of 0.08.',... '',... ' 5) Enter any variable name you wish in the box marked',... ' Data name:. For example: dry.'); set(XID.demo(2),'userdata',3); set(XID.demo(3),'userdata',1); elseif number==3 text=str2mat(... ' 6) By pressing the button ''More'', you get access',... ' to more book-keeping options. You may, e.g., enter',... ' any text you wish in the box marked Notes.',... ' For example, %This is my first attempt at using ident.',... '',... ' 7) Now press the Import button and the data will be represented', ... ' as an icon in the ident figure. To close the Import Data dialog,', ... ' press the dialog''s Close button.'); text=str2mat(text,... '',... ' Notice that the data also fills the Working Data and the',... ' Validation Data icons. More about that later.') ; set(XID.demo(2),'userdata',4); set(XID.demo(3),'userdata',2); elseif number==4 text=str2mat(... ' 8) Click on the Time Plot checkbox beneath Data Views to open',... ' a plot figure.',... ' 9) Examine the plot. The plot figure has a number of options accessible', ... ' via the menu bar which you may wish to explore now.',... ' 10) Note for example that Zoom under the Style menu is active.',... ' You can draw rectangles in the plot with the left mousebutton down.',... ' When you release the button, the area within the rectangle expands'); text=str2mat(text,... ' to fill the axes. Double-click to return to the original plot limits.',... ' Click the left/right (alt-click) mouse button to zoom in/out. Type',... ' help zoom at the MATLAB command line for more information.'); set(XID.demo(2),'userdata',5); set(XID.demo(3),'userdata',3); elseif number==5 text=str2mat(... ' 11) To remove the constant levels in the data sequences, return to', ... ' the ident window and select Remove means from the Preprocess', ... ' popup menu. Notice that a new data set will be inserted into the', ... ' Data board.',... '',... ' 12) The new data set has also automatically been plotted in the',... ' Time plot figure. You may have to choose Autorange under the',... ' Options menu to see the new plots. To unplot a data set, just',... ' click on its icon to deselect it. We shall work further with',... ' this new data set (with a ''d'' appended to its name), so drag',... ' and drop it onto the Working Data icon',... ' at the center of the ident figure.'); set(XID.demo(2),'userdata',6); set(XID.demo(3),'userdata',4); elseif number==6 text=str2mat(... ' 13) Now double-click on a data set icon. (On slower machines, use the',... ' right mousebutton if you have a three/two button mouse. With one',... ' button use the usual alt selection key sequence e.g. ctrl-click.)',... ' A new dialog box will open containing information about the data.',... '',... ' At this point the name of the data can be changed. The color of',... ' its icon can also be changed. Try this, or just close the',... ' dialog by pressing its Close button.'); set(XID.demo(2),'userdata',7); set(XID.demo(3),'userdata',5); elseif number==7 text=str2mat(... ' 14) Now select the Select range... option from the Preprocess', ... ' popup menu in the ident figure. A new figure will open.',... '',... ' 15) You can use the editable text fields to select a subset of the data.', ... ' Or, draw a rectangle (with the mouse button down) to select a portion',... ' of the data to be used for estimation purposes. For example, choose',... ' the interval from 1 to 50 seconds. Note that if you make a mistake,',... ' you can erase the rectangle and restart by pressing the Revert button.',... '',... ' 16) Press Insert in the dialog box, and the selected data range',... ' is added as a new data set to the Data board.'); set(XID.demo(2),'userdata',8); set(XID.demo(3),'userdata',6); elseif number==8 text=str2mat(... ' 17) Repeat once more to select a portion of the data set to be',... ' used for validation purposes. For example, select the interval',... ' from 50 seconds to the end. Insert this data set. Close the',... ' dialog by pressing its Close button.',... ' ',... ' Make sure that you now have the two data sets dryde and drydv',... ' on the data board.'); set(XID.demo(2),'userdata',9); set(XID.demo(3),'userdata',7); elseif number==9 text=str2mat(... ' 18) In the Data board of the ident figure, select a data set to',... ' work with for estimation by dragging it to the Working Data',... ' icon. The data set to be used for validation purposes should be',... ' dragged and dropped onto the Validation Data icon.',... '',... ' 19) Use the third set (dryde) for estimation purposes and the fourth',... ' data set (drydv) for validation purposes. Drag them to the icons',... ' as described in step 18.'); set(XID.demo(2),'userdata',10); set(XID.demo(3),'userdata',8); elseif number==10 text=str2mat(... ' 20) To estimate the step response (or impulse response) of the',... ' system, select Correlation model... from the Estimate popup menu',... ' in the ident figure. Note this adds a model to the Model board.',... ' Close the dialog.',... '',... ' 21) To view the result, click on the Transient Response',... ' checkbox beneath Model Views in the ident figure.',... '',... ' 22) Examine the various menu items in the plot that appears.'); set(XID.demo(2),'userdata',11); set(XID.demo(3),'userdata',9); elseif number==11 text=str2mat(... ' 23) To estimate the frequency response of the system, select',... ' Spectral model... from the Estimate popup menu in the', ... ' ident figure. Press Estimate in the dialog box that opens.',... ' Note that this adds a model to the Model board. Close the dialog.',... ' 24) To view the result, click on the Frequency Response ',... ' checkbox beneath Model Views in the ident figure.',... '',... ' 25) Examine the various menu items in the plot that appears. Apart',... ' from Style and Options, these are the same as in the other plots.'); set(XID.demo(2),'userdata',12); set(XID.demo(3),'userdata',10); elseif number==12 text=str2mat(... ' 26) Now, let us compute parametric models. Select Parametric models...', ... ' from the Estimate popup menu.', ... '',... ' A new dialog will open. The idea is to enter model structure',... ' information into the Parametric Models dialog and then generate a', ... ' model by pressing the Estimate button.',... '',... ' To help you with the model structures, the Order Editor',... ' provides more information on the different types of supported structures.'); set(XID.demo(2),'userdata',13); set(XID.demo(3),'userdata',11); elseif number==13 text=str2mat(... ' 27) As a default, a fourth order ARX type difference equation model has',... ' been selected. Press Estimate, and the model will be computed',... ' and inserted into the Model board as an icon. Note that the',... ' corresponding curves will be added in the open Model Views.',... ' 28) Open the Order Editor by pressing the button Order Editor',... ' in the Parametric Models dialog.',... ' Select the orders na=2, nb=2, and nk=3, by using the',... ' popup menus in the model structure editor dialog. This indicates',... ' a second order model with a deadtime of 3 samples (i.e. 0.24 sec).'); text=str2mat(text,... ' 29) Press Estimate to compute another model.'); set(XID.demo(2),'userdata',14); set(XID.demo(3),'userdata',12); elseif number==14 text=str2mat(... ' 30) Let us now look at the transient response. It contains',... ' three curves. These are color coded along with',... ' the corresponding model icons in the Model board.',... '',... ' 31) Clicking on the model icons in the ident figure the associated',... ' curves are added or removed from the open views.',... '',... ' 32) Using the middle mouse button (alt selection or ctrl-click),',... ' you can click on a line in a Model View and the corresponding',... ' Model name will be displayed.'); set(XID.demo(2),'userdata',15); set(XID.demo(3),'userdata',13); elseif number==15 text=str2mat(... ' 33) Choose menu item Show 99% confidence intervals under the main menu',... ' Options in the Transient Response plot.',... ' Curves corresponding to confidence intervals will now be',... ' added to the plot. Picking the same menu item again disables',... ' the confidence curves.'); set(XID.demo(2),'userdata',16); set(XID.demo(3),'userdata',14); elseif number==16 text=str2mat(... ' 34) There are several other views by which the models can be evaluated',... ' These are listed under Model Views. It is better',... ' to open and close the view figures by clicking on the checkboxes,',... ' rather than to iconify the figures.',... '',... ' 35) Pick the Model Output view. This shows a comparison between',... ' the models'' simulated outputs and the measured output, when applied',... ' to the validation data set. The (color coded) numbers at the side of',... ' the plot show how much (in %) of the output variation was reproduces',... ' by the model.'); set(XID.demo(2),'userdata',17); set(XID.demo(3),'userdata',15); elseif number==17 set(XID.demo(2),'enable','on'); text=str2mat(... ' 36) The parametric model dialog allows a wide range of different',... ' model structures. Use the popup menu in the Parametric Models dialog',... ' to select State Space: (n). Here it is sufficient to',... ' select the order n of the state-space model, and then press Estimate.',... '',... ' 37) If you don''t know what a good order is, you can enter a range of',... ' orders, e.g. 1:10. Then pressing Estimate opens a dialog',... ' within which different orders can be estimated by just clicking',... ' accordingly along the x-axis and then inserting the chosen models.'); set(XID.demo(2),'userdata',18); set(XID.demo(3),'userdata',16); elseif number==18 text=str2mat(... ' This is the last page of the demo.',... '',... ' There are many other features to be explored. As you need to use more',... ' of the functionality, simply open the appropriate dialog box and press', ... ' the help button for more information. Good luck!'); set(XID.demo(2),'userdata',19,'enable','off'); set(XID.demo(3),'userdata',17); elseif number==19 close(gcf) end if number<19,set(XID.demo(1),'string',text);end set(Xsum,'UserData',XID);