clab_help.m

matlab例程

      'Graphics Window in CtrlLAB provides easy to use graphics processing ',...
      'facilities which include zooming, legend adding, curve editting and other ',...
      'useful functions.  These functions can be called from the {\bfOptions} menu ',...
      'items.',' ',...
      '  {\bfAxis} and {\bfGrid} menu item: toggle on/off grid and axis.  Also the curves',...
      '      can be held to allow more to be drawn in the same axis.',' ',...
      '  {\bfCursor} menu item: to use mouse to locate points on the curves.',' ',...
      '  {\bfLegends} menu item: to allow user to add, edit legends/arrows/lines to the',...
      '      curves.  Legends can be manipulated also using this menu item.',' ',...
      '  {\bfZoom} menu item: to allow user to zoom the curves.  By selecting the',...
      '      menu item, one can drag on the curve for zooming.',' ',...
      '  {\bfEdit Plot} menu item: allow the user to modify existing curves.  One can',...
      '      delete and modify colors/lne styles of the existing curves.',' ',...
      '  {\bfSave as EPS File} menu item: allow the user to save current figure to an',...
      '      EPS (Encaptulated PostScript) file.',' ',...
      '  {\bfPlot Preference} menu item: Allow user to specify more graphics properties',...
      '      and model combinations.  More details can be obtained from the {\itHelp} ',...
      '      information of its dialog box.');
case 10,
   hlp_Title='Lead/lag Cascade Compensation';
   hlp_Contents=str2mat(...
      'The general form of the controller is',' ',...
      '      G_c(s)=K_c(s+z_{c1})(s+z_{c2})/[(s+p_{c1})(s+p_{c2})], ',' ',...
      'with z_{c1}\leq p_{c1}, and z_{c2}\geq p_{c2}.  Also z_{c2}=p_{c2} for lead compensation and ',...
      'z_{c1}= p_{c1} for lag compensation.',' ',...
      'One can select the {\itCompensator Type} listbox to determine the compensator ',...
      'format.  If {\it Auto} is selected, then it will be determined automatically.',' ',...
      'One has to specify the crossover frequency \omega_c, the phase margin \gamma and the',...
      'the steady-state error tolerance K_v to design a compensator.  When these ',...
      'parameters are specified, one can just press {\itDesign} button to design the ',...
      'controller.  Or the {\it Cancel} button allows one to terminate the design ',...
      'process.  Some times one can press {\itMaximise} \omega_c button to design a ',...
      'controller for the given \gamma.',' ',...
      'Please note that the controller thus designed may not satisfactorily meet the',...
      'original \omega_c and \gamma specifications.  It is adviced to check the design results ',...
      'before it is used.');
case 11,
   hlp_Title='LQ Optimal Controller Design';
   hlp_Contents=str2mat(...
      'To design a controller which minimises the LQ (linear quadratic) criterion',' ',...
      '       J = \int_0^\infty[x^T(t) Q(t)x(t)+u^T(t) R(t)u(t) dt',' ',...
      'where x(t) is the state variable, Q \leq 0 and R>1 are weighting matrices.  One ',...
      'can select R=1 in CtrlLAB.  The control signal can be written as u(t)=-K x(t), ',...
      'with K a constant matrix satisfying the algebraic equation',' ',...
      '      -P A - A^T P  + P B R^{-1} B^T P - Q = 0',' ',...
      'and (A,B) are the state space equation of the plant model.  ',' ',...
      'Two types of controller are allowed in CtrlLAB: ',...
      '   1) state feedback control',...
      '   2) observer-based feedback control',' ',...
      'In the dialog box, one should enter the weighting matrix Q in MATLAB ',...
      'matrix form.  Also, select a control type.  If observer-based method is ',...
      'selected, then, the expected pole positions of the observer can be used',...
      'in a vector form.');
case 12,
   hlp_Title='Pole-Placement Controller Design';
   hlp_Contents=str2mat(...
      'Suppose that the eigen polynomials of the plant and expected closed loop ',...
      'model can be written as ',' ',...
      '   a(s)=det(sI-A)=s^n+\Sigma_{i=0}^{n-1}a_is^i,   \alpha(s)=\Pi_{i=1}^n (s-\mu_i)=s^n+\Sigma_{i=0}^{n-1} \alpha_i s^i',' ',...
      'The control signal can be written as u(t)=-K x(t), with K a constant vector',' ',...
      '   K = (\alpha-a)^T L^{-1} C^{-1}',' ',...
      'where (a-\alpha)^T=[(a_0-\alpha_0),\cdots, (a_{n-1}-\alpha_{n-1})], C=[b, Ab, \cdots, A^{n-1}b], and L is ',...
      'a Hankel matrix composed with coefficients of a(s).',' ',...
      'Two types of controller are allowed in CtrlLAB: ',...
      '   1) state feedback control',...
      '   2) observer-based feedback control',' ',...
      'In the dialog box, one should enter the expected pole positions in MATLAB',...
      'vector form.  Also, select a control type.  If observer-based method is ',...
      'selected, then, the expected pole positions of the observer can be used',...
      'in a vector form.');
case 13,
   hlp_Title='PID Controller Design';
   hlp_Contents=str2mat(...
      'PID (Proportional-Integral-Derivative) controller is very useful in',...
      'process control.  The general form of the controller is ',' ',...
      '   G_c(s)=K_p [1+T_d s + 1/(T_i s)]',' ',...
      'and the variation of it can be written as',' ',...
      '   G_c(s)=K_p (1+T_d s) ',' ',...
      'and the derivative action is performed in the feedback path.  The variables ',...
      'T_i, T_d can also be omitted to compose the P, PI controller as well.  ',...
      'Type type of the controller can be set in the {\bfController type} menu item.',' ',...
      'To design the PID controller, the plant model is often fitted into a first ',' ',...
      'order model with delay G_m(s)=k e^{-\tau s}/(T s+1), using different methods which',...
      'can be set in {\bfFirst-order identification Method} menu item.',' ',...
      'Various of algorithms are available in CtrlLAB and {\bfOne-shot} menu designs',...
      'PID controller with no extra specifications.  Details of other methods can be ',...
      'found in the Help facility in different methods.');
case 14,
   hlp_Title='PID Controller Design with Specifications';
   hlp_Contents=str2mat(...
      'Three PID algorithms are provided in the group:',' ',...
      '  {\bfChien (CHR) tuning}: Implement the Chien-Hrones-Reswick algorithm.  ',...
      '      Two overshoots are allowed such that 0% and 20% overshoots.  Selecting ',...
      '      an overshoot and the click {\bfDesign} button, the controller can be',...
      '      designed.',...
      '  {\bfModified Ziegler-Nichols tuning}: The two parameters, r_b and \phi_b, are ',...
      '      expected from the user, which means that to design a controller to move',...
      '      the original negative real axis crossing point to the one with gain of r_b ',...
      '      and phase angle of \phi_b.',...
      '  {\bfInternal model control}: A filter constant T_f is expected from the user',...
      '      to design a PID controller.',' ',...
      '{\itReferences}:',...
      '   [1] Astrom K J, PID controllers: theory, design and tuning, Research ',...
      '      Triangle Park, Instrument Society of America, 1995',...
      '   [2] D Xue and D P Atherton, Feedback Control Systems (Lecture Notes)');
case 15,
   hlp_Title='Optimal PID Controller Design';
   hlp_Contents=str2mat(...
      'Optimum PID controller design techniques are allowed in CtrlLAB through',...
      'the {\bfOptimum tuning} and {\bfGain phase tuning} menu items.  These design ',...
      'are based on the first-order plus delay approximation to the plant model, ',...
      'and the methods available can be selected from the {\bfFirst-order }',...
      '{\bfIdentification Method} menu.',' ',...
      '   For {\bfOptimum tuning} menu, three optimum criteria are available:',' ',...
      '      (1) ISE, which is \int_0^\infty e^2(t)dt',...
      '      (2) ISTE, which is \int_0^\infty t^2 e^2(t)dt',...
      '      (3) IST2E, which is \int_0^\infty t^4 e^2(t)dt',' ',...
      'where e(t) is the error signal enters the controller.',' ',...
      '   For {\bfGain phase tuning} menu, optimum PID controller can be designed',...
      'based on the gain/phase margin of the plant model.  ',' ',...
      '{\itReferences}:',...
      '  [1] Zhuang  M,  Atherton D P. Automatic tuning of optimum PID controllers.',...
      '      Proceedings IEE, Pt D, 1993, 140: 216-224',...
      '  [2] D Xue and D P Atherton, Feedback Control Systems (Lecture Notes)');
case 16,
   hlp_Title='LQG Controller Design';
   hlp_Contents=str2mat(...
      'LQG (Linear Quadratic Gaussian) controller design are allowed in CtrlLAB ',...
      'for the plant model',' ',...
      '    dx(t)/dt=A x(t)+B u(t)+\Gamma w(t), y(t)=C x(t)+v(t)',' ',...
      'where w(t) and v(t) are Gaussian signals with ',' ',...
      '    E[w(t)w^T(t)]=\Xi \geq 0, E[v(t)v^T(t)]=\Theta >0',' ',...
      'The optimisation criterion is defined as',' ',...
      '    J=\int_0^\infty z^T(t) Q z(t)+u^T(t) R u(t) dt',' ',...
      'where z(t)=M x(t).  Assume that R=1, and one should supply matrix Q, ',...
      'as well as the variance \Xi and \Theta, then click {\itDesign} button to find a LQG ',...
      'controller.');
case 17,
   hlp_Title='LQG/LTR Controller Design';
   hlp_Contents=str2mat(...
      'LQG (Linear Quadratic Gaussian) controller design are allowed in CtrlLAB ',...
      'for the plant model',' ',...
      '    dx(t)/dt=A x(t)+B u(t)+\Gamma w(t), y(t)=C x(t)+v(t)',' ',...
      'where w(t) and v(t) are Gaussian signals with ',' ',...
      '    E[w(t)w^T(t)]=\Xi \geq 0, E[v(t)v^T(t)]=\Theta >0',' ',...
      'The optimisation criterion is defined as',' ',...
      '    J=\int_0^\infty z^T(t) Q z(t)+u^T(t) R u(t) dt',' ',...
      'where z(t)=M x(t).  Assume that R=1, and one should supply matrix Q, ',...
      'as well as the variance \Xi and \Theta, then click {\itDesign} button to find a LQG ',...
      'controller.',' ',...
      'LTR (Loop Transfer Recovery) technique is useful to recover the loop transfer',...
      'function under observer-based technique.  An large value of q should be used',...
      'to recover the loop gain.');
case 18,
   hlp_Title='H-norm Based Controller Design';
   hlp_Contents=str2mat(...
      'Two kinds of weightings are supported:',' ',...
      '   1) {\bfMixed sensitivity} problem: with ',...
      '       T_{yu}(s)=[W_1 S(s); W_2 G_c(s) S(s); W_3 T(s)];',...
      'where S(s) and T(s) are sensitivity and complementary senitivity',...
      'functions.  and W_1(s), W_2(s), and W_3(s) are the weighting functions. ',...
      'One may enter the weighting functions for W_1-W_3(s) and also to display ',...
      'their Bode diagrams using the related push buttons.',' ',...
      '   2) {\bfSensitivity} problem: with T_{yu}(s)=W_1 S(s), and W_1(s) selected from ',...
      'the standard transfer functions of the order and natural frequency \omega_n of the',...
      'user''s choice.   For the standard model selection, the ITAE I, ITAE II and ',...
      'Butterworth model are allowed.',' ',...
      'Three such controller are allowed:',' ',...
      '   1) H_2 Controller,  2) H_\inftyController, and 3) Optimal H_\inftyController',' ',...
      'The Robust Control Toolbox of MATLAB should be used to complete the ',...
      'design of the robust controllers.  All the controller designed are observer-',...
      'based.');
case 19,
   hlp_Title='Plot Preference Options';
   hlp_Contents=str2mat(...
      'In this dialog box, the settings for the graphics windows in CtrlLAB can be',...
      'specified.  Once everything one wants to modify is specified, one can press',...
      '{\itChange} button to confirm the changes, and press the {\itDefault} button to set',...
      'the properties back to default ones.  ',' ',...
      '{\itGrid} group: tanggle the grids to the plot with on/off options.',' ',...
      '{\itBox} group: tanggle the boxes on the plot with on/off options.',' ',...
      '{\itApply} group: Allow the user to select whether he wants to set the specified ',...
      'changes to the {\itCurrent} graph, or the {\itAll} graphics windows.',' ',...
      'In {\itPlot Color} group, one can set the background color, and the color of',...
      'the figures.  Once can also set the color back to default.',' ',...
      'The {\itLoop} group: allow the user to set the system to be studied to {\itOpen Loop}',...
      'or {\itClosed Loop}.',' ',...
      'The {\itCombinations} group: One can set the model to be analysed to {\itUncom-}',...
      '{\itpensated} and {\itCompensated} using the multiple selectable boxes.  If the ',...
      'controller dose not exist, the latter is grayed.');
case 20,
   hlp_Title='Using the Zooming Facilities';
   hlp_Contents=str2mat(...
      'The {\bfZooming} menu allows the user to do zoomings to the current graphics',...
      'window.  The following menus are useful.',' ',...
      '{\bfZooming} menu allows one to use mouse button to drag the range one ',...
      '    interested.  Also once selected, one can use left and mouse button to click',...
      '    the graph to enlarge or reduce the graph.',' ',...
      '{\bfX-axis Zooming} and {\bf Y-axis Zooming} menus allows one to perform ',...
      '    zooming functions to x- and y-axis only.',' ',...
      'The {\bfFull} menus allows one to set the current axes in an automatic way, ',...
      'according to the actual data set in the graph.',' ',...
      'The {\bfUser Define} menu brings out the current dialog box, and from here',...
      'one can specify the zooming ranges using keyboard.  One can spefify ',...
      '{\itX-Axis} and {\itY-Axis} separately.  One can either select the {\itCurrent} or to select ',...
      '{\itDefine}, where one can set the lower and upper values in the edit boxes.',' ',...
      'Please note that, for graphics windows with more than one figures, such as ',...
      'the case in Bode diagram, one should select the one he is interested in by ',...
      'clicking the figure (an effective way is to click the axis labels), before the ',...
      'zooming function can be applied.');
case 21,
   hlp_Title='Using the Legend Facilities';
   hlp_Contents=str2mat(...
      'The {\bfLegend} menu allows the user to add and edit legends to the graphs.  The',...
      '{\itlabels, cursor positioning, line} and {\itrrows} can be added to the plot.  The',...
      'following menu items are provided in the menu:',' ',...
      '{\bfNew Legend} shows a legends editting box, and you can enter any text in the',...
      '    box, and then click {\bf Apply} button, then move the mouse to the position ',...
      '    you want to add the legend, and click the mouse button.  The legend will ',...
      '    then be added.',' ',...
      '{\bfEdit Legend}, {\bfMove Legend} and {\bfDelete Legend} menus allows one to edit ',...
      '    legends of his choice, using the mouse.',' ',...