\% ********* List of classes of dynamical systems ********* \topic ClassList {list of classes of dynamical systems} CONTENT currently supports the following classes:\par { \li 1 {\bf generic autonomous ODEs}\par {\bf generic iterated maps}\par {\bf generic PDEs on the interval}\par {\bf generic autonomous DAEs}\par } \par \end \% ClassList \% ********** Classes of dynamical systems ********** \topic Classes {class of dynamical system} The term {\bf class of dynamical systems} refers to particular common properties of dynamical systems. In {\bf Content} the notion of class is used to establish a link between the properties and a set of numerical algorithms for analysis of dynamical systems. Class also defines the way in which functions specified by the user are interpreted by {\bf Content}. For example, by saying that you will study a dynamical system belonging to the class of {\bf generic autonomous ODEs} you implicitly specify that functions you give are right-hand sides of ODEs and that algorithms for bifurcation analysis of autonomous ODEs are applicable to you dynamical system.\par You change the class by the \jump kh_menuSelect {Select|Class} command from the \jump kh_MainWin {Main window.} \include ClassList \par See description of the \jump *curclass {Current class.} \end \% Classes \% ********** Dynamical system ********** \topic DynSys {dynamical system} A dynamical system in {\bf Content} can be specified as belonging to one of the supported \jump Classes {class} of dynamical systems. The name of the current class is shown in the \jump kh_MainWin {Class field} in the Main window. You may change current class through the \jump kh_menuSelect {Select|Class} menu in the Main window.\par \include *curclass \par \end \% DynSys \% *********** Continuer generator ********** \topic conti {continuer} A {\bf continuer} generates a sequence of points \par \qc{$y^{(1)},y^{(2)},\ldots,y^{(k)},\ldots$} \par approximating a curve that is implicitly defined by the system of $n$ scalar equations in $n+1$ dimensional space: \par \qc{$F(y)=0,\;\; F:{\f b* R}^{n+1} \rightarrow {\f b* R}^{n}.$} \par Given a point $y^{(k)}$ located on the curve and the normalized tangent vector $v^{(k)}$ to the curve at this point, the next point $y^{(k+1)}$ and its associated tangent vector $v^{(k+1)}$ are computed in CONTENT as following. \par {\bf Prediction:} \par \qc{$Y^{0}=y^{(k)}+h_{k}v^{(k)},\:\: V^{0}=v^{(k)}.$} \par where $h_{k}$ is the current {\bf step size}. \par \qc {\picture 0.5 conti.fig} \par {\bf Corrections:} \par Iterate for $i=0,1,2,\ldots$, I_{max}: \par \qc{$A=F_{y}(Y^{i}),\:\: B=\matrix 2 c 0 0 {A}{[V^{i}]^{T}},$} \par \qc{$R=\matrix 2 c 0 0 {AV^{i}}{0},$} \par \qc{$W=V^{i}-B^{-1}R,\:\: V^{i+1}=\frac{W}{||W||},$} \par \qc{$R=\matrix 2 c 0 0 {F(Y^{i})}{0},$} \par \qc{$Y^{i+1}=Y^{i}-B^{-1}R.$} \par If \par \qc{$||F(Y^{i})|| < \varepsilon_{F}$ and $||Y^{i+1}-Y^{i}|| < \varepsilon_Y$} \par then \par \qc{$y^{(k+1)}=Y^{i+1},\:\: v^{(k+1)}=V^{i+1}$.} \par Recomputation of the Jacobian matrix $A$ and update of the vector $V$ can be suppressed after $I_{maxNewt}$ iterations. \par \par {\bf Step size control:} Set, $h_{k+1}=sh_{k},$ where \par \qc{$ s= \system 3 cl 1.2 1 {1.3} {i<3,} {1.0} {3\leq i < I_{max},} {0.5} {i=I_{max}.} $} \par If $h_{k+1} > H_{max}$, $h_{k+1}=H_{max}$. If $h_{k+1}\par {\li 1 InitStepsize} \> Initial stepsize\par {\li 1 CurrentStepsize} \> Size of the current step along the curve\par {\li 1 MinStepsize} \> Minimal stepsize $H_{min}$\par {\li 1 MaxStepsize} \> Maximal stepsize $H_{max}$\par {\bf Corrector data}\>\par {\li 1 MaxIter} \> Maximal number $I_{max}$ of corrections\par {\li 1 MaxNewtonIter} \> Maximal number $I_{maxNewt}$ of corrections with the recomputation of the Jacobian matrix\par {\li 1 VarTolerance} \> Accuracy $\varepsilon_Y$ with respect to variables\par {\li 1 FuncTolerance} \> Accuracy $\varepsilon_F$ with respect to defining functions\par {\li 1 TestTolerance} \> Accuracy for zero location of test and user functions\par {\li 1 MaxIterTest} \> Maximal number of bi-sections for zero location\par {\bf Stop data}\>\par {\li 1 MaxNumPoints} \> Maximal number $N_{maxPoint}$ of points along the curve in one direction\par {\li 1 BeforeClosureCheck} \> Number of points to skip before closure check\par {\bf Output data}\>\par {\li 1 BetweenAdaptation} \>Curve type dependent parameter meaning the number of points between adaptation.\par {\li 1 OutLevel}\> Output index: 0 - no output during corrections; 1 - output at each correction.\par \par The parameters can be modified via the \jump kh_GeneratorWin {generator window}.\end \%conpar