simulateNonHybrid

Method of continuousTimeMdss.

Integrate a non-hybrid mdss model as a DAE with ode15i.

Source: src/simulator/@continuousTimeMdss/simulateNonHybrid.m

What is solved

A non-hybrid model has no binaries and no inequalities, so it is a plain index-1 differential-algebraic system

$0 = F(\dot x, x, u(t), y) = \Phi_{\mathrm{eq}}\, m(\dot x, x, u, y),$

integrated with MATLAB’s fully-implicit ode15i. The solver state is $w=[x;\,y]$ with slope $\dot w=[\dot x;\,\dot y]$ ; the binary guess is empty and held fixed.

Residual and Jacobian handles

odeHandle returns $F$ at $(\dot w, w, u(t))$ with the inputs linearly interpolated. jacobiHandle returns the two blocks the fully-implicit solver needs:

$\frac{\partial F}{\partial w} = \big[\,J_{F,x}\ \ J_{F,y}\,\big], \qquad \frac{\partial F}{\partial \dot w} = \big[\,J_{F,\dot x}\ \ 0\,\big],$

the $\dot y$ block being zero because algebraic variables are not differentiated. (The handles are referenced directly, not through anonymous wrappers, so the Jacobian reports nargout == 2 as ode15i requires.)

Consistent initial conditions

$(\dot x_0, y_0, \dot y_0)$ are obtained from eventTotalDerivative with an empty event index (no mode enumeration); a non-zero algebraic guess is retried if the first attempt returns NaN/Inf. After applySolverOptions, solve runs over the input time span and the trajectory is stored in sim.time and sim.variableTrajectory; sim.eventTime is empty since there are no events.

Input arguments

Argument	Description
`sim`	The `continuousTimeMdss` object, with the model, initial state, input trajectory and solver options set.

Output arguments