With gain or switch by instance you can act on simulation parameter during run.

Suman,

"Optimum" is not an equation. 

It is an Interpretive Observation,

attained after the fact of sweeping values. 

Sounds like you want "the answer" with one shot. 

I like Jean's idea of "switch by instance". 

In that line of thinking, I use a manual approach

by running three circuits at one time, plotting together,

to help narrow down the tolerances,

to the point where an optimized compromise can be selected. . 

Method Description:

Using ngSPICE, I use a single signal source, 

into parallel circuits, 

sweep the circuits, 

and plot the bode or transient together, as one. 

Comment:

This is only an elementary way of aproximating  the results,

but can come very close to showing related values of interest.  

Paul suggests a ramp-block:

If a ramp-block were available, it would produce numerous results.

Conclusion ( for my work ):

For me, having 17 audio filter stages in sequence

doing a myriad of different functions, at high-Q resonance, 

involves much inter-stage interaction,

and the potential results of sweeping component values

would produce Exponentially Numerous Results.  

I understand the physics of how the circuits work 

... so, the simple 3 circuit approach, and repeated runs, 

helps me narrow down the better parameters

and helps me explore the circuit tolerances.

This helps narrow down the results

to the point where an optimized compromise can be selected.

Hope that helps, Glen

Yup you can do it

Sathish,

Please explain your comment.  

My procedures are being posted today 

on my website in the "Electronics" section. 

under "Signal Phase Delay" and click down. 

Link has been updated (170316).

Glen

http://www.geocities.ws/glene77is/GC_ET_AFX_PhaseD.html

Hello Suman,

Any ODE solver in Simulink or PSPICE behaves as a dynamical system. When you change a parameter during simulation is similar to producing a disturbance in a dynamical system. A transient response will occur. To avoid confusion between the own circuit dynamics and that of the solver, you should vary the parameter (component value) slowly, as compared to the maximum solver time step.

I designed a function with which I could change block parameters right during the model run. Maybe, you find it useful.