1. The default is defined by the gencost, either a polynomial or 
piecewise-linear cost on P.
2. Set the costs defined by gencost to zero, i.e. use a linear function with a 
zero coefficient.
3. It is not essential to make the distinction between control and state 
variables in an optimization problem. All of the elements of x are optimization 
variables. It is true however, that the power flow equations enforce 
dependencies between the injections and the voltages and angles.
4. The default is pretty good for most problems. Knitro is probably one of the 
best, but no solver is likely to be the best for all problems.
5. This question is outside the scope of this forum. I suggest you find a good 
textbook. The short answer is DC is a linear model that is easier to solve and 
is adequate for some purposes. AC is non-linear, harder to solve, but 
represents the physics better.

-- 
Ray Zimmerman
Senior Research Associate
B30 Warren Hall, Cornell University, Ithaca, NY 14853
phone: (607) 255-9645





On Jun 19, 2013, at 12:25 AM, linhomeperson <[email protected]> wrote:

> Dear all,
>  
> I tried to run OPF with runopf function in MATPOWER.
> 1, What is the default objective function ( e.g. "runopf(case30)" )? In the 
> manual, the objective function is fp and fq, however, what is the detailed 
> expression of function fp and fq.
> <Catch.jpg>
> 2, how can I replace fp and fq with my own user-defined function.
> from the manual, I can define fu. <CatchB056.jpg>, can I make f(x)=0?
>  
> 3, the solution is <Catch127A.jpg>.  From my understanding, the problem is to 
> minimize the cost by setting proper Pg and Qg (In x, Pg and Qg are control 
> variable while theta and Vm are not). Is that correct?
>  
> 4, Is there a best solver in MATPOWER, thanks?
>  
> 5,  What is the difference between DC and AC OPF? What is the physical 
> meaning of ignoring theta and Qg? In a real power system, why do we need AC 
> OPF?
>  
>  
> Thanks,
>  
> Lin

Reply via email to