Ahmad,
Can you be more specific with your question. How big of a system are you
planning to solve? Do you need to perform the CPF run within some stipulated
time?
CPF basically solves a series of power flow solutions given some transfer
direction. Roughly, its computational complexity equals the number of power
flow solutions (CPF steps) + an additional overhead for computing the tangent
vector. There are a couple of things one can do to run the CPF faster:
i) Increase the step-size (mpopt.cpf.step): A larger
step-size would result in fewer CPF steps and in turn fewer power flow
solutions. However, increasing the step-size also means that the CPF may need
to do additional iterations at each step.
ii) Use step adaptivity (mpopt.cpf.adapt_step): The CPF will
adaptively change the step-size when this option is turned ON. It will take
larger steps on the flat portion of the PV curve and small steps for larger
curvatures. One can limit the max. and min. step with the options
mpopt.cpf.step_max and mpopt.cpf.step_min.
Shri
From: <[email protected]> on behalf of Ahmad Sadiq
Abubakar <[email protected]>
Reply-To: MATPOWER discussion forum <[email protected]>
Date: Friday, April 12, 2019 at 12:44 AM
To: MATPOWER discussion forum <[email protected]>
Subject: CPU time
Hi all,
What do you think will be the effect of CPF run of a real power systems with
large number of buses on CPU time?
I mean MATPOWER's continuation power flow (CPF)
Kind regards
Ahmad.
