Hi Carlos,

This is a complex formulation, so I’m not 100% sure of the answer. The 
TransMask input parameter employed by filter_ramp_transitions() to create the 
“individual trajectories” case only affects the definition of the ramp 
reserves, not the storage handling. The storage is still not handled 
independently for the individual trajectories. I believe that prevents it from 
being identical to a classical 2-stage stochastic program. If there are no 
storage units, it might be identical, but I’m not completely certain.

Carlos Murillo-Sánchez, do you have thoughts on this?

   Ray


On Aug 14, 2020, at 2:26 PM, Carlos Ferrandon Cervantes 
<ferrand...@gmail.com<mailto:ferrand...@gmail.com>> wrote:

Hello everyone:

I've followed the example "most_ex7_suc.m", especifically on the individual 
trajectories case. I am aware that what the code does in this case (correct me 
if I'm wrong) is to optimise  the cost of a set of probability weighted 
scenarios (and transitions) with certain costs and constraints over the 
planning horizon. Let's assume this is the case without contingencies, for the 
sake of simplicity. Would it be correct to say that this specific case is 
similar for a two-stage stochastic programming approach? My reasoning is that 
we commit on the first stage for the start up and shut down events; then, over 
the planning horizon we deal with the redispatch (if necessary), load-following 
ramp reserve, or fixed reserve (mdi.FixedReserves), given the probability of 
the scenarios previously defined.

Also, some months ago I had the problem in MOST that on the results, when we 
had certain unit disconnected (hydro units in my case), we would expect it to 
be off (u=0), but it happened to appear on (u=1). Well, the problem was that I 
had set the hydro unit with a minimum of zero, so for the solution it wasn't a 
problem to update this value since it didnt affect the results. Since hydro 
units dont usually run at that minimum level (could be the case if it's running 
as a synchronous condenser), I updated this value. Now this is sorted! So I 
wanted to share it with the community. Thanks for the heads up you gave me in 
that email!.

Hope you are all ok in these days,


--
Carlos Ferrandon


Reply via email to