Dear All,
I’m working on microgrid scheduling using GAMS considering the PV uncertainty and the gams documentation for stochastic programming is recommending to use LINDO solver for making a random variable normally distributed but I saw other codes using Cplex solver with multiple scenarios to consider uncertainty in a random variable.
which one of these two ways work in power systems scheduling?
I highly appreciate your cooperation,
Thank you,
Anan
Eventually, any continuous distribution is discretized before solving the stochastic optimization model. You can do this in the GAMS code and set up a discrete distribution then you can use e.g. DE with Cplex or you let Lindo do it internally and then you need to use Lindo’s algorithms.
-Michael
thank you so much for your prompt reply.
I don’t have distribution yet, I will work with hourly output power for one year and I’m looking for a way to fit this data into a distribution normal or discrete.