I would like to model a system where a storage operator has to decide on actions
at the beginning of each time step while facing uncertainty about the
realizations of photovoltaic generation and consumption profiles.
It’s a multi-stage stochastic problem with 96 time steps (96 stages, 15 minutes each time step, one day).
I have a huge number of scenarios: 10^96.
I think L-shaped methods, Lagrangian decomposition, progressive hedging technique and other decomposition methods will not help in solving this problem, since I have a huge number of scenarios.
Kindly, I have two questions:
1. Do you think such a problem can be solved within a reasonable time?
2. Is there any alternative method to approximate the expected cost functions of future decisions without using scenarios?
3. Which method do you recommend to solve this problem?
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering