Hello everyone, I am new to AIMMS and I think I like it because of its graphical user friendliness. I want to use AIMMS to optimize the charging profile of a number of vehicles in a decentralized way such that the original central problem is decomposed to be solved independently as a subproblem by each vehicle. This is to be done for vehicles 1 to N over a number of time steps. I discovered this is difficult for me because the objective function in AIMMS is usually a scalar. Please, how can I do this? thanks
Best answer by FaridView original
Do you want to solve the problem for each vehicle independently, or all the vehicles part of one big problem and you want to create sub problems for each vehicle as decompositions ?
All the vehicles are part of one general problem and I want to decompose this general problem for each vehicle as subproblem to solve independently and simultaneously. Hope it is clearer now. Thank you.
Decomposing could mean many things, an algorithmic approach (like Bender’s decomposition for example), or you could formulate a model and solve that for each truck in a loop by changing the data.
I believe you need to first gain more theoretical knowledge about decentralized power planning before trying to implement a decentralized method in AIMMS. The following article solves a similar problem to what you mentioned:
The implementation of a decentralized solution in AIMMS is not difficult. The resources in AIMMS Academy will help you in actual implementation.