Batch acquisition for parallel bayesian optimization—application to hydro-energy storage systems scheduling

HIGHLIGHTS

SUMMARY

    Using these approximations, evaluation of the expected profit is fast and classical optimization methods (e_g, dynamic programming, genetic algorithms) can be applied to optimize the UPHES scheduling within a reasonable time budget. This motivates the authors to investigate the use of surrogate models to (partially) replace the time-consuming simulator, in conjunction with parallel evaluations of the expected profit. And despite the potentially larger number of simulations gained by parallelization, the final outcome of the algorithms is often worse with high batch sizes (8 or 16 compared to 2 or 4) within a fixed . . .

     

    Logo ScioWire Beta black

    If you want to have access to all the content you need to log in!

    Thanks :)

    If you don't have an account, you can create one here.

     

Scroll to Top

Add A Knowledge Base Question !

+ = Verify Human or Spambot ?