The promising results obtained from the example presented in the previous
section seem to justify a combination of the two optimization strategies. As a
global optimizer the *very fast simulated re-annealing* algorithm was used due to its better overall
performance compared to the

The following algorithm was implemented to perform a combined optimization.

- Perform a configurable number of initial evaluations with the
*very fast simulated*algorithm.*re-annealing* - Start the gradient-based optimizer and use the best value of the
*very fast simulated*optimization so far.*re-annealing* - Wait for the gradient-based optimizer to exit and update the best target
of the
*very fast simulated*algorithm if the target was improved.*re-annealing* - Terminate the optimization task if the truncation condition is
fulfilled.
- Perform a configurable number of evaluations with the
*very fast simulated*algorithm. If the target was improved continue at step 2, otherwise at step 4.*re-annealing*