create
Creates a Random Restart solver that utilizes Simulated Annealing for its inner optimization phases.
Architecture Note on Auto-Calibration: If temperatureConfiguration is set to TemperatureConfiguration.AutoCalibrate, the inner SA solver will dynamically recalculate a new starting temperature at the beginning of every single restart. This ensures the initial temperature is perfectly tuned to the local landscape of each new random starting point.
Return
A RandomRestartSolver wrapping a dynamically configuring SimulatedAnnealing inner solver.
Parameters
The formal definition of the optimization problem.
The builder responsible for constructing the simulation model.
The total number of macro-iterations (restarts) the outer solver should perform.
An optional MutableMap specifying the starting coordinates for the first restart. All subsequent restarts will automatically generate random feasible starting points.
The temperature strategy applied to each inner SA run. Defaults to AutoCalibrate.
The cooling strategy applied to each inner SA run.
The stopping threshold for each inner SA run.
The maximum number of SA steps per restart.
The default number of simulation replications to run per evaluation.
A cache to store evaluated solutions across all restarts.
An optional cache for individual simulation replication data.
Optional parameters defining the simulation run properties.
the random number stream number for the outer random-restart driver; 0 (the default) means the next available stream
the provider of random number streams for the outer driver; defaults to a fresh RNStreamProvider
selects and configures parallel evaluation of multi-point requests; mutually exclusive with concurrentRestarts greater than 1
the number of restarts allowed to run at the same time, each on its own worker with private evaluation resources (and its own copy of the cooling schedule); 1 (the default) preserves the sequential behavior
stream-block size and optional confirmation stage for concurrent restarts