Witness Optimizer: 
 
WITNESS Optimizer significantly reduces the time spent experimenting, by automatically finding the optimum solution to satisfy chosen performance criteria. Using the latest in sophisticated mathematical techniques, it offers an easy-to-use interface and the presentation of optimal results in a selection of useful and innovative tables and graphs.

Within a simulation model, an optimization algorithm typically works in the following way:

1. An initial set of parameter values is chosen and one or more replication experiments is carried out with these values.

2. The results are obtained from the simulation runs and the optimization algorithm then chooses another parameter set to try.

3. The new values are set and the next experiment set is run.

4. Steps 2 and 3 are repeated until either the algorithm is stopped manually or a set of defined finishing conditions are met.

The WITNESS Optimizer module is seamlessly integrated with the standard WITNESS system. The Optimizer can be used with any WITNESS model to guide the user towards the most successful options for a process by identifying the effect of changes to model parameters. Success is typically measured in terms of throughput or profitability (taking into account cost).

The WITNESS Optimizer offers a wide choice of options for experimentation. These include:

• Random solutions—to enable an appreciation of the shape of the solution space.

• All combinations—this option should indeed be chosen if time allows as it covers all possibilities.

• Min/Mid/Max—tests the extremes and mid points of all parameter settings. Covers all options for non-range parameters.

• Hill Climb—a simple algorithm. Fast but prone to get stuck in local optima.

• Adaptive Thermostatistical SA—The main algorithm, a variant of simulated annealing with extra adaptive nature. Includes some elements of tabu search. Developed by Lanner in conjunction with optimization experts specifically to tackle simulation experimentation.

• Six Sigma Algorithms—An adaptation of the SA algorithm to ensure that the amount of change from a defined AS-IS situation is limited. Useful for determining the best improvements that can be made without wholesale redesign of a process.

WITNESS Optimizer presents its reports as the algorithm is running. The main tables and charts can be interacted with during the run to view, sort, copy, paste and analyze. Optimizer not only searches for the optimum solution but also usually completes the analysis quickly. Using intelligent algorithms, which measure and react to the success of each analysis path, the best solution is often found by trying fewer than 1% of all possible combinations.