Does it matter what values are in the adjustable cells when I click Start for an optimization?

 

Answer:
Yes—but how much depends on the optimization engine.

Genetic Algorithm: Starting values matter a lot. You should begin with a feasible solution (all constraints met). If the initial values violate constraints, the optimizer spends extra time finding feasible solutions before focusing on the main optimization. Also, starting near a local optimum may cause the algorithm to converge there rather than the global optimum. Trying different feasible starting points can help find better solutions.

OptQuest Engine: Starting values matter very little. The engine uses its own methods to generate feasible solutions even if the initial values don’t meet constraints. If you leave the Engine setting as Automatic, the software usually picks the best engine for your problem.

Tip: With complex models, you can experiment with different feasible starting points to see if you find a better optimum with the genetic algorithm.

 

 

 

Related Article:

Debugging RISKOptimizer and Evolver Models