Excel Solver uses classical linear programming and
nonlinear programming methods designed for problems where every locally
optimal solution is also globally optimal.
In contrast, OptQuest uses a combination of metaheuristic
procedures from methods such as Tabu Search, Neural Networks, and
Scatter Search. All of these methods are effective for models that
Excel Solver can solve, as well as for models with local solutions
that are not globally optimal, which is usually true in the real world.
OptQuest has three major advantages over the Excel
Solver:
- OptQuest finds the optimal solution in an environment
of uncertainty
- OptQuest will not get trapped in local optimal
solutions
- OptQuest can handle nonlinear relationships difficult
to describe with mathematical formulas
Uncertainty
Most models have uncertainty and variability in them,
such as uncertain supplies, demands, and prices, or variable costs,
flow rates, and queuing rates. OptQuest lets you define the uncertainty
for each spreadsheet value you need to, in the form of assumptions
or decision variables.
Excel's Solver has no way of handling uncertainty
or variability.
Local solutions
Many real-world problems have nonlinear components.
The nonlinear components can create a solution space with many local
optimal solutions that might be significantly inferior to the true
optimal solution. OptQuest is designed to find global solutions
for all types of objectives, especially complex objectives with
many local, inferior solutions.
Excel Solver's method of finding a solution will
get trapped in the first solution it finds, whether it is a local
solution or the true optimal solution.
Nonlinear relationships
Many real-world problems have nonlinear relationships
that cannot be described by equations or formulas used in mathematical
programming. Such problems require simulation. OptQuest can handle
these types of relationships, since it uses simulation.
Excel's Solver can only handle relationships that
you can specify by an equation or mathematical formula.
