Content area
Full Text
Keywords
Logistics, Quality management, Information resources management
Abstract
New software products are now available that offer solutions to operational problems that could not be handled by traditional linear programming. The problem of quality performance described could previously only have been handled by writing a special purpose simulation program. Now it can be solved in a spreadsheet environment using Evolves and RISKOptimizer software. Evolves can identify a global rather than a local optimal solution when non-linear relationships are present. RISKOptimizer takes the process one step further. It combines @RISK simulation capablity with Evolves s optimization algorithms to handle stochastic variables (uncertain demand and prices).
RISKOptimizer
Linear programming is a powerful analytical tool used to obtain solutions for various business problems expressed in terms of linear equations. Unfortunately, linearity may be difficult to preserve in modeling business situations. Applying linear programming techniques to a non-linear system of equations can lead to a local rather than a global optimal solution. Moreover linear programming is of no avail if a model contains IF statements, lookup tables, and other spreadsheet features.
Palisade Corporation (www.palisade.com) has recently introduced Evolver and RISKOptimizer to handle problems that cannot be solved by traditional linear programming. Evolver can identify a global rather than a local optimal solution when non-linear relationships are present. Evolver examines an analytical landscape of hills and valleys. In maximizing an objective cell, Evolver starts at some location and begins a process of climbing up the slope to reach its peak. At the same time, Evolver examines other locations to ensure that the hill it is climbing is also the highest hill in the domain. Evolver progresses to a solution by intelligent trial and error through its genetic algorithms. The genetic algorithms use past simulation results to generate a new set of values for the adjustable cells in its search for the optimal value for the objective cell.
RISKOptimizer takes the process one step further. Both linear programming and Evolver assume that demand or price for a product is discrete and, therefore, certain. Suppose that demand or price for a product is stochastic, that is, uncertain. In other words, future demand is not 100 per day, but 100 per day plus or minus 5 percent. RISKOptimizer combines @RISK...