If Solver stops with a solution (set of values for the decision variables) that is different from what you expect, or what you believe is correct, please follow the suggestions below. You can usually narrow down the problem to one of a few possibilities.
· Be sure to read the message in the Solver Results dialog, and check the explanation of the message in this Help topic.
· Consider the possibility that the solution found by Solver is correct, and that your expectation is wrong. This may mean that what your model actually says is different from what you intended.
· If you receive the message “Solver could not find a feasible solution,” read the topic Understanding the Feasibility Report.
· If you receive the message “The linearity conditions required by this LP Solver are not satisfied,” read the topic Understanding the Linearity Report.
· If your model includes integer, binary or alldifferent constraints, read the topic Integer Constraints and the Integer Optimality Tolerance.
· Select the check box to Show trial solutions in the Solver Options dialog and re-solve. Solver will pause with the message “Solver paused, current solution values displayed on worksheet.” Click Continue to see the path towards the solution taken by Solver.
· Read the topic Problems with Poorly Scaled Models. This is one of the most common causes of unexpected error messages, suboptimal solutions.
· Read the topics on Limitations on Smooth Nonlinear Optimization, affecting the GRG Solving method, and Limitations on Non-Smooth Optimization, affecting the Evolutionary Solving method.
Depending on the message, the solution values in the decision variable cells may be a globally optimal solution, a locally optimal solution, only an approximate solution, or no solution (for example if no feasible solution exists).