Determining the appropriate values for Minimum and Maximum constraints in the optimizer model is a critical and challenging part of the optimization process. Finding numerical values for current recipe specification requires gathering relevant data and this might be difficult sometimes. Without any objective restrictions (nutritional limitations, Feed inventory availability, etc.), we often make rough estimates. Because of the uncertainty of the true value of a constraint, it is important to find out how the solution derived from the optimizer would change (if at all) if the constraints assigned were changed to other plausible values. This process is referred to as sensitivity analysis.
We can apply such analysis to the various ingredients included in the solution by simply varying one value in the constraints of a given amount and examine the impact that the change has on the model’s results. This is known as one-way sensitivity analysis, as only one parameter is changed at one time. The analysis could, of course, be repeated on different parameters at different times.
NDS introduces a new sensitivity analysis for solutions of the Nonlinear Optimizer. The purpose of the new sensitivity is to help the user to get better solutions in terms of optimality or, in case of already optimal solutions, obtaining better objectives functions (e.g. lower cost).
The key concept behind the sensitivity is the capability of setting new constraints for the feeds in the specifications in order to improve the accuracy of the solutions; every time a feed amount in a solution is set on its minimum or maximum bounds, moving the constraint of a certain amount beyond its current value could give a chance to get better solutions when running a new optimization. Moreover in the sensitivity screen the user can choose a new amount for the feed beyond its constraint (minimum or maximum) and preview the solution values after this change.
After performing an optimization and choosing a solution you can open the sensitivity screen with the Sensitivity button in the top ribbon of the optimization screen.
Note that for each solution there is a specific sensitivity, as each solution can have its feed amounts set to different minimum or maximum; moreover changing the same feed amount in different solutions can lead to different values for the model parameters of each solution, due the intrinsic nonlinearity of the biological model.