As already noted in chapter 5 engineering design is almost always concerned with most problems that have multiple, often conflicting goals and a host of demanding constraints. We turn now to the use of surrogate based approaches to design improvement when dealing explicitly with multiple objectives. We note in passing that in some cases the designer may be able to reduce problems with multiple goals to single objective problems by some suitable weighting function that combines the goals of interest. In aerospace applications this can often be something like weight, payload capacity or cost, provided suitable conversions to a common form can be devised. When such an approach can be taken the problem reverts to a single objective search and the techniques already discussed can be applied. Sometimes however the
correct weighting to apply between goals is not obvious or the designer does not wish to commit to a fixed weighting while carrying out design searches: this leads to the concept of Pareto optimality and sets of designs that must be found and considered simultaneously.
Optimization Codes
%d bloggers like this:
