We show ways in which differential evolution, a member of the genetic/evolutionary family of global opti- mization methods, can be used for the purpose of discrete optimization. We consider several nontrivial problems aris- ing from actual practice, using differential evolution as our primary tool to obtain good results. We indicate why meth- ods more commonly seen in discrete optimization, such as integer linear programming, might be less effective for prob- lems of the type we will consider.
展开▼
