Abstract
A new approach to the problem of minimizing a quartic (fourth order) cost function, subject to linear constraints, is presented. A “satisficing range” is introduced for each variable (i.e., a range in which all values of the variable are considered to entail the same minimal cost), leading to a new linear formulation (“range programming”). Improvements in results over previous approaches are reported for a specific example (due to Goodman), and should be expected in general.

This publication has 0 references indexed in Scilit: