Two New Approximate Solution Techniques for a Moving Target Problem When Searcher Motion is Constrained

摘要

The objective of this thesis is to develop and test two new approximate solution techniques for a moving target problem in discrete time and space where both the searcher and the target have constraints on their paths. The first technique is an application of the heuristic Local Search Method and the second technique is an application of the Frank-Wolfe Method. The motivation for looking at approximate methods is that the problem is NP-complete and optimal solution techniques become impractical for large size problems. Experiments showed that the Local Search Method approach is an efficient technique for obtaining approximate computation solutions. However, the Frank-Wolfe Method approach does not perform well for the problem. Keywords: Target detection; Paths; Fortran; Algorithms. (Author)