Section: New Results

Two phase solution for an intelligent moving target search problem based on a 0–1 linear model

We developed a generic discrete model for the moving, intelligent target problem. Our objective is to maximise the probability of detection of the moving target with respect to target and searcher’s constraints. The solution method proposed in [15] is composed of two stages. The first one aims at providing a large-scale strategy based on an Integer Linear Program approach. As a direct solution of this problem is not practically possible, we use a decomposition of the problem into a searcher’s strategy on one side, and the target’s strategy on the other side. A good strategy for the searcher is determined using a sliding window procedure. Concerning the target, our approach consists in simulating some of the target’s possible strategies and considering each of these possibilities as an independent and deterministic entity. The second stage is dedicated to adjusting the large-scale strategy provided by stage 1. Numerical results are presented so as to assess the impact of our approach.