An iterated greedy heuristic for collaborative Human-UAV search of missing tourists

Abstract

This paper investigates a problem of collaborative human-unmanned aerial vehicle (UAV) search for missing tourists. A mathematical model including detection stage and rescue stage is constructed to describe practical operation scenarios. The objective function is formulated using conditional probability theory by introducing the weighted probability that tourists being rescued at an allowable time limit. To solve this combinatorial optimization problem, an iterated greedy heuristic (IGH) is designed to search solution space. IGH starts with an initialization procedure based on problem-specific knowledge, then a neighborhood perturbation-based local search method is implemented on perturbed solution to intensify exploitation capability. Meanwhile, an acceptance criterion is presented to avoid premature convergence. Ultimately, Taguchi method is utilized to calibrate parameters of IGH for best performance. Comparison with other algorithms on objective function value and search performance validates the effectiveness of our proposed algorithm.