Environmental effects on MCM tactics planning

Abstract

A successful anti-submarine warfare search planning tool, the Genetic Range-dependent Algorithm for Search Planning (GRASP), is adapted and evaluated for the purpose of planning near-optimal reconnaissance plans for the mine counter-measures community. High-fidelity range- and azimuth-dependent sonar performance predictions over a high-resolution grid are ingested by a genetic algorithm, which uses Monte Carlo simulation and Bayesian detection statistics, to evaluate and refine proposed search paths against a given target distribution. In essence, GRASP simulates a Darwinian evolution of reconnaissance paths to obtain the statistically best path based on Cumulative Detection Probability (CDP). In previous proof-of-concept work, GRASP provided a ladder-like solution in an environment where such solution was expected by search theory. Further, as the environment was perturbed slightly (just outside the bounds for which search theory can determine the optimal path), GRASP produced a path that was a variant on, and an improvement over, the intuitively expected path. In the present work, the efficiencies of acoustically blind and acoustically sensitive mine clearance strategies are compared in a real environment. The GRASP results reveal improvement in search coverage obtained by exploiting the environment with respect to sonar performance.