Maritime Dynamic Resource Allocation and Risk Minimization using Visual Analytics and Elitist Multi-Objective Optimization
Loading...
Authors
Raha, Mayamin Hamid
Issue Date
2023
Type
Thesis
Language
Keywords
Intensity Maximization , Multi-Objective Optimization , Resource Allocation , Risk Minimization , Threat Evaluation , Visualization
Alternative Title
Abstract
Enhancing the safety of protected regions around Navy vessels is one of the most challenging research topics in maritime domains. Robust tactical resource allocation depends on understanding of how the placements, configurations, orientations of multiple assets affect both the area and intensity of coverage around the ships. Towards this end, we built a unique resource allocation problem where we apply a randomized genetic algorithm for searching through a space of 2144 possible parameters representing area coverage and orientation of 6 tactical assets. Our elitist genetic algorithm yielded a maximum fitness value of 90%, 98%, 100% within 50, 150 and 300 generations respectively. Moreover, we put forward a distinctive constrained dynamic resource allocation problem specific to USS Arleigh Burke Destroyer model (DDG-51), where the assets are defenses and coastal guards having binoculars. To solve this, we have used a cross-generational elitist selection based evolutionary algorithm (EA) where our objective is to maximize area of coverage and minimize risksimultaneously. It is a non-deterministic polynomial-time hard (NP-Hard) problem which required searching through a space of 248 parameters and resulted in a fitness value of 98% within 35 generations. Furthermore, we present two novel visualization techniques addressing both types of resource allocations.