Simulation model of an autonomous underwater vehicle for design optimization

In this research, a stochastic simulation based design tool that facilitates design in a dynamic environment is developed. The simulation includes a full dynamic model of an Autonomous Underwater Vehicle (AUV) that was developed to evaluate the effectiveness of such a craft given a set of physical attributes. These attributes include but are not limited to: speed, mass, moments of inertia, control gains in the auto-pilot, and target detection capabilities. The effectiveness of the AUV is based upon the probability that it can successfully complete a given mission and how quickly it can complete this mission. The model is coupled with the Applied Research Lab's unclassified AUV problem that can compute weights, speeds, and efficiencies based on propulser types, sonar configurations, and various other subsystems. In order to use this model in an optimization framework a mission was selected. This mission was to hit an oncoming torpedo before the torpedo was able to hit its target. The objective of this mission was to maximize the probability of successfully hitting the torpedo before the torpedo reaches its own target. In order to calculate this probability, the simulation was run with many different starting configurations including: different speeds of the oncoming torpedo, evasive maneuvers of the oncoming torpedo, and also various spacial orientations between the AUV and the targeted torpedo. This paper includes a detailed description of the simulation model, the development of the multidisciplinary design problem, and results obtained from the optimization of this problem.