Autonomous mission management and intelligent decision for AUV

This paper presents an intelligent decision algorithm and an on-board architecture designed to enable an autonomous underwater vehicle (AUV) to carry out a survey mission autonomously. The connotation of global model in decision algorithm is analyzed, including environment information, mission information and Self-states of AUV. The survey mission of multi-objective operation areas in an unsafe zone including forbidden areas and obstacles is defined. Based on these, the intelligent decision algorithm including global path optimization and speed optimization is researched. Operational research method is used to optimize the path and genetic algorithm is used to optimize the speed. The on-line mission management architecture is designed to call the decision algorithm and supervise the implementation of the autonomous mission based on Petri net formalism. The feasibility and the algorithm effectiveness of the architecture and intelligent decision algorithm are checked by lake experiment in nominal and degraded situations.