Design And Simulation Of Behavior-Based Modular Decision-making Control System For Autonomous Underwater Vehicle

Since mid-1990s, autonomous underwater vehicle (AUV) has been more widely used in marine and scientific exploration and the military field. However, no matter what field AUV has been used in, the control system is the key of its ability to successfully complete the various tasks. This paper describes a behavior-based modular control system of AUV which integrates the behavior-based control method and modularization idea. It is expected to control AUV effectively in the limited sensor resources.First of all, the decision-making control system introduced in this paper is behavior-based, which has the majority characteristics of behavior-based control method. For better characterizations of the various behaviors and more logically building of the decision-making control system, the entire system is composed of two kinds of behaviors: Rational Behavior and Perceptual Behavior. Rational Behavior is advanced behavior which packages complex algorithm, all of the Rational Behaviors constitute the Rational Behavior System; the algorithm of Perceptual Behavior is simple and high real-time, all of the Perceptual Behaviors constitute the Perceptual Behavior System; these two systems cooperate with each other to ensure that AUV can complete the mission safely and effectively.Secondly, the modularity is introduced into the Behavior-Based Decision-making Control System. Not only all the behaviors have been modularized, but also Basic Behavior Modules are introduced into the system to modularize all of Rational Behaviors and Perceptual Behaviors. Key features of the system include: easy addition and removal of behaviors, ability to execute non-conflicting behaviors on parallel, and conducive to analysis, simulation and compiling and so on.Finally, this paper selects a few typical behaviors, combining the control characteristics of the test platform AUV"C-RANGER", using SIMULINK, VRML and FUZZY toolbox in MATLAB so as to complete a simple and efficient simulation platform. The entire building process of the simulation platform fully shows the advantages of construction, principle analysis and simulation of the system after modularity.