Research On AUV Behavior Control Method Based On Interval Programming Model

With the rapid development of Autonomous Underwater Vehicle(AUV)platforms,sensors,communications,and intelligent algorithms,AUVs have been widely used.Behavior control is one of the key technologies of AUVs,and it is also the basis for completing underwater tasks.Since the Interval Programming(IVP)model is an action selection control algorithm based on the behavioral architecture,it can flexibly plan AUV behaviors,has the characteristics of small calculation amount and strong real-time decision-making ability.Once the model was put forward,it has received extensive attention from scholars.In this paper,the distributed control architecture of AUV,IVP model architecture based on behavior,wrap-around waypoint tracking,and real-time obstacle avoidance are studied accordingly.First,in order to build an AUV system with a high degree of autonomy,this paper designs an IVP model architecture based on behavior to solve the problem of AUV behavioral competition.At the same time,the distributed behavior control architecture of AUV was built,and the whole AUV was divided into two parts: the body section and the load section.The load section provided decision-making instructions such as heading,speed and depth,while the body section executed control decisions and fed navigation information such as position back to the load section,to ensure the real-time decision-making ability of the whole control architecture.Secondly,in view of the behavior control problem of AUV when detecting static sinking targets in a complex shallow sea environment,the traditional "lawn mower" behavior mode is time-consuming and difficult to process data in real time,etc.,designed a wrap-around waypoint tracking behavior control method.This behavior takes the static sinking target as the center,and plans a wrap-around tracking path point.When the AUV is far away from the tracking path point,define the proximity attractor to guide the AUV to approach the tracking path point;When the distance deviation between the AUV and the tracking path point is small,define the tracking attractor to quickly realize the consistency of the AUV and the track at the tracking path point.In addition,the arrival conditions of each path point and the transition process between path points are defined,and the behavioral utility function is established to evaluate the control action of the AUV.At the same time,the behavioral architecture was loaded on the actual AUV platform for lake test,which verified the feasibility of the behavioral algorithm.Finally,aiming at the problem of AUV avoiding obstacles in unknown environments,traditional behavior control methods have a large amount of calculation and poor adaptability to unknown environments,etc.,a real-time obstacle avoidance behavior is designed.In this behavior method,the included Angle between the AUV and the barrier is defined as the repulsive element,and the real-time obstacle avoidance behavior is modeled based on behavioral kinematics method.In addition,considering the uncertainty of the AUV’s own sensor in sensing the size of obstacles,a buffer polygonal area is set to achieve a balance between AUV maneuvering safety and feasible paths.On this basis,the obstacle avoidance behavior utility function is established to evaluate the control actions of the AUV at each moment,and the IVP model is used to coordinate control among multiple competing behaviors.