Coordinated Path-following Control Of Multiple Unmanned Underwater Vehicles For Region Search

With the sequentially promotion of China’s marine strategies and the fast-changing of unmanned technology,the unmanned underwater vehicle(UUV),one intelligent marine equipment,revolutionizes the war mode in military fields and spawns a huge industrial chain in civilian fields.Compared with the individual system,an UUV swarm system has remarkable advantages in terms of flexibility,fault tolerance and cooperation,which has become the inevitable trend of artificial intelligence and the core domain of advanced marine equipments.In this context,the coordinated path-following control of UUVs has been investigated in this dissertation,which can provide theoretical support and technical guidance for large-scale,omnidirectional and autonomous underwater region search.By virtue of the Navigation-Guidance-Control structure widely used in the unmanned system control,this dissertation has constructed a double-layer distributed control framework consisting of kinematics guidance and dynamics control.Within the above framework,coordinated path-following guidance methods and control methods have been respectively proposed by combining with multi-agent system theory and advanced control theory.By the aid of different coordinated guidance and control strategies,an efficient technical scheme related to region search is successfully creatively developed.Main contributions are as follows.(1)Taking high-efficiency coordinated search into consideration,a coordinated path-following control scheme for multiple UUVs along multiple parameterized paths is created,where the desired paths are planned.Within the guidance level,the path-parameter interaction based surge-heading guidance method is developed for under-actuated UUVs subject to underwater communication bandwidth constraints,where the path update protocol is incorporated into the light-of-sight guidance,thereby ensuring that vehicles enable to track the multiple paths synchronously.Within the control level,to enhance the approximation accuracy of the existing neural network,the scale factors and translation factors based wavelet neural network approximator is developed for estimating unknown disturbances;in addition,the adaptive disturbance compensation based dynamics controllers are further proposed and thus make tracking errors asymptotically stable.Moreover,a solution to the region-searching is creatively addressed,where boustrophedon paths based global path-planning method is developed using nonlinear functions,thus ensuring the completed coverage for one region.Finally,simulation results are deployed to illustrate the efficacy of the proposed coordinated path-following control scheme as well as the region search strategy.(2)Taking swarm self-organization of deployment and retrieval into consideration,a coordinated path-following control scheme for multiple UUVs along one parameterized path is also created,where vehicles can achieve dynamic self-organization.Within the guidance level,the formation pattern guided by one path and the finite-time sideslip observer based surge-heading guidance law are developed,respectively.Worthy of mention is that compared with the existing work,the proposed sideslip observer overcomes the assumption that the sideslip angle must be regarded as a small constant.And the proposed guidance method can make a fleet of UUVs track the predefined path with desired formation patterns.Within the control level,model parameter perturbations and unknown external disturbances are dealt simultaneously,and are subtly identified by the proposed fuzzy logic system.Compared with the existed work where the fuzzy approximator is based on tracking errors,fuzzy weights in this dissertation are updated using prediction errors,which can improve the system transient performance.Furthermore,this dissertation considers the self-healing and extended abilities for a swarm system under complex marine environments.The virtual structure and local information interaction based dynamic self-organizing reconstruction strategy is first developed which includes the swarm reconstruction under homogeneous formation transformation and heterogeneous formation transformation,the swarm reconstruction under invalid agents and new agents,thereby enhancing system stability and invulnerability.Finally,simulation results are deployed to illustrate the efficacy of the proposed coordinated path-following control scheme as well as the dynamic self-organizing reconstruction strategy.(3)Taking flocking migration among different regions into consideration,a novel coordinated path-following control scheme for multiple UUVs with unfixed pattern constraints and obstacle constraints is developed in this dissertation.Within the guidance level,a self-organization swarm model with center position and member dispersion is first constructed for multiple UUVs,thus enhancing flexibility and adaptability of the coordinated control system.And the neighbor information interaction based swarm center observer is employed to implement swarm aggregation.By combining with the proposed swarm model and the surge-heading guidance method,the coupling guidance scheme based swarm velocity guidance laws are successfully developed.In addition to this,an improved artificial potential field using ring-shaped repulsion is incorporated into the guidance laws,thus forcing vehicles to bypass obstacles smoothly.Within the control level,to solve the unknowns in dynamics model,unknown state observers using the differential inclusion based finite-time control theory are developed under the reasonable assumptions that unknowns acting on vehicles are bounded and differentiable.Besides,by virtue of asymmetric barrier Lyapunov functions,the finite-time disturbance observer based surge-heading controllers are proposed for under-actuated UUVs subject to motion constraints,thereby guaranteeing the tracking performance.Finally,through simulation results of coordinated path-following and obstacle avoidance,it is demonstrated that the proposed control methods are feasible.This dissertation deeply researches the problem of coordinated path-following control for multiple UUVs aiming at region search.To be specific,surge-heading guidance laws are designed to synchronize vehicles’ attitudes,and disturbance observer based active anti-disturbance rejection dynamics controllers are developed to drive UUVs to track the desired parameterized paths.By virtue of the existing virtual prototype based visual simulation system,simulation studies related to coordinated control of an UUV swarm are deployed,including the coordinated path-following along multiple paths and one path under fixed formation pattern,the coordinated path-following along single paths under unfixed formation pattern,the boustrophedon paths based underwater region search and the dynamic self-organizing reconstruction.Simulation results exhibit that the above coordinated guidance and control scheme have an excellent prospect and contribute positively to equipment industrialization of unmanned swarm systems.