Research On UUV Formation Cooperative Control In Complex Environment

Due to the enhancement of human awareness of the protection of the marine environment,and the demands for the development and utilization of marine resources,Unmanned Underwater Vehicles(UUV)have gained a lot of attention.Many research institutes,technology corporations,and industry and academic research experts all over the world are interested in UUV,also because of its small size,strong maneuverability,and wide range of task operation capabilities.However,complicated missions in a complex and changing marine environment with a single UUV may be difficult.A multi-UUV formation can broaden the scope of a single UUV mission,improve system redundancy,and improve the efficiency of executing complex and laborious operations.Therefore,UUV formation control has become an important research topic in the field of UUV control.In this dissertation,the formation control method in the complex environment of multiple UUVs based on consistency is investigated,the main research contents are as follows:Firstly,to deal with the nonlinear strongly coupled UUV mathematical model,the model is transformed into a second-order affine model through coordinate transformation.Aiming at the problem of large delay in underwater acoustic communication of UUV formations,an UUV formation rendezvous control method based on information consistency is proposed,and the formation keeping control law of follower UUVs is designed.The stability of the system and the system convergence conditions are given.The simulation experiment verifies the effectiveness of the distributed control law designed for the UUV formation system in discrete time without and with communication delay.In addition,in view of the problem of random loss of UUV formation communication information,an UUV formation tracking control method based on a semi-Markov process considering time-varying communication topologies are proposed,and the formation tracking control law of follower UUVs is also designed,respectively.The stability of the UUV formation system with and without communication delay and the system convergence conditions are given,and the precise 3D trajectory tracking control of the UUV formation is realized.Secondly,for the tracking control problem of UUV formation in a complex unknown ocean disturbance environment,a UUV formation tracking control method based on aggregated cooperative learning via distributed Gaussian process regression algorithm is proposed.The GPR method is used to learn about the unknown marine environment.Each follower UUV can share state information and the prediction results of environmental dynamics from GPR at the same time.Using the characteristics of information exchange between multiple UUVs,each follower UUV can simultaneously share state information and the prediction results of Gaussian process regression.An aggregated cooperative learning via distributed Gaussian process regression based on Gaussian process posterior variance is designed.Then we propose a decentralized UUV formation tracking controller based on this algorithm,analyze the stability of the system and give a guaranteed unified error bound.and simulation experiments verify the effectiveness of the data-driven distributed controller designed,wherein the UUV formation system is disturbed by an unknown environment.The results show that accurate 3D trajectory tracking control of UUV formation can be realized.Next,in order to improve the aggregated cooperative learning algorithm via distributed Gaussian process regression based on Gaussian process posterior variance and reduce the computational complexity of the algorithm,an aggregated collaborative learning algorithm based on correlation-aware distributed Gaussian process regression is designed.The non-centralized UUV formation tracking controller is designed,the stability of the system is analyzed and a guaranteed unified error bound is also given.The effectiveness of the distributed controller is verified by simulation experiments,and the precise three-dimensional trajectory tracking control of the UUV formation under unknown environmental interference is achieved.Finally,in order to solve the problem of excessive dependence on the predicted value given by adjacent UUVs in the aggregated GPR algorithm,an independent distributed localization learning algorithm,and a formation tracking controller based on this algorithm considering the marine environment interference is designed.The stability of the system is analyzed and a unified error boundary is given.Finally,the effectiveness of the proposed controller is verified by simulation.Then,in order to improve the accuracy of the learning and prediction results of local GPR,a fusion collaborative learning algorithm based on dynamic consistent distributed independent GPR is proposed,and a distributed UUV formation tracking controller is designed based on this algorithm.Stability and a unified error boundary are given.Simulation experiments verify that the learning algorithm based on dynamic consistency can effectively improve the effect of UUV formation control and UUV formation can achieve precise formation tracking control.To sum up,this paper starts by analyzing the motion model of UUV,establishes the control and communication framework of the UUV formation system,and studies the problems of formation rendezvous with communication delay and formation tracking control with timevarying communication topology under stable environment for multi-UUV.And the problem of formation tracking control in the case of environmental interference based on GPR is also investigated.A series of feasible and effective control schemes are proposed to ensure that multiple UUVs can achieve precise formation control tasks in a specific formation.