Cooperative Formation Control Of Unmanned Surface Vehicles With Prescribed Performance

Unmanned surface vehicle(USV),also called as autonomous surface craft/vessel/ship,is a kind of intelligent floating vessel with autonomous voyage in open waters,which can be deployed in waters that is unacceptable to send a manned vessel,e.g.,high threat environments.The USV has many outstanding advantages,such as small size,high speed,and flexible maneuverability.The most recent years have witnessed the increasing deployment of multiple USVs instead of a single USV to perform various maritime tasks such as ocean resource exploration,offshore topographic survey,and environmental sampling.Limited energy and coverage of the USV make it challenging to accomplish these tasks within reasonable time and cost using a single USV.The robustness against the system faults is another weakness comparing with multi-vehicle system.Consequently,cooperation of a group of vehicles is more desirable when executing these missions,where the coordinated operation of the vehicle group is an important part of the cooperation.In general,the coordinated operation could be divided into guidance,navigation and control of the vehicle group.Thus,cooperative control of multiple USVs is practically motivated,in which cooperative formation control of multi-vehicle system has attracted considerable attention in control and marine engineering communities during the past two decades.However,the problem of cooperative formation control for multiple USVs with prescribed performance is challenging under limited sensing capability of the onboard sensors mounted on the USVs,safety requirement,and complex maritime environment,which motivates this thesis with the main contents given below.Firstly,the platoon formation control problem for multiple fully actuated USVs is investigated,in the presence of modeling uncertainties and time-varying external disturbances.The control objective is to make the vehicular platoons proceed along a given trajectory while maintaining a desired line-of-sight(LOS)range between each vehicle and its predecessor.To provide transient performance specifications on the formation errors consisting of LOS range error and angle error,the exponentially decaying functions of time are employed to constrain the formation errors.Using prescribed performance control(PPC)methodology,neural network(NN)approximation,disturbance observers,dynamic surface control(DSC)technique,and Lyapunov synthesis,an adaptive formation controller is proposed,which guarantees collision avoidance and connectivity maintenance between two consecutive vehicles during motion.Secondly,the decentralized leader-follower formation control problem for a group of fully actuated USVs with prescribed performance and collision avoidance is addressed.The collision avoidance constraint imposed on the follower and its leader is incorporated into the prescribed performance constraints imposed on the formation errors by appropriately selecting the performance functions.In the kinematic design,the DSC technique is introduced to avoid using the leader’s unavailable acceleration.To compensate for the uncertainties and disturbances,the adaptive control technique is employed to estimate the uncertain parameters including the upper bounds of the disturbances and the NNs are applied to estimate uncertain nonlinear dynamics.Consequently,a decentralized adaptive formation controller is developed,which guarantees the prescribed performance constraint and avoids collision between each vehicle and its leader.Thirdly,the leader-follower formation control problem for a group of underactuated USVs with non-diagonal inertia matrix subject to prescribed performance constraints is investigated,in the presence of modeling uncertainties and external time-varying disturbances.In practical applications,some USVs equipped without independent sway actuator belong to a class of underactuated systems,which means that the USVs have three degrees of freedom to be controlled but have only two independent control inputs.Note that the nonzero off-diagonal terms in the system matrix result in yaw moment control acting directly on the sway and yaw dynamics,which causes difficulty in simultaneous stabilization of the sway and yaw dynamics using only one control input.To overcome difficulties raised by underactuation and nonzero off-diagonal terms in the system matrices,a novel transverse function control approach is developed to introduce an additional control input in backstepping procedure.No communication is required among the USVs,but every USV is only equipped with on-board sensors to measure the LOS range and the relative bearing angle with respect to its leader.The connectivity preserving constraint arisen from the limited sensing capability and the collision avoidance constraint resulting from the safety requirement are imposed on the LOS range and the relative bearing angle between every follower and its leader.These constraints are subsequently incorporated into the PPC-based formation control design.The adaptive control technique is employed to estimate the upper bounds of the leader’s velocity and the upper bounds of the external disturbances.Based on transverse function control approach,backstepping procedure,DSC technique,and PPC methodology,an adaptive formation controller is then developed,which achieves satisfaction of prescribed performance specifications on the formation errors and guarantees collision avoidance as well as connectivity maintenance between every follower and its leader.Finally,the problem of cooperative learning from adaptive neural formation control for a group of underactuated USVs with modeling uncertainties is investigated,where the formation errors are subject to prescribed performance constraints.To cope with the inherent modeling uncertainties,NNs are typically employed to obtain NN-based adaptive formation control strategies.However,these NN-based adaptive formation controllers have to readapt to the modeling uncertainties even when the same control task is performed,since the learning capability of NNs is unexploited.In this sense,the NN-based adaptive formation controllers are computationally expensive due to the lack of knowledge on system dynamics.Thus,the problem of cooperative learning from adaptive neural formation control for multiple uncertain underactuated USVs is investigated.A coordinate transformation is introduced to overcome the difficulties caused by off-diagonal system matrix.Under limited communication range,the connectivity preservation as well as collision avoidance among the initial connected vehicles are achieved by guaranteeing the inter-vehicle distances converge to small neighborhoods of the desired distance.Meanwhile,the convergence of bearing angles to small neighborhoods of desired bearing angles avoids the possible controller singularity problem arising from underactuation and achieves the predefined formation shape.The prescribed performance constraints are imposed on the formation errors to improve the transient and steady-state performances.Using the deterministic learning theory,the modeling uncertainties are locally accurately identified/learned by the localized radial basis function(RBF)NNs along the union of all vehicles’ state orbits in a cooperative way.The learned knowledge is stored in constant RBF networks and is reutilized to develop experience-based formation control protocol to improve the control performance including reduction of the computational burden.Comparative simulations are carried out to verify the improvement of control performance with knowledge reuse,when performing a same task.