【作者】 牟春晖；

【导师】 边信黔；

【作者基本信息】 哈尔滨工程大学 ， 控制理论与控制工程， 2011， 博士

【摘要】 早期的编队控制只要求UUV间的相对位置或距离达到期望值,对单个UUV的路径没有加以限制,而多UUV在进行海底管线检测、协作建立海底地图等特殊任务时,UUV必须沿着指定的路径运动,显然,传统的编队控制律不能解决此类特殊的编队问题。本文针对同构的某型无人水下航行器(Unmanned Underwater Vehicle, UUV)在完成路径跟踪(直线路径和曲线路径)的基础上对多UUV的协调编队控制进行了研究。主要研究内容如下：首先,依据多UUV的协调路径跟踪控制任务,提出了一种具有“时空解耦”特性的多UUV分散控制体系结构。该体系结构将空间任务(路径跟踪)和时间任务(协调编队)解耦。第一步,UUV执行各自的路径跟踪控制律,完成空间上的控制任务,即保证各UUV运动收敛到指定路径上；第二步,通过UUV之间的信息交换设计协调控制律,调整UUV的前向运动速度,完成时间上的控制任务。另外,介绍了多UUV系统协调编队时水声通信的应用；在通信拓扑结构表示上提出了图论表示法。该部分内容为后续章节控制策略和控制器的设计奠定了理论基础。其次,基于上述设计的体系结构,研究了空间直线路径跟踪下多UUV的协调编队控制问题。在路径跟踪控制器设计中,忽略横摇的影响,基于视距导航法建立了单个UUV的路径跟踪误差模型,将UUV的位置误差跟踪控制分别转化为UUV的航速误差艏向角误差和纵倾角误差的渐近稳定控制,并结合级联控制方法和反馈线性化控制方法分别对其设计了控制器,该方法设计的控制器简单,易于工程实现。针对领航者法中因领航者失效造成全局崩溃的问题,提出了基于图论的分散式队形控制方法,该方法提供了UUV之间的状态信息反馈通道,UUV根据所得状态信息,可交互式调整自身的速度来设计速度协调控制器。相对于领航者法中领航者不能得到跟随者的状态反馈信息,其更能有效的保证UUV间的信息交互和队形控制。再次,由于直线路径跟踪控制是在平衡点附近的邻域内的镇定控制,要求位置上实现跟踪即可,而曲线路径跟踪控制中不仅要求位置上实现跟踪,而且速度的矢量方向必须沿路径的切向方向。两者的控制器设计要求不同,为此,提出了基于虚拟参考目标点“牵引”的路径跟踪控制策略,对UUV设计的动力学路径跟踪控制器和运动学路径跟踪控制器分别实现了速度和位置的跟踪,并利用Lyapunov直接法和输入-状态稳定性定理证明了路径跟踪子系统的稳定性、以及系统状态和控制输入的有界性,该方法提高了路径跟踪控制的精度。针对水下声通信受到通信带宽的限制问题,提出了只交换单变量协调状态的信息交换方式,该信息交换方式相应地减少了UUV间的通信量,满足了水下声通信带宽窄的限制条件。在速度协调控制器设计中,利用图论描述多UUV间的通信拓扑形式,建立协调误差模型,通过对虚拟参考目标点的整体协调控制,且.考虑了UUV执行机构的饱和特性,设计了含有双曲正切函数的协调编队控制器,从而保证了系统的控制品质。最后,针对理想环境的局限性,研究了具有环境约束的多UUV的协调路径跟踪控制问题,主要考虑了海流干扰对单个UUV路径跟踪控制的影响和水声通信约束对多UUV间协调控制的影响。针对UUV路径跟踪过程中未知海流的干扰,基于Lyapunov稳定性理论设计了一种海流观测器来估计未知恒定海流流速,并应用LaSalle不变集原理对路径跟踪控制子系统进行了收敛性分析,证明了该闭环子系统是全局渐近稳定的。针对多UUV间通信的通信约束(即通信延迟)问题,提出了基于逻辑通信的控制策略,通信中每个UUV通过估计模块来估计群集中其它UUV的协调状态,并同步更新状态信息逐步实现多UUV间的协调。更多还原

【Abstract】 The early formation control of multiple UUVs requires only the relative position or distance information among the UUVs to be the expected values, the path of single UUV wasn’t restricted and when multi-UUV perform special tasks, such as subsea pipeline inspection, seabed maps’establishment, UUV must follow a specified path. Obviously, the traditional formation controller can not solve such a special formation problem. In this paper, formation coordinated control is discussed based on the path following (straight path and curved path) for a certain type of UUV. The main contents are as follows:Firstly, based on the coordinated path following control tasks of multi-UUV, we propose a decentralized control structure which has the time-space decoupling characteristic. Using this set-up, path following (in space) and inter-vehicle coordination (in time) are almost decoupled. The first step, a path following control law is derived that drives each vehicle to its assigned path, which is aimed at achieving space task. The second step, through the exchange of information among UUVs to design a coordinated controller aimed at achieving time task. In addition, the application of under acoustic communication is described, and graph theory notation is applied to communication topology. It has laid a theoretical foundation for controllers designed.Secondly, the problem of coordinated formation control of multi-UUV under the straight line path following is discussed based on the above architecture. For the path following control problem of a single UUV, the line-of-sight method is proposed to establish the space movement error models of UUV. Then the problem of position error’s stabilization control is turned into the asymptotic stability design of the speed error, heading error and pitch error. These controllers are designed based on feedback linearization which is convenient for engineering realization, respectively. To solve the problem of the leader failure caused the global collapse in the leader-follower method, a decentralized formation control method which is based on graph theory is proposed. This method provides the feedback channels of status messages between the multi-UUV. Vehicle coordination is achieved by adjusting the speed of each of the vehicles along its path, according to information on the status of the remaining vehicles only. Relative to the leader-follower, in which the leader can not get the followers’ status feedback information, this method has more effective information exchange between the multi-UUV and formation control.Then, the problem of coordination formation control of curve path following is discussed because of the limitations that straight line path following can not meet in practical applications. In the spatial curve path following control, the roll degree of freedom can not be ignored, and need to be considered during the manoeuvre motion. So, the "pulling" control strategy, which is based on virtual reference target, is presented. The dynamics path following controller and the kinematics path following controller have achieved the position and the speed tracking of the UUV, respectively. The Lyapunov direct method and the input-to-state stability theorem proved the stability of path following subsystem and the bound of system status and control inputs in the design of speed coordination controller. This method improves the path following control precision. To solve the underwater acoustic communication bandwidth constraints problem, only to exchange one variable coordination state of information exchange is presented. This method is corresponding to reduce the amount of communication to meet the narrow bandwidth of underwater acoustic communication constraints. Using graph theory describes the communication topology of multi-UUV and established coordination error model in the formation control. Moreover, the actuators saturation of UUV has been considered, and the formation controller contains double curve tangent function to ensure the quality of the control system.Finally, the problem of multi-UUV’s coordination path following control which has limitations of environment is discussed, considering the impacts of ocean current interference and limitation of acoustic communication. The nonlinear adaptive path following controller is designed based on backstepping under disturbance of ocean current, which compensates for the disturbance of unknown currents. LaSalle’s invariance principle is applied to prove a set of global convergence of path following subsystem. The available nonlinear part is eliminated in feedback linearization, which is avoided in this method. In the case of acoustic communication limited, logic-based communications strategy is proposed based on consensus of multi-agent, so that each UUV achieve the purpose of coordination between mulli-UUV by updating their own status estimate information. The results of simulation has shown the effectiveness of the controllers of path following and coordination formation under the condition of limited environment.更多还原