【作者】 毕凤阳；

【导师】 黄文虎； 张嘉钟；

【作者基本信息】 哈尔滨工业大学 ， 一般力学与力学基础， 2010， 博士

【摘要】 自主水下航行器(Autonomous Underwater Vehicle, AUV)在军用和民用领域有着广泛的应用前景,是当前国内外研究的一个热点。随着AUV应用范围的增加,对其自主性的要求也随之增加,而增加AUV自主性的关键之一就是提高其控制系统的性能。然而出于节约成本和减轻重量的考虑,很多AUV仅具有较其运动自由度更少的控制装置,是欠驱动系统,欠驱动AUV控制问题是目前国内外研究的一个难点。AUV的动力学是强非线性的,难以获得其足够精确的水动力系数,易受到海流等海洋环境的干扰,这些不确定性要求控制器有很强的鲁棒性,这使得一般的控制器很难胜任AUV的控制任务。因此对具有模型不确定性欠驱动AUV的非线性鲁棒控制研究是非常重要的。本文对具有模型不确定性欠驱动AUV的定深、水平面轨迹跟踪和位置跟踪的非线性鲁棒控制策略作了深入系统的研究。建立了欠驱动AUV控制模型。给出了AUV六自由度运动学和动力学方程,根据AUV平移速度对地可测或仅对周围水可测、粘性阻尼系数的标称值是否已知这四种情况,建立了四个考虑参数不确定性和外界干扰的欠驱动AUV水平面运动模型。利用小扰动方法建立了垂直面运动模型。介绍了一些相关的控制理论和方法,并给出了一些相关的基本概念和定理。针对具有模型不确定性欠驱动AUV的定深控制问题,设计了一种模糊滑模控制器。为了获得更好的控制性能,模糊控制输入采用了压缩宽度隶属度函数,模糊控制输出采用了扩张宽度隶属度函数。为了更好地削弱抖振现象,在该模糊滑模控制器的基础上,设计了两种控制策略对该模糊滑模控制器进行性能优化,一是设计一定的模糊规则自适应地调整模糊控制器的比例因子,二是基于遗传算法对该模糊滑模控制器的参数进行了性能优化。为了验证所设计的控制器和控制策略的有效性和鲁棒性,对具有控制输入时滞、较大参数不确定性和未建模动态的欠驱动AUV系统进行了相应的数值仿真。针对具有参数不确定性和外界干扰的欠驱动AUV的水平面轨迹跟踪问题,根据不同情况,提出了三种控制策略:一是对于小曲率轨迹跟踪问题,根据横向速度远小于纵向速度这一动力学特征,使用Backstepping方法提出了一种控制策略,其中设计了一种自适应控制律来补偿完全未知的粘性阻尼系数,使用滑模控制和模糊滑模控制方法来补偿别的参数不确定性和外界干扰,并基于李雅普诺夫稳定性理论证明了闭环控制系统的稳定性;二是假设期望偏航角速度满足持续激励条件,基于级联方法提出了一种解耦控制策略,首先建立跟踪误差系统并将其解耦成两个独立的子系统,使用反馈线性化方法和滑模变结构控制方法各设计了一种控制律分别稳定两个子系统,最后根据级联系统的稳定性理论证明了闭环控制系统是全局K指数稳定的;三是对于一般情况下的欠驱动AUV轨迹跟踪问题,使用Backstepping方法设计了一种鲁棒轨迹跟踪控制器,其中使用传统滑模和准滑模控制方法补偿参数不确定性和外界干扰,推导出了滑模控制增益与参数不确定性、外界干扰和系统状态的关系,分析了正负质量和附加质量不确定性对控制系统性能和控制成本的不同影响,基于李雅普诺夫稳定性理论证明了闭环控制系统的稳定性。为了验证这三种控制策略的有效性和鲁棒性,分别对具有参数不确定性和外界干扰的欠驱动AUV模型进行了相应的数值仿真。针对具有参数不确定性和外界干扰的欠驱动AUV水平面位置跟踪问题,使用Backstepping和级联方法,为前面所建立的四个水平面运动模型各提出了一种控制策略;其中,使用Backstepping方法设计了鲁棒位置跟踪控制器,为了补偿完全未知的粘性阻尼系数,设计了一种自适应控制律,并使用滑模变结构控制方法来补偿其他参数不确定性和外界干扰;AUV平移速度仅对周围水可测情况下,为了估计完全未知的海流速度,设计了两种观测器,这两种海流观测器都能确保海流速度估计误差指数收敛到零;基于李雅普诺夫稳定性理论和一些级联系统稳定性定理证明了四个闭环控制系统的全局稳定性、以及系统状态和控制输入的有界性。仿真结果表明所提出的控制策略能够很好地实现欠驱动AUV的水平面位置跟踪控制,对完全未知的粘性阻尼系数和海流速度有很好的自适应性,对其它参数不确定性和外界干扰具有很强的鲁棒性。更多还原

【Abstract】 Autonomous underwater vehicles (AUVs) have recently become an intense area of domestic and international research because of their potential military and civil applications. As the application range of AUVs expands, it becomes more and more important to develop AUV autonomy. One of key technologies is to improve its control performance. However, the problem of AUV control continues to pose challenges to system designers since many AUVs are underactuated systems, with fewer control inputs than the degrees of freedom, out of the need to reduce the actuator cost and weight. In addition, the dynamics of AUVs are highly nonlinear, the hydrodynamic coefficients are not precisely known, and the vehicle is often disturbed by the ocean environments. As such the development of nonlinear robust control strategies for underactuated AUVs is of considerable important.In this dissertation, nonlinear robust control strategies are deeply studied for the depth, planar trajectory and position tracking of underactuated AUVs with model uncertainties.Control models are built for underactuated AUVs. The six degrees of freedom kinematic and dynamic equations of AUV motion are given. According to that the translational velocities relative to the Earth or surrounding water are measured, and whether the nominal values of the viscous damping coefficients are known or not, four horizontal motion models are built for underacutuated AUVs with parameter uncertainties and external disturbances. A vertical motion model is built using a small disturbance analysis. Some related control theories and methods are introduced, and some related control concepts and theorems are given. To solve the depth control problem of underacutuated AUVs with model uncertainties, a fuzzy sliding mode controller is presented. The design method of the controller offers a systematical means of constructing a set of shrinking-span for fuzzy inputs and dilating-span membership functions for fuzzy output. To avoid chattering better, two control strategies are proposed. One is to design some fuzzy rules to adaptively tune scaling factors of the fuzzy controller. Another is to optimize parameters of the fuzzy controller by genetic algorithm. Numerical simulations are presented for the underactuated AUV systems with control input delay, large parameter uncertainties and unmodeled dynamics to illustrate the effectiveness and robustness of the proposed control strategies.To solve the trajectory tracking control problem of underacutuated AUVs with parameter uncertainties and external disturbances in the horizontal plane, three control strategies are proposed according to different cases. The first control strategy is developed using the backstepping approach for the problem of small curvature trajectory tracking where the sway velocity is relatively much smaller than the surge velocity. An adaptive control strategy is proposed for the unknown constant viscous damping coefficients. For compensating other parameter uncertainties and external disturbances, different control laws based on sliding mode and fuzzy sliding mode control methods are presented. Stability proof of the closed-loop control system is given by the Lyapunov stability theory. The second control strategy is developed by using the cascade approach where the reference yaw velocity satisfies the persistence of excitation condition. The tracking error system is formed, and then decoupled into two separate subsystems. A feedback linearization and sliding mode controller are presented respectively to stabilize the two subsystems. The closed-loop control system is proved to be globally K exponentially stable by stability criteria for cascade system. The third control strategy is developed by using the backstepping approach for the general trajectory tracking control problem. For compensating parameter uncertainties and external disturbances, different control laws based on classical and quasi SMC methods are presented. The relation is derived between parameter uncertainties, external disturbances, system states and sliding mode gains. Different effects of positive and negative mass uncertainties on the control systems are analyzed. Stability proof of the closed-loop control system is given by the Lyapunov stability theory. Numerical studies for different cases are presented to demonstrate the effectiveness and robustness of the above proposed three control strategies.To solve the position tracking control problem of underacutuated AUV with parameter uncertainties and external disturbances in the horizontal plane, four control strategies are proposed respectively for the above four horizontal models. The robust position tracking controllers are proposed by the backstepping approach. If the translational velocities relative to surrounding water are measured, two observers are proposed to estimate unknown constant ocean current velocities, which can ensure that estimation errors are globally exponentially stable. It is proved that the above proposed four control strategies are globally stable, and system states and control inputs are bounded, by the Lyapunov stability theory and stability criteria for cascade system. Simulation results show that the proposed control strategies are effective, adaptive to unknown ocean current velocities and viscous damping coefficients, and strongly robust against parameter uncertainties and external disturbances.更多还原