【作者】 刘振业；

【导师】 施小成；

【作者基本信息】 哈尔滨工程大学 ， 控制理论与控制工程， 2013， 博士

【摘要】 作为典型的两栖高性能船舶，全垫升气垫船在军事和民用领域得到了广泛的应用。特殊的航行机理使全垫升气垫船能够脱离水面（地面）航行，但也造成其操纵性与常规水面船有明显差别。由于船体与航行表面之间的阻力很小，受环境扰动时，航向极易改变。回转时，船体横倾，一舷漏气，容易出现大横倾-大侧滑的危险航行状态。在水面航行时，波浪会造成围裙底部泄流面积、气垫体积的周期性改变，以致气垫压力大范围变化，进而造成船体垂向加速度剧烈改变，使全垫升气垫船适航性较差。针对全垫升气垫船难以控制的特点，研究其安全、高效、精确的航行控制方法是有必要的，因此，本课题基于自抗扰控制策略(Active Disturbance Rejection Control, ADRC)研究了全垫升气垫船航行控制问题。首先，建立了全垫升气垫船6自由度运动数学模型。根据船模水池实验及风洞试验，得到了水动力模型及空气动力模型；考虑增压风扇吸入的空气需被加速到全垫升气垫船航行速度，得到了空气动量力模型；根据垫升系统结构，建立了气垫力模型，将各自由度的合外力（矩）进行合成，进而得到了全垫升气垫船6自由度运动数学模型，为后续安全航行控制问题奠定了基础。然后，深入研究了具有动态补偿线性化功能、不依赖被控对象精确数学模型的自抗扰控制策略。根据一般状态观测器理论，对扩张状态观测器(Extended State Ovbserver,ESO)的稳定性及误差范围进行了分析。在跟踪微分器(Tracking Differentiator, TD)设计过程中，引入了fsun ()函数，有效地消除了系统状态的颤振现象。分析了非线性状态误差反馈控制律(Nonlinear State Error Feedback，NLSEF)的稳定性。为后续针对全垫升气垫船航行控制特点而对自抗扰控制策略作出的改进奠定了基础。其次，研究了基于ADRC的全垫升气垫船航向及航迹控制方法。设计了全垫升气垫船ADRC航向控制器、航迹引导控制器。根据自抗扰控制理论中的安排过渡过程思想，提出了回转率-航向协调控制策略，使全垫升气垫船以安全的回转率进行航向控制，避免了侧滑角经常进入安全限界。针对全垫升气垫船对环境扰动敏感，运动非线性强，运动数学模型难以准确建立等特点，提出了一种基于支持向量回归机的参数自整定自抗扰控制算法，明显增强了所设计航向控制器、航迹引导控制器的自适应性，有效地改善了控制效果。再次，研究了基于ADRC的全垫升气垫船航向-横倾解耦控制方法。针对全垫升气垫船航向控制过程中，经常出现大横倾-大侧滑危险航行状态，提出了一种航向-横倾ADRC解耦控制方法，解决了航向控制通道（执行机构为空气舵）和横倾控制通道（执行机构为艏喷管）的控制耦合问题。提出了一种动态面ADRC控制算法，利用动态面控制算法代替自抗扰控制策略中的NLSEF，这样既提高了航向-横倾解耦控制的效率，又解决了动态面控制算法对系统精确模型的高度依赖。最后，研究了垫升系统运动及静态特性，设计了基于ADRC的气垫压力控制方法。在基本假设条件下，建立垫升系统的非线性压力流量方程组。在垫升系统数学模型的基础上，研究了全垫升气垫船进行典型运动时的垫升系统运动特性，并研究了升沉稳定性及升沉阻尼（升沉位置导数及速度导数）。建立了垫升系统控制模型，提出了一种基于ADRC的垫升风机进气量控制方法，以控制气垫压力。利用ESO估计各气垫压力，利用fal ()函数反馈结构估计垂向加速度，设计了NLSEF，并实现了垂向加速度补偿。仿真结果表明，所设计带有加速度补偿的ADRC气垫压力控制方法，有效地改善了全垫升气垫船的适航性及航行稳定性。更多还原

【Abstract】 Air cushion vehicle (ACV) has been widely used in military and civilian fields, as atypical amphibious high performance ship. With the special navigation mechanism, ACV islifted from water (or ground), which makes its maneuverability significantly different fromconveantional surface ships. With little resistance between the surface and the hull, headingcan be easily changed under the environment disturbance. When ACV is turning, the hullheeling makes air leakage from one side, sometimes large heeling can induce large sideslipdangerously. When ACV is sailing on the water, the wave can cause the effusion area underthe skirt and cushion volume change periodically, cushion pressure and heave accelerationvary severely, so the seaworthiness of ACV is poor. Obviously ACV is difficult to control, itis necessary to study safety, efficiency, precise navigation control methods, therefore, thispaper studies the navigation control problem of ACV based on Active disturbance rejectioncontrol (ADRC).Firstly, a six-degree motion mathematical model of ACV is established. Hydrodynamicsand aerodynamics moel is obtained from the tank and tunnel experimental data. Consideringthe sucked air from the fan need to be accelerated to the same speed of ACV, the airmomentum force model is established. Cushion force model is set up based on the cushionsystem. By synthetizing the external force (torque) on each DOF, the motion mathematicalmodel of ACV is finished, it’s the foundation of the navigation control problem.Secondly, the theory of ADRC is well studied, which can achieve linearized dynamiccompensation and does not depend on the accurate mathematical model of the controlledobject. According to the theory of general state observer, the stability and error range ofextended state ovbserver (ESO) are analysised. In the design of tracking differentiator (TD),fsun()function is used to eliminate the flutter of system state. The stability of the nonlinearstate error feedback (NLSEF) control law is studied. The work above is the foundation for theimprovement of ADRC for the different navigation control assignment of ACV.Thirdly, the motion control method of ACV is studied based an ADRC. The heading andtrack guidance controllers of ACV are designed. Based on the theory of arrangement fortransitional process from ADRC, the coordinated control strategy of heading and turning isproposed, with method, ACV can achieve heading control under the safety turning rate, andthe sideslip angle is restricted in its safety boundary. To solve the problem that ACV issensitive to environment disturbance, its motion nonlinearity is obvious and its motion mathematical model of ACV is not so accurate, ADRC with control parameters self-tuningbased on support vector regression is proposed. With the adaptive ADRC, the control effectsof heading and track guidance are improved.Fourthly, heading-heeliing decoupling control of ACV is studied based on ADRC. In theheading control process of ACV, sometimes, large heeling can cause severe sidelipdangerously. To solve the problem above, ADRC decoupling control method is used, thedecoupling between heading control passage (The actuator is air rudders.) and heeling controlpassage (The actuator is nozzle.) is weakened. Dynamic surface ADRC is proposed, theNLSEF is replaced by dynamic surface control, then, the control efficiency ofheading-heeling decoupling control is improved and the high dependence on accurate systemmathematical model of Dynamic surface control is reduced.Finally, the static and motion characteristics, control method based on ADRC is studiedfor the cushion system. Under the basic assumptions, the nonlinear pressure-flow equations ofcushion system are established. With the mathematical mode of cushion system, the motioncharacteristics for the typical motion of ACV, the heave stability and damping (the heavelocation derivative and speed derivative) are studied. The control model of cushion system isestablished, ADRC is used to control the air into the fan to control the pressures of cushions.ESOs are used to estimate pressures of cushions and heave acceleration, NLSEF is designedwith compensation of heave acceleration. The method proposed effectively improves theseaworthiness and sailing stability of ACV.更多还原