【作者】 王俊；

【导师】 裴海龙；

【作者基本信息】 华南理工大学 ， 控制理论与控制工程， 2013， 博士

【摘要】 高超声速飞行器是一个强耦合、不确定、时变以及具有独特气动特性的高阶非线性系统，对其再入飞行中烧蚀预测及控制的研究至今仍具有严峻的挑战。然而，由于高超声速飞行器的再入环境复杂、气动热烧蚀严重、涉及多学科的交叉以及对飞行控制要求较高等特点，吸引着国内外学者的广泛研究，以及在各国军事领域中的战略价值和应用前景同样备受关注。为此本文在高超声速飞行器的再入飞行建模与特性分析、气动热烧蚀预测和烧蚀控制策略等方面开展较为深入的研究。本文主要研究工作的内容和创新点包括以下几个方面：1.在参照国内外公开发表的文献基础上，介绍了有关高超声速飞行器的研究现状，并重点对气动热烧蚀、故障与烧蚀控制以及先进飞行控制技术的研究现状进行了阐述与分析。2.分析了高超声速飞行器的动力学和气动特性，建立了六自由度再入飞行的非线性数学模型及降阶模型，以及弹道式三自由度再入非线性模型，并通过仿真验证了所建模型具有复杂的非线性、耦合性及时变性等特点。3.利用已建立的数学模型，针对再入过程预测热防护材料烧蚀深度和温度的动态变化问题，根据设计的热防护罩几何参数研究了计算烧蚀质量、气动热和热辐射的工程估算方法，提出了基于Newton-Raphson和TDMA动态解析烧蚀的算法并应用一维非线性热传导方程模拟了烧蚀过程，阐述了控制舵面对烧蚀的影响以及故障条件下构建新气动平衡的优化方法，通过烧蚀预测的结果与热平衡积分法相比较，说明了该算法既合理可靠，又实现了再入全过程表面热防护材料的烧蚀和表面温度的动态变化，为估算烧蚀的质量以及动态舵面实时烧蚀控制提供了较为精准的预测。4.在研究了舵面故障下重构气动平衡的基础上，针对无约束控制舵面的故障导致烧蚀量突变问题考虑了烧蚀预测与容错控制相结合的方法。首先建立了控制舵面的故障模型，提出了自适应鲁棒输出反馈控制器设计方法。该方法实现了非参数化时变故障的自适应处理，并在反馈环节中直接应用估计的增益，从而提高了闭环系统的稳定性和鲁棒性。为了解决扰动与故障的耦合问题，构造了扰动观测器和故障重构控制器的设计方案，该方案不仅有利于评估故障的严重程度也为实施故障调节提供了一种有效的观测途径。仿真结果表明将自适应输出反馈控制器、重构气动系数和烧蚀预测算法相结合，确保了动态烧蚀预测轨迹的跟踪，有效减少了故障产生的烧蚀增量，并验证了该烧蚀控制结构是一种被动烧蚀控制的方法。5.由于无约束控制舵面故障导致再入飞行轨迹的突变，使得动态烧蚀的增量无法及时预测，而约束控制舵面即等效于限制了烧蚀的增量，为此提出了结合烧蚀预测和基于控制分配算法的在线鲁棒模型预测控制的方法。该方法应用控制分配算法能够有效地对健康冗余的控制舵面进行再分配，并利用离线鲁棒观测器改善闭环系统的抗扰动能力。通过设计无偏模型预测控制器，不仅实现了系统状态和控制舵面的约束并且解决了模型不匹配的问题。该方案不仅保证了系统的鲁棒性和稳定性，也通过限定控制舵面的最大最小偏转角度避免了故障所引起的不可预测性烧蚀，并且能够在线修正烧蚀增量达到了主动控制烧蚀的目的。本文最后，在总结全文的基础上，提出了有待进一步研究和探索的一些问题。更多还原

【Abstract】 Hypersonic vehicle as a high order nonlinear system has the characteristics of strongcoupling, uncertainty, time-varying and unique aerodynamics. So it is a severe challengeto predict and control ablation during reentry flight. However, hypersonic’s special char-acteristics such as complex reentry environment, Serious aerodynamic thermal ablation,multidisciplinary cross and higher flight control requirements make it attracts the exten-sive research of scholars at home and abroad, and also has focused on significant strategicvalue and application prospect for military applications in the countries. Therefore, thereentry flight modeling and characteristic analysis of hypersonic vehicle, aerodynamicthermal ablation prediction and ablation control strategy is introduced and in-depth re-search in this dissertation. The main research content and primary innovation of thisdissertation can be summarized as follows:1. Based on the research contributions at home and abroad, the background of thehypersonic vehicle is introduced. And then, the research status of the aerodynamic abla-tion, fault and ablation control and advanced flight control technology is mainly describedand analyzed.2. A very specific analysis of the kinetic equation and aerodynamic characteristicsof the hypersonic vehicle is given, then a six degree of freedom nonlinear mathematicalmodel and reduced-order model of reentry flight is established, and then a three degree offreedom nonlinear mathematical model of ballistic reentry flight is also built. The resultsof simulation verify that the establishment of the model has the characteristics of thecomplex nonlinear, coupling and time-varying, and so on.3. By using the established mathematical model, the predicted problem of thermalprotection material subject to the dynamic change of the ablative depth and tempera-ture is presented in the reentry process. According to the designed geometric parameterof the heat shield, the engineering prediction method for calculating the ablation mass,the aerodynamic heating and thermal radiation is studied, and then the algorithm ofdynamic analytic ablation is proposed based on the Newton-Raphson and TDMA, anda one dimensional nonlinear heat conduction model is employed to simulate the processof the ablation. Furthermore, The influence of control surface on the ablation and the optimization method of the building a new aerodynamic blance under the fault conditionis elucidated. The results of the proposed method are reliable and reasonable by compar-ing with heat balance integral method, and is able to realize that the dynamic change ofsurface protection material ablation and surface temperature in the whole reentry, andalso provides more accurate prediction for the estimation of the ablation mass and thereal-time ablation control of dynamic control surface.4. It is studied on the basis of reconstitution aerodynamic balance under the condi-tion of actuator fault, the combination of ablation prediction and fault-tolerant controlmethod subject to unpredictable ablation problem result from the fault of the uncon-strained control surface is considered. A adaptive output feedback controller is presentedbased on control surface fault model. By this controller, the non-parametric time-varyingfailures can be realized by the way of adaptive process, and then the estimated gain canbe directly used in the feedback loop. Therefore, The stability and robustness of theclosed-loop system is enhanced. In order to solve the problem of disturbance decoupledfault, the disturbance observer and fault reconstruction control is designed. This ap-proach can be beneficial to evaluate the severity of the fault and provides an efcientobservation method to implement fault accommodation. Simulation results show thatthe combination of adaptive output feedback controller, reconfiguration aerodynamic co-efcients and ablation prediction can ensure the dynamic ablation to track the predictivetrajectory, efectively reduce the ablation increment caused by faults, and also verify thatthis ablation control framework is a kind of passive ablation control method.5. The unpredictable of dynamic ablation increment in time is because that thesaltation of reentry trajectory resulting from the fault of unconstraint control surface.But, constraint control surface is equivalent to limit the ablation increment. Hence, themethod of ablation prediction combined with robust model predictive control with on-line control allocation is proposed. This approach is capable of redistributing the controlsurface among healthy actuators in a stable manner by control allocation scheme, and theof-line robust observer is applied to improve the ability against disturbances of the closed-loop system. By designing the ofset-free model predictive controller, the constraints ofsystem states and the control surface can be realized and the model mismatch problemis also resolved. This control framework can not only guarantee the robustness andstability of the system, but also by limiting the maximum-minimum angular deflection of the control surface to avoid the unpredictability ablation cased by faults, and achievethe purpose of active control with online correction ablation increment.Finally, the main results of the dissertation are concluded and some issues for futureresearch and exploration are proposed.更多还原