【作者】 王涛；

【导师】 肖建；

【作者基本信息】 西南交通大学 ， 交通信息工程及控制， 2007， 博士

【摘要】 论文以高性能感应电机调速系统为研究对象，针对感应电机模型中变量的非线性耦合，模型参数不确定性和外界干扰等问题，在转子磁链观测、电机调速闭环控制器设计等方面，进行了深入的理论分析和计算机仿真研究。本文研究具有参数不确定性和外部扰动的一类系统，即线性变参数系统。系统的参数变化状态方程由多胞矩阵描述，利用H₂／H_∞范数条件与系统状态空间实现的线性矩阵不等式（LMI）之间的等价性，获得系统的鲁棒性能。然后，基于鲁棒加权函数法所设计的控制器，实现了对输入参考信号的有效跟踪并能有效抑制外界干扰。最后，提出了一种同时考虑分段及参数变化率的LMI设计方法，讨论了参数分段、参数变化率与系统稳定性及L，增益的关系，以降低系统的保守性。仿真说明了系统具有鲁棒稳定性和对干扰的抑制作用，验证了设计方案的有效性。针对转子转速大范围变化下，特别在低速运行时，定子的电阻变化影响下，如何获得精确的转子磁链问题，论文采用线性变参数多胞输出反馈控制器设计理论，提出了一种新的转子磁链观测方法。该方法引入了多胞技术，设计了含极点配置的随变参数自调整的输出反馈控制器；同时，将感应电动机作为线性变参数系统，转子转速与定子电阻作为系统的变参数，利用鲁棒控制理论，通过求解线性矩阵不等式组，设计出了系统控制器，即转子磁链观测器，它具有鲁棒稳定性能和良好动态特性。实验及仿真验证了所设计观测器的有效性，观测器的性能有了很大的提高。基于无源性控制理论，证明了感应电机转子磁链子系统的无源性，找到系统能量耗散特性方程中的“无功力”，表明它不会影响系统的稳定性，得出无需转子磁链观测反馈即能稳定跟踪转子磁链的参考值的控制系统设计方法，它是一种间接磁场定向控制方法；同时考虑电机转子电阻在运行中发生未知变化，设计了相应自适应调节器使系统对转子电阻呈现鲁棒性，由于其控制律是全局定义的，不存在输入输出线性化解耦奇点问题。仿真验证了转矩、转子磁链及转速的渐近跟踪性能。基于耗散性与稳定性之间的等价关系和投影原理，提出了具有L₂增益的非线性自适应鲁棒器并应用于感应电机控制。针对感应电机转子电阻与负载转矩的不确定性，在设计控制器时，引入了与不确定性参数相对应的可调参数，利用投影定理设计了自适应控制律，在线修正参数值，该自适应控制器设计方案保证了感应电机系统的无源性；同时耗散性分析中采用了L₂增益算法，使系统不但对不确定项具有鲁棒性，而且对外部干扰具有抑制作用，从而保证了整个闭环系统的动态品质。数值仿真验证了设计方法的正确性。针对感应电机复杂耦合非线性模型，提出了一种应用二次型最优控制理论设计跟随调节器的控制方法。首先比较了3种转子磁场定向线性解耦控制方法各自的特点，提出了一种将电机数学模型中转子磁链幅值与转子角速度完全线性化解耦的方法；再按照线性系统理论对磁链幅值Φ与转子角速度ω的两个线性的一阶微分方程求解；最后，应用二次型最优控制理论设计跟随调节器。该方法解决了矢量控制非线性耦合问题，设计的控制律没有奇异点问题，算法简单易于理解。计算机仿真验证了系统具有良好的动静态性能。针对含有不确定项的非线性系统的控制问题，提出了一种基于Lyapunov理论的鲁棒稳定和跟踪控制算法并应用到感应电机控制中。针对转子磁场定向的感应电机控制中，转子电阻具有时变性且难以测量，它的不确定性会影响到电机的动态性能，该方法将感应电机的数学模型分解为标称子系统和不确定子系统，通过设计非线性反馈补偿控制律，可以实现预期的控制目标，从而有效地克服了不确定项的测量和影响问题。仿真结果表明，论文所提出的方法是可行性的，能够保证即使在转子电阻偏离其标称值时，转子磁链和转子转速具有稳定跟踪性能。更多还原

【Abstract】 In this dissertation, researches are focused on high performance induction motor drive system. Aimed to solve the control problems of nonlinearity coupling of electrical variables, model parameter uncertainty and external disturbance existing in induction motor, in-depth theory analysis, simulations are performed concerning the rotor flux observer, the close-loop controllers in the system. Robust control performance with quick response is obtained.Parameter uncertainty and exogenous disturbances in the system, which is called linear parameter-varying system, is studied in the dissertation. State space equation of parameter variation is described by polytope matrixes. Based on the equivalence between H₂/H_∞norm condition and linear matrix inequality, the robust of the system is obtained. Meanwhile on the basis of robust weighted function method, the controller is designed. Then input signal can be effectively tracked and external disturbance is outstandingly restrained. The relationship of subsection and changing rate of parameter between stability and L₂-gain of system is introduced and analyzed. Experiment and simulation results show the robust of the system is obtained and the method designed is valid.The variation of the stator resistance during operation degrades the performance of the flux observer at low speed segment. A new approach to the design of robust rotor flux observer of induction motor is presented. A polytope technique is introduced in the paper. Polytopic self-scheduled output feedback controller with limiting point disposition in linear parameter-varying system is acquired. The induction motor is regarded as a linear parameter-varying system of which the flux observer is a controller. The rotor speed and the stator resistance are considered variable parameters especially. Based on robust control theory, a flux observer, which is of H_∞performance and good dynamic characteristics, is designed with a linear matrix inequality approach. Meanwhile the adverse influence of external disturbances such as measurement noises is suppressed. Simulation results demonstrate the validity of the flux observer designed. The performance of flux observer based on above method is improved to a great degree.Passivity-based control（PBC） theory to the control of induction motor is introduced. First the passivity of motor rotor flux subsystem is proved. Workless force has no influence on the energy balance equation and the asymptotic stability. Then it is discussed that the system can stably track the reference rotor flux without rotor flux observer. The controller developed is an indirect field-oriented one. At last an adaptive PBC strategy is proposed with regard to the rotor resistance change. The design of the control law is globally defined without singularity. Simulink example demonstrates the torque, rotor flux magnitude and rotor speed can be asymptotically tracked.Based on the equipollence relationship of passivity and stability and direct project tlieorem, an nonlinear adaptive L2-gain robust controller is designed. It is a novel approach which is applied to the control of induction motor. Aimed at the uncertainty of rotor resistance and load torque, adjustable parameters corresponding to uncertain ones are introduced. According to direct project theorem, the adaptive control law is acquired by adjusting parameters on-line. Meanwhile the L₂-gain algorithm is adopted as passivity analyzed. The designed control scheme ensures the total system is not only robust to rotor resistance and load torque, but also restrains external disturbance. So the dynamic quality of closed loop system is guaranteed. The correctness of the design method is verified by numerical simulations.Induction motor is a nonlinear system with complicated coupling. On the basis of quadratic optimal control theory, the method of the track adjuster is designed. At first, three linearization and decoupling control methods of induction motor rotor field-orient control are compared in the paper. By selecting rotor flux rotorΦ=φ_rα² +φ_rβ², a new method is designed to make rotor flux and speed decoupled into two independent subsystems and realized the complete linearization using feedback control. The design of the control law is globally defined without decoupling singularity. At last, the control of system can be performed to asymptotically track rotor flux and speed using the linear quadratic optimal control theory. The computations are simple and allow more insight and understanding. Simulink example denotes the system possesses good dynamic and static performance.Aimed at solve the control problems of uncertainty existing in nonlinear system, the robust stability and robust tracking control law based on Lyapunov theory is introduced in the paper. The mathematical model of this system is divided the nominal and uncertainty subsystems. Nonlinear feedback compensation to realize the expectative objective is designed. Then theory above is applied to control problem of a rotor field-oriented induction motor, because of the rotor resistance is subjected to variability and hard to be measured, which more or less deteriorate the performance. Simulation results show the feasibility of the method. The designed scheme guarantees rotor flux and speed to track stably their reference even if the rotor resistance varies from its nominal value.更多还原