永磁同步电机伺服系统自适应迭代学习控制

发布时间：2018-06-21 10:38

本文选题：永磁同步电机 + 不确定性　；参考：《沈阳工业大学》2017年硕士论文

【摘要】：永磁同步电机(PMSM)以其体积小、结构简单、转动惯量低、功率密度高等优点在工业机器人、高精度数控机床等伺服控制领域获得了广泛的应用。迭代学习控制(ILC)适用于执行重复任务的伺服控制系统,理论上可以达到完美的跟踪效果。然而实际系统中存在的各类扰动、建模误差、参数时变等不确定性会对ILC的收敛性和系统的跟踪性产生不利影响。因此针对PMSM伺服系统在ILC过程中处理不确定性问题方面的缺陷进行研究,采用自适应控制与ILC相结合的方法,即自适应迭代学习控制(AILC),其兼具了ILC解决重复跟踪问题和自适应控制解决系统不确定性问题上的双重优势,提高了PMSM伺服系统的跟踪精度,并加快了系统的收敛速度。首先,介绍了PMSM的结构及分类,根据坐标变换建立了PMSM在dq坐标系下的数学模型。采用di=0控制建立了PMSM矢量控制系统并分析了摩擦转矩、齿槽转矩、模型误差、参数时变等不确定因素对控制系统的影响,建立了系统的状态方程。其次,针对PMSM伺服系统受模型不确定性影响导致的跟踪精度下降,误差发散等问题,提出了一种参数自适应迭代学习律。其主要是在PD反馈控制的基础上增加自适应迭代项,通过沿迭代轴进行参数学习,辨识出控制律未知参数。然后提出一种改进型自适应迭代学习律,相当于在前一种学习律的基础上增加了对未知参数的时域估计,充分利用了时域和迭代域的信息。基于Lyapunov稳定性理论,分析了两种方案的收敛性。仿真结果表明,在存在模型不确定性的情况下,AILC比传统型ILC收敛速度快,跟踪精度高,可有效改善系统的性能。最后,针对PMSM伺服系统运行过程中受动态参数不确定性影响的问题,提出一种L1自适应反馈与ILC相结合的方法。其中L1自适应控制器用来处理系统的动态参数不确定性,并对时域上的外加扰动进行补偿以便于学习控制器可以在标称系统上设计。而学习控制器用来提高系统对周期性输入的跟踪能力,并对摩擦、齿槽转矩等迭代域重复不确定性进行补偿。仿真结果表明,该方案对于减小系统动态参数不确定性对系统的影响,提高系统的性能有良好的效果。
[Abstract]:PMSM (permanent Magnet synchronous Motor) has been widely used in the servo control fields such as industrial robot, high precision CNC machine tool and so on because of its advantages of small size, simple structure, low moment of inertia and high power density. Iterative learning control (ILC) is suitable for servo control systems that perform repetitive tasks and can achieve perfect tracking results in theory. However, uncertainties such as disturbances, modeling errors and time-varying parameters in the actual system will adversely affect the convergence of ILC and the tracking of the system. Therefore, aiming at the shortcomings of PMSM servo system in dealing with uncertain problems in ILC process, the method of combining adaptive control with ILC is adopted. That is adaptive iterative learning control (AILC), which has the dual advantages of ILC to solve the repeated tracking problem and adaptive control to the uncertainty problem of the system, improves the tracking accuracy of the PMSM servo system, and accelerates the convergence speed of the system. Firstly, the structure and classification of PMSM are introduced, and the mathematical model of PMSM in dq coordinate system is established according to coordinate transformation. The di=0 vector control system is established, and the influence of the uncertain factors such as friction torque, slot torque, model error and time-varying parameters on the control system is analyzed, and the state equation of the system is established. Secondly, a parameter adaptive iterative learning law is proposed for PMSM servo system, which is caused by the uncertainty of the model, such as the decrease of tracking accuracy and the divergence of error. On the basis of PD feedback control, the adaptive iteration term is added, and the unknown parameters of the control law are identified by learning the parameters along the iterative axis. Then an improved adaptive iterative learning law is proposed which is equivalent to adding the time domain estimation of unknown parameters on the basis of the former learning law and making full use of the information in the time domain and iteration domain. Based on Lyapunov stability theory, the convergence of two schemes is analyzed. The simulation results show that AILC has faster convergence speed and higher tracking accuracy than traditional ILC in the presence of model uncertainty and can effectively improve the performance of the system. Finally, a method of combining L1 adaptive feedback with ILC is proposed to solve the problem of dynamic parameter uncertainty in PMSM servo system. The L1 adaptive controller is used to deal with the dynamic parameter uncertainty of the system and to compensate the external disturbance in the time domain so that the learning controller can be designed on the nominal system. The learning controller is used to improve the tracking ability of the system to the periodic input, and to compensate for the repeated uncertainties in the iterative domain such as friction, slot torque and so on. The simulation results show that this scheme can reduce the influence of dynamic parameter uncertainty on the system and improve the performance of the system.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM341;TM921.541

【参考文献】