【作者】 陆华才；

【导师】 陈子辰； 徐月同；

【作者基本信息】 浙江大学 ， 机械制造及其自动化， 2008， 博士

【摘要】 随着装备制造技术的不断进步,传统的进给系统越来越不能满足高速、高精密机械加工的要求,而永磁直线同步电机(PMLSM)直接进给技术以高速、高精密、快响应、大行程等优异的特性,在高速、高精密机械加工设备中得到了广泛的应用。但现有永磁直线同步电机进给系统中都需要机械位置传感器,位置传感器的存在不仅增加了成本,降低了系统可靠性,而且还带来了安装和维护的困难。因此研究成本低、性能良好的无位置传感器永磁直线同步电机进给系统成为近年来的一个研究热点。本文结合国家自然科学基金资助项目“无位置传感器的高速精密直接进给系统研究(No.50475101)”为背景,以解决永磁直线同步电机进给系统无位置传感器控制关键问题为主要目的,建立永磁直线同步电机进给系统无位置传感器控制系统,并采用理论研究、计算机仿真与实验研究相结合的研究方法,系统地进行算法分析、比较和优化及相应的技术基础研究与实验研究。第一章。阐述本论文研究的背景与重要意义,详细介绍永磁直线同步电机进给系统无位置传感器矢量控制的研究及发展现状,并指出永磁直线同步电机矢量控制与无位置传感器技术存在的关键问题,概述论文研究的主要内容。第二章,在分析永磁直线同步电机空间矢量变换规律基础上,建立永磁直线同步电机的d-q轴模型和α-β轴模型,提出PMLSM无位置传感器矢量控制整体方案。同时,利用Matlab/Simulink软件,构建PMLSM无位置传感器矢量控制仿真系统的基本模块,为后续的研究奠定基础。第三章,分析进给系统初始位置估计对于无位置传感器PMLSM进给系统矢量控制的重要性及对电机平稳启动的必要性。提出利用绕组磁饱和效应检测进给系统初始位置的方法,并用有限元方法分析求解PMLSM的绕组电感饱和特性,得到电感(电流响应)和位置的关系。阐述施加电压脉冲矢量检测初始位置方法的实现过程。实验结果表明,该方法检测精度可以达到7°电角度,完全满足平稳启动的要求。第四章,提出中高速时利用基于反电势积分的估计算法,低速时利用状态增广的扩展卡尔曼滤波估计算法,并利用加权的方法实现两种算法的平滑过渡,取长补短地构成一种复合估计算法。仿真结果表明,构建的复合估计算法在全速范围内具有较好的动、静态性能,受电机参数影响较小,能较好地估计电机速度和位置。第五章,在分析PMLSM进给系统各种控制方法优缺点的基础上,针对无位置传感器控制反馈信号是通过估计得到的特点,设计无位置传感器PMLSM进给系统滑模变结构速度控制器和滑模变结构最优位置控制器,并分析所设计的速度和位置控制器的稳定性与抗扰动能力。仿真结果表明,设计的控制器具有很强的鲁棒稳定性和较好的抗扰动性能。第六章,构建无位置传感器PMLSM进给系统软、硬件系统,重点阐述电压电流检测装置的构成和工作原理、高速数字信号处理器(DSP)及其外围电路的构成和原理,以及采用DSP实现无位置传感器PMLSM进给系统速度和位置估计算法的软件实现方法。进行无位置传感器进给系统速度和位置估计实验,并和实测值进行比较和分析。结果表明,速度与位置估计值和实测值基本一致。第七章,对本论文的研究工作和研究成果进行总结,展望未来的研究工作。更多还原

【Abstract】 The traditional feeding system can not satisfy the need of high speed and precision machining increasingly. On the contrary, the feeding system of permanent magnet linear synchronous motor (PMLSM) has the advantages in high speed, high precision, fast response and long journey, so it is widely applied in high speed and precision machining equipments. However, PMLSM feeding systems with mechanical position sensor in existence have many disadvantages such as high cost, low system dependability, and inconvenience of installation and maintenance. As a result, researching for the position sensorless control system with excellent performance has been a research hotspot recently. This dissertation has project background of the National Natural Science Foundation which is titled "Research on high speed and high precision direct drive system without position sensor (No. 50475101)". And the research theme is resolving the key problems of position sensorless control of PMLSM feeding system. A position sensorless control system of PMLSM feeding system is established, the estimation algorithm, optimization and control as well as some key technologies of the permanent magnet linear synchronous motor system are studied systematically by combining theoretical analysis, computer simulation with experiment.In chapter 1, the background and the significance of the research are stated. The development trend and the current research situations of sensorless control of the PMLSM feeding system are introduced in detail. The keystone of the vector control of PMLSM and the essential sensorless control are pointed out. And the main content of this dissertation is proposed.In chapter 2, based on analyzing the space vector transfer laws of the PMLSM, the d-q axis mathematics model and theα-βmathematics model of PMLSM are built. The sensorless position control project of PMLSM is constituted. Simultaneously, the basic simulating modules of PMLSM sensorless control are established using the software Matlab/Simulink.In chapter 3, the significance of estimating the feeding system initial position for PMLSM sensorless position control and the necessity for starting smoothly are analyzed. A method of detecting initial position using saturation effect of the winding inductance is put forward, which is verified through electromagnetic field two-dimension finite element method, and a relationship between winding inductance (current response) and mover position. Then, the process of estimating initial position by injecting voltage impulse vector is expounded. Experimental results show that the precision of the method can arrive 7 electrical degrees, which can satisfy requirements of PMLSM starting smoothly.In chapter 4, an estimated method is put forward, which is using back-EMF to estimate the speed and position at middle-high speed, and using augmented Extended Kalman Filter at low speed, and obtaining a compound checking method using weighted method to merge the forward ways. Simulations demonstrate that the compound method has high performance at dynamic and static state. And the method can estimate the mover speed and position accurately.In chapter 5, based on analyzing merits and demerits of control strategies for PMLSM vector control, a sliding mode speed controller and a sliding mode position controller are designed, considering that the feedback signal is obtained by estimation not by mechanical sensors. Then the stabilization and anti-disturb ability of the designed sliding mode controller are analyzed theoretically. Simulating results show that the performances of strong robustness to external disturbance.In chapter 6, the hardware and software for PMLSM sensorless position control system are built. The composition and pricnciple of the measurement devices of currents and voltages, the fundamentals of the digital signal processor (DSP) and its peripheral circuits, the ways of sensorless estimation software using DSP are represented. Experiments of speed and position estimation without position sensor are made. Experimental results show that the position sensorless control has good performance, and the speed and position of the feeding system can be estimated.In chapter 7, the study contents and conclusions of the dissertation have been summarized. Then, the further research works have been forecast.更多还原