电励磁双凸极电机起动系统的控制优化研究

发布时间：2018-03-30 08:20

本文选题：电励磁双凸极电机　切入点：转矩脉动抑制　出处：《南京航空航天大学》2016年硕士论文

【摘要】：电励磁双凸极电机(Doubly Salient Electro-magnetic Machine,DSEM)具有结构简单可靠、制造方便、气隙磁通调节灵活等优点,且可实现起动/发电一体化,因而在航空电源领域有着广泛的应用前景。但DSEM作为起动机运行时存在转矩脉动较大的问题,且需位置传感器提供准确的位置信息以实现换相,这都降低了起动系统的性能,也限制了其应用范围。因此本课题致力于DSEM起动控制系统的优化研究,研究内容包括控制策略的优化和无位置传感器低速起动控制技术。本文以一台18KW 12/8极结构的DSEM为研究对象,解析了电机的数学模型,阐述了电动运行时转矩的产生机理;并通过有限元软件对其基本电磁特性进行了分析,为研究电机控制策略的优化奠定基础。本文从电机换相、非导通相环流问题以及转矩特性方面对DSEM的四种控制策略进行详细分析,并得到如下结论:三状态控制策略下电机换相时因续流而总会向电源回馈电能;三状态提前角度控制策略下非导通相环流问题较为严重;六状态控制策略下非导通环流问题可忽略,且转矩特性较优;九状态控制策略可使电机获得最优的转矩特性,但是同时也增加了变换器下管的电流应力。本文还研究了DSEM初始位置检测技术,基于脉冲注入基本原理研究了四种初始位置检测方法:1)针对峰值电流响应法在串联电感曲线交叉位置易判断出错的问题,将全周期内串联电感进行重新分区,并优化转子位置判断逻辑,可避免因电流检测精度较低引起的初始定位失败问题;2)通过拟合响应电流峰值与转子位置的函数关系可得到精确的转子位置;3)提出一种基于非导通相端电压的相自感比较法及其改进算法,可与电流响应法互为备份;4)提出了一种基于相自感线性分区模型的精确初始定位算法,该方法不会受母线电压变化的影响。本文还提出了基于电压脉冲注入的低速无位置起动控制方法,实现了电机转子位置闭环起动;分析了电枢反应对理想换相点的影响,指出电枢反应虽然造成电感波形畸变,但并不会使得理想换相点产生偏移;同时还对该方法进行了误差分析,指出电机实际换相位置与理想换相点的最大偏差为一个组合脉冲的宽度。最后,介绍了实验的硬件系统和软件程序的流程图,并进行了实验验证。
[Abstract]:Doubly Salient Electro-magnetic Machine (DSEM) has the advantages of simple and reliable structure, convenient manufacture, flexible air gap flux regulation, and can realize the integration of starting and generating, so it has a wide application prospect in the field of aviation power supply.However, DSEM has the problem of large torque ripple when it is running as starter, and needs position sensor to provide accurate position information to realize commutation, which reduces the performance of starting system and limits its application scope.Therefore, this thesis is devoted to the optimization research of DSEM starting control system, including the optimization of control strategy and the low speed starting control technology without position sensor.In this paper, a 18KW 12 / 8 pole structure DSEM is taken as the research object, the mathematical model of the motor is analyzed, the mechanism of the torque generation during the electric operation is expounded, and the basic electromagnetic characteristics of the motor are analyzed by the finite element software.It lays a foundation for the optimization of motor control strategy.In this paper, the four control strategies of DSEM are analyzed in detail from the aspects of commutation, non-conducting circulation and torque characteristics. The conclusions are as follows: under the three-state control strategy, the electric energy is always fed back to the power supply when the motor commutates because of the continuous current;The non-conducting circulation problem is more serious under the three-state advance angle control strategy; the non-conductive circulation problem is negligible under the six-state control strategy and the torque characteristic is better; the nine-state control strategy can make the motor obtain the optimal torque characteristic.But it also increases the current stress of the converter.In this paper, the DSEM initial position detection technique is also studied. Based on the basic principle of pulse injection, four initial position detection methods: 1) the peak current response method is easy to judge the error in the crossover position of the series inductance curve.The series inductors are rezoned throughout the period, and the rotor position judgment logic is optimized.The problem of initial positioning failure caused by low accuracy of current detection can be avoided. By fitting the functional relationship between the peak value of response current and the position of the rotor, an accurate rotor position can be obtained.) A phase based on the terminal voltage of non-conducting phase is proposed.Self-Inductive comparison method and its improved algorithm,An accurate initial location algorithm based on phase inductance linear partition model is proposed, which is independent of bus voltage variation.In this paper, a low speed positionless starting control method based on voltage pulse injection is proposed to realize the closed loop starting of the motor rotor position, and the effect of armature reaction on the ideal commutation point is analyzed, and it is pointed out that the armature reaction results in the distortion of inductance waveform although the armature reaction results in the distortion of inductance waveform.At the same time, the error analysis of the method is carried out, and it is pointed out that the maximum deviation between the actual commutative position of the motor and the ideal commutation point is the width of a combined pulse.Finally, the hardware system and software flow chart of the experiment are introduced and verified.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM352

【参考文献】