毕业论文外文翻译--基于DSP高速无刷直流电机控制使用直流环节电压控制.DOC

1、基于DSP高速无刷直流电机控制使用直流环节电压控制一个基于DSP高速度传感器控制无刷直流电机(无刷直流)汽车使用直流环节电压控制方案被提出了。无刷直流电机的运行在一个高速度范围、驱动系统可以有一个比较轻体积小,在同一输出等级。在现有的无传感器控制方案,通常采用PWM(脉宽调制)技术作为一个速度控制。然而,由于PWM技术和变频变换不能履行独立,明显的变换延迟存在于高速地区。另一方面，使用的直流母线电压控制方案，变频器操作与方波120传导速度控制是通过调节斩波直流环节逆变器的输入电压实现。利用这项技术,因为电压控制和变换就可以实现独立,延迟不存在运算可以交换甚至在一个高速地区。此外，以有一个波形相

2、位目前类似的矩形波和终端电压更有效率的处理在位置检测电路。实际应用变换议题延迟的一个高速度的无传感器控制进行了讨论。整个控制系统的实施应用DSP芯片的无刷直流电机TMS320C240和有效性的比较验证了仿真和实验。关键词 无刷直流电机、无传感器控制、DSP控制。1.介绍在许多工业领域,需要安装一个轴传感器可能会大幅度增加推动成本以及复杂的电机配置1。特别是,为电动机建在一个完全密封压缩机、轴传感器是难以运用由于传感器可靠性降低高温需要额外的导线。此外,这些传感器,尤其是霍尔传感器,温度敏感,限制了电机运行大约75以下1。一个绝对速度传感器通常限于大约6000转速与旋转需要一个特殊的外部电路。同

3、时,传感器的精度也会受到安装的准确性。要克服这些弊端,无位置传感器无刷直流电机控制技术提出了一个1 5。有两类位置检测方案,即,该方法利用电机的反电势2,该方法基于检测间隔进行随心所欲的二极管3。在现有的无传感器控制方案、PWM技术技术通常用于一个速度控制。然而,由于PWM技术和变频变换不能履行独立,明显的变换延迟的一个高速度可能存在的区域。最近，以提高驱动器E的效率，并提供所需的电流波形，一个传感器控制计划使用准电流源逆变器已提出6。这样的电路装置被称为一个变量直流环节逆变器7。在该方案中,逆变频率控制供应电流有三相矩形脉冲宽度120度及马达速度控制电压调节采用降压斩波器作为降压转换器。然而

4、，一些优势的直流母线电压超过传统的两相PWM在高转速传感器控制计划控制计划都没有得到解决。本文提出了一种基于DSP高速无刷直流电机无位置传感器控制使用直流环节电压控制方案。无刷直流电机推在一个重量轻在相同的额定功率。控制高速无刷直流电机无转轴侦测元件传感器、基于DSP开发利用TMS320C240控制器。使用直流母线电压控制计划，逆变器的操作与方波120度传导间隔和速度控制是通过调节斩波直流环节逆变器的输入电压来实现。利用这项技术,因为电压控制和变换就可以实现独立,如运算可以交换延迟传统PWM方法二段式激励是不存在的。甚至在一个高速地区,将讨论在以后的部分。转子位置信息利用反电动势检测电压从终端

5、电机和逆变器的开关顺序的2。反电动势的感觉到用于集成电路和比较得到变换信号。检测变换信号用于申请适当的下一个序列,得到了转速逆变器在DSP。计算速度的数字控制，控制算法和控制器的输出应用到斩波器。实际应用议题变换时延的激励方案二段式PWM高速进行了论述,并对直流环节电压的优势控制方案在高速度传感器控制提及。整个控制系统的实施应用DSP芯片的无刷直流电机TMS320C240和有效性的比较验证了仿真和实验。2、无刷直流电机的无传感器控制一个无刷直流电机本文认为由永磁体安装对转子表面和三相集中而流离失所的定子120度。定子电流励磁方案段提供的地方只有两三个阶段都很兴奋在任何紧急的时间和一阶段在120

6、年期间进行8。这激励方案不需要死亡的时间电力设备的发动的软铁转子,即使它没有持续的扭矩。永磁类型都有径向非磁化转子。这种类型在展,可以有效进行非全相利用得到的转子位置信息。转子位置信息通常得到间接检测方法利用电机反电动势无刷电机无位置传感器控制1 _4。在文献2的基础上,从转子位置估计的整合反电动势波形。该方法是众所周知的提供等优点减少开关噪声灵敏度和自动调节的开关瞬间不相移30度。因此,该检测方案本文采用。速度的信息可从衍生工具检测信号的位置。自从变换信号输入DSP每隔60度期内,时钟在DSP台TC数量和计数的期间是一个60度,机械转子速度可计算转速如下:在P是大量的增长极。3、传感器存在的

7、问题的速度控制方案在现有的无传感器控制方案,二段式激励技术是PWM(脉宽调制)通常用于一个速度控制。基于此方法执行PWM(脉宽调制),脉宽调制方案的经典歌曲了单极和双相性精神交换的方法。在单极开关的方法,PWM技术是叠加在那两人中的一个主动开关在国家,而其他开关仍在状态。另一方面一方面,在双极切换方法,这两个积极执行PWM开关在同一时间内。自从单极开关有一个优势的减少开关损耗,这个方案是首选的4。此外，基于位置的脉宽调制叠加，单极开关的方法是分类为持续的阶段，将相位调制，上下开关开关脉宽调制，脉宽调制。在PWM调制方式进行的阶段,每一个开关被执行PWM技术在第一个60度程度的活跃时间和保留在国

8、家的期间第二个60度区在间,去相PWM调制方式,反之亦然3,4。在上面的开关PWM调制方式、PWM(脉宽调制)被执行的时候,只有在上部之一两个活跃的开关,在较低的开关PWM调制方式,反之亦然。根据基于其使用的PWM调制方式,该控制技术可能导致减刑延迟或者一个不规则的开关频率的电力设备在高速度传感器控制。图1显示PWM开关周期和PWM方案2相励磁整流在瞬间之间的关系。在图1,T年代和f年代表示PWM技术转换期间和频率,分别。图1(一)说明情况理想的变换。作为古雷中可以看出,如果运算可以交换的瞬间同步,与去年底PWM开关期间, 可以得到一个理想的换相逆变器序列的变化没有任何延迟。然而,由于运算可以

9、交换即时以同步进行,与去年底目前PWM周期开始下一个逆变器顺序图1(b),这是一般使用方法。这个结果在一个不受欢迎的变换延迟和最大的价值这次延误PWM技术转换期间来。如果开关频率被选择作为16千赫,最高价值的变换将是62.5秒的延迟。尽管这些减刑延迟可以忽略一个速度范围，它具有重大影响的相电流响应和驱动器的性能在高速度。 例如,当一个两极电动机转速为50000转/分,60度间隔200秒。这就可以减少运算可以交换延迟增加PWM开关频率。然而事实上,这些开关频率不能增加无极限,因为增加的开关损耗。同时,开关频率的商用电力设备是少于20 kHz。因此,为了避免不良变换延迟,接下来的逆变器应用序列一旦

10、变换信号中断发生。那么,现在的PWM周期必须终止和新型PWM周期的同步变换中断信号必须开始了。在上部和下部开关PWM(脉宽调制)方案,这可能会得到一个不规则的开关频率大大高于f年代高职条件下如图1(c)。在持续的和持续相PWM方案，这种不规则的开关频率不会发生以来阶段执行PWM不断改变每60度间隔。因此，对正在进行的和持续的PWM方法在图1（c）计划阶段，可以是一个很高的速度传感器控制的首选方式。不过,仍然有一个问题。在高速度,只有少数的PWM脉冲可以用于速度控制在60度间隔。因为一个60度区间的两极成为200秒内每秒电机50000转/分,如果开关频率被选择作为16千赫,一定数量的PWM脉冲在

11、60仅为3.2,导致不平等的PWM脉冲数3或4在60度间隔。除非解决脉冲宽度相当高,这可能导致速度脉动在稳态和降解精度的位置信号的检测。这个问题比较严重的地区以更高的速度,可以有效克服，通过控制电压和频率直流环节电压独立的控制方案。图1 PWM开关周期之间的关系和运算可以交换即时:(a)理想的变换,变换(b)的情况下延迟和(c)不规则的情况下切换频率。9Electric Power Components and Systems, 30:889900, 2002Copyright c 2002 Taylor & Francis1532-5008/ 02 $12.00 + .00DO I: 10.

12、1080/ 15325000290085190 DSP-Based High-Speed SensorlessControl for a Brushless DC Motor Using a DC Link Voltage ControlKYEONG-HWA KIMMYUNG-JOONG YOUNDepartment of Electrical EngineeringKorea Advanced Institute of Science and TechnologyTaejon, KoreaA DSP-based high speed sensorless control for a brus

13、hless DC (BLDC) motor using a DC link voltage control scheme is presented. By operating the BLDC motor in a high speed range, the drive system can have a small size and be light weight at the same output rating. In the existing sensorless control schemes, the PW M technique is generally used as a sp

14、eed control. However, since the PWM and inverter commutation cannot be performed independently, a significant commutation delay may exist in a high-speed region. On the other hand, using the DC link voltage control scheme, the inverter is operated with the squarewave of 120 conduction and the speed

15、control is achieved by regulating the DC link input voltage of the inverter through the chopper. By using this technique,since the voltage control and commutation can be achieved independently, a commutating delay does not exist even in a high speed region. Also, the phase current can have a wavefor

16、m similar to the rectangular wave and the terminal voltage is more e -cient to deal with in the position detection circuits. The practical implementation issues concerning the commutation delay in a high speed sensorless control are discussed. The whole control system is implemented on a BLDC motor

17、using DSP TMS320C240 and the eectiveness is veried through the comparative simulations and experiments.Keywords brushless DC motor, sensorless control, DSP control1. IntroductionIn many industrial elds, the installation of a shaft sensor may signi cantly increase the drive cost as well as complicate

18、 the motor configuration 1. In particular, for a motor built in a completely sealed compressor, a shaft sensor is difficult to apply due to the degradation of the sensor reliability in high temperature and the need for extra lead wires. Furthermore, these sensors, particularly Hall sensors, are temp

19、erature sensitive, limiting the operation of the motor to below about 75C 1. An absolute sensor is generally speed limited to about 6000 rpm and a resolver needs a special external circuit. Also, the sensor accuracy may be affected by the accuracy of the mounting. To overcome these drawbacks, sensor

20、less control techniques for a BLDC motor have been proposed 1_5. There are two categories of position detection schemes, namely, the method using the back EMF of the motor 2 and themethod based on the detection of the conducting interval of free-wheeling diodes 3.In the existing sensorless control s

21、chemes, the PWM technique is generally used for a speed control. However, since the PWM and inverter commutation cannot be performed independently, a signi cant commutation delay may exist in a high speed region. Recently, to improve the drive effciency and provide the desired current waveform, a se

22、nsorless control scheme using a quasi-current source inverter has been proposed 6. Such a circuit arrangement is known as a variable DC link inverter 7.In this scheme, the inverter frequency is controlled to supply three-phase rectangular current with a pulse width of 120 and the motor voltage for t

23、he speed control is regulated by using a step-down chopper acting as a buck converter. However, some advantages of the DC-link voltage control scheme over the conventional 2-phase PWM scheme in the high speed sensorless control have not been addressed.This article presents a DSP-based high speed sen

24、sorless control for a BLDC motor using a DC link voltage control scheme. By driving the BLDC motor at high speed, the overall drive system can have a small size and a light weight at the same power rating. To control the BLDC motor at high speed without a shaft sensor, a DSP-based controller is deve

25、loped using TMS320C240. Using the DC link voltage control scheme, the inverter is operated with the squarewave of 120conduction interval and the speed control is achieved by regulating the DC link input voltage of inverter through the chopper. By using this technique, since the voltage control and c

26、ommutation can be achieved independently, the commutating delay such as in the conventional 2-phase excitation PWM methods does not exist even in a high speed region, which will be discussed in the later section. The rotor position information is detected using the back EMF from the terminal voltage

27、s of the motor and the switching sequence of the inverter 2. The sensed back EMF is used in the integration and comparison circuits to obtain the commutation signals. The detected commutation signals are used to apply the proper next sequence of inverter and obtain the rotational speed within a DSP.

28、 The calculated speed is controlled by a digital PI control algorithm and the controller output is applied to the chopper. The practical implementation issues concerning the commutation delay of the 2-phase excitation PWM schemes at high speed are discussed and some advantages of the DC link voltage

29、 control scheme in a high speed sensorless control are mentioned. The whole control system is implemented on a BLDC motor using DSP TMS320C240 and the eectiveness is veri ed through the comparative simulations and experiments.2. Sensorless Control of BLDC MotorA BLDC motor considered in this paper c

30、onsists of permanent magnets mounted on the rotor surface and three-phase concentrated stator windings displaced by 120 . The stator currents are supplied by the 2-phase excitation scheme where only two of the three phases are excited at any instant of time and one phase is conducted during 120 peri

31、od 8. This excitation scheme does not require dead time of the power devices, and furthermore, the unconducting open-phase can be usefully utilized to obtain the rotor position information. The rotor position information are generally obtained from the indirect detection method using the motor back

32、EMF 1_4. In 2, the rotor position has been estimated from the integration of the back EMF waveform. This method is known to provide the advantages such as the reduced switching noise sensitivity and automatic adjustment of the switching instants without the phase shift of 30 degrees. Thus, this dete

33、ction scheme is employed in this paper. The speed information can be obtained from the derivative of the detected position signals. Since the commutation signals are fed into a DSP every 60period, if the counter clock in DSP is TC and the number of count during 60 degrees is a, the mechanical rotor

34、speed can be computed in rpm as follows:where P is the number of poles.3. Problems of Existing Sensorless Speed Control SchemesIn the existing sensorless control schemes, the 2-phase excitation PWM technique is generally employed for a speed control. Based on the method executing the PWM,PWM schemes

35、 can be classified as the unipolar and bipolar switching methods.In the unipolar switching method, the PWM is superimposed on one of the two active switches in on state, while the other switch remains on state. On the other hand, in the bipolar switching method, the two active switches execute the P

36、WM at the same time. Since the unipolar switching has an advantage of the reduced switching loss, this scheme is generally preferred 4. Also, based on the position that the PWM is superimposed on, the unipolar switching method is classi ed as the on-going phase PWM, off-going phase PWM, upper switch

37、 PWM, and lower switch PWM schemes. In the on-going phase PWM scheme, each switch executes the PWM during the rst 60 degrees of active interval and is held in on state during the second 60interval, and in the off-going phase PWM scheme, vice versa 3, 4.In the upper switch PWM scheme, the PWM is exec

38、uted only on the upper one of two active switches, and in the lower switch PWM scheme, vice versa. Depending on the used PWM scheme, this control technique may cause a commutation delay or an irregular switching frequency of the power devices in a high speed sensorless control.Figure 1 shows the rel

39、ation between the PWM switching period and commutating instant in the 2-phase excitation PWM scheme. In Figure 1, Ts and fs denote the PWM switching period and frequency, respectively. Figure 1(a) shows a case of the ideal commutation. As can be seen in the gure, if the commutating instant is synchr

40、onized with the end of the PWM switching period, an ideal commutation can be obtained without any delay in the inverter sequence change. However,since the commutating instant depends on the rotor position, it does not usually coincide with the end of the PWM period. In this case, the commutation can

41、 be performed synchronized with the end of the present PWM period to start a next inverter sequence as Figure 1(b) , which is the normally used method. This results in an undesirable commutation delay and the maximum value of this delay becomes the PWM switching period. If the switching frequency is

42、 chosen as 16 kHz,the maximum value of the commutation delay will be 62.5 sec. Even though this commutation delay can be neglected for a medium speed range, it has significant in uences on the phase current response and drive performance at high speed since the 60-degree interval that the commutatio

43、n arises in is relatively small. For example, when a 2-pole motor is rotating at 50,000 rpm, 60-degree interval becomes 200 sec. This commutating delay can be reduced by increasing the PWM switching frequency. In practice, however, the switching frequency cannot be increased without limit because of

44、 the increased switching loss. Also, the switching frequency of commercially available power devices is less than 20 kHz. Thus, to avoid an undesirable commutation delay, the next inverter sequence has to be applied as soon as the commutation signal interrupt occurs. Then, the present PWM period has

45、 to be terminated and the new PWM period synchronized with the commutation interrupt signal must be started. In the upper and lower switch PWM schemes,this may yield an irregular switching frequency much larger than f s under a high duty condition as shown in Figure 1(c) . In the on-going and off-go

46、ing phase PWM schemes, this irregular switching frequency does not occur since the phase executing the PWM is continually changed every 60interval. Thus, the on-going and off-going phase PWM schemes with the method in Figure 1(c) can be a preferred way for a high speed sensorless control. Neverthele

47、ss, there is still a problem. At high speed, only a few PWM pulses can be used for the speed control during a 60interval. Since a 60interval of a 2-pole motor becomes 200 sec at 50,000 rpm,if the switching frequency is chosen as 16 kHz, the number of PWM pulses during 60 is only 3.2, which results i

48、n an unequal number of PWM pulses 3 or 4 during a 60 interval. Unless the resolution of the pulse width is considerably high, this may result in a speed ripple at steady state and degrade the accuracy of the position signal detection. This problem is more serious at a higher speed region and can be effectively overcome by controlling the voltage and frequency independently by the DC link voltage control scheme.Figure 1. Relation between the PWM switching period and commutating instant:(a) Ideal commutation, (b) Case of commutation delay, and (c) Case of irregular switchin