滑膜观测器物理意义_电机控制中的滑模控制(滑模观测器设计)

本文探讨了电机无位置传感器控制中的滑模观测器设计,特别是针对永磁同步电机(PMSM)。通过分析电机的数学模型,介绍了如何利用反正切函数的滑模观测器算法来估算转子位置。文章还讨论了滑模观测器的构造,以及为确保系统稳定性所需的条件,并提出了高频开关信号、相位延迟等问题的解决方案。
摘要由CSDN通过智能技术生成

本文小结一下,电机无位置控制中的一种--滑模观测器的设计,还有自适应观测器的设计等其他内容,等有时间之后再来总结。现代永磁同步电机控制原理及MATLAB仿真

永磁同步电机无位置传感器控制系统的研究与设计-陈凯-华中科技大学

目前关于电机无传感器的控制的方式大体上是可以分为:电机的无传感控制的概览

如果描述不对,或者是还有别的更好的方法,欢迎留言或者私信,共同进步。在这里仅仅为了说明滑模观测器的应用,故采用简单的传统滑模观测器算法中的基于反正切函数的转子位置估计算法。

首先,要明白一下,什么叫做无传感器控制?在自动控制原理中,我们学过了闭环控制的诸多好处,在处理复杂、干扰量多的情况时,通常是采用闭环进行控制。比如在BLDC的控制中,为了节约成本,通常是采用霍尔元器件来定位转子位置,尽管它的精度低,但是由于BLDC是方波控制,无伤大雅;但是在PMSM的正弦波控制中,通常采用编码器传感器来反馈电机的转子位置。而传感器往往是整个控制系统中价格较昂贵的器件,因此处于节约成本的考虑,采用无传感器控制。

基于反正切函数的滑模观测器系统框图(基于静止坐标系 )

因为还有基于同步旋转坐标系的。三相PMSM矢量控制系统框图基于反正切函数的滑模观测器系统框图

从该控制框图可以知道,基于反正切函数的滑模观测器控制系统同三相PMSM矢量控制的区别在于,使用SMC框图,代替了原先的Clark和Park(需要知道转子的位置

equation?tex=%5Ctheta_r )变换。

二. 永磁同步电机的数学模型分析

SMC的输入量为由

equation?tex=u_%7B%5Calpha+%5Cbeta%7D

equation?tex=i_%7B%5Calpha+%5Cbeta%7D (无输入

equation?tex=%5Ctheta_r),下面推导SMC中的滑模观测器的设计部分:

可以先参考一下,永磁同步电机的动态模型帆布鞋:永磁同步电机的动态模型​zhuanlan.zhihu.com15ec876b445883da657f17518a8ab77a.png

首先在dq坐标系下 ,对于PMSM,其电压方程为:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_d%5C%5C+u_q+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26-w_eL_q%5C%5C+w_eL_d%26R_s%2BpL_q+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_d%5C%5C+i_q+%5Cend+%7Barray%7D+%5Cright%5D+%2B+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+0%5C%5C+w_e%5Cpsi_f+%5Cend+%7Barray%7D+%5Cright%5D

有时间再详细将这一部分整理出来;如果是直接对上式进行Park逆变换,

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_%7B%5Calpha%7D%5C%5C+u_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26-sin%5Ctheta%5C%5C+sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_d%5C%5C+u_q+%5Cend+%7Barray%7D+%5Cright%5D

即在上面的等式中,同时左乘变换矩阵

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26-sin%5Ctheta%5C%5C+sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D 可以得到:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26-sin%5Ctheta%5C%5C+sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_d%5C%5C+u_q+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26-sin%5Ctheta%5C%5C+sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26-w_eL_q%5C%5C+w_eL_d%26R_s%2BpL_q+%5Cend+%7Barray%7D+%5Cright%5D++%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_d%5C%5C+i_q+%5Cend+%7Barray%7D+%5Cright%5D+%2B+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26-sin%5Ctheta%5C%5C+sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+0%5C%5C+w_e%5Cpsi_f+%5Cend+%7Barray%7D+%5Cright%5D

其中的难点计算量在于等式右边的第一项的计算,我们可以将

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_d%5C%5C+i_q+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26sin%5Ctheta%5C%5C+-sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_%7B%5Calpha%7D%5C%5C+i_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D 代入到上面的式子中进行计算,则简化为计算:

equation?tex=+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26-sin%5Ctheta%5C%5C+sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26-w_eL_q%5C%5C+w_eL_d%26R_s%2BpL_q+%5Cend+%7Barray%7D+%5Cright%5D++%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+cos%5Ctheta%26sin%5Ctheta%5C%5C+-sin%5Ctheta%26cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D+

可以得到,该变换的2*2的矩阵中的元素为:

equation?tex=A_%7B11%7D%3D%28R_s%2BpL_d%29cos%5E2%5Ctheta-w_eL_dsin%5Ctheta+cos%5Ctheta%2Bw_eL_qcos%5Ctheta+sin%5Ctheta%2B%28R_s%2BpL_d%29sin%5E2%5Ctheta%5C%5C+%3DR_s%2BpL_%7B%5Calpha%7D

其中

equation?tex=%E4%BB%A4L_0%3D%5Cfrac%7BL_d%2BL_q%7D%7B2%7D

equation?tex=%2CL_1%3D%5Cfrac%7BL_d-L_q%7D%7B2%7D ,

equation?tex=%E5%88%99%E6%9C%89L_%7B%5Calpha%7D%3DL_0%2BL_1cos2%5Ctheta

同理可以得到变换到静止坐标系 下

equation?tex=%5Calpha+%5Cbeta 坐标系下的数学模型为:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_%7B%5Calpha%7D%5C%5C+u_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_%7B%5Calpha%7D%26pL_%7B%5Calpha+%5Cbeta%7D%5C%5C+pL_%7B%5Calpha+%5Cbeta%7D%26R_s%2BpL_%7B%5Calpha%7D+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_%7B%5Calpha%7D%5C%5C+i_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D+%2B+w_e%5Cpsi_f+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+-sin%5Ctheta%5C%5C+cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D

其中,

equation?tex=L_%7B%5Calpha%7D%3DL_0%2BL_1cos2%5Ctheta%5C%5C+L_%7B%5Cbeta%7D%3DL_0-L_1cos2%5Ctheta%5C%5C+L_%7B%5Calpha+%5Cbeta%7D%3DL_1sin2%5Ctheta%5C%5C+L_0%3D%5Cfrac%7BL_d%2BL_q%7D%7B2%7D%5C%5C+L_1%3D%5Cfrac%7BL_d-L_q%7D%7B2%7D+

由于其中包含了转子位置信息

equation?tex=2%5Ctheta 项,增加了转子位置信息估计的难度。为了便于计算,可通过适当的变换将

equation?tex=2%5Ctheta 项消除,它的出现是因为电感矩阵不对称导致的。因此需要将电感矩阵变换为对称的矩阵。

三. 永磁同步电机的数学模型变换与分析

将原始的数学模型

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_d%5C%5C+u_q+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26-w_eL_q%5C%5C+w_eL_d%26R_s%2BpL_q+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_d%5C%5C+i_q+%5Cend+%7Barray%7D+%5Cright%5D+%2B+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+0%5C%5C+w_e%5Cpsi_f+%5Cend+%7Barray%7D+%5Cright%5D

中的电感矩阵变为对称的矩阵,即

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26-w_eL_q%5C%5C+w_eL_q%26R_s%2BpL_d+%5Cend+%7Barray%7D+%5Cright%5D ,则容易将永磁同步 电机的数学模型化简为:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_d%5C%5C+u_q+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26-w_eL_q%5C%5C+w_eL_q%26R_s%2BpL_d+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_d%5C%5C+i_q+%5Cend+%7Barray%7D+%5Cright%5D+%2B+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+0%5C%5C+w_e%5Cpsi_f%2B%28L_d-L_q%29%28w_ei_d-pi_q%29+%5Cend+%7Barray%7D+%5Cright%5D

则按照同样的变换规则,即左乘Park逆变换矩阵,可以得到:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_%7B%5Calpha%7D%5C%5C+u_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%26w_e%28L_d-L_q%29%5C%5C+-w_e%28L_d-L_q%29%26R_s%2BpL_d+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_%7B%5Calpha%7D%5C%5C+i_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D+%2B+%5Bw_e%5Cpsi_f%2B%28L_d-L_q%29%28w_ei_d-pi_q%29%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+-sin%5Ctheta%5C%5C+cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D

同时,可以定义该数学模型中的扩展反电势为:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+E_%7B%5Calpha%7D%5C%5C+E_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Bw_e%5Cpsi_f%2B%28L_d-L_q%29%28w_ei_d-pi_q%29%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+-sin%5Ctheta%5C%5C+cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D

则只要是能够估算出扩展反电势,就可以利用反正切函数来求解得到转子的位置信息。

四. 理解滑模观测器的构造

为了简化计算过程,这里以隐极电机(也叫作表贴式电机),即有

equation?tex=L_d%3DL_q%3DL_s ,关于这方面的称呼区别,请参考 前面的文章。变换之后的等效数学模型化简为:

equation?tex=%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_%7B%5Calpha%7D%5C%5C+u_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D%3D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+R_s%2BpL_d%260%5C%5C+0%26R_s%2BpL_d+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_%7B%5Calpha%7D%5C%5C+i_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D+%2B+%5Bw_e%5Cpsi_f%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+-sin%5Ctheta%5C%5C+cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D

将这个矩阵方程写成电流的状态变量方程,即

equation?tex=%5Cfrac%7Bd%7D%7Bdt%7D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_%7B%5Calpha%7D%5C%5C+i_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D%3D%5Cfrac%7B1%7D%7BL_d%7D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+-R_s%260%5C%5C+0%26-R_s+%5Cend+%7Barray%7D+%5Cright%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+i_%7B%5Calpha%7D%5C%5C+i_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D+%2B%5Cfrac%7B1%7D%7BL_d%7D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+u_%7B%5Calpha%7D%5C%5C+u_%7B%5Cbeta%7D+%5Cend+%7Barray%7D+%5Cright%5D++%2B+%5Bw_e%5Cpsi_f%5D+%5Cleft%5B+%5Cbegin+%7Barray%7D%7Bccc%7D+-sin%5Ctheta%5C%5C+cos%5Ctheta+%5Cend+%7Barray%7D+%5Cright%5D

定义误差

equation?tex=E_%7B%5Calpha%7D%3Di_%7B%5Calpha%7D

竟然是 忘记 了自己已经小结过一篇了帆布鞋:表贴式永磁同步电机滑模控制的小结​zhuanlan.zhihu.com15ec876b445883da657f17518a8ab77a.png

粘贴过来

将上述方程写成状态方程的形式 :

equation?tex=%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+%5Cfrac%7Bdi_%5Calpha%7D%7Bdt%7D%5C%5C%5Cfrac%7Bdi_%5Cbeta%7D%7Bdt%7D+%5Cend%7Barray%7D+%5Cright%5D%3D+-%5Cfrac%7BR_s%7D%7BL_s%7D+%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+i_%5Calpha%5C%5Ci_%5Cbeta+%5Cend%7Barray%7D+%5Cright%5D+%2B%5Cfrac%7B1%7D%7BL_s%7D+%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+u_%5Calpha%5C%5Cu_%5Cbeta+%5Cend%7Barray%7D+%5Cright%5D+-%5Cfrac%7B1%7D%7BL_s%7D+%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+e_%5Calpha%5C%5Ce_%5Cbeta+%5Cend%7Barray%7D+%5Cright%5D

这时,

equation?tex=i 是我们的状态变量,定义

equation?tex=i_d 是我们的期望值,误差

equation?tex=e%3Di_d-i ,

则有

equation?tex=%5Cdot%7Be%7D%3D%5Cdot%7Bi_d%7D-%5Cdot%7Bi%7D ,代入

equation?tex=%5Cdot%7Bi%7D%3D-%5Cfrac%7BR_s%7D%7BL_s%7Di%2B%5Cfrac%7B1%7D%7BL_s%7Du-%5Cfrac%7B1%7D%7BL_s%7DE ,可得:

equation?tex=%5Cdot%7Be%7D%3D%5Cdot%7Bi_d%7D-%5B-%5Cfrac%7BR_s%7D%7BL_s%7Di%2B%5Cfrac%7B1%7D%7BL_s%7Du-%5Cfrac%7B1%7D%7BL_s%7DE%5D+%5C%5C+%3D%5Cdot%7Bi_d%7D%2B%5Cfrac%7BR_s%7D%7BL_s%7Di%2B%5Cfrac%7B1%7D%7BL_s%7DE-%5Cfrac%7B1%7D%7BL_s%7Du

equation?tex=u+%3D+L_s%5Cdot%7Bi_d%7D-R_si_d%2B%5Crho%5Cfrac%7B%7Ce%7C%7D%7Be%7D 代入上述误差方程中可得:

equation?tex=%5Cdot%7Be%7D%3D-%5Cfrac%7BR_s%7D%7BL_s%7De%2B%5Cfrac%7B1%7D%7BL_s%7D%28E-%5Crho%5Cfrac%7B%7Ce%7C%7D%7Be%7D%29

误差方程稳定性分析:

1)假设e<0,则要求

equation?tex=%5Cdot%7Be%7D >0:从上面的表达式可以看出

equation?tex=-%5Cfrac%7BR_s%7D%7BL_s%7De >0,

equation?tex=%5Cfrac%7B%7Ce%7C%7D%7Be%7D%3D-1 ,则要求

equation?tex=%5Crho%3E%7CE%7C 则可以成立;

2)假设e>0,则要求

equation?tex=%5Cdot%7Be%7D <0:从上面的表达式可以看出

equation?tex=-%5Cfrac%7BR_s%7D%7BL_s%7De <0,

equation?tex=%5Cfrac%7B%7Ce%7C%7D%7Be%7D%3D+1 ,则要求

equation?tex=%5Crho%3E%7CE%7C 则可以成立;

这样就可以得到滑模成立的条件是

equation?tex=%5Crho%3E%7CE%7C

再由误差方程,可以得到,我们实际上构造的定子电流滑模观测器为:

equation?tex=%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+%5Cdot%7Bi_%7B%5Calpha+d%7D%7D%5C%5C%5Cdot%7Bi_%7B%5Cbeta+d%7D%7D+%5Cend%7Barray%7D+%5Cright%5D%3D+-%5Cfrac%7BR_s%7D%7BL_s%7D+%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+i_%7B%5Calpha+d%7D+%5C%5Ci_%7B%5Cbeta+d%7D+%5Cend%7Barray%7D+%5Cright%5D+%2B%5Cfrac%7B1%7D%7BL_s%7D+%5Cleft%5B%5Cbegin%7Barray%7D%7Bccc%7D+u_%5Calpha%5C%5Cu_%5Cbeta+%5Cend%7Barray%7D+%5Cright%5D+-%5Cfrac%7B%5Crho%7D%7BL_s%7Dsgn%28e%29++++

在得到滑模观测器之后,后续待解决的问题是

<1>高频开关信号:使用低通滤波器

<2>相位延迟:引入角度补偿

<3>后续的改进:使用饱和函数等方法

评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值