异步电机无速度传感器模型预测控制

发布时间：2018-02-14 17:21

本文关键词： 异步电机模型预测控制无速度传感器全阶观测器　出处：《中国电机工程学报》2014年15期 　论文类型：期刊论文

【摘要】：模型预测控制(model predictive control,MPC)是近年来在交流电机控制领域得到广泛关注的一种优化控制方法。对于两电平逆变器驱动的异步电机系统,传统MPC需要对定子磁链和定子电流进行7次预测得到使目标函数最小的最佳电压矢量,计算量较大,不利于在线实施。通过解析推导定子电流和定子磁链之间的关系,该文提出一种改进的MPC,只需对定子磁链进行预测,省却了计算相对复杂的定子电流预测,使算法复杂性和计算量得到显著降低。另外在MPC的基础上,为了提高系统的可靠性和降低硬件成本,采用速度自适应全阶观测器实现了无速度传感器运行。文中考虑了数字控制延迟对MPC的影响,并提出了补偿方法,采用预励磁的方法获得较大的启动转矩并减小启动电流。最后在两电平逆变器异步电机平台上进行仿真和实验,结果表明,该文提出的无速度传感器模型预测控制,在较宽的速度范围内都具有良好的动静态性能。
[Abstract]:Model predictive control (model predictive control, MPC) in recent years is a kind of optimal control method of widespread concern in the field of AC motor control system for asynchronous motor. Two level inverter drive, the traditional MPC of stator flux and stator current of 7 times forecast obtains the minimum target function of optimal voltage vector, the large amount of calculation that is not conducive to the implementation of online. The relation between the derivation of stator current and stator flux, this paper proposes an improved MPC, only need to predict the stator flux and avoids the complicated stator current prediction, the algorithm complexity and computation are significantly reduced. In addition on the basis of MPC, in order to improve the reliability of the system and reduce the cost of hardware, realized speed sensorless speed using adaptive full order observer. The digital control delay of MPC. Ring, and put forward the compensation method, using the method of pre excitation obtaining high starting torque and reduce the starting current. Finally, simulation and experimental platform in two level inverter induction motor. The results show that the proposed sensorless model predictive control in a wide range of speed has good dynamic and static performance.

【作者单位】：北京市电力电子与电气传动工程研究中心(北方工业大学);
【基金】：国家自然科学基金(51207003,51347004) 北京市科技新星计划(xx2013001) 教育部留学回国人员科研启动基金重庆市自然科学基金(cstc2012jjB107) 军用特种电源军队重点实验室开放课题(MSPS2012-03,MSPS2013-02)的资助
【分类号】：TM343

【参考文献】