改进型MRAS无速度传感器的无轴承异步电机矢量控制
Vector control of speed sensorless bearingless induction motors using improved MRAS

无轴承异步电机（Bearingless Induction Motor, BIM）的转子磁链电压模型中含有纯积分环节，其积分初值和累计误差会影响磁链观测精度，进而使转速估计产生严重失真。为了实现BIM无速度传感器运行，本文借鉴模型参考自适应法（model reference adaptive system, MRAS）基本结构，将改进二阶广义积分器与锁频环结合以代替原有纯积分器，提出了一种新的基于MRAS的BIM无速度传感器控制方法，构建了BIM转子磁链定向无速度传感器矢量控制系统。并且，基于Matlab/Simulink的仿真验证和基于dSPACE的实验结果表明:与传统电压模型观测方法相比，所提出的转子磁链电压模型有效避免了纯积分环节带来的直流偏移和积分初值影响，有着更好的观测效果。同时，基于无轴承异步电机转子磁链定向无速度传感器矢量控制系统，电机能稳定悬浮运行，估算转速和实测转速具有很好的一致性。
Because of the pure integrator of rotor flux voltage model in bearingless Induction Motor (BIM), the initial value and cu-mulative error of integral affects the observation precision of the magnetic linkage, and makes the speed estimation serious-ly distorted. In order to realize the speed sensorless operation of BIM, this paper learns from the basic structure of model reference adaptive system (MRAS), and adopts second-order generalized integrator with frequency locked loop to replace the original pure integrator. That is, a novel BIM speed sensorless control method based on MRAS is proposed, and a vec-tor control system of BIM speed sensorless sensor with rotor flux oriented is constructed. Furthermore, the simulation re-sults of Matlab / Simulink system and the experimental results of dSPACE show that the proposed rotor flux linkage volt-age model effectively avoids the influence of DC offset and integral value caused by pure integral comparing with the tradi-tional voltage model observation method. Moreover, the motor can be stably suspended on BIM speed sensorless control system, and the estimated speed is highly consistent with the measured speed