基于MRAS的无刷双馈电机无位置传感器鲁棒控制  

Position sensorless robust control of brushless doubly fed generator based on the MRAS observer

作　　者：陈晗 张永昌 蒋涛 杨长山 何俊辉 CHEN Han;ZHANG Yongchang;JIANG Tao;YANG Changshan;HE Junhui(School of Electrical and Electronic Engineering,North China Electric Power University,Beijing 102206,China;Shenzhen INVT Electric Co.,Ltd.,Shenzhen 518106,China)

机构地区：[1]华北电力大学电气与电子工程学院,北京102206 [2]深圳市英威腾电气股份有限公司,深圳518106

出　　处：《电机与控制学报》2024年第10期24-33,共10页Electric Machines and Control

基　　金：国家自然科学基金(52077002)。

摘　　要：为取消级联无刷双馈电机系统中成本高、布线难且脆弱的位置传感器,在模型参考自适应系统(MRAS)方法的基础上,提出了一种改进的基于控制绕组电流的无位置控制方法。通过将参考模型和可调模型转换到同步坐标系下,减少了磁链观测器的使用,并将参考模型和可调模型的误差进行线性化,在简化控制器参数设计的同时提高了动态性能。将其与基于扩张状态观测器的鲁棒预测控制相结合,实现了无位置传感器高性能闭环控制。实验结果表明,与传统的转速观测器和模型参考自适应方法相比,所提方法在不同转速下稳态脉动和电流谐波更小;动态阶跃响应速度略慢但估计的转速位置信息更精确;在参数变化时能快速准确地计算出位置和转速,表现出参数强鲁棒性,具有实用性。To remove the fragile position sensor with high cost,and complex wiring in the cascaded brushless doubly-fed generator(CBDFG)system,an improved position sensorless method based on the control winding current was proposed,which employed the principle of model reference adaptive system(MRAS).The method converts the reference model and the adjustable model to the synchronous coordinate system to reduce the use of the flux observer,and linearizes the error between them since simplifying the parameter design also improving the dynamic performance.It is combined with robust predictive control based on extended state observer to realize sensorless closed-loop control of high performance.The experimental results demonstrate that,compared with the rotor speed observer method or the conventional MRAS method,the improved MRAS method has smaller steady-state pulsation and current THD at different speed.While the dynamic step response speed is marginally slower,the speed and position estimation are more precise.Furthermore,when the parameters change,the position and speed information can swiftly and accurately be calculated,which exemplifies a high level of parameter robustness and practicality.

关 键 词：级联无刷双馈电机 无位置传感器 预测控制 模型参考自适应 参数鲁棒性 扩张状态观测器 

分 类 号：TM343[电气工程—电机]