电工技术学报
论文 |
分段式双三相永磁直线同步电机的无模型电流预测控制
周世炯1,2, 李耀华1,2, 史黎明1,2, 张明远1,2, 刘进海1,2
1.中国科学院电力电子与电力驱动重点实验室(中国科学院电工研究所) 北京 100190;
2.中国科学院大学 北京 100049
Model-Free Predictive Current Control of Segmented Dual Three-phase Permanent Magnet Linear Synchronous Motor
Zhou Shijiong1,2, Li Yaohua1,2, Shi Liming1,2, Zhang Mingyuan1,2, Liu Jinhai1,2
1. Key Laboratory of Power Electronics and Electric Drive Institute of Electrical Engineering Chinese Academy of Sciences Beijing 100190 China;
2. University of Chinese Academy of Sciences Beijing 100049 China
全文: PDF (1900 KB)  
输出: BibTeX | EndNote (RIS)      
摘要 

分段式结构的双三相永磁直线同步电机(PMLSM),动子穿入和穿出定子分段时,传统的无差拍预测电流控制(DPCC)易受反电势动态扰动的影响,无法实现精确的电流跟踪和推力控制。同时,该电机模型的参数失配会增大DPCC电流和推力控制的难度。针对以上问题,本文提出了一种基于无模型电流预测控制(MFPCC)的电流控制策略。首先,构建了分段式双三相PMLSM的超局部模型。其次,基于该超局部模型设计了预测电流控制器,并设计扰动观测器以补偿反电势动态扰动和参数失配的影响。最后,实验验证了本文所提MFPCC能够抑制分段式双三相PMLSM的反电势和参数失配的扰动,有效提升参数鲁棒性、电流跟踪性能和降低推力波动。

服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
周世炯
李耀华
史黎明
张明远
刘进海
关键词 双三相永磁直线同步电机超局部模型扰动观测器预测电流控制    
Abstract

During the segmented dual three-phase permanent magnet linear synchronous motor (PMLSM), the deadbeat predictive current control (DPCC) is vulnerable to the back electromotive force (EMF) fluctuation, which cannot achieve exact current tracking and thrust control when the mover enters and exits the stator segments. At the same time, DPCC depends on motor models, while the model of segmented dual three-phase (PMLSM) is complex. Besides, there are additional harmonic subspace inductors in its harmonic subspace, which is easy to aggravate the degradation of current control when parameters are mismatched. To solve these problems, a model-free predictive current control (MFPCC) method is proposed in this paper.
Firstly, the influence of parameter mismatch and back EMF fluctuation in the conventional DPCC is analyzed, and it is deduced that parameter mismatch causes the errors in the control voltage, and the additional inductance in the harmonic subspace aggravates the parameter variation. Secondly, this paper takes the parameter mismatch and back EMF fluctuation as the lumped disturbance. Then, a model-free predictive current controller is designed based on an ultra-local model, which is divided into a known linear term related to the current and an unknown part related to the parameter mismatch and back EMF. Finally, a disturbance observer is designed to observe the unknown part in this ultra-local model, and its stability is demonstrated.
Hardware-in-the-loop experiments prove that the proposed method is independent of motor parameters and has strong robustness. 1) When the speed control is open-loop and the flux linkage ψf changes in the DPCC control, the q-axis current cannot effectively track the given value, which thrust fluctuation is about 5.25%. When the Ld, Lq and Lxy of the controller increase to 200% of the original value, the q-axis current of the DPCC has obvious fluctuations, and the thrust fluctuation range is about 2%. The q-axis current of the MFPCC proposed in this paper can strictly track the given value and the thrust fluctuation range is only 1.25%, which proves that the MFPCC has good parameter robustness. 2) When the speed control is closed-loop and the speed reference suddenly increases to 12m/s within 0.75s, the q-axis current tracking performance of DPCC has overshoot and cannot effectively track the reference value iqref, due to the inductance and flux linkage parameters become 200% and 50% of the original respectively. However, the q-axis current tracking performance of MFPCC is smooth, and the tracking speed is fast. After parameter mismatch, the time of MFPCC following the slope is 0.36s, which is similar to 0.35s that of DPCC. The time of MFPCC following step signal is 0.53s, which is smaller than 0.6s that of DPCC speed tracking time. 3) The total harmonic distortion of DPCC is twice that of MFPCC.
The conclusions are as follows. (1) The parameter mismatch of the segmented dual three-phase PMLSM and the dynamic disturbance of back EMF will introduce disturbance to the current control, deteriorate the current tracking performance, and lead to thrust fluctuation. In addition, the mismatched harmonic subspace inductance parameter Lxy after vector space decomposition (VSD) further affects the current performance. (2) The MFPCC proposed in this paper uses an ultra-local model to simplify the system model, which is divided into a known linear part and an unknown part including back EMF and parameter mismatch. The disturbance observer accurately estimates the unknown part. (3) MFPCC has strong robustness to parameter mismatch. It can effectively suppress the influence of back EMF dynamic disturbance caused by mover motions. The proposed method has good tracking performance of q-axis current in the whole movement, small thrust and velocity fluctuation, and fast dynamic response.

Key wordsDual three-phase permanent magnet linear synchronous motor    ultra-local model    disturbance observer    predictive current control   
    
PACS: TM359.4  
基金资助:

磁悬浮车辆及轨道关键技术(E125320402)

通讯作者: 史黎明 男,1964年生,研究员,博士生导师,研究方向为特种电机和驱动控制、磁悬浮技术、电能无线传输技术。E-mail:limings@mail.iee.ac.cn   
作者简介: 周世炯 男,1995年生,博士研究生,研究方向大功率电力电子变换技术与直线电机驱动控制。E-mail:shijiongz@mail.iee.ac.cn
引用本文:   
周世炯, 李耀华, 史黎明, 张明远, 刘进海. 分段式双三相永磁直线同步电机的无模型电流预测控制[J]. 电工技术学报, 0, (): 20235420-20235420. Zhou Shijiong, Li Yaohua, Shi Liming, Zhang Mingyuan, Liu Jinhai. Model-Free Predictive Current Control of Segmented Dual Three-phase Permanent Magnet Linear Synchronous Motor. Transactions of China Electrotechnical Society, 0, (): 20235420-20235420.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.222131          https://dgjsxb.ces-transaction.com/CN/Y0/V/I/20235420