电工技术学报  2023, Vol. 38 Issue (14): 3708-3722    DOI: 10.19595/j.cnki.1000-6753.tces.221611
“轻量化高效高可靠新能源电机与控制系统”专题(特约主编:吴立建 研究员) |
双三相永磁同步电机低共模电压模型预测电流控制
孙全增, 张志锋
沈阳工业大学电气工程学院 沈阳 110870
Low Common-Mode Voltage Model Predictive Current Control for Dual Three-Phase Permanent Magnet Synchronous Motor
Sun QuanZeng, Zhang ZhiFeng
School of Electrical Engineering Shenyang University of Technology Shenyang 110870 China
全文: PDF (3613 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 针对双三相永磁同步电机模型预测控制中存在共模电压较大、xy平面上谐波电流较大和控制器计算负担严重的问题,该文提出一种低共模电压的模型预测电流控制方法。首先,将虚拟电压矢量作为输入控制集,通过具有小共模电压的虚拟电压矢量对xy平面进行控制。然后,在每个采样周期中,通过计算参考电压矢量位置来减少候选电压矢量的数量。再通过无权重系数的价值函数对候选电压矢量进行评估,选择出最优电压矢量。所提方法有效地减少了xy平面上的谐波电流,消除了价值函数中的权重系数;减少了控制器的计算负担和双三相永磁同步电机驱动系统中固有的共模电压。最后,通过与两种现存的传统模型预测控制方法进行对比实验,证明了所提方法的有效性。
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
孙全增
张志锋
关键词 模型预测电流控制双三相永磁同步电机共模电压虚拟电压矢量权重系数    
Abstract:Compared with three-phase motors, dual three-phase permanent magnet synchronous motors (DT-PMSM) have better fault tolerance, higher efficiency, and higher torque density. Therefore, it is widely used in high-reliability fields, such as electric vehicles, ships, and aerospace.
The model predictive control (MPC) method has the characteristics of fast dynamic response and simple structure, compared with field-oriented control and direct torque control methods. Therefore, it has been intensively studied in recent years. However, when the traditional MPC is applied to the DT-PMSM, larger harmonic currents and severe computational burden are the existing issues. In addition, the common-mode voltage generated by the driving system can shorten the life of the motor and produce electromagnetic interference.
Therefore, a model predictive current control method with low common-mode voltage is proposed. The proposed method consists of three steps. First, the voltage vectors with small common-mode voltage are analyzed. The virtual voltage vectors are synthesized as the input control set to suppress the common-mode voltage and reduce the harmonic current. Then, to avoid the ergodic prediction voltage vector, the position of the reference voltage vector is obtained by the deadbeat current control method to reduce the candidate voltage vector. Finally, the cost function of the unweighting factor is designed. The candidate voltage vectors are evaluated, and the optimal voltage vector is selected. It solves the issue that the weighting factor is difficult to design without theoretical basis.
An experiment is designed to compare the traditional model predictive current control method (MPCC-T), the traditional model predictive current control method with reduced common-mode voltage (RCMV-36) and the proposed MPCC-VVV-RCMV method. Firstly, the phase currents of the three methods are analyzed by the fast Fourier transform algorithm. The total harmonic current distortion rates of MPCC-T, RCMV-36, and the proposed method are 6.63 %, 7.48 %, and 3.68 %, respectively. The fifth harmonic distortion rates of the three methods are 5.71 %, 6.57 %, and 1.06 %, which proves that the proposed method provides higher-quality phase currents and reduces the harmonic current in the x-y plane. Secondly, the common-mode voltage under the three methods is tested. The negative peak value of the common-mode voltage under the MPCC-T method is (-1/2)Udc, and the positive peak value is (1/6)Udc. However, with RCMV-36 and the proposed method, the negative peak value of the common-mode voltage is (-1/6)Udc, and the positive peak value is (1/6)Udc. Which indicates that the proposed method can effectively reduce the common-mode voltage. In addition, considering that the motor parameters will change during operation, the influence of the non-ideal prediction model on the control performance is also studied. The results show that the output current waveform keeps well despite the change of resistance and inductance. Therefore, when the motor parameters change within a specific range, the proposed method is robust. Finally, the total execution time of the three methods in the controller is compared. The total execution time of MPCC-T, RCMV-36, and the proposed method in the controller is 33 μs, 95 μs, and 16 μs, respectively, which shows that the proposed method can reduce the computational burden on the controller.
The proposed method is compared with two existing traditional methods. The following conclusions are obtained: (1) The proposed MPCC-VVV-RCMV method can effectively reduce the common-mode voltage and the harmonic current. (2) The proposed MPCC-VVV-RCMV method eliminates the weight factors in the cost function, avoids the tedious adjustment task, and reduces the computational burden. (3) The shortcoming of the proposed MPCC-VVV-RCMV method is the high average switching frequency.
Key wordsModel predictive current control    dual three-phase permanent magnet synchronous motor    common-mode voltage    virtual voltage vector    weighting factor   
收稿日期: 2022-08-22     
PACS: TM351  
基金资助:国家自然科学基金资助项目(61603263)
通讯作者: 张志锋 男,1981年生,教授,博士生导师,研究方向为现代电机控制技术、多相电机控制、电动汽车电驱动控制系统等。E-mail: zzf_sut@126.com   
作者简介: 孙全增 男,1997年生,博士研究生,研究方向多相电机控制。E-mail: 1074287890@qq.com
引用本文:   
孙全增, 张志锋. 双三相永磁同步电机低共模电压模型预测电流控制[J]. 电工技术学报, 2023, 38(14): 3708-3722. Sun QuanZeng, Zhang ZhiFeng. Low Common-Mode Voltage Model Predictive Current Control for Dual Three-Phase Permanent Magnet Synchronous Motor. Transactions of China Electrotechnical Society, 2023, 38(14): 3708-3722.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.221611          https://dgjsxb.ces-transaction.com/CN/Y2023/V38/I14/3708