[1]徐宇柘,曹彦萍,钟锐,等.基于LSSVM的开关磁阻电机转子位置估算[J].东南大学学报(自然科学版),2013,43(2):307-311.[doi:10.3969/j.issn.1001-0505.2013.02.015]
 Xu Yuzhe,Cao Yanping,Zhong Rui,et al.Rotor position estimation of switched reluctance motor based on LSSVM[J].Journal of Southeast University (Natural Science Edition),2013,43(2):307-311.[doi:10.3969/j.issn.1001-0505.2013.02.015]
点击复制

基于LSSVM的开关磁阻电机转子位置估算()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
43
期数:
2013年第2期
页码:
307-311
栏目:
电气工程
出版日期:
2013-03-20

文章信息/Info

Title:
Rotor position estimation of switched reluctance motor based on LSSVM
作者:
徐宇柘1曹彦萍1钟锐1屈严2彭富林2
1东南大学国家专用集成电路系统工程技术研究中心, 南京210096; 2东南大学电子科学与工程学院, 南京210096
Author(s):
Xu Yuzhe1 Cao Yanping1 Zhong Rui1 Qu Yan2 Peng Fulin2
1 National ASIC System Engineering Research Center, Southeast University, Nanjing 210096, China
2 School of Electronic Science and Engineering, Southeast University, Nanjing 210096, China
关键词:
开关磁阻电机 最小二乘支持向量机 位置估算
Keywords:
switched reluctance motor least squares support vector machine position estimator
分类号:
TM352;TM307
DOI:
10.3969/j.issn.1001-0505.2013.02.015
摘要:
针对开关磁阻电机(SRM)中位置传感器在恶劣环境下存在脱落、故障的隐患,提出了基于最小二乘支持向量机(LSSVM)的SRM转子位置估算方法.利用改进的电压脉冲法获取电机磁特性数据,将交叉验证法与凑试法相结合以选取合适参数,并对LSSVM进行训练.与改进的反向传播神经网络(BPNN)的位置估算相比,LSSVM的训练时间缩短了72%,训练样本的均方误差减小了7.8%,测试样本均方误差减小了19.6%,且测试样本误差较训练样本减小64.5%.在Simulink中利用S函数建立仿真系统,结果表明,在高速、低速、负载突变的情况下,LSSVM位置估算模块的平均误差在0.1°以内,说明其训练速度快,泛化能力强,精度高,鲁棒性好.
Abstract:
To avoid the problem that a position sensor easily falls off or becomes breakdown when a switched reluctance motor(SRM)operates in adverse environment, a method based on least squares support vector machine(LSSVM)for the rotor position estimator of the SRM is proposed. LSSVM is trained by using the measured prototype characteristics which are obtained from improved DC pulse method. The optimal parameters are chosen by combining the cross-validation method and the trying method. Compared with the position estimator by using improved back-propagation neural network(BPNN), the training time of LSSVM estimator is reduced by 72%; the mean squared error(MSE)of the training sample and that of the test sample decrease by 7.8% and 19.6%, respectively. Moreover the error of the test sample decrease 64.5% compared with the training sample. The simulation system of LSSVM estimator is built in Simulink by using S function. The experimental results show that the mean error of position estimator using LSSVM is less than 0.1° when SRM operates under high speed, low speed and load sudden change condition, indicating the fast training, high generalization performance, improved accuracy and good robustness of LSSVM estimator.

参考文献/References:

[1] Soares F, Branco P J C. Simulation of a 6/4 switched reluctance motor based on Matlab/Simulink environment [J]. IEEE Transactions on Aerospace and Electronic, 2001, 37(3): 989-1009.
[2] Mese E, Torrey D A. An approach for sensorless position estimation for switched reluctance motors using artificial neural networks [J]. IEEE Transaction on Power Electronics, 2002, 17(1): 66-75.
[3] Fahimi B, Emadi A, Sepe R B. Four-quadrant position sensorless control in SRM drives over the entire speed range [J]. IEEE Transactions on Power Electronics, 2005, 20(1): 154-163.
[4] Krishnan R. Switched reluctance motor drives: modeling, simulation, analysis, design, and applications [M]. New York: CRC Press, 2001: 1-25.
[5] Ehssai M, Husaln I, Kulkami A B. Elimination of discrete position sensor sad current in switched reluctance motor drives [J]. IEEE Transactions on Industry Applications, 1992, 28(1): 128-135.
[6] Afjei E, Cheshmehbaygi H M, Nouri H. Detecting the rotor position by employing pulse injection technique and digital pulse width modulation decoder in switched reluctance motor [C]//IEEE International Electric Machines & Drives Conference. Antalya, Turkey, 2007: 40-44.
[7] Khalil A, Underwood S, Husain I, et al. Four-quadrant pulse injection and sliding-mode-observer-based sensorless operation of a switched reluctance machine over entire speed range including zero speed [J]. IEEE Transactions on Industry Applications, 2007, 43(3): 714-723.
[8] Cheok A D, Wang Z. Fuzzy logic rotor position estimation based switched reluctance motor DSP drive with accuracy enhancement [J]. IEEE Transactions on Power Electronics, 2005, 20(4): 908-921.
[9] Mese E. A rotor position estimator for switched reluctance motors using CMAC [J]. Energy Conversion and Management, 2003, 44(4): 1229-1245.
[10] Cristianini Nello, Shawe-Taylor John. 支持向量机导论[M]. 李国正, 等译. 北京: 电子工业出版社, 2004.
[11] He Z, Xia C, Zhou Y, et al. Rotor position estimation for switched reluctance motor using support vector machine[C]//IEEE International Conference on Control and Automation. Guangzhou, 2007: 1683-1687.
[12] Qu B, Song J, Jia H, et al. Nonlinear modeling of switched reluctance machine [C]//International Conference on Electrical Machines and Systems(ICEMS). Beijing, China, 2011:1-4.
[13] 宋建成,郑建斌,曲兵妮,等.开关磁阻电机的最小二乘支持向量机建模与仿真[J].电机与控制学报, 2010, 14(5): 32-36.
  Song Jiancheng, Zheng Jianbin, Qu Bingni, et al. Modeling and simulation for switched reluctance motor based on least squares support vector machine [J]. Electric Machine and Control, 2010, 14(5): 32-36.(in Chinese)
[14] Carstensen C E, Fuengwarodsakul N H, de Doncker R W. Flux linkage determination for correct modeling of switched reluctance machines-dynamic measurement versus static computation [C]//IEEE International Electric Machines & Drives Conference. Antalya, Turkey, 2007: 1317-1323.
[15] Zhong R, Wang Y B, Xu Y Z. Position sensorless control of switched reluctance motors based on improved neural network [J]. IET Electric Power Applications, 2012, 6(2): 111-121.

相似文献/References:

[1]仲兆平,王春华,李睿,等.喷动流化床射流穿透度智能拟合[J].东南大学学报(自然科学版),2010,40(1):139.[doi:10.3969/j.issn.1001-0505.2010.01.026]
 Zhong Zhaoping,Wang Chunhua,Li Rui,et al.Intelligent fitting of jet penetration depth in spout-fluid bed[J].Journal of Southeast University (Natural Science Edition),2010,40(2):139.[doi:10.3969/j.issn.1001-0505.2010.01.026]
[2]刘闯,朱学忠,刘迪吉.开关磁阻电机的非线性电感参数法仿真分析[J].东南大学学报(自然科学版),2001,31(4):106.[doi:10.3969/j.issn.1001-0505.2001.04.025]
 Liu Chuang,Zhu Xuezhong,Liu Diji.Simulation Analysis of Non-Linear Inductance Parameters Method for SRM[J].Journal of Southeast University (Natural Science Edition),2001,31(2):106.[doi:10.3969/j.issn.1001-0505.2001.04.025]
[3]蒋全,周鹗.开关磁阻电机静特性的有限元分析[J].东南大学学报(自然科学版),1989,19(6):93.[doi:10.3969/j.issn.1001-0505.1989.06.013]
 Jiang Quan Zhou E(Department of Electrical Engineering).Finite-Element Analysis of Static Characteristics of Switched Reluctance Motors[J].Journal of Southeast University (Natural Science Edition),1989,19(2):93.[doi:10.3969/j.issn.1001-0505.1989.06.013]
[4]乔宗良,周建新,周卫庆,等.基于预数值计算的除雾器叶片结构优化设计[J].东南大学学报(自然科学版),2013,43(1):76.[doi:10.3969/j.issn.1001-0505.2013.01.015]
 Qiao Zongliang,Zhou Jianxin,Zhou Weiqing,et al.Method for optimum design of wave-plate demister based on numerical computation[J].Journal of Southeast University (Natural Science Edition),2013,43(2):76.[doi:10.3969/j.issn.1001-0505.2013.01.015]
[5]马欢,司风琪,李岚,等.间接空冷系统变工况快速计算模型[J].东南大学学报(自然科学版),2015,45(1):79.[doi:10.3969/j.issn.1001-0505.2015.01.015]
 Ma Huan,Si Fengqi,Li Lan,et al.Fast calculation model of indirect air-cooled system under off-design condition[J].Journal of Southeast University (Natural Science Edition),2015,45(2):79.[doi:10.3969/j.issn.1001-0505.2015.01.015]

备注/Memo

备注/Memo:
作者简介: 徐宇柘(1982—),男,博士生;钟锐(联系人),男,博士,副研究员,ray@seu.edu.cn.
基金项目: 国家自然科学基金资助项目(51277026, 61201034)、江苏省自然科学基金资助项目(BK2010167).
引文格式: 徐宇柘,曹彦萍,钟锐,等.基于LSSVM的开关磁阻电机转子位置估算[J].东南大学学报:自然科学版,2013,43(2):307-311. [doi:10.3969/j.issn.1001-0505.2013.02.015]
更新日期/Last Update: 2013-03-20