|
|
|
| Online Generator Excitation Current Estimation Based on Model-Data Hybrid Drive |
| Liu Mulin1, Jiang Tong1, Xu Yongjin2, Wang Rui3 |
1. Skate Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources North China Electric Power University Beijing 102206 China;
2. GE (Shanghai) Power Technology Co. Ltd Beijing 100600 China;
3. Huainan Pingwei Electric Power Generating Co. Ltd Huainan 232089 China |
|
|
|
Abstract Monitoring excitation current is essential for synchronous generators, especially huge turbo generators. By observing trends in excitation current, plant personnel can promptly assess the generator's operational status. Moreover, various internal rotor faults can be detected online by comparing the theoretically estimated excitation current with the measured values. Traditional model-based methods for calculating excitation current are fast, but the accuracy is often unsatisfactory. On the other hand, finite element simulation-based methods offer precise calculations but are too slow for online monitoring systems. This paper proposes an online estimation method based on model-data hybrid drive, leveraging the strengths of generator models and machine learning techniques.
This method integrates the concept of dynamic Potier reactance into traditional online estimation methods. The reactance, modeled and estimated through machine learning, replaces the previously fixed Potier reactance value in theoretical excitation current calculations. The method has been validated through over 1 400 simulations in PSCAD and more than 4 400 experiments on a 644 MW turbo generator unit. During validation with PSCAD simulation data, the model-based method’s error ranges from 0% to -6% between theoretical and measured excitation currents, while the proposed method’s error is within ±1% with a minimal exceeding ±2%. Similarly, operational unit data show the error ranges from -3% to -5% for the model-based method, and the proposed method's error is predominantly within ±1.3% with rare deviations of about 2%. Additionally, the average time of this method is 113 milliseconds, meeting the real-time requirements of online systems. Finally, fault diagnosis experiments are conducted for 499 inter-turn short-circuit fault conditions of the same generator, confirming the feasibility and effectiveness of this method for online fault detection.
The following conclusions are drawn. (1) Compared to the model-based online excitation current estimation methods, the proposed method introduces dynamic Potier reactance, fully considering the saturation and distortion of the generator’s internal magnetic field. The accuracy of theoretical excitation current estimation under a wide range of operating conditions is enhanced. (2) The input to the machine learning prediction model only consists of basic measured data from the synchronous generator (including stator voltage, stator current, and power factor angle). In contrast, the finite element simulation-based methods require numerous parameters and substantial computational resources. This method offers excellent adaptability and computational speed. Additionally, the overall operation time is very short, meeting the real-time requirements of online systems. (3) The data containing magnetic field characteristics is input into the dynamic Potier reactance model to extract features. This method better reflects the relationship between excitation current variations and generator operating conditions machine learning models. Accurate theoretical excitation current estimation can also facilitate online detection of rotor inter-turn short-circuit faults without additional detection equipment. Furthermore, it can be combined with other data to analyze the generator’s operational status.
|
|
Received: 19 July 2024
|
|
|
|
Corresponding Authors:
刘沐霖 男,1999年生,博士研究生,研究方向为同步电机参数识别、故障检测及寿命预测技术。E-mail: ncepu_lml@163.com
|
|
|
|
[1] 沈小军, 李梧桐, 乔冠伦, 等. 同步发电机励磁系统模型参数离线辨识自动寻优方法[J]. 电工技术学报, 2018, 33(18): 4257-4266.
Shen Xiaojun, Li Wutong, Qiao Guanlun, et al.Automatic optimization method for model parameters off-line identification of synchronous generator excitation system[J]. Transactions of China Electro-technical Society, 2018, 33(18): 4257-4266.
[2] Cai Xiuhua, Cheng Ming, Zhu Sa, et al.Thermal modeling of flux-switching permanent-magnet machines considering anisotropic conductivity and thermal contact resistance[J]. IEEE Transactions on Industrial Electronics, 2016, 63(6): 3355-3365.
[3] 赵耀, 陆佳煜, 李东东, 等. 基于机电信号融合的电励磁双凸极电机绕组匝间短路故障诊断[J]. 电工技术学报, 2023, 38(1): 204-219.
Zhao Yao, Lu Jiayu, Li Dongdong, et al.A fault diagnosis strategy for winding inter-turn short-circuit fault in doubly salient electro-magnetic machine based on mechanical and electrical signal fusion[J]. Transactions of China Electrotechnical Society, 2023, 38(1): 204-219.
[4] Platero C A, Blázquez F, Frías P, et al.New on-line rotor ground fault location method for synchronous machines with static excitation[J]. IEEE Transactions on Energy Conversion, 2011, 26(2): 572-580.
[5] 梁郑秋, 郝亮亮, 周艳真, 等. 基于卷积神经网络的核电多相无刷励磁系统旋转整流器故障诊断[J]. 电工技术学报, 2023, 38(20): 5458-5472.
Liang Zhengqiu, Hao Liangliang, Zhou Yanzhen, et al.Fault diagnosis of rotating rectifier in nuclear multi-phase brushless excitation system based on con-volutional neural network[J]. Transactions of China Electrotechnical Society, 2023, 38(20): 5458-5472.
[6] Regan R H, Wakeley K.Rotor monitoring and protection for large generators[C]//1995 Seventh International Conference on Electrical Machines and Drives, Durham, UK, 1995: 203-207.
[7] Tian Pengfei, Platero C A, Gyftakis K N.On-line turn-to-turn protection method of the synchronous machines field winding[C]//2019 IEEE 12th Inter-national Symposium on Diagnostics for Electrical Machines, Power Electronics and Drives (SDEMPED), Toulouse, France, 2019: 69-74.
[8] Platero Gaona C A, Pardo Vicente M &, Rebollo L E, et al. System and method for protecting against faults between turns in excitation windings of synchronous machines with static excitation: US20200335965[P].2020-10-22.
[9] 罗玉东, 徐余法, 陈亚新, 等. 基于ExcelVBA驱动Maxwell的汽轮发电机励磁电流计算[J]. 电机与控制应用, 2016, 43(11): 61-66.
Luo Yudong, Xu Yufa, Chen Yaxin, et al.Calculation of excitation current for turbo generator based on excel VBA drive Maxwell[J]. Electric Machines & Control Application, 2016, 43(11): 61-66.
[10] IEEE Std115 IEEE Guide Test procedure for Syn-chronous Machines[S].
[11] 姚维康, 徐余法, 黄厚佳, 等. 基于有限元法的无刷交流励磁机空载特性分析及额定励磁电流计算[J]. 电机与控制应用, 2019, 46(3): 58-63.
Yao Weikang, Xu Yufa, Huang Houjia, et al.No-load characteristics analysis and rated excitation current calculation of brushless AC exciter based on finite element method[J]. Electric Machines & Control Application, 2019, 46(3): 58-63.
[12] 胡笳, 罗应立, 刘晓芳, 等. 汽轮发电机暂态分析中计及励磁电流集肤效应的时步有限元模型[J]. 中国电机工程学报, 2008, 28(30): 90-95.
Hu Jia, Luo Yingli, Liu Xiaofang, et al.An improved time-stepping finite element model considering skin effect of excitation current for turbo-generator transient analysis[J]. Proceedings of the CSEE, 2008, 28(30): 90-95.
[13] Kentli F, Birbir Y, Onat N.Examination of the stability limit on the synchronous machine depending on the excitation current wave shape[C]//IEMDC 2001. IEEE International Electric Machines and Drives Conference, Cambridge, MA, USA, 2002: 528-532.
[14] 张雅晖, 杨凯, 李天乐. 一种利用融合相关谱的异步电机故障诊断方法[J]. 电机与控制学报, 2021, 25(11): 1-7.
Zhang Yahui, Yang Kai, Li Tianle.Fault diagnosis method of asynchronous motors using fusion correlation spectrum[J]. Electric Machines and Control, 2021, 25(11): 1-7.
[15] 龚夕霞, 李焱鑫, 卢琴芬. 模块化永磁直线同步电机考虑制造公差的推力鲁棒性优化[J]. 电工技术学报, 2024, 39(2): 465-474, 513.
Gong Xixia, Li Yanxin, Lu Qinfen.Thrust robustness optimization of modular permanent magnet linear synchronous motor accounting for manufacture tolerance[J]. Transactions of China Electrotechnical Society, 2024, 39(2): 465-474, 513.
[16] 王恒泓, 王激尧, 徐炜,等. 基于领域对抗网络的永磁同步电机初始位置估计[J]. 电工技术学报, 2025, 40(2): 425-438.
Wang Henghong, Wang Jiyao, Xu Wei, et al.Initial position estimation of surface permanent magnet synchronous motor base on domain-adversarial neural networks[J]. Transactions of China Electrotechnical Society, 2025,40(2): 425-438.
[17] 张健, 张钦, 黄晓艳, 等. 基于加速退化数据和现场实测退化数据的电机绝缘剩余寿命预测模型[J]. 电工技术学报, 2023, 38(3): 599-609.
Zhang Jian, Zhang Qin, Huang Xiaoyan, et al.Motor insulation remaining useful life prediction method based on accelerating degradation data and field degradation data[J]. Transactions of China Electro-technical Society, 2023, 38(3): 599-609.
[18] Guillén C E G, de Porras Cosano A M, Tian Pengfei, et al. Synchronous machines field winding turn-to-turn fault severity estimation through machine learning regression algorithms[J]. IEEE Transactions on Energy Conversion, 2022, 37(3): 2227-2235.
[19] 苟智德, 孙力, 富立新, 等. 迭代法求取保梯电抗Xp[J]. 电机与控制学报, 2007, 11(2): 153-157.
Gou Zhide, Sun Li, Fu Lixin, et al.Determination of Potier reactance Xp from the iterative method[J]. Electric Machines and Control, 2007, 11(2): 153-157.
[20] IEC 60034-4 General Requirement for Rotating Electrical Machine. Part 104[S].
[21] 刘晓芳, 康锦萍, 罗应立, 等. 汽轮发电机在饱和与磁场畸变时负载励磁电流计算的新方法[J]. 电机与控制学报, 2010, 14(8): 7-12.
Liu Xiaofang, Kang Jinping, Luo Yingli, et al.A new method of calculating turbine generator load field currentunder iron saturation and magnetic field distortion[J]. Electric Machines and Control, 2010, 14(8): 7-12.
[22] 杨建华, 高军. 考虑磁饱和影响的同步发电机励磁电流计算[J]. 电力系统及其自动化学报, 2009, 21(2): 104-108.
Yang Jianhua, Gao Jun.Field current calculation of synchronous generators taking into account the magnetic saturation[J]. Proceedings of the Chinese Society of Universities for Electric Power System and Its Automation, 2009, 21(2): 104-108.
[23] 李峰, 王琦, 胡健雄, 等. 数据与知识联合驱动方法研究进展及其在电力系统中应用展望[J]. 中国电机工程学报, 2021, 41(13): 4377-4390.
Li Feng, Wang Qi, Hu Jianxiong, et al.Combined data-driven and knowledge-driven methodology research advances and its applied prospect in power systems[J]. Proceedings of the CSEE, 2021, 41(13): 4377-4390. |
|
|
|
2025, Vol. 40 
