广域测量环境下发电机组模型辨识研究
|
摘要
电力系统数学模型是电力系统规划、运行及控制的基础。当前电网的快速发展对模型参数的准确性要求越来越高。WAMS的出现,将电力系统运行中数据实时采集并上传到调度中心,其中各种扰动下的暂态数据可用来对模型参数的辨识。论文首先对WAMS及其在系统辨识中的应用做了介绍,然后,针对发电机、励磁系统、原动机及其调速系统分别建立状态空间模型,并详细分析了用非线性最小二乘法辨识的可行性,最后结合电科院CEPRI-36节点算例仿真,模拟广域测量环境下采集的数据,对同步发电机组模型参数辨识进行研究。取得的主要成果如下：
在前人研究的基础上,着重分析发电机参数在不同扰动类型下的辨识结果,包括励磁参考电压阶跃扰动、调速系统功率给定扰动和负荷阶跃扰动。其中前两种属于人为扰动,而负荷扰动为电力系统自然扰动。论文用非线性最小二乘法对发电机五阶模型参数进行辨识,结果显示,在以上人为扰动和系统自然扰动情况下都能成功辨识。
传统方法对励磁系统和调速系统参数的辨识,一般并不涉及对调差系数辨识。论文尝试将调差系数与模型其他参数一起进行辨识。结果显示用非线性最小二乘法也能够在上述三种扰动情况下对包括调差系数在内的参数进行辨识,这将会使在电力调度中心端确定励磁系统调差率和调速系统调差特性成为可能,对电网运行具有重要意义。
针对WAMS无选择地采集大量数据的特点,大部分时间的采样数据波动不大而难以进行辨识,论文尝试用某些电气量指标衡量其波动情况,确定动态数据具有足够的扰动强度能够用于辨识,论文从采样数据的采样周期、总时间跨度和扰动电气量的选择三个方面进行分析,分别定量给出了辨识发电机、励磁系统、原动机和调速系统所用的动态数据满足的扰动强度指标。
Power system mathematical model is the basis of power system planning, operation and control. With the rapid development of the current grid, we need the model parameters to be more and more accurate and precise. The advent of WAMS real time data acquisition in power system operation can be real-time collection and upload it to the dispatching center, including the transient data in a variety of disturbances, these data may be used for identification of model parameters. The paper first introduces the WAMS and its application in system identification. Then, for generator model, excitation system, prime mover and governor system, the state space model is established respectively and analysis in detail the feasibility of identification using the nonlinear least square method. Finally, combined with the CEPRI-36buses system example, data collected in the simulation of wide area measurement environment to study the power system model parameter identification, the main results are summarized as follows:
Based on the previous results, the identification results of the generating under different disturbance types are emphatically analyzed. The disturbances include the excitation reference voltage step disturbance, the power given disturbance and load step disturbance. The former two disturbances are artificial imposed disturbances, and the last one is system natural disturbance. The fifth-order generator model parameters are identified by nonlinear least squares method. The results show that the generator model parameters can be identified successfully under these three kinds of disturbances.
The traditional methods for excitation system and governor identification system are usually not related to the identification of the difference coefficients. The paper attempts to identify the difference coefficients with other parameters. The results show that the parameters can also be identified using the nonlinear least squares method under the three disturbance parameter identification. Then, it is possible that the voltage adjustment and governor system difference coefficient can be got at the power dispatch center sides, which have great significance to the power grid.
As the characteristics of the power system WAMS have acquired large amounts of data which has no choice, most of the data fluctuations are small and difficult to identify, the paper attempts to use indicators to measure the fluctuations of some electrical quantities to determine the dynamic data using to identify with sufficient fluctuations in intensity. It is analyzed from three aspects of the data sampling period, the total time and the choice of the electrical quantities, for generator model excitation system, prime mover and governor system, the dynamic data using identifying model parameters have to meet the disturbance intensity index respectively and quantitatively.
在前人研究的基础上,着重分析发电机参数在不同扰动类型下的辨识结果,包括励磁参考电压阶跃扰动、调速系统功率给定扰动和负荷阶跃扰动。其中前两种属于人为扰动,而负荷扰动为电力系统自然扰动。论文用非线性最小二乘法对发电机五阶模型参数进行辨识,结果显示,在以上人为扰动和系统自然扰动情况下都能成功辨识。
传统方法对励磁系统和调速系统参数的辨识,一般并不涉及对调差系数辨识。论文尝试将调差系数与模型其他参数一起进行辨识。结果显示用非线性最小二乘法也能够在上述三种扰动情况下对包括调差系数在内的参数进行辨识,这将会使在电力调度中心端确定励磁系统调差率和调速系统调差特性成为可能,对电网运行具有重要意义。
针对WAMS无选择地采集大量数据的特点,大部分时间的采样数据波动不大而难以进行辨识,论文尝试用某些电气量指标衡量其波动情况,确定动态数据具有足够的扰动强度能够用于辨识,论文从采样数据的采样周期、总时间跨度和扰动电气量的选择三个方面进行分析,分别定量给出了辨识发电机、励磁系统、原动机和调速系统所用的动态数据满足的扰动强度指标。
Power system mathematical model is the basis of power system planning, operation and control. With the rapid development of the current grid, we need the model parameters to be more and more accurate and precise. The advent of WAMS real time data acquisition in power system operation can be real-time collection and upload it to the dispatching center, including the transient data in a variety of disturbances, these data may be used for identification of model parameters. The paper first introduces the WAMS and its application in system identification. Then, for generator model, excitation system, prime mover and governor system, the state space model is established respectively and analysis in detail the feasibility of identification using the nonlinear least square method. Finally, combined with the CEPRI-36buses system example, data collected in the simulation of wide area measurement environment to study the power system model parameter identification, the main results are summarized as follows:
Based on the previous results, the identification results of the generating under different disturbance types are emphatically analyzed. The disturbances include the excitation reference voltage step disturbance, the power given disturbance and load step disturbance. The former two disturbances are artificial imposed disturbances, and the last one is system natural disturbance. The fifth-order generator model parameters are identified by nonlinear least squares method. The results show that the generator model parameters can be identified successfully under these three kinds of disturbances.
The traditional methods for excitation system and governor identification system are usually not related to the identification of the difference coefficients. The paper attempts to identify the difference coefficients with other parameters. The results show that the parameters can also be identified using the nonlinear least squares method under the three disturbance parameter identification. Then, it is possible that the voltage adjustment and governor system difference coefficient can be got at the power dispatch center sides, which have great significance to the power grid.
As the characteristics of the power system WAMS have acquired large amounts of data which has no choice, most of the data fluctuations are small and difficult to identify, the paper attempts to use indicators to measure the fluctuations of some electrical quantities to determine the dynamic data using to identify with sufficient fluctuations in intensity. It is analyzed from three aspects of the data sampling period, the total time and the choice of the electrical quantities, for generator model excitation system, prime mover and governor system, the dynamic data using identifying model parameters have to meet the disturbance intensity index respectively and quantitatively.
引文
[1]沈善德.电力系统辨识[M].北京：清华大学出版社,1993
[2]张爽.基于PMU的广域电网相量测量系统概述[J].广东电力电力,2007,20(12)：26-29
[3]鞠平,代飞,金宇清等.电力系统广域测量技术[M].机械工业出版社,2008
[4]李丹,韩福坤,郭子明等.华北电网广域实时动态监测系统.电网技术[J],2004,28(23)：52-56
[5]徐兴伟,穆刚,王文等.东北电网广域动态测量系统[J].电网技术,2004,30(17)：70-73
[6]李颖,贺仁睦,徐衍会等.广东电网基于PMU的负荷模型参数辨识研究[J].南方电网技术,2009,3(1)：16-19
[7]李丹,韩福坤,郭子明等.华北电网广域实时动态监测系统[J].电网技术,2004,28(23)：52-56
[8]Hope G. S. Measurement of transfer function of Power system components under Operation conditions [J]. IEEE Transactions on Power Apparatus and Systems,1977,59(6):96-99
[9]方崇智,萧德云.过程辨识[M].北京：清华大学出版社,1988
[10]米增强,陈志忠,南志远.同步发电机动态参数辨识[J].中国电机工程学报,1998,18(2)：100-105
[11]王良,沈善德,朱守真等.基于EE模型的励磁系统参数辨识时域法[J].电力系统自动化,2002,26(8)：25-28
[12]曹浩军,张承学,单勇.基于连续时间方法的励磁系统参数辨识[J].电力系统自动化,2004,28(17)：49-54
[13]戴义平,邓仁纲,刘炯等.基于遗传算法的汽轮机数字电液调速系统的参数辨识研究[J].中国电机工程学报,2002,22(7)：101-104.
[14]许津津,马进,唐永红等.基于改进DE算法的负荷建模参数辨识[J].电力系统保护与控制,2009,37(24)：37-45
[15]徐丽娜.神经网络控制[M].哈尔滨：哈尔滨工业大学出版社,1999
[16]米增强,张宁,刘国平等.基于WAMS的同步发电机组动态等值研究[J].华北电力大学学报,2008,35(2)：30-34
[17]Kang Lin, Elias Kyriakides, Gerald T. Heydt. Experience with Synchronous Generator Parameter Identification using a Kalman Filter[J]. IEEE Transaction on Power Systems, 2010.4
[18]E. Ghahremani, M. Karrari, O. P. Malik. Synchronous generator third-order model parameter estimation using online experimental data[J]. IEEE, The Institution of Engineering and Technology.708-718,2008
[19]Robert Eriksson, Lennart. Royal Institute of Technology, Sweden System Identification Techniques for Obtaining Linear Models of Large Power Systems with Controllable Devices from Noisy Measurements[J]. IEEE, Symposium-Bulk Power System Dynamics and Control,2010,4
[20]金宇清,岳陈熙,甄威等.基于功角测量的同步发电机参数辨识频域法[J].电力系统自动化,2007,31(4)：7-11
[21]王正风,葛斐,戴长春.基于PMU的电力系统建模及参数辨识[J].继电器,2007.35：378-382
[22]郭小龙,胥国毅,薛安成等.基于PMU的励磁系统参数辨识研究[C].中国高等学校电力系统及其自动化专业第二十四届学术年会论文集,2008
[23]陈金猛.基于PMU的同步发电机与线路参数在线辨识研究[D].华北电力大学硕士论文,2008.1
[24]A. Phadke, M. Ibrahim, T. Hlibka. Fundamental basis for distance relaying with symmetrical components[J]. Power Apparatus and Systems, IEEE Transactions on,1977,96(2):635-646
[25]A. Phadke, J. Thorp, M. Adamiak. A New Measurement Technique for Tracking Voltage Phasors, Local System Frequency, and Rate ofChange of Frequency[J]. Power Apparatus and Systems, IEEE Transactions on,1983(5):1025-1038
[26]Z. Huang, B. Kasztenny, V. Madani et al. Performance Evaluation of Phasor Measurement Systems[C]. Power and Energy Society General Meeting,2008 IEEE 2008, Pittsburgh, PA 1-7
[27]韩松,何利铨,邱国跃WAMS研究、建设与应用的新进展[J].电测与仪表,2011.48(04)：1-8
[28]M. Lixia, C. Muscas, S. Sulis. On the Accuracy Specifications of Phasor Measurement Units[C]. Instrumentation and Measurement Technology Conference(I2MTC),2010 IEEE May 2010, Austin, TX 1435-1440
[29]E. Johansson, K. Uhlen, A. Leirbukt et al. Coordinating Power Oscillation Damping Control Using Wide Area Measurements[C]. Power Systems Conference and Exposition, 2009. PSCE 09. IEEE/PES 2009, Seattle, WA:1-8
[30]M. Hojo, K. Abe, Y. Mitani et al, Real-time Power System Monitoring at Demand Sides by Campus Wide Area Measurement System[C]. Industrial Electronics,2009, IECON 09.35th Annual Conference of IEEE 2009, Porto 3609-3614
[31]S. Boroczky, E. Gentle, Real-time Transient Security Assessment in Australia at EMMCO [C]. Power Systems Conference and Exposition,2009. PSCE 09. IEEE/PES 2009, Seattle, WA 1-6
[32]陈实,许勇,王正风.电网实时动态监测技术及应用[M].北京：中国水利水电出版社,2010
[33]卢志刚,郝玉山,康庆平等.电力系统实时相角监控系统研究[J].电力系统自动化,1997,21(9)：17-19
[34]陈凯,胡志坚,王卉等.介损角的非同步采样算法及其应用[J].电网技术,2004,28(18)：58-61
[35]赵凤山,王旭东.用单片机测量同步发电机功角的方法[J].黑龙江电力技术,1995,17(5)：279-281
[36]李大路,李蕊,孙元章.混合量测下基于UKF的电力系统动态状态估计[J].电力系统自动化,2010,34(17)：17-21
[37]李春艳,孙元章,彭晓涛等.采用广域测量信息反馈的广域PSS参数设计[J].电力系统自动化,2009,33(18)：6-11
[38]赵茜茜.基于光与测量的电力系统频率稳定预测算法研究[D].西南交通大学硕士论文,2010
[39]王英涛.基于WAMS的电力系统动态监测及分析研究[D].中国电力科学研究院,博士论文,2006
[40]段俊东,郭志忠,魏成江.电压稳定临界域的描述及在线近似确定[J].继电器,2007,35(17)：47-51
[41]戚军,江全元,曹一家.一种简单实用的分散非线性励磁控制方法[J].电力系统自动化,2008,32(20)：32-36
[42]鞠平.电力系统建模理论与方法[M].北京：科学出版社,2010
[43]王锡凡,方万良,杜正春.现代电力系统分析[M].北京：科学出版社,2003
[44]刘取.电力系统稳定性及发电机励磁控制[M].北京：中国电力出版社.2007
[45]杨琳,何仁睦,马进等.基于实测的自并励励磁系统参数辨识[J].继电器,2005,22(6)：31-34
[46]邰彬.基于免疫算法的水轮机调速系统参数辨识的研究[D].华中科技大学硕士论文,2005
[47]杜晓勇.水电站动力系统模型辨识[D].郑州大学硕士学位论文,2006
[48]刘友波,刘俊勇等.定义于受扰轨迹的电力系统暂态冲击测度[J].电力系统保护与控制,2012,40(1)：29-36
[49]倪以信,陈寿孙,张宝霖.动态电力系统的理论和分析[M].清华大学出版社,2002
[50]杨冠成.电力系统自动装置原理[M].中国电力出版社,2007
[51]王信杰.电力系统暂态稳定建模的研究[D].郑州大学硕士学位论文,2007
[2]张爽.基于PMU的广域电网相量测量系统概述[J].广东电力电力,2007,20(12)：26-29
[3]鞠平,代飞,金宇清等.电力系统广域测量技术[M].机械工业出版社,2008
[4]李丹,韩福坤,郭子明等.华北电网广域实时动态监测系统.电网技术[J],2004,28(23)：52-56
[5]徐兴伟,穆刚,王文等.东北电网广域动态测量系统[J].电网技术,2004,30(17)：70-73
[6]李颖,贺仁睦,徐衍会等.广东电网基于PMU的负荷模型参数辨识研究[J].南方电网技术,2009,3(1)：16-19
[7]李丹,韩福坤,郭子明等.华北电网广域实时动态监测系统[J].电网技术,2004,28(23)：52-56
[8]Hope G. S. Measurement of transfer function of Power system components under Operation conditions [J]. IEEE Transactions on Power Apparatus and Systems,1977,59(6):96-99
[9]方崇智,萧德云.过程辨识[M].北京：清华大学出版社,1988
[10]米增强,陈志忠,南志远.同步发电机动态参数辨识[J].中国电机工程学报,1998,18(2)：100-105
[11]王良,沈善德,朱守真等.基于EE模型的励磁系统参数辨识时域法[J].电力系统自动化,2002,26(8)：25-28
[12]曹浩军,张承学,单勇.基于连续时间方法的励磁系统参数辨识[J].电力系统自动化,2004,28(17)：49-54
[13]戴义平,邓仁纲,刘炯等.基于遗传算法的汽轮机数字电液调速系统的参数辨识研究[J].中国电机工程学报,2002,22(7)：101-104.
[14]许津津,马进,唐永红等.基于改进DE算法的负荷建模参数辨识[J].电力系统保护与控制,2009,37(24)：37-45
[15]徐丽娜.神经网络控制[M].哈尔滨：哈尔滨工业大学出版社,1999
[16]米增强,张宁,刘国平等.基于WAMS的同步发电机组动态等值研究[J].华北电力大学学报,2008,35(2)：30-34
[17]Kang Lin, Elias Kyriakides, Gerald T. Heydt. Experience with Synchronous Generator Parameter Identification using a Kalman Filter[J]. IEEE Transaction on Power Systems, 2010.4
[18]E. Ghahremani, M. Karrari, O. P. Malik. Synchronous generator third-order model parameter estimation using online experimental data[J]. IEEE, The Institution of Engineering and Technology.708-718,2008
[19]Robert Eriksson, Lennart. Royal Institute of Technology, Sweden System Identification Techniques for Obtaining Linear Models of Large Power Systems with Controllable Devices from Noisy Measurements[J]. IEEE, Symposium-Bulk Power System Dynamics and Control,2010,4
[20]金宇清,岳陈熙,甄威等.基于功角测量的同步发电机参数辨识频域法[J].电力系统自动化,2007,31(4)：7-11
[21]王正风,葛斐,戴长春.基于PMU的电力系统建模及参数辨识[J].继电器,2007.35：378-382
[22]郭小龙,胥国毅,薛安成等.基于PMU的励磁系统参数辨识研究[C].中国高等学校电力系统及其自动化专业第二十四届学术年会论文集,2008
[23]陈金猛.基于PMU的同步发电机与线路参数在线辨识研究[D].华北电力大学硕士论文,2008.1
[24]A. Phadke, M. Ibrahim, T. Hlibka. Fundamental basis for distance relaying with symmetrical components[J]. Power Apparatus and Systems, IEEE Transactions on,1977,96(2):635-646
[25]A. Phadke, J. Thorp, M. Adamiak. A New Measurement Technique for Tracking Voltage Phasors, Local System Frequency, and Rate ofChange of Frequency[J]. Power Apparatus and Systems, IEEE Transactions on,1983(5):1025-1038
[26]Z. Huang, B. Kasztenny, V. Madani et al. Performance Evaluation of Phasor Measurement Systems[C]. Power and Energy Society General Meeting,2008 IEEE 2008, Pittsburgh, PA 1-7
[27]韩松,何利铨,邱国跃WAMS研究、建设与应用的新进展[J].电测与仪表,2011.48(04)：1-8
[28]M. Lixia, C. Muscas, S. Sulis. On the Accuracy Specifications of Phasor Measurement Units[C]. Instrumentation and Measurement Technology Conference(I2MTC),2010 IEEE May 2010, Austin, TX 1435-1440
[29]E. Johansson, K. Uhlen, A. Leirbukt et al. Coordinating Power Oscillation Damping Control Using Wide Area Measurements[C]. Power Systems Conference and Exposition, 2009. PSCE 09. IEEE/PES 2009, Seattle, WA:1-8
[30]M. Hojo, K. Abe, Y. Mitani et al, Real-time Power System Monitoring at Demand Sides by Campus Wide Area Measurement System[C]. Industrial Electronics,2009, IECON 09.35th Annual Conference of IEEE 2009, Porto 3609-3614
[31]S. Boroczky, E. Gentle, Real-time Transient Security Assessment in Australia at EMMCO [C]. Power Systems Conference and Exposition,2009. PSCE 09. IEEE/PES 2009, Seattle, WA 1-6
[32]陈实,许勇,王正风.电网实时动态监测技术及应用[M].北京：中国水利水电出版社,2010
[33]卢志刚,郝玉山,康庆平等.电力系统实时相角监控系统研究[J].电力系统自动化,1997,21(9)：17-19
[34]陈凯,胡志坚,王卉等.介损角的非同步采样算法及其应用[J].电网技术,2004,28(18)：58-61
[35]赵凤山,王旭东.用单片机测量同步发电机功角的方法[J].黑龙江电力技术,1995,17(5)：279-281
[36]李大路,李蕊,孙元章.混合量测下基于UKF的电力系统动态状态估计[J].电力系统自动化,2010,34(17)：17-21
[37]李春艳,孙元章,彭晓涛等.采用广域测量信息反馈的广域PSS参数设计[J].电力系统自动化,2009,33(18)：6-11
[38]赵茜茜.基于光与测量的电力系统频率稳定预测算法研究[D].西南交通大学硕士论文,2010
[39]王英涛.基于WAMS的电力系统动态监测及分析研究[D].中国电力科学研究院,博士论文,2006
[40]段俊东,郭志忠,魏成江.电压稳定临界域的描述及在线近似确定[J].继电器,2007,35(17)：47-51
[41]戚军,江全元,曹一家.一种简单实用的分散非线性励磁控制方法[J].电力系统自动化,2008,32(20)：32-36
[42]鞠平.电力系统建模理论与方法[M].北京：科学出版社,2010
[43]王锡凡,方万良,杜正春.现代电力系统分析[M].北京：科学出版社,2003
[44]刘取.电力系统稳定性及发电机励磁控制[M].北京：中国电力出版社.2007
[45]杨琳,何仁睦,马进等.基于实测的自并励励磁系统参数辨识[J].继电器,2005,22(6)：31-34
[46]邰彬.基于免疫算法的水轮机调速系统参数辨识的研究[D].华中科技大学硕士论文,2005
[47]杜晓勇.水电站动力系统模型辨识[D].郑州大学硕士学位论文,2006
[48]刘友波,刘俊勇等.定义于受扰轨迹的电力系统暂态冲击测度[J].电力系统保护与控制,2012,40(1)：29-36
[49]倪以信,陈寿孙,张宝霖.动态电力系统的理论和分析[M].清华大学出版社,2002
[50]杨冠成.电力系统自动装置原理[M].中国电力出版社,2007
[51]王信杰.电力系统暂态稳定建模的研究[D].郑州大学硕士学位论文,2007