采用最大-最小蚁群算法的励磁系统参数辨识  

Excitation System Parameter Identification via Maximum-minimum Ant System

作　　者：毛晓明[1] 蔡永智[1] 赵勇[2] 

机构地区：[1]广东工业大学自动化学院,广州510006 [2]中国南方电网科学研究院,广州510080

出　　处：《电力系统及其自动化学报》2015年第5期51-55,共5页Proceedings of the CSU-EPSA

基　　金：广东省自然科学基金项目(2014A030313509;S2013010012431;S2013040013776);广东省教育厅育苗工程项目(2013LYM_0019)

摘　　要：为获得发电机励磁系统准确的模型参数,对BPA(bonneville power administration)软件中励磁系统典型仿真模型进行深入分析,得到模型参数与励磁系统大、小阶跃响应特性的关联关系。结合实际情况,对需辨识参数进行筛选,对参数取值范围进行限制。采用最大-最小蚁群算法对参数进行辨识,先求得影响发电机空载电压小干扰阶跃响应特性的主要参数,再得到影响发电机空载电压大干扰阶跃响应特性的主要参数。BPA计算得到的辨识模型仿真曲线与实测数据吻合良好,仿真结果表明了辨识方法的有效性。In order to acquire the accurate parameters of generator excitation systems, the typical simulation models used in bonnevitle power administration BPA software package are analyzed in depth. The relationship between the parameters and system small and large-signal voltage-step response is investigated. Parameters needed to be identified are selected and their ranges of values are specified according to practical status. Selected parameters are identified via the maximum-minimum ant system, and the parameters which mainly impact on generator small-signal voltage step-response are initially obtained, then the parameters dominating generator large-signal voltage step-response are subsequently acquired. Simulation curves of the identified model obtained by BPA fit well with the field data, and results show the effectiveness of the developed method.

关 键 词：励磁系统 参数辨识 最大-最小蚁群算法 电压阶跃响应 

分 类 号：TM743[电气工程—电力系统及自动化]