CORC  > 清华大学
基于MIMO辨识的复杂系统多控制器最优设计
陆超 ; 金小明 ; 李鹏 ; 吴小辰 ; LU Chao ; JIN Xiaoming ; LI Peng ; WU Xiaochen
2010-06-10 ; 2010-06-10
关键词电力系统稳定 预测误差 辨识 最优控制 阻尼控制 power system stability prediction error identification optimal control damping control TM571
其他题名Optimal design of multiple controllers based on MIMO identification for large scale power system
中文摘要由于互联电网低频振荡的全局性质,各种阻尼控制器必须协调设计,但对于实际大规模电网,作为设计基础的系统模型难以获取,提出了一种基于辨识的最优协调设计方案。通过注入的带宽信号(0.1~3Hz,已包含整个机电振荡模式频率范围)激发动态稳定控制相关的系统特性,采用预测误差方法(PEM)辨识得到复杂系统的多输入多输出(MIMO)降阶模型。通过引入额外的状态量和输出量改进传统的线性最优控制设计方法,使其控制器结构中可包含领先-滞后环节,而不仅仅是比例环节,并给出了扩展后的系统方程及控制器参数计算方法。上述方案通过2条直流线路的调制控制器设计进行了验证,其中MIMO降阶模型辨识时曲线拟合度可达80%以上。; The multiple damping controllers must be coordinately designed to avoid the low-frequency oscillation of inter-connected power systems. However,it is difficult to mathematically model a complicated power system. An optimal design approach based on system identification is proposed. A special injection signal,0.1~3 Hz covering all the electro-mechanical oscillation modes,is injected to excite the dynamics of system and the prediction error method is employed to identify the MIMO (Multiple Inputs and Multiple Outputs) order-reduced system model. Additional state and output variables are introduced to improve the traditional linear optimal control technique,which enable the controller to include both lead-lag and proportion blocks. The extended system equations and controller parameter calculation method are given. This approach is verified by the modulation controller design for two HVDC links. The curve fitting degree of MIMO order-reduced model identification is more than 80%.; 国家科技支撑计划项目(2006BAA02A17); 国家重点基础研究发展计划资助项目(2004CB217907)~~
语种中文 ; 中文
内容类型期刊论文
源URL[http://hdl.handle.net/123456789/62054]  
专题清华大学
推荐引用方式
GB/T 7714
陆超,金小明,李鹏,等. 基于MIMO辨识的复杂系统多控制器最优设计[J],2010, 2010.
APA 陆超.,金小明.,李鹏.,吴小辰.,LU Chao.,...&WU Xiaochen.(2010).基于MIMO辨识的复杂系统多控制器最优设计..
MLA 陆超,et al."基于MIMO辨识的复杂系统多控制器最优设计".(2010).
个性服务
查看访问统计
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。


©版权所有 ©2017 CSpace - Powered by CSpace