| 基于MMTMV方法的多变量时变扰动系统性能评估 |
| |
| 引用本文： | 刘春平, 王昕, 王振雷. 基于MMTMV方法的多变量时变扰动系统性能评估. 自动化学报, 2015, 41(5): 928-935. doi: 10.16383/j.aas.2015.c140719 |
| |
| 作者姓名： | 刘春平 王昕 王振雷 |
| |
| 作者单位： | 1.华东理工大学化工过程先进控制和优化技术教育部重点实验室 上海 200237;;;2.上海交通大学电工与电子技术中心 上海 200240 |
| |
| 基金项目： | 国家重点基础研究发展计划(973计划) (2012CB720500),国家自然科学基金(61134007, 61222303), 上海市科技攻关项目(12dz1125100),上海市科学技术委员会(13111103800),上海市自然科学基金(14ZR1421800),流程工业综合自动化国家重点实验室开放课题基金(PAL-N201404)资助 |
| |
| 摘 要： | 工业控制系统性能评估关系到工业生产安全与企业绩效. 工业过程扰动频繁且存在时变特性,多变量时变扰动系统缺少统一有效的性能评估方法. 本文提出了基于多模型混合时变最小方差(Multi-model mixing time-variant minimum variance, MMTMV)的多变量系统性能评估方法. 首先,根据扰动作用起止时间设定混合权重,基于多模型混合的思想利用混合权重与每一时变扰动特性设计多变量MMTMV控制器. 然后,利用多变量MMTMV控制器得出各被控变量输出方差,将多变量MMTMV控制器下被控变量的平均方差作为性能评估的基准. 最后,本文通过在裂解过程和精馏过程中的控制性能评估应用,验证了多变量MMTMV评估方法的有效性.
|
| 关 键 词： | 多变量 时变扰动 多模型混合 最小方差 |
| 收稿时间： | 2014-10-24 |
| 修稿时间： | 2015-01-19 |
Performance Assessment of MIMO System with Time-variant Disturbances Based on MMTMV Method |
| |
| LIU Chun-Ping, WANG Xin, WANG Zhen-Lei. Performance Assessment of MIMO System with Time-variant Disturbances Based on MMTMV Method. ACTA AUTOMATICA SINICA, 2015, 41(5): 928-935. doi: 10.16383/j.aas.2015.c140719 |
| |
| Authors: | LIU Chun-Ping WANG Xin WANG Zhen-Lei |
| |
| Affiliation: | 1. Key Laboratory of Advanced Control and Optimization for Chemical Processes, East China University of Science and Technology, Shanghai 200237;;;2. Center of Electrical &Electronic Technology, Shanghai Jiao Tong University, Shanghai 200240 |
| |
| Abstract: | Performance assessment of industrial processes is related to industrial production safety and benefit. It is widely acknowledged that most of the industry processes are time-varying and influenced by several disturbances. There is no available method to assess the performance of the MIMO system with time-variant disturbances. A multi-model mixing time-variant minimum variance (MMTMV) method of MIMO is proposed in this paper to solve the problem. First, a multivariable MMTMV controller is obtained based on the properties of time-variant disturbances and the weights gained by the action times of disturbances. Then each output variance of the MIMO system employing the multivariable MMTMV controller is derived and the average variance of all controlled variables is utilized as the benchmark to assess the performance of the MIMO system with time-variant disturbances. Finally, the effectiveness of MMTMV is verified in the CPA of a cracking process and a distillation process. |
| |
| Keywords: | Multivariable time-variant disturbances multi-model mixing minimum variance |
|
| 点击此处可从《自动化学报》浏览原始摘要信息 |
|
点击此处可从《自动化学报》下载全文 |
|
