基于被控变量离线设计与在线更新的实时优化技术研究

项目名称： 基于被控变量离线设计与在线更新的实时优化技术研究

项目编号： No.61304081

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 叶凌箭

作者单位： 浙江大学宁波理工学院

项目金额： 26万元

中文摘要： 本项目以设计控制系统的被控变量为途径，研究化工生产过程在不确定扰动下的实时优化问题。围绕被控变量离线设计和在线更新这两个工作重点，提出一套新的实时优化实现方法并提炼出系统性的解决方案。主要包括：通过分析系统的最优状态必须满足一阶最优性必要条件这一特征，提出一种基于数值方法的被控变量离线设计思路；建立全局最优被控变量的搜索流程和快速算法,实现控制系统的自主优化控制；探索一种在线更新被控变量的策略，进一步提升控制系统在不同扰动工况下的实时优化性能；设计一种基于在线更新被控变量的新型分层递阶控制系统，给出该控制系统的设计思路和工作流程；引入过程监测方法对系统偏离最优状态的程度进行检测和评估，建立更新被控变量的必要条件。项目的研究成果将为生产过程的高效运行与优化操作提供重要的理论、技术支撑。

中文关键词： 化工过程；控制结构；被控变量；实时优化；反馈控制

英文摘要： Chemical processes are often operated under various uncertain disturbances, which may lead to large economic losses for plant operations. This research application aims to study real-time optimization (RTO) problem by means of designing appropirate controlled variables of control systems. The contents focus on the two parts, namely off-line design and on-line update of controlled variables, and proposes a series of new RTO methods and extracts a systemetic solution of RTO challenge. Specifically, the contents of this reasearch includes the following points: Via analysing the common feature of optimal statuses, namely the necessary conditions of optimality should be satisfied, an off-line design strategy for controlled variables based on numerical approach is proposed; The searching procedure of global optimal controlled variables and fast algorithm will be studied, the self-optimizing performance will be achieved for the control system; An on-line update strategy for controlled variables is proposed to further improve the RTO performance of control system; A novel hierarchy control structure based on on-line update of controlled variables is proposed, the design outlines, as well as the working procedure will be given; In the optimization layer of new hierarchy control system, the process monitoring methods will

英文关键词： chemical process；control structure；controlled variable；real-time optimization；feedback control