项目名称: 广义Hammerstein模型辨识及其工业应用:调节阀粘滞问题的诊断与补偿
项目编号: No.61074105
项目类型: 面上项目
立项/批准年度: 2011
项目学科: 电工技术
项目作者: 王建东
作者单位: 北京大学
项目金额: 10万元
中文摘要: 系统辨识是研究如何从观测数据中建立系统的数学模型,在工程和科学领域拥有十分广泛的应用空间。Hammerstein模型是将简单的非线性静态模块和线性动态模块连接起来,其辨识问题一直深受国内外研究者和工程技术人员的关注。本项目将放宽目前相关研究中对Hammerstein模型的组成模块和实验条件做出的限制,引入带有一步记忆步长的迟滞模块和连续时间线性模块,研究广义Hammerstein模型的辨识方法和理论,从而准确地描述更多的实际工业系统,力图推进非线性系统辨识领域的研究进程。本项目还将基于广义Hammerstein模型辨识的研究结果,探索对调节阀的粘滞问题进行诊断与补偿的新方法,有效地指导工业生产的优化运行和预知维修。
中文关键词: 系统辨识;非线性系统;Hammerstein模型;调节阀粘滞
英文摘要: System identification is about the theory and practice of building mathematical models of dynamic systems from experimental data. It has a broad application in various areas of science and engineering. Hammerstien systems are a class of block-oriented nonlinear models, where a static nonlinearity is followed by a linear time-invariant dynamic subsystem in a cascade connecion. The objectives of the proposed object are to generalize Hammestein systems with hysteresis blocks and continuous time models, and to relax some restrictions on nonlinearity structures in earlier studies. By doing so, it is expected to extensively increase the applicability of Hammerstein systems, while contributing significantly to the progress of the researches on nonlinear system identification. The proposed project is motivated by the detection and compensation of control valves with stiction in closed-control loops via the identification of generalized Hammerstein system, in order to provide the guide to device manetinance and increase the efficiency and the reliability of control systems.
英文关键词: System identificaiton; nonlinear system; Hammerstein model; Control valve stiction