量测滞后下的发酵过程状态在线估计方法研究

项目名称： 量测滞后下的发酵过程状态在线估计方法研究

项目编号： No.61503019

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 自动化技术、计算机技术

项目作者： 赵利强

作者单位： 北京化工大学

项目金额： 20万元

中文摘要： 发酵过程生物参量的在线检测对于发酵过程在线监测、优化控制具有重要的意义。本项目从非线性高斯滤波方法的研究出发，提出一种具有较强数值稳定性的高阶容积卡尔曼滤波（CKF）算法，并结合鲁棒处理算法和噪声自适应估计器，实现用于非线性状态估计的自适应鲁棒高阶CKF方法；基于量测滞后信息模型和不完全量测滤波性能评价指标，研究量测滞后下非线性系统的滤波策略和方法，结合自适应鲁棒高阶CKF滤波算法，实现量测滞后下的非线性系统滤波；研究量测滞后下的发酵过程混合模型建模方法，在混合模型可观测性条件保证下，结合不完全量测滤波理论与方法，给出量测滞后下的发酵过程状态在线估计方法，实现发酵过程生物参量的在线检测。本项目将现代信号处理、系统建模、人工智能等理论与方法应用于发酵过程生物参量在线检测的研究，最终为发酵过程监测与过程优化控制技术及其应用提供支持。本项目将在医药、食品、轻工、农业等领域具有很好的应用前景。

中文关键词： 容积Kalman滤波；量测滞后；不完全量测；混合模型；发酵过程

英文摘要： It’s of important significance to research biological parameters on-line measurement for on-line monitoring and optimal control of fermentation process. The work starts from the research of nonlinear Gauss filtering algorithm, and one high-degree cubature Kalman filter (CKF) with strong numerical stability will be proposed. By combining the robust algorithm and noise adaptive estimator, the adaptive robust high-degree CKF method for nonlinear state estimation will be achieved. The nonlinear filtering strategy and method with delayed measurements will be studied based on information model of delayed measurements and performance evaluation index of incomplete measurements filtering. Combined with the adaptive robust high-degree CKF algorithm, the filtering of nonlinear system with delayed measurements will be developed well. The hybrid modeling method for fermentation process with delayed measurements will be explored. In the guarantee of observability conditions of the hybrid model, the state on-line estimation method for fermentation process with delayed measurements will be established based on incomplete measurements filtering theory and method, and the biological parameters on-line measurement for fermentation process will be realized. The modern signal processing theory and technology, system modeling method, and artificial intelligence theory and method are applied to the research of the biological parameters on-line measurement for fermentation process, and the research achievements of this project will provide theoretical support for on-line monitoring and optimal control of the fermentation process. This project will have a very good application prospect in the fields of medicine, food, light industry, agriculture, etc.

英文关键词： cubature Kalman filter;delayed measurements;incomplete measurements;hybrid model;fermentation process