MATLAB环境下的系统辨识仿真工具箱

介绍了在MATLAB环境下构建一种新的系统辨识仿真工具箱(SIST:SystemIdentificationToolbox)的原理与实现,讨论了在MATLAB环境下工具箱的数据和算法结构形式,以及利用GUIDE完成图形界面的设计,用MATLAB与C 混编技术实现物理系统的I/O接口和系统辨识模型与算法、报告文件(report)和帮助文档等功能模块的编程。这种在MATLAB环境下开发的系统辨识仿真工具箱,具有功能丰富、可扩展性好、使用方便、操作直观等特点。还给出了实际使用例,证实工具箱的有效和可靠,可作为辨识研究和教学的辅助工具。     

Process modeling and optimizing control based on sparse nonuniformly sampled data

<正> In this paper,a process modeling and related optimizing control for nonuniformly sampled ( NUS)systems are addressed.By using a proposed nonuniform integration filter and subspace method estimation,an identification method of NUS systems is developed,based on which either an output soft sensor or ahidden state estimator is developed.The optimizing control is implemented by replacing the sparsely-measured/immeasurable variable with the estimated one.Examples of optimizing control problem are given.The proposed optimizing control strategy in the simulation examples is verified to be very effective.