Basic Research on Fusion of Control Theories towards A New Generation of System Control Engineering

下一代控制技术的基本问题是控制理论的整合，它覆盖了一个广泛的区域：内部非线性控制系统的建模、控制法的设计和安装以及计算功率的开发。这个研究项目很成功，为研究人员带来了关于讨论的更有意义的领域，提出了他们自己的新的研究结果，并为集成控制技术创造了新的方向。八次研究会议和一次有组织的控制会议在项目周期内将计划纳入实践;项目的贡献被总结为遵循：·与robust控制理论兼容的建模，·与隐式和混合系统的建模，·无限尺寸系统的robust控制，·非全息系统的控制和分析，·非线性H* 控制的理论和应用。·对样本数据系统的稳定性分析，以及基于线性矩阵不规则的优化理论的控制理论;用于建模，用于标识robust控制的主题，以及设置成员身份识别被讨论。为了控制，与非线性的理论进行交易，这些理论是常见于实用的控制系统中的。本报告包括非全息系统和非线性H* 控制的控制和分析。为了利用控制理论来开发计算能力，一个条件是为了确保样本数据控制的稳健性。此外，线性矩阵不确定性和信息几何之间的联系也得到了分析。“控制理论的整合刚刚开始，但它确实是以它的方式进行的。”

A fundamental issue of control technologies of the next generation is the integration of control theories that cover a wide range of areas : modelling of intrinsically nonlinear control systems, design and installation of control laws, and exploitation of computational power. This research project has been succesful in bringing researchers on the above mentioned areas in on discussions, presenting new research results of their own, and creating new directions for integrated control technologies. Eight research meetings and an organized session of a control conference during the project period materialized the plans into practice.Contributions of the project are summarized as follows :・modelling compatible with robust control theories,・modelling with implicit systems and hybrid systems,・robust control for infinite dimensional systems,・control and analysis of nonholonomic systems,・theories and applications of nonlinear H* control,・stability analisys of sample data systems, and・control theories based on optimization such as linear matrix inequalities.As for modelling, topics such as identification for robust control, and set membership identification were discussed. As for control, theories dealing with nonlinearities that are common in practical control systems are discussed. This includes control and analysis of nonholonomic systems and nonliner H* control. As for control theories exploiting computational power, a condition for robust stability of sample data control is considered. Also a connection between linear matrix inequalities and information geometry is analyzed. The integration of control theories has been just started, but it certainly is on its way.