Neo-Robust Control Theory and Its Design Tool-Toward a Super HighPerformance Robust Control-

在具有模型不确定性的系统控制中，如何最大限度地提高控制性能是一个非常重要的问题。在传统的鲁棒控制理论中，几乎都只利用了不确定性信息的一部分。因此，提高系统的控制性能是相当困难的。为了解决这一问题，本研究建立了一种新的鲁棒控制理论，该理论不仅利用了增益信息，而且利用了不确定性的相位信息。该理论是由一个包含不确定性的增益信息和相位信息、鲁棒稳定性条件、鲁棒灵敏度条件和鲁棒带宽条件的对象集模型组成的。此外，通过理论和基准系统证明，与传统的鲁棒方法相比，新理论能够显著提高鲁棒性能。该理论具有极大地提高鲁棒控制系统在工业应用中的性能的潜力。

In the control of systems with model uncertainty, it is an extremely important issue how to enhance the control performance to the limit. In the conventional robust control theories, almost all only make use a part of the uncertainty information. For this reason, it is rather difficult to improve the control performance of the system.To resolve this problem, this research established a new robust control theory, which utilizes not only the gain information, but also the phase information of the uncertainty. This theory is composed of a model for the set of plant which includes both the gain information and the phase information of the uncertainty, the robust stability condition, the robust sensitivity condition and the robust bandwidth condition. Further, it is proved both theoretically and via a benchmark system, that the new theory is able to enhance the robust performance significantly, as compared with conventional robust approaches. This theory has a potential of greatly improving the performance of robust control systemsin industrial applications.