Uncertainty modeling methods capable of capturing phase and gain information

• Published in: IFAC-PapersOnLine 2023-01, Vol.56 (2), p.1871-1876
• Main Authors: Oki, Y., Liu, K.Z., Koiwa, K., Zanma, T.
• Format: Article
• Language: English
• Subjects: gain and phase; hard disk drive; Robust control; uncertainty model
• ISSN: 2405-8963; 2405-8963
• DOI: 10.1016/j.ifacol.2023.10.1904

In practice, the plant model is usually built from its frequency responses obtained from identification experiments. In this sense, it is desirable to build an uncertainty model from the frequency responses. Up to date, there are mainly two kinds of uncertainty models: norm-bounded and passive. For general systems, the range of uncertainty is inevitably enlarged by these two uncertainty models, thus leading to design conservatism. This paper proposes two novel uncertainty models, together with sophisticated modeling algorithms. These models are able to reduce the enlargement of uncertainty's range to the utmost, and can be easily combined with existing robust design methods. It can be interpreted as a fine capture of uncertainty's gain and phase ranges. A case study demonstrates the superiority of the proposed models.