Robust Model Matching Design Method Which Ensures Stability

Yi‐Sheng Zhong,Ryozaburo Tagawa,Toshio Eisaka
DOI: https://doi.org/10.1002/ecjc.4430720910
1989-01-01
Abstract:AbstractThe design method entitled “robust model matching” was recently proposed for the design of practical robust control systems. By this method, if the overall system is stable, any desired robustness in the steady‐state and transient characteristics can be guaranteed in spite of parameter variations for minimum‐phase plants. However, when using this method it is always necessary to employ a trial‐and‐error process to ascertain stability. In this paper a robust model matching design method which guarantees stability is proposed for the case when the parameter variations are such that the plant remains a minimum‐phase plant without a change in the sign of the gain. Moreover, only the controlled variable and the operating quantity are measurable. By this approach, any desired robustness in the steady‐state and transient characteristics can be achieved while guaranteeing robust stability without the necessity of a trial‐and‐error process.In this paper, first the necessary conditions a robust filter must fulfill to ensure stability are presented. Next, the degree relationships that certain polynomials concerning the model matching controlled must fulfill to design a robust filter are clarified. Further, it is proven that control system stability is ensured by the design of an appropriate robust filter without narrowing the class of realizable reference models. Finally, the proposed design method is set forth and an example design problem is given.
What problem does this paper attempt to address?