| Summary: | 碩士 === 國立成功大學 === 航空太空工程學系 === 89 === In general, the task of a control system synthesis begins with the estima-tion of a system model. Unfortunately, a parametric system model is often difficult to obtain, especially from analytically means. In addition, system models that are estimated from input and out plant test data are mostly with unknown modeling error. As a result, trial-and-error design process becomes inevitable in most control system design. In order to conduct a control system synthesis, or DBCS, technique has been developed. A DBCS identify and aut-henticate a sought controller directly from plant test data when the plant model remains unidentified. Moreover, the zero of a dynamic system plays an important role in determining the response behavior. And despite a vast design techniques in relocating the system pole, few is capable of changing the location of system zeros. For discrete time systems, though, placement of certain zeros of the system can be made possible by modifying the input holding pattern from zero-order-hold. However, previous methods for this design require explicit knowledge of a parametric plant model. In this work, we combined the zero placement formulas with DBCS technique to achieve a data-based zero placement design.
|
|---|
