Adaptive optimization of a lead compensator based upon input command frequency during operations of X-Y positioning table

Due to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to digital@library.tamu.edu, referencing the URI of the item.Includes bibliographical references.Issued also on microfiche from Lange Micrographics.The premise for this research is a desire to improve the performance of systems that have complex command inputs. For such systems, a fixed controller may not provide the necessary means to optimize the system's performance. To remedy this problem, a supervisory controller is implemented that will consider the command input, and update the gains of the controller based on a performance index. The performance index is comprised of the various costs associated with operation of the system. Each cost has a control engineer-specified weight. These costs, unlike those in conventional optimal control, are not restricted to the system's states and input, but include any other cost associated with the system that the control designer can express or approximate mathematically. The optimization of this performance index results in optimal design specifications. The supervisory controller is comprised of three functions; the Input Characteristic Analysis (ICA), the Design Specification Function (DSF) and the Controller Parameter Computation (CPC). The ICA measures the characteristics of the input. The DSF uses the characteristics from the ICA and the performance index to determine the optimal design specifications. The CPC then updates the controller parameters based on the specifications from the DSF. As a demonstration, a supervisory controller is developed for an X-Y positioning table which tracks a frequency varying sinusoid. The results show that this control scheme allows costs normally only considered by the control engineer to be accounted for. It also allows the control engineer to predictably tune the performance of the supervisor using the cost weights of the performance index. Lastly, this supervisory control allows the most appropriate controller to be utilized, as opposed to having one fixed controller which may not apply to every input performed by the system