Subspace based direct adapitve control for a class of nonlinear systems

Abstract

This project exploits subspace prediction methods in order to apply a novel direct "predictive" control design technique, which can be viewed as an extension of model free subspace based linear quadratic Gaussian (LQG) control, and in the class of adaptive control. The main purpose of this study is to design a much simpler control approach for a wastewater treatment plant using a data driven direct adaptive predictive controller based on subspace identification of prediction matrices. The general direct control design problem requires the engineer to collect experimental data, and choose a performance objective. With these design choices, it is then possible to calculate a control law that optimizes expected future performance. Recently, there has been significant interest in developing a direct control design methodology producing a more reliable and automated control design technique. Effective control of wastewater treatment plants (WWTPs) has been receiving rising attention during the last decade due to increasing concern about environmental issues. In this sense, the importance of studies concentrating on control and simulation of WWTP is remaining intact. Activated sludge process is commonly used in biological wastewater treatment. The applied methodology of this project is supposed to regulate the substrate concentration and dissolved oxygen concentration to specified values, compensate the disturbances may occur in the load influents as well as to track any mechanistic or kinetic parameter variation immediately in the shortest possible time. Hence, direct adaptive predictive control (DAMPC) can provide simplicity, good performance and stability robustness of an activated sludge process.