Joint Research on Modeling- and Optimization-based Control

点击翻译按钮获取中文摘要

The object of the research is integrate the ideas of system modeling, optimization and control and develop a new approach for these fields.1.Local modeling and controlLocal modeling is known as a useful approach for nonlinear system identification. Since the system model is composed of several local models corresponding to local operating regions, selection of local operating regions is crucial. An algorithm for automatic selection of suitable operating regimes by using Kullback's discrimination information (KDI) and Akaike's information criterion (AIC). Numerical simulation results demonstrate the effectiveness of the proposed method. Professor Foss (Norwegian University of Science and Technology), international co-operating researcher of this research, addresses the control of the nonlinear systems modeled by this approach.2.Nonlinear system identification using Genetic ProgrammingNonlinear system identification based on Genetic Programming has been developed. This approach provides the system model structure, which plays an important part in system identification. Wiener and Hammerstein models are identified by this approach.