The Controllability Analysis Of Nonlinear Controlled Systems Based On Data

The analysis and control of nonlinear systems is an important problem in complex control theory, but also a difficult problem. For nonlinear systems, there is yet no complete and systematic theory. There have been some methods with problems such as complexity, no hierarchy, quasi-linear control and space measure destruction. Therefore, based on the analysis of kinematics, a comprehensive method which can be used on the research of controllability of nonlinear system is established in this dissertation.A mode of approaching element is constructed in this dissertation. A data sample is acquired through sampling. Based on the data sample, a model of approaching element is established with LS-SVR, which is used for the approximation of nonlinear systems. In the element space, important concepts such as quasi-working element, controllable element, controllable domain and the controllability of the cut set for approximation element space are defined. A transfer matrix is set up for nonlinear controlled systems based on the element model of approaching. The analysis on controllability of nonlinear systems can carry out based on the transfer matrix. The element model of approaching integrates the quantity space, approximation space and symbol space which helps conduct an effective analysis and research on the nonlinear systems.