Multiple Model Adaptive Control And Application For A Class Of Nonlinear System

Research on control science and control engineer is the foundamental project of industry production and aerospace, there are many tools to analysis of linear systems, but most of control system is the form of nonlinear. For the analysis of nonlinear system, all the methods before are no longer efficient.The multiple model control is a better tool for analysis of nonlinear system, research on multi-model control of nonlinear system has significance theoretical purport and practical value. In this paper, we research the design and application problem of multiple model and adaptive controller in a class of nonlinear system, the main contents are as follows:Discussed the fundamental problem of multiple model adaptive control, such as the controller structure, the method of parameter identification, and the switch strategy.Then studied the control method for nonlinear system based on the theory of feedback linearization, recursive designed the controller and Lyapunov function of the whole nonlinear system. Furthermore, studied the stabilization problem of nonlinear systems with unknown parameters, combined with the ideas of recursive design and method of adaptive control in the process of backstepping design, finally construct such a traditional adaptive controller. In addition, considered the multiple model and adaptive control of nonlinear system with a class of parameter strict feedback, when the system parameters are unknown or the operation condition have been changed dramatically, the transient characteristic of traditional adaptive controller maybe disappointing, and the output time of asymptotic tracking will be longer, for such situation, multiple model adaptive controller has been designed, and the the boundedness and stability of algorithm has been proved at the same time. Moreover, for nonlinear system with a class of parameter and feedback strictly, we studied the affect of system’s switching strategy, in order to guarantee using the each model information effectively under the premise of system transient response, this paper proposes a method of no-switching and multiple model adaptive control, it’s proved that the no-switching algorithm has better performance under the situation with less number of model. Simulation results proposed to illustrate the effectiveness of the developed multiple models adaptive controller.According to the various advantages of multiple model control, it’s studied the practical application of multi-model control, for variable speed and variable pitch wind power generation system, there is given the scheme of multiple model control for the nonlinear of wind turbines and hysteresis characteristics. Compared with the traditional PID control strategy, that’s highlighted the effectiveness and advanced of multiple model algorithm. At the same time, the simulation result under the Matlab Simulinkenvironment is given, and the algorithm is ported to the wind power control system of Siemens PLC1200.