Research On The Technical Problems Of The Continuous-time Syetem Model Parameters Direct Identification Based On Modulation Function Method

Continuous time parameter model is the best way to describe the system dynamic characteristics, and direct identification method for continuous-time model is the best identification method for dynamical systems. As the direct identification method can not directly measure the system input and output signals of every order derivative, by using modulation function method, this thesis solves the differential by parameter modulation integral which avoids solving the differential by itself. The studies of this thesis are as follow:(1) First it detailed describes the nature and characteristics of several modulation functions, and constitute the direct Least Square algorithms and scalable algorithms model of the linear continuous-time dynamic parameter system. Furthermore it introduces a noise model to instead the input and output noise which reduce the computation and improves the identification accuracy. By using hermite modulation function, combined with the noise model, identification simulation for linear continuous parameter model has been carried out. Next the first, second, third and fourth order of the noise model was taken to used for identification simulation, finally the model parameters are solved by using extended least squares method.(2) Second the equivalent identification models of Hammerstein nonlinear and bilinear continuous-time system are formulated. Next the equivalent discrete identification model has been established by digital modulating integral in modulating function. Then by selecting Hermite and Hartley function respectively, along with window shifting and frequency index changing modulation function, the model parameters are estimated using the least-squares algorithm. The conclusions deduced in simulation examples of linear system are proved to fit nonlinear system as well.(3) Third the method proposed in this thesis are used in 2/2 and 1/2 order dynamical parameter of closed loop HAGC to do the parameter identification of the model.The study of parameter identification method models based on modulation function of linear and nonlinear continuous-time system enriches the achievements of system identification. And it can applied to large-scale industry equipments.