| Process identification is a theory and methodology that is focused on the model building of production process, which is aiming at improving the performance of control system as well as a core technology to serve as a basement for design of advanced control and reality of optimizational control. To adapt to the needs of application of advanced control, process identification combined with the design of controller design technology. In this paper, theory and arithmetic for process identification as well as four base elements are introduced briefly. In allusion to multivariable identification, based on a great deal of theory analysis, simulations, some important conclusions are made. Finally, for the purpose of practicality of the foregoing research, a spftware for IMC is designed in a structure mode. The main job is as follows:1,Make a introduction to the asymptotic identification in detail,discuss several issues which is realated to its reality,such as model selection,model validation and so on.2,Do some research on the step identification deeply. An improved algorithms so-called second-order Tayor approximation was proposed, which is proved to achieve a better performance of delay time estimation. 3,Extended the asymptotic identification to the situation of closed-loop and gave its reality arithmetic combined with a indirect closed-loop identification method so-called as two-stage method.4,Discussed identification issues on intergral and dead time process, designed the experiment signnal, identification experiment. Results of simulation prove the availability of asymptotic identification to identify intergral and dead time process.5,Based on the model gained through those methods methioned above, aiming at the internal model control(IMC) which is based on the model, developed a IMC software package which is faced to distributed control system combined with the model idetification methods which are researched in this paper. It established a basic work for further application. |
PDF Full Text Request