Research on Second-order Feedback Filter for Two-degree-of-freedom Internal Model Control
|Course||Control Theory and Control Engineering|
|Keywords||internal model control feedback filter intelligent control robustness|
With the development and application of the new and high technology, the requirement for control system’s performance is higher and higher. At the same time, the application of intelligent control is more and more extensive and it has become one of the indispensable tools in industry control field and much attention has been paid to it increasingly. In addition, as a kind of new-type control method, the internal model control(IMC) has many advantages, such as good robust performance, simple structure, only a few online tuning parameters and being easy to design and tune. The recent research shows a tendency to combine intelligent control with IMC. This paper begins with a brief description of the recent, including the design and improvement of the controller, the improvement of the structure of internal model control and the connection of internal model control with other control methods. Then we explain the basic principle of internal model control, analysis the fundamental properties of internal model control and explains the controller design method of two-degree-of-freedom internal model control. On this basis, we summarize the current research situation of the most commonly-used first-order feedback filter in the IMC structure. There is only one regulation parameter for first-order filter and the coordinated performance is not excellent. Therefore we propose the second-order feedback filter structure for IMC and analyse its properties. Since it has two tuning parameters, its performance is superior to the first-order feedback filter and it is proved by the simulation results. Furthermore, continuing the study of second-order feedback filter, we proposed a new intelligent method online tune the parameters of the second-order filter to treat by using the system error and the change rate of system error and compared itsperformance with that of the first-order-filter.