Internal Model Control For System With Constraints

In the background of fast development for social economy, the development of modern industrial control has been moving forward steadily. In actual industrial control field, multiple input multiple output systems(MIMO) exist widely, at the same time, the constraints are also very common. The implementation of the conventional control may encounter many restrictions, leading to it can’t meet the production requirements. So the control for the constrained multivariable complex system has become the focus in these years. Seeking the solutions of constraint problem in system is the basic element to explore innovation in this paper.In the developing course of internal model control(IMC), a kind of classic saturated IMC structure based on the coprime decomposition has better control effect in terms of dealing with constraints. Inspired by the train of thought, this topic is mainly studied how to design the compensation to reduce the influence on the system performance of the constraint, aiming at providing a new solution for the constraint system control.For setting controller parameters, this paper is based on performance indexes of NPSO setting method of optimization algorithm. The multivariable decoupling generally can be divided into static decoupling and dynamic decoupling, and the latter effect is usually better than the former. This paper adopts the diagonal matrix decoupling which belongs to the latter method, to get the general ideal object. The overall architecture is based on internal model control (IMC) which has a unique processing to the system time delay and can get ideal control effect by setting the filter parameters. For the controlled object with input amplitude constraints resulting in serious decline the system tracking performance, this paper puts forward two kinds of compensator design method. First is the design of optimal compensator based on internal model controller coprime decomposition, the second is a new type of internal model control scheme based on feedback compensation. The design of the two plans are both divided into two steps, first of all, design classic internal model controller, and then add the compensation. The essence is to extend the constraint boundary value time of the controller output, making the system response in a state of relatively rapid response in a longer period of time and improving the tracking performance of the system. Through the simulations which use ITAE and ISE index to analyze the system performance, it shows that the proposed methods in this paper have a short adjusting time and fast response while guarantee the system robustness.