Design of a data-driven internal model controller

Abstract

In recent years competition in the industrial world has become intense. As a result, the necessity for highly effective control techniques has increased. Furthermore, simple control structure and high robustness are also desirable. To meet these demands, model-driven type control theories type have been proposed. If these controllers are able to describe the model of the process accurately, then perfect control is theoretically possible. In this paper, internal model control (IMC) is the core proposal option though there are some controllers of the model driven type. However, it is difficult to obtain good control performances, since most process systems are nonlinear. This paper thus proposes a new IMC based on the data-driven(DD) approach as a control system that can also be adjusted to a nonlinear system. In taking the DD approach, a suitable model can be automatically generated. As a result, a preferable control performance can be obtained even for nonlinear systems. The behavior of the newly proposed control scheme is examined on a simulation example.