A design of multiloop PID controllers with neural-net based decoupler

In process industries, PID control schemes have been widely used due to their simple structure and ease of comprehending the physical meanings of control parameters. However, since most processes are considered as nonlinear multivariable systems with mutual interactions, good control performance can not be obtained by simply using PID control schemes. In this paper, a design method for neural-net based PID controllers is proposed for a nonlinear multivariable system with mutual interactions. The proposed method consists of a decoupler generated by the sum of a static decoupler and a neural-net based decoupler, and multiloop PID controllers. Finally, the effectiveness of the proposed control scheme is evaluated for a simulation example.