Internal Model Control Strategy For Nonlinear System

Almost all highly complex chemical reaction processes with a clear nonlinear, although nonlinear theory has made great progress in recent years, but algorithms in the model analysis and synthesis is very complex and not easy to realize, therefore, research on a simple algorithm and easy industrial implementation of control method for nonlinear system, both in theory and in practice have value of theory and application research significance.This paper proposes an internal model control method based on multi-model according to the large working point of nonlinear process. Using multi-model makes the nonlinear system with transparent simplified model and controller, compared with other methods of the nonlinear method, computational complexity is greatly reduced. Aiming at the large working point of nonlinear process, using multiple local models to describe the object characteristics under different conditions, at the same time, aiming at a plurality of local mathematical model to design the corresponding partial internal model controller, and through the design of membership function to determine the proportion of each local model for different conditions, finally weight the local mathematical models to get the global model. At the same time design the multi-model IMC controller by weighting the local controllers to get the global controller, achieve global stability control requirements. By changing the load carried on the industrial field and analyzing the validity of the simulation results prove the method effectively.At the same time, this paper uses Wiener model to describe a class of block-oriented nonlinear processes, so we can put a nonlinear system represented as the form of a linear model and a nonlinear static model. This paper aiming at Wiener model, proposes the use of support vector machines to obtain α-order inverse model of the nonlinear part of Wiener model, then series connection it with Wiener model to get the base linearization pseudo-linear model, and then using internal model control strategy to control it. Through the simulation results of the motor control and pH neutralization process control, verify internal model control method adopting Wiener model to describe the dynamic process and based on support vector machines to obtain α-order inverse model can be good nonlinear control method, the control performance is good, and has good robustness when the model parameters change.