Modeling of K-250 Compressor Using NARX and Hierarchical Fuzzy Model

Due to the increasing use of compressors in the industry, it is important to determine a mathematical model for the compressor to design a control system, analysis and simulation of the computer. Also, in recent years, smart modeling such as neural network and fuzzy network have been considered by researchers for their more realistic performance, and their types have been used for modeling. Smart methods have high capability to communicate between input and output data. In this paper, modeling of K-250 compressor at Isfahan smelter company based on smart models of fuzzy neural network is presented. The Nonlinear Auto Regressive With exogenous input (Narx) and hierarchical fuzzy network are presented. For modeling, the system has been tested and the input and output data of the compressor using compressor sensors and image processing are used to convert the data into the required data in the modeling, then the above algorithms of the compressor model will be achieved with the help of software, MATLAB. The results of modeling Which NARX performed better than hierarchical fuzzy. Among the two models presented in this paper, the NARX model shows a better response than the hierarchical fuzzy network in all cases and in all aspects of the performance criteria.