PERFORMANCE EVALUATION OF A FLOW-LINE BY IMPROVING ANFIS SUGENO WITH FIS MAMDANI

Abstract

This paper presents the Fuzzy Logic method in order to evaluate the performance of a real flow-line. The flow-line comprises a number of machines that are arranged in series and the flow of products is continuous. The objective of this paper consists in the production rate prediction of a manufacturing system using the Fuzzy Logic method through ANFIS SUGENO and FIS MAMDANI. Input data are processed in ANFIS SUGENO to view the construction mode of the input membership functions in order to achieve our own model of FIS MAMDANI. Input variables' membership functions are modified by removing the values which exceed the representation range, after which the inference rules and the representation range of the output variables are to be established. The forms of membership functions are very important because they influence the prediction accuracy. A case study is carried out to demonstrate the performance and flexibility of the proposed approaches.

Keywords

• estimation
• flow-line
• ANFIS SUGENO
• FIS MAMDANI