Application of fuzzy classification and fuzzy pattern recognition for distributed production and global supply chain

Abstract

Global Corporations have to manage distributed production over the whole world. Therefore global supply chains are needed. This paper discusses the problem how global production plants and their supply chains can be classified. The classification focuses on demand and supply of production and supply chain. The problem of forecasting the demand of a global supply chain is introduced. The difficulty of aggregated planning is examined. The problem of supply and demand synchronization of predictable variability is shown. Objective of the paper is to show solutions of the mentioned problems by using fuzzy classification and fuzzy pattern recognition methods. The approach is to use the classification methods fuzzy c-means (FCM) and Improved Fuzzy Clustering (IFC). Supply and demand patterns can be found with fuzzy pattern recognition. Therefore the methods Multi Feature Pattern Recognition and Fuzzy Inference System Type-2 (FIS 2) with neural network methods are introduced. The solution of the mentioned approach is realized by the application PROCAS (Process Optimization, Control, Analysis and Simulation). PROCAS uses a data warehouse database for multidimensional fuzzy classification data and Business Intelligence (BI) functionalities. The key result is that fuzzy classification and fuzzy pattern recognition applications improve the planning and operating of supply and demand in a distributed production and a global supply chain.