Data mining based configuration of cyclically interlinked production systems

Abstract

Cyclically interlinked production systems are commonly used production facilities. During the bidding phase manufacturers of these systems assure customers about the performance and operating efficiency despite numerous uncertainties. Up to now, there has been no planning approach that could hedge these results. Recently however, a method for configuring cyclically interlinked production systems during the bidding phase has been developed in order to resolve this deficit. Data mining models are built to predict performance figures of production systems, whereas, characteristic curves are used to predict the operating efficiency of a system configuration. This paper introduces the challenge of bid proposal management and explains the developed method. © 2010 CIRP.