DEVELOPMENT OF A STOCHASTIC HIGH-LEVEL PETRI NET SYSTEM FOR FMS PERFORMANCE EVALUATION

Abstract

Flexible manufacturing system (FMS) was developed to fill-up the gap between transfer line (high production volume, low variety) and computer Numerically Controlled machines (low production volume, high variety). A typical FMS would consist of process stations, material, handling and storage, computer control systems and so forth. These highly capital intensive systems can make the products which are competitive in world markets. To guarantee profitable returns on investment on such competitive manufacturing systems, it is essential to have effective design strategies. Decision making is involved at various stages of FMS. viz. planning, design and operation. Evaluation of performance measures (e.g. machine utilization, work-in-process etc) can aid in this decision process. Various techniques like Markov chain, queueing network, simulation technique etc. have been used to model the FMS. Recently, Petri nets have become a popular modelling tool for FMS because they can efficiently model several types of system behaviours such as concurrency, synchronization, conflict, communication among the subsystems. In this work,stochastic high-level Petri net. (SHLPN) is used to model the flexible manufacturing system with three machines and two part types. They are extensions of generalised stochastic Petri nets. The main advantage of modelling homogeneous systems with HLPNs is that resulting models are simpler, more intutive and have smaller number of states. For the modelled PN system, Markov chain is obtained, which is further processed to evaluate various performance measures.