Performance modeling of FMS with flexible process plans - a Petri Net approach

A. Jain
2006 International Journal of Simulation Modelling  
This paper presents modeling and performance analysis of FMS when flexible process plans for each part type are available using Petri nets (PN) modeling approach. PN have been applied successfully for modeling of discrete event dynamic systems such as FMS that are characterized by conflicts, concurrency, synchronization and deadlocks. An example FMS consisting of four machines with individual input and output buffer has been taken into consideration and its PN model construction is explained.
more » ... ion is explained. System performance is evaluated in two manufacturing environments (i.e. virtual batch manufacturing and virtual line manufacturing). The analysis will assist the planner in selecting optimum set of operating parameters (such as dispatching rule, number of pallets released to the system etc.) for a given production order to achieve the desired performance measure. Several performance measures such as makespan, mean flow time, maximum flow time and variance of flow time have been used to evaluate system performance. Jain, Jain, Singh: Performance Modeling of FMS with Flexible Process Plans -A Petri Net ... The capital-intensive nature of FMS, coupled with their immense strategic potential demands careful attention to decision regarding their design and operation. Design decisions include the specification of system requirements, selection of resources and their configuration. Operational decisions include planning, scheduling and real time control of system operation. These decisions require use of system model that represents the relevant system features. Analysis of model reveals information regarding system behavior and serves as input to decision-making process. A variety of modeling approaches such as mathematical programming approach [3, 4, 5], markov chain approach [6, 7] , queuing networks [8, 9, 10], and computer simulation [11, 12, 13, 14, 15, 16] exist for design and operational analysis of manufacturing system. While these models provided insight into system behavior, they introduced many restrictive assumptions and tended to be computationally complex, making it difficult to model and evaluate manufacturing system in dynamic situation [17]. Petri nets (PN) modeling approach has been successfully employed to model, control and analyze the discrete event dynamic system that are characterized by concurrency or parallelism, asynchronous process, deadlocks, conflicts and event driven process and thus suitable for modeling, control and performance analysis (it provides a means of determining system characteristics) of FMS [18, 19, 20, 21] . However, it has been reported in literature that none of the present PN models incorporate variable process sequencing and a few incorporate variable routing [22] . PN is a graphical and mathematical modeling tool. As a graphical tool, it serves as a visual modeling technique and as a communication aid for describing models. As a mathematical tool, it can be used to set up state equations, algebraic equations, simulations and other mathematical models. PN models FMS in terms of places, which represent the state of resources or parts in the system and transitions, which represent events or activities in the system. PN models can be analyzed either by invariant analysis or simulation. Invariant analysis is a formal method of analyzing system properties such as liveness and bounded ness while simulation analyzes system model by generating actual behavior of the FMS. From operational decisions viewpoint, PN is having following advantages over other modeling approaches (such as markov chains and computer simulation etc) [22] :
doi:10.2507/ijsimm05(3)2.064 fatcat:tnvlyamgnrbw3bjyhbgusucsm4