267943 Supply Chain Management Optimization Using a Hybrid Simulation Based Optimization Approach

Thursday, November 1, 2012: 12:30 PM
325 (Convention Center )
Fnu Nihar and Marianthi G. Ierapetritou, Department of Chemical and Biochemical Engineering, Rutgers, The State University of New Jersey, Piscataway, NJ

Supply Chain Management Optimization using a Hybrid Simulation based Optimization Approach

 

Nihar, Marianthi Ierapetritou

Department of Chemical and Biochemical Engineering, Rutgers University, Piscataway, NJ

A supply chain is a network of suppliers, production facilities, warehouses and markets designed to acquire raw materials, manufacture and store and distribute products among the markets. The entire process is driven by the demand generated at the markets. Since there are different entities involved and they work towards their individual interests, the optimum performance of the whole network is often not achieved. It has been shown that greater efficiency and reduced costs can be achieved through proper coordination among the entities in terms of material, financial and information flow [1, 2].

In order to gain competitive edge in a global economy, companies aim at optimizing their overall supply chain. In a typical supply chain, there are a large number of entities. The interacting behavior of these entities among themselves is complex. This makes the problem of optimization large and complex. Mathematical modeling has been very widely used to formulate these problems. Several mathematical models use mixed integer linear programming (MIP) and mixed integer non-linear programming to solve the SC optimization problems [3, 4]. Stochastic models have also been developed for supply chain networks in uncertain environments [5]. However, these approaches lack realism since they are not able to capture the complex relations among the different entities comprising the network. Simulation models are another category of models that have been used to solve such problems. They help mimic complex relations among the different entities of the supply chain. Agent based models, a class of simulation models, treat the different entities as autonomous agents, each having its own set of behavioral rules. It has been shown that using this approach, it is possible to govern the interaction among the different entities of the whole network and thus reach at improved solutions [6]. Compared to the optimization based methods, limited work has been done using this approach.

Analytic and simulation models can be regarded as the two extremes of a range of mathematical models available for modeling [7]. Hybrid simulation/analytic models which are combinations of these two types of models enable us to take advantages of both of them and can be very useful [8, 9]. Such models are useful for stochastic optimization as well since they can mimic real systems including their stochastic and nonlinear behaviors. In the recent years, such models have been used to explore the supply chain domain. It is an active area of research and rapid progress is being made.

In this work, we demonstrate the use of a hybrid modeling approach to solve a small scale supply chain management problem. The problem has been formulated as a mathematical model minimizing the overall cost incurred over a number of planning periods. The objective of using the hybrid approach is to overcome the computational complexity involved in solving the large scale mixed integer nonlinear problem (MINLP) and to obtain a solution which represents supply chain reality more closely. The proposed framework involves the integration of the solution strategies of separate optimization and simulation models. The optimization model solves the planning and scheduling problem while the simulation model is used to verify the quality of the solution obtained by comparing it with real world scenarios. An iterative solution procedure has been used such that there is communication between the two models in terms of some decision variables in every cycle. The optimization model of the problem has been developed in GAMS while the simulation model has been developed using the JAVA Repast tool. Various scenarios involving different number of planning periods and different sets of deterministic demand have been used to investigate the proposed framework. The effectiveness of the approach has been shown in terms of the computational effort required and the quality of the solution achieved.

Considering the increasing environmental concerns and associated legislations these days, the proposed framework has been extended to account for environmental impact through a multi-objective optimization approach.  In addition to the economic objective which minimizes the overall operation cost of the supply chain, an environmental objective which minimizes the environmental cost associated with the manufacturing and transportation processes of the network is also considered. The consideration of an environmental objective at the stage of supply chain management is critical in order to make a holistic assessment. Incorporating the environmental aspects at a later stage might lead to decreasing the environmental impact locally while causing an overall increase in the environmental costs. The simulation model is able to assess the various environmental aspects of the supply chain and decide whether the solution provided by the optimization model meets the environmental requirements or not. The presence of the two objective functions helps maintain the environmental performance of the supply chain while minimizing the overall cost.

References:

1.            Stadtler, H., Supply chain management and advanced planning--basics, overview and challenges. European Journal of Operational Research, 2005. 163(3): p. 575-588.

2.            Varma, V.A., et al., Enterprise-wide modeling & optimization--An overview of emerging research challenges and opportunities. Computers & Chemical Engineering, 2007. 31(5-6): p. 692-711.

3.            Li, Z. and M.G. Ierapetritou, Production planning and scheduling integration through augmented Lagrangian optimization. Computers & Chemical Engineering, 2010. 34(6): p. 996-1006.

4.            Timpe, C.H. and J. Kallrath, Optimal planning in large multi-site production networks. European Journal of Operational Research, 2000. 126(2): p. 422-435.

5.            Petrovic, D., R. Roy, and R. Petrovic, Modelling and simulation of a supply chain in an uncertain environment. European Journal of Operational Research, 1998. 109(2): p. 299-309.

6.            Behdani, B., et al., Agent based model for performance analysis of a global chemical supply chain during normal and abnormal situations, in Computer Aided Chemical Engineering. 2009, Elsevier. p. 979-984.

7.            Shanthikumar, J.G. and R.G. Sargent, A Unifying View of Hybrid Simulation/Analytic Models and Modeling. Operations Research, 1983. 31(6): p. 1030-1052.

8.            Lee, Y.H., S.H. Kim, and C. Moon, Production-distribution planning in supply chain using a hybrid approach. Production Planning & Control, 2002. 13(1): p. 35-46.

9.            Young Hae, L. and K. Sook Han. Optimal production-distribution planning in supply chain management using a hybrid simulation-analytic approach. in Simulation Conference, 2000. Proceedings. Winter. 2000.


Extended Abstract: File Not Uploaded
See more of this Session: Supply Chain Optimization I
See more of this Group/Topical: Computing and Systems Technology Division