461232 Supply Chain Planning and Scheduling Approach for Multiproduct Multistage Continuous Plants under Uncertainty

Monday, November 14, 2016: 1:08 PM
Carmel II (Hotel Nikko San Francisco)
Adrian M. Aguirre1, Songsong Liu2 and Lazaros G. Papageorgiou1, (1)Centre for Process Systems Engineering, Centre for Process Systems Engineering, University College London, London, United Kingdom, (2)School of Management, Swansea University, Swansea, United Kingdom

This paper presents a novel MILP-based (Mixed Integer Linear Programming) approach for the planning and scheduling of multiproduct multistage continuous plants in a supply chain network with product demands and price elasticity under uncertainty. For this, we consider a supply chain network where multiple echelon nodes, such as, plants, distribution centers (DCs) and markets, are interconnected to satisfy those customer’s demands (Liu et al., 2012). In here, multiproduct continuous plants are in charge to produce several products in an efficient way to supply the DCs, while DCs are in charge to store and deliver these products to the final markets. In general continuous plants, several products must be processed in simultaneous by following a series of production stages. In this particular problem we are considering multiproduct multistage continuous plants, with a single production unit per stage where sequence-dependent changeovers occur in each processing unit between different products. Another important issue of the real supply chain networks is the uncertain behaviour of the system. In this work, two uncertain parameters, the demand and the price elasticity, are taken into account. In this context, planning decisions must be done across the supply chain to deal with the production, storage and delivery issues of the final products. 

A hybrid discrete/continuous precedence-based model was proposed for this problem by using the main ideas of the Traveling Salesman Problem (TSP) and global precedence representation (Mendez et al., 2006). For this, the original TSP of Liu et al. (2008, 2009) was reformulated to easily tackle changeover issues without generating additional binary variables. In order to deal with the uncertainty, we adapt the Model Predictive Control (MPC) approach proposed in Liu et al. (2012) for this particular problem. Despite of the efficiency of this solution technique, the final solution reported by this method could be far from the global optimal when many of the decision variables are successively fixed iteration by iteration. Due to this, a Local-Search (LS) approach was developed to improve the solution of the MPC by rescheduling successive products from the current schedule (Castro et al., 2011). The effectiveness of this alternative solution technique has been demonstrated by solving large scale instances of the problem.

References:

  1. Castro, P. M., Harjunkoski, I., & Grossmann, I. E. (2011). Greedy algorithm for scheduling batch plants with sequence‐dependent changeovers. AIChE journal, 57(2), 373-387.

  2. Liu, S., Pinto, J. M., & Papageorgiou, L. G. (2008). A TSP-based MILP model for medium-term planning of single-stage continuous multiproduct plants. Industrial & Engineering Chemistry Research, 47(20), 7733-7743.

  3. Liu, S., Pinto, J. M., & Papageorgiou, L. G. (2009, June). Medium-term planning of multistage multiproduct continuous plants using mixed integer optimisation. In 19th European Symposium on Computer Aided Process Engineering: ESCAPE-19: June 14-17, 2009, Cracow, Poland (Vol. 26, p. 393). Elsevier.

  4. Liu, S., Shah, N., & Papageorgiou, L. G. (2012). Multiechelon supply chain planning with sequence‐dependent changeovers and price elasticity of demand under uncertainty. AIChE Journal, 58(11), 3390-3403.

  5. Méndez, C. A., Cerdá, J., Grossmann, I. E., Harjunkoski, I., & Fahl, M. (2006). State-of-the-art review of optimization methods for short-term scheduling of batch processes. Computers & Chemical Engineering, 30(6), 913-946.

 


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