Multi-objective and multi-period optimization of a regional timber supply network with uncertainty

To assess the impacts of uncertainty and environmental objectives on the configuration of timber supply networks, we develop a generic multi-period, mixed-integer fuzzy linear programming model with demand uncertainty and two objectives of minimizing total transportation cost and greenhouse gas (GHG) emissions. We then use the triangular fuzzy number method to define the uncertain demands and convert the model into its equivalent auxiliary crisp counterpart. To derive Pareto solutions more efficiently, we propose the nondominated sorting genetic algorithm (NSGA-II) to solve the model. Finally, we apply the model framework and solution method to a real-world case of regional timber supply in Fujian, China, to demonstrate their applicability. The simulation results of the model show that trade-offs exist between total cost and GHG emissions and that the proper selection of the number and locations of distribution centers can help reduce both the cost and GHG emissions. Demand uncertainty and supply fluctuations across different time periods can increase the cost and GHG emissions. Our empirical results provide useful insights into the design and management of regional timber supply networks, and our generic model is applicable to the analysis of regional supply networks of other products or materials besides timber.