An uncertain furniture production planning problem with cumulative service levels

To investigate how the loss averse customer's psychological satisfaction affects the company's furniture production planning, we establish a furniture production planning model under uncertain environment, where customer demand and production costs are characterized by mutually independent uncertain variables. Based on prospect theory, customer's psychological satisfaction about stockout performance is measured by cumulative service levels in our model. In the framework of uncertainty theory, the proposed uncertain model can be transformed into an equivalent deterministic form. However, the transformed model is a nonlinear mixed integer programming problem, which cannot be solved by conventional optimization algorithms. To cope with this difficulty, a chemical reaction optimization algorithm integrated with LINGO software is designed to solve the proposed production planning problem. In order to verify the effectiveness of the designed hybrid chemical reaction optimization (CRO) algorithm, we conduct several numerical experiments via an application example and compare with a spanning tree-based genetic algorithm (hst-GA). The computational results show that our proposed CRO algorithm achieves better performance than hst-GA, and the results also provide several interesting managerial insights in production planning problems.