The capacitated dynamic lot size problem with variable technology

Abstract

Traditionally, the technological coefficients in production models were assumed to be fixed. In recent years however, researchers have used the learning curve model to represent nonlinear technological coefficients, and the dynamic lot-sizing problem with learning in setups has received attention. This article extends the research to consider capacity restrictions in the single-level, multi-item case. The research has two goals, first, to analyze the effects of setup learning on a production schedule, and second, to investigate efficient ways of solving the resulting nonlinear integer model. Previously derived algorithms do not address the issue of capacity; thus a heuristic is developed and its solution is compared with the optimal solution, where possible, or to a lower bound solution.