Meta-Heuristics for Dynamic Lot Sizing: a review and comparison of solution approaches

摘要: Proofs from complexity theory as well computational experiments indicate that most lot sizing problems are hard to solve. Because these so difficult, various solution techniques have been proposed solve them. In the past decade, meta-heuristics such tabu search, genetic algorithms and simulated annealing, become popular efficient tools for solving combinatorial optimization problems. We review specifically developed problems, discussing their main components representation, evaluation, neighborhood definition operators. Further, we briefly other approaches, dynamic programming, cutting planes, Dantzig-Wolfe decomposition, Lagrange relaxation dedicated heuristics. This allows us compare techniques. Understanding respective advantages disadvantages gives insight into how can integrate elements several approaches more powerful hybrid algorithms. Finally, discuss general guidelines illustrate with examples.