A study on production and distribution planning problems using hybrid genetic algorithm

Rapid development in computer and network technology these days has created an environment in which decisions for manufacturing companies can be made in a much broader perspective. Especially, better decisions on production and distribution planning (PDP) problems can be made taking advantage of real time information from all the parties concerned. However, since the PDP problem – a core part of the supply chain management – is known to be the so-called NP-hard problem, so heuristic methods are dominantly used to find out solutions in a reasonable time. As one of those heuristic techniques, many previous studies considered genetic algorithms. A standard genetic algorithm applies rules of reproduction, gene crossover, and mutation to the pseudo-organisms so the organisms can pass along beneficial and survival-enhancing traits to a new generation. When it comes to representing a chromosome on the problem, it is hard to guarantee an evolution of solutions through classic algorithm operations alone, for there exists a strong epitasis among genes. To resolve this problem, we propose a hybrid genetic algorithm based on Silver–Meal heuristic. Using IMS-TB (Intelligent Manufacturing System Test-bed) problem sets, the good performance of the proposed algorithm is demonstrated.