Genetic Algorithm In The Erp Production Planning System, Applications And Research

In the production management of manufacture, the result of the single item incapacitied lot-sizing planning is not very good. The main reason is that the plan is beyond the capacity of product. This problem is considered in the capacitied lot-sizing planning. This is the optimize of assembled which usually be resolved by intelligent optimize algorithm. The simulate annealing, artificial nerve cell network, genetic algorithm, tabu search and the commix of the above algorithm are the main method to resolve the optimize problem. People proved many new ideas in the course of practice. In this thesis we resolved the problem using the multi-level annealing genetic algorithm.Usually the genetic algorithm is the method to resolve the capacitated lot-sizing problem. This algorithm is better than the W-W dynamic optimize and lot-for-lot. It has the merit of the other two. To a resolve of a typical capacitated lot-sizing problem, the produce cost of the best result before is not very good. The ideas of genetic algorithm and simulate annealing are combined in this article. The mechanism of Boltzmann, tactic of immigration and the punish function are used to the genetic algorithm. The best result be remained all the time, so the algorithm could not only get out from the local optimize result but also remain the decline trend in the course of optimizing. We designed a enterprise resource planning system for the company YinGe. In the system, we use the multilevel simulate annealing genetic algorithm to solve the capacitied lot sizing problem, and the result is very good.