| Simulated annealing genetic algorithm(SAGA) is formed by the genetic algorithm(GA) and the simulated annealing algorithm(SA), which combines the advantages of both of them. It not only effectively prevents the premature convergence of GA, but also improves the search capability of SA, thus providing faster convergence speed. Since the development of SAGA, it has been widely applied to various fields.Improving the fitness values, the crossover and mutation operators to make crossover probability and mutation probability adjusted adaptively with the change of fitness value, which is called improved SAGA and used to solve the premature convergence and slow convergence problems. Experiments show that the improved SAGA is effective in improving the convergence and search capability.Inventory is the basis for maintaining the efficient operation of the enterprise, as well as a lifelong project. Having a good inventory level can be useful for reducing the risk of losing profits. In recent years, electronic products with short life cycle — such as mobile phone, update more and more quickly and user demands are highly fluctuating, the increase of uncertainty leads to a greater risk of the enterprise. Consequently, it raises higher requirements and challenges for the inventory management of the enterprise. Demand forecasting is the key technology of maintaining a good inventory levels. In this paper, multivariate gray model(MGM(1, N)) is applied to forecasting.The defects of MGM(1, N) in initial conditions, background values and 1-AGO are corrected and the improved SAGA is applied to optimize the parameters of the amendment MGM(1, N). Experiments show that the model accuracy is improved.Based on an enterprise of products, this paper researches the inventory optimization model in a replenishment cycle. The model is established by the lowest cost and maximum average profit, and the improved SAGA is used to solve, analyze and optimize the model. |
PDF Full Text Request