改进神经网络在粮油配送中心选址中的应用

Application of improved Neural Network in grain and oil distribution center location

在建立以控制成本为核心的粮油配送中心选址模型的基础上,分析了选址过程中的影响因素;结合人工神经网络与选址影响因素之间的特点,引入了经遗传算法优化的神经网络模型。为解决遗传算法易陷入局部最优的问题,提出了一种在变异算子中增加禁忌表的方法。实例求解表明,与专家评审的选址决策相比,该算法能够有效解决粮油配送中心的选址问题。

On the basis of building a modle of costs control in the location of grain and oil distribution center,this paper analyzes the affecting factors of the location process,combining with the characteristics of the Artificial Neural Network and the impact factors of location,it introduces a Neural Network model by the genetic algorithm optimization.In order to solve the problem of genetic algorithm in local optimum easily,this paper increases a tabu search in a mutation method.The example proves that it can solve the problems of grain and oil distribution center location,comparing with the expert assessment of the location decision-making methods.