Integrated Optimization And Simulation Of Supply Chain Understochastic Demand

As the increasingly fierce market competition,supply chain management has become a main management mode of modern enterprises.Supply chain management is based on reasonable contro,and achieve maximum profit with the minimum costs.In reality,however,customer demand is affected by environment and a herd mentality.There is no any clear discipline.Therefore,in view of the random demand of multi-stage supply chain inventory management problems,design efficient inventory management model is very significant.In order to control the optimization of inventory cost,this paper uses genetic algorithm and BP neural network simulation and optimize inventory to effectively reduce the inventory cost.First,this paper introduces a kind of random demand based on genetic algorithm under the environment of supply chain inventory control method.This section of the traditional genetic algorithm is improved,and in the crossover operation adopted more pay 245 fork,a randomized crossover probability.The algorithm could make individual two cross not repeat,when two individuals meet after the crossover probability that each gene for unconditional cross.This improves the ability of search the offspring of high fitness.Mutation in each subsequent generation are adopted to decrease the mutation rate,the method of random mutation rates set.Genetic algorithm is used in this section,to strengthen the global search and avoid premature phenomenon.This method can be used in more than one under the influence of uncertain factors,the optimal portfolio choice.Assessment criteria,the more was the combination of precise.At the same time,on the basis of considering the cost and the shortage of punishment,the algorithm improved obviously the inventory control.Secondly,this paper use neural network applying to the random demand of inventory control of supply chain optimization.This paper applies BP neural network to supply chain inventory control optimization under stochastic demand.And uses the BP neural network method to the supermarket within a period of time the sales record for the sample data,the analysis process of training neural network model for inventory control of BP,and verify that the BP neural network adaptive ability,fault tolerance and nonlinear relations ability,ensure the accuracy of inventory forecast.Finally,we proposed the BP neural network algorithm for the inventory of goods based on the control model.The results show that the control model can accurately and efficiently control the supermarket commodity inventory,can provide decision support for reasonable control of inventory,and effectively improve the efficiency of inventory control.