Simulation Analysis Of Time-Delay Electronic Commerce Closed-Loop Supply Chain Based On Neural Network MPC

When the supply chain system operates,there are many uncertainties in the supply chain,especially the time lag in the supply chain,which greatly affects the operation of the supply chain system.Traditional commerce is less affected by time lag because of its fixed structure,and e-commerce is greatly affected by time lag because of its network structure.And due to the frequent outbreak of the epidemic,the impact on the supply chain has deepened,making the supply chain almost paralyzed during the epidemic.The bullwhip effect is used as a measurement index in supply chain management to observe the amplification effect of time delay on the supply chain system.Seek a feasible way to optimize supply chain management,so as to inhibit the bullwhip effect generated in supply chain management,reduce enterprise operating costs and enhance core competitiveness.In order to further reduce the bullwhip effect caused by time lag,neural networks are used to train the historical data of the supply chain,and the complex mapping relationship between historical data is found through the neural network.The traditional model predictive control and neural network are combined to solve the problem of excessive bullwhip effect in the supply chain system,considering that the development of the epidemic presents a short,frequent and fast characteristics,by establishing a dual model method,when the epidemic occurs,the model switches from the normal model to the abnormal model,so as to achieve the purpose of reducing the bullwhip effect.In the SIMULINK environment of MATLAB,the material balance equation and spatial state model are constructed to solve the impact of time lag by analyzing the material information flow in the supply chain system.The dual recycling channel supply chain model,the third-party recycling channel supply chain model and the dual model switching supply chain model were established respectively,and the traditional model predictive control and neural network and model predictive control were combined to simulate the supply chain model at the same time,and the results proved that compared with the traditional model predictive control,the combination of neural network and model predictive control can tend to dynamic stability faster,which can keep the bullwhip effect at a small level.Simulation analysis proves that the combination of neural network and model predictive control can have a good suppression effect in the early stage of supply chain operation.When the epidemic occurred,the model was switched,and the fluctuation of the bullwhip effect was not large,and the control of the bullwhip effect was completed.In the operation of the supply chain system,there are many uncertain factors in the supply chain,especially the time delay in the supply chain,which makes the operation of the supply chain system greatly affected.Traditional commerce is less affected by time delay because of its fixed structure,while e-commerce is greatly affected by time delay because of its network structure crisscross.Moreover,due to the frequent outbreaks of the epidemic,the impact on the supply chain was deepened and the supply chain was nearly paralyzed during the epidemic.Bullwhip effect is used as a measurement index in supply chain management to observe the amplification effect of time delay on supply chain system.Seek a feasible method to optimize the supply chain management,to restrain the bullwhip effect in supply chain management,to reduce the operation cost of enterprises and improve the core competitiveness.