Switching Model of Multi-Channel Closed-Loop Supply Chain with Time Delay and Neural Network Predictive Control Method

A solution combining a neural network and model predictive control is proposed for the problem of supply chain switching under the influence of an epidemic that causes uncertainty in the supply chain and leads to inventory instability in companies. In this solution, the neural network uses the Neural Net Fitting toolbox in MATLAB to fit the past data. The model predictive control is to predict the data afterward. The supply chain is modeled in the SIMULINK and simulates sinusoidal and stochastic consumer demand. Experimental results show that the model can switch quickly when an epidemic occurs, demonstrating the effectiveness of combining neural networks with model predictive control.