Dynamic relationship network and international management of enterprise supply chain by particle swarm optimization algorithm under deep learning

Abstract The traditional enterprise decision evaluation model based on neural network has the problems of mismatch with the optimal solution and slow convergence speed. In order to enable companies to make decisions that are in line with changes in the market, the particle swarm optimization (PSO) algorithm is used to optimize deep learning neural networks. Firstly, the model parameter setting is improved, and the inertia weight strategy of normal distribution attenuation is combined. On this basis, a normal distribution decay inertial weight particle swarm optimization (NDPSO) is proposed. The inertia weight of the optimized algorithm maintains a large value in the initial stage, which makes the PSO algorithm maintain a large step size in the optimization process and a small value in the later stage. Through experimental analysis, the trend parameter of the best normal distribution of the algorithm is obtained as 0.4433 and then using the detection function, the NDPSO algorithm is tested by two types of test functions. The NDPSO algorithm is compared with the optimization results of other algorithms which are optimized on the Sphere function. The minimum value of 554.29, the average value of 2032.11, and the standard deviation of 918.47, all of them are at the leading level. Taking into account other experimental results, it is proved that the normal distribution decay inertia weight can balance the global search and local development capabilities from the perspective of parameter improvement. It can speed up the convergence with ensuring the convergence accuracy. The improved PSO algorithm has certain optimization capabilities for neural network models. The use of optimized neural network models can enable companies to make decisions in line with changes in the market and optimize the dynamic relationship network of the company's supply chain, which is of great significance to the implementation of the company's international management.