Abstract: In order to help cold chain food production enterprises to select the best cold chain logistics service providers quickly, a rough PSO-BP neural network model was constructed by integrating rough set and particle swarm algorithm on the basis of traditional BP neural network. The model uses rough set to eliminate the redundant information in the original data and make the input index more compact. Particle swarm optimization is used instead of gradient descent to train the weights of the neural network, so that the output results are not easily trapped in local minimal and the generalization ability of the network is enhanced. Finally, an example is given to verify the validity and feasibility of the model. The results show that while improving the operation speed, the prediction error of the model is 40.94% of the BP neural network model, and the prediction result is more accurate and reliable, which provides a new method for the cold chain food production enterprises to quickly select the best cold chain logistics service provider.

Key words: BP neural network, choice of cold chain logistics provider, rough set, particle swarm optimization