Abstract: Running optimization of pipeline is a complex integral programming question and it is usually solved through dynamic programming algorism and genetic algorism. However, these algorisms are difficult to be used and the results obtained from which are often deviated far from the true optimal answers. Aiming at resolving the optimal model of pipeline running, the article adopts the PSO algorism to calculate a real pipeline case. The results show that the PSO algorism has the advantages in high precision and fast speed of iteration, which can be used in the study of optimal running of pipelines.