Correction method for parameters of liquid pipeline simulation model based on particle swarm optimization algorithm

In view of the deviation between the simulation model and the actual pipeline system due to the deviation ofliquid pipeline model parameters, a pipeline model parameter correction method was proposed. At first, a fluid dynamicsmodel was established based on the fluid motion characteristics in the pipeline to analyze the reasons for the inaccuracyof the pipeline model. Then, the influence of the inaccuracy of pipeline parameters such as length, inner diameter, elevationdifference and roughness on the accuracy of the model was analyzed. Thus, the parameter correction problem was transformedinto an optimization problem. A parameter correction model was established, and the optimization problem was solved by particleswarm optimization algorithm. Finally, the accuracy of the corrected results, the adaptability of the corrected model, and theapplicability of the correction method were verified based on the laboratory experiments and simulation data. The verificationresults show that: In terms of accuracy, the average relative error of the parameter correction results based on 26 test data sets iswithin 2%. As for adaptability, after the correction of pipe length and inner diameter, the precision is improved for more than84.6% and 96.1% of the models respectively. In view of applicability, the relative error of the parameter correction results basedon 10 simulation models is within 1%. After parameter correction, the adaptability of the models is improved obviously andthe adaptive rate is reduced from 0.04%-0.53% to 0.01%-0.18%.