Research on Parameter Tuning and Optimization Method for Natural Gas Gathering and Transportation Systems Based on Least Squares

As a core surface facility in natural gas development, the operational efficiency of a gas gathering and transportation system depends largely on the precise configuration of its parameters. However, the complex structure of such a system and the large number of components make real-time data acquisition and accurate monitoring difficult under all operating conditions, so numerical simulation has become the standard way to represent its behavior. At present, the input parameters of professional simulation software still rely heavily on manual empirical settings, and parameter accuracy is therefore difficult to guarantee. If the manually set parameters deviate significantly from the actual operating conditions, the reliability of the simulation results is seriously compromised and subsequent decisions may be misleading. To address this problem, the present work uses Pipesim as the simulation platform and formulates parameter calibration as a weighted least-squares optimization problem. For a network with a fixed topology, the key component parameters are calibrated jointly. The objective is to minimize the deviation between the simulated and the field-measured values of the selected operating variables; pressure and flow rate are used in the case study, while the temperature term is retained in the general formulation. The Pipesim solver is proprietary and does not expose analytical derivatives of its outputs with respect to the calibrated parameters, so the model is solved with particle swarm optimization (PSO) combined with a continuation-projection encoding that handles continuous, discrete-gear and Boolean parameters. On the network examined here, the proposed procedure reduces the calibration effort and outperforms manual tuning in both accuracy and time. It therefore provides a practical reference for the design and operation management of gas gathering and transportation systems.