Abstract: In order to determine the structural layout of oil and gas gathering pipeline network reasonably, well group optimization model was established with the well-to-station yield distance sum as the objective function, and the well-to-station connection relation and the location of metering station as the optimization variables. And then it was solved by the little-window ant colony algorithm. The well-to-station connection relation was converted into the routing selection. Heuristic factor was calculated using the yield and distance of each pipe section, the yield distance sum corresponding to various routing schemes is taken as the evaluation index of information accumulation. Constraint conditions (e.g. well pattern, gathering radius, treatment capacity of metering station) are satisfied by controlling the state transition process of ants so as to avoid the infeasible solution. The example calculation result shows that the optimization result derived from ant colony algorithm is better than that from the existing genetic algorithm in terms of yield distance of pipeline network and total length of pipeline. It can provide technical support for the optimization design of oil and gas gathering pipeline networks in the future.