A low carbon operation optimization model for natural gas pipeline networks is proposed based on different energy structures to reduce carbon emissions. SSA (sparrow search algorithm) , WOA (whale optimization algorithm) , and AHA (artificial hummingbird algorithm) were compared using test functions. Results show that AHA has better optimization performance. The Shanxi-Beijing natural gas pipeline network was taken as an example. A carbon emission optimization model with the goal of minimizing carbon emissions was established. AHA was used to solve the model. Results show as follows. The energy consumption optimization scheme (objective 1 scheme) tends to use natural gas, and the carbon emission optimization scheme (objective 2 scheme) tends to use electricity. The objective 1 scheme reduces energy consumption by 21. 75% , while the objective 2 scheme reduces energy consumption by 21. 32% . The objective 1 scheme reduces carbon emissions by 21. 18% , and the objective 2 scheme reduces carbon emissions by 24. 06% . Therefore, the carbon emission optimization scheme minimizes carbon emissions from the pipeline network but cannot minimize energy consumption. In conclusion, the optimization model established in this paper can realize the functions of pipe network energy consumption optimization and carbon emission optimization and help to build a smart pipe network.