| After the establishment of China Oil & Gas Pipeline Network Corporation,the national gas pipeline network was able to be dispatched in a unified manner,showing the characteristics of multiple sources,multiple users,multiple paths and large scale,and even a complex structure of interweaving between ring structures.Optimization techniques for pipeline networks with linear and branch topologies are relatively mature,but these methods cannot be directly applied to the problem of optimizing the operation scheme of that with cyclic topologies.In this thesis,in order to minimize the energy consumption and energy cost,the flow distribution and process operation scheme of gas pipeline networks with cyclic topologies are optimized,which is important to improve the economic benefits of gas transmission in the networks and reduce the operational energy consumption.In this thesis,the MINLP model is developed for flow distribution and process operation scheme optimization of gas pipeline network with the number of compressors operating at the compressor stations,the outlet pressure of the compressor stations,and the flow of the pipelines as decision variables,"minimum energy consumption" or "minimum energy cost" as objective function,and large-scale linear and nonlinear constraints such as hydraulic and thermal characteristics of pipeline,flow balance of nodes,and compressor characteristics are taken into account.The model is general and can be applied to large scale complex gas pipeline network of arbitrary topology.It can accurately describe the problem,and it can take into account both flow distribution scheme optimization and process operation scheme optimization without splitting the problem level.Since there are integer variables in the model,three hybrid algorithms are designed in this thesis,including Enumeration-IP Algorithm,SA-IP Algorithm and B&B-IP Algorithm.The optimization procedures are written using the Julia programming language and the Ju MP mathematical modeling language.The optimization effects of these three algorithms are compared and analyzed in terms of computational time and solution accuracy,and the best algorithm is determined to be B&B-IP Algorithm.In this thesis,the optimization results of the self-programmed program are verified using the simulation software,named TGNET.The results show that the average value of the relative error between the optimization results of the self-programmed program and the software simulation results is within 3% in absolute value.In this thesis,it is found that: 1)when the flow of pipeline network is low,it is appropriate to adopt the operation mode of "high outlet pressure and fewer compressor stations operating at high pressure ratio";when the flow of pipeline network is high,it is appropriate to adopt the operation mode of "high outlet pressure and more compressor stations operating at low pressure ratio".2)In order to minimize the equivalent energy consumption,since the electric motor is significantly more efficient than the gas turbine,the gas should be distributed as much as possible to the compressors driven by the electric motor.3)In order to minimize the primary energy consumption or the energy cost,since the difference in the impact produced by the driver is not obvious,attention should be paid to the fit of the compressors in the pipeline network.If the difference in the range of inlet flow in the high efficiency area of compressors has a greater impact on the gas distribution scheme than the driver difference,more gas should be allocated to the compressors with the higher inlet flow in the high efficiency area as much as possible,so that the compressors in the pipeline network can operate at the more efficient operating point.In this thesis,mathematical model and algorithm for flow distribution and process operation optimization of large-scale complex natural gas pipeline networks are proposed,and the optimization laws of flow distribution of gas pipeline networks are analyzed and summarized,which have certain guiding significance for engineering practice. |