Research On Control And Application Of Automatic Distribution System Of Natural Gas Pipeline

With the advent of the new century,it was owing to the rapid developement of the economy and the continuous improvements with regard to the quality of life that natural gas long-distance pipelines,which are viewed as energy arteries,have played an irreplaceable in ensuring the people's production and living needs.Given the continuous increases in gas consumption and the successive construction of various sub-transmission stations,traditional methods of manual distribution can no longer meet the actual demands,and new requirements are put forward for natural gas sub-transmission.In this article,accordind to the current situation of natural gas pipeline's distribution systems and the characteristics of different users,a method with respect to an automatic distribution control,which is aimed to meet the needs of users,is investigated,in order to solve the problem of precise distribution and efficient control of natural gas pipelines.And the main tasks and contents are summarized in the following three parts.(1)With regard to the problems that it is difficult to establish the effective mechanism model belonging to the regulating valves in the processes of natural gas transmission.According to massive empirical datas in the processes of distribution,the accuracy of control,and the different needs of control required for users in the processes of various types of distribution,an PSO-SVM data-driven model which is provided for the control valves in the processes of distribution is built.Constrasted with the test samples,the effectiveness of the data-driven model and algorithm is testified.(2)With regard to the problems which is involved with the complicated logic,incompatibility and low control effect for the various types of distribution between users.In this article,on the basis of the data-driven model which is associated with the pressure and flow rate of the control valves existing in the distribution of different users,the expert cascade PID control strategy is utilized and a variety of expert modes and regional expert rules are devised for various actual conditions in the processes of distribution.According to the operational results,the adjustment accuracy concerning the pressure or flow rate of various types of distribution between users is effectively improved,and the intelligent start-up,precise distribution and quantitative shutdown of the whole-process distribution station are achieved through the control strategy.(3)The optimizational software platform is investigated and devised by the data-driven model and control methods in the processes of automatic distribution.The sofeware platform mainly consists of some interfaces in regard to databases,model libraries,the management of model parameters,the settings,monitoring and setting parameters in the control processes and so on.The design and implementation of the software platform provide a scientific basis for the optimization of model parameters and efficient control concerning the processes of distribution.