Study On Optimal Control Of Natural Gas Flowmeter Calibration Station

With the rapid increase of natural gas trade volume and consumption,flowmeters are more and more used in metering handover trade,the task of flowmeter calibration is aggravated,and domestic natural gas metering calibration institutions are running at high load.The establishment of the Pipe China,the use of X+1+X business model,the separation of upstream and downstream,the number of metering handover will be further increased.The increasing calibration task puts forward higher requirements for the production efficiency and control optimization of the natural gas flowmeter calibration station,which requires the natural gas flowmeter calibration station to introduce intelligent technology to improve the production level.According to the process characteristics of natural gas flowmeter calibration station,the hydraulic characteristics of natural gas flowmeter calibration station are analyzed in this paper.Making full use of the pipeline operation data collected by SCADA system,this paper puts forward the on-line data usage scheme,the process pipeline topology simplification scheme of natural gas flowmeter calibration station and the valve flow coefficient correction algorithm.Through the equivalent and simplified simulation model generation mechanism model control scheme of the natural gas flowmeter calibration station,the qualitative and quantitative optimal control scheme for the calibration process is given.Furthermore,based on SCADA data,a BP neural network model for calibrating flow regulation is constructed,and K-FOLD cross-validation algorithm and genetic algorithm are used to optimize the BP neural network control model.Then,according to the model predictive control theory,the self-adaptability of the control algorithm is improved to form a control closed loop.For the mechanism model,the closed-loop control effect is achieved through open-loop optimization;for the BP neural network model,the whole calibration process is closed-loop through the constructed control logic.Finally,the control algorithm test platform of natural gas flowmeter calibration station is constructed by using hydraulic simulation software SPS.Combined with the actual situation of Nan Jing branch of national station of petroleum & natural gasflow measurement,the control scheme is deployed.The results show that the combination of mechanism control model and neural network control model can realize the accurate control of the field calibration process and improve the calibration efficiency.It can be used as a reference for the intelligence and application of natural gas flowmeter calibration station.