Fully automated controller for the real-flow calibration of natural gas flowmeters based

Abstract

With the rapid development of the natural gas industry, the natural gas measurement handover volume is increasing year by year, and the calibration of natural gas flowmeter is becoming more and more important. This paper defines and demonstrates the framework of a fully automated controller for natural gas flowmeter calibration stations, from conceptualization to field test. When the calibrated flowmeter is installed on bench, the controller will identify the corresponding process by success tree algorithm. It checks whether the current process is appropriate, and automatically switch the best process according to the diameter of the calibrated meter. In the regulation stage of different calibrated flow rates, the controller performs predictive control on different flow rates according to the actual hydraulic condition by the calculation of hybrid driven model based on neural network model and mechanism model. This process gives the initial position of regulating valves with fast and efficient movement. Under a mild PID control law, the actual flow will go to the setting flow rates neither overshoot nor fall back. Finally, the control scheme is demonstrated in three original case studies, one in the simulation environment built by hydraulic simulation software, and the other two in real systems with different complexity. The algorithm is verified by simulation software and the control scheme is implemented in the actual environment. The results show that the intelligent calibration process based on model predictive control can increase the calibration efficiency. The integrated system based on hybrid driven can promote the intelligent development of natural gas calibration stations.