Energy-Saving for a Velocity Control System of a Pipe Isolation Tool Based on a Reinforcement Learning Method

다국어 초록 (Multilingual Abstract)

The pipe isolation tool (PIT) demonstrates remarkable advantages in safety and efficiency compared with traditional plugging devices. However, its utilization in plugging operations is limited by the operation duration. In addition, the existing energy recovery system has low energy saving efficiency. In this paper, a real-time control energy-saving system of the PIT was designed based on a reinforcement learning algorithm. First, an experimental device for energy-saving was designed. Secondly, the energy distribution scheme of a hydraulic pump and accumulator based on experimental data was proposed. Finally, the reinforcement learning algorithm was used to adjust the opening of the hydraulic pump and the accumulator valves in real time during the plugging process to improving energy saving efficiency. The results verify that the energy saving efficiency of the PIT control system based on reinforcement learning could reach 23.71%, which satisfies the objectives of energy-saving and environmental applicability.

참고문헌 (Reference)

1 Liang, Z., "Speed simulation of bypass hole PIG with a brake unit in liquid pipe" 42 : 40-47, 2017

2 Mirshamsi, M., "Speed control of inspection pig in gas pipelines using sliding mode control" 77 : 134-140, 2019

3 Edd Tveit, J. A., "Remote controlled (Tether-Less) high pressure isolation system" 2000

4 Tveit, E., "Remote Controlled(Tether-Less)high pressure isolation System" Society of Petroleum Engineers 8 : 2000

5 Hao, Y., "Potential energy directly conversion and utilization methods used for heavy duty lifting machinery" 155 : 242-251, 2018

6 Shen, P., "Optimal energy management strategy for a plug-in hybrid electric commercial vehicle based on velocity prediction" 155 : 838-852, 2018

7 Chiriboga J, T. M. W. L., "Input-output feedback linearization control of a load-sensing hydraulic servo system" IEEE 2002

8 Kong, Z., "Implementation of real-time energy management strategy based on reinforcement learning for hybrid electric vehicles and simulation validation" 12 (12): e0180491-, 2017

9 Hendrix, M. H. W., "Experiments and modeling of by-pass pigging under low-pressure conditions" 71 : 1-13, 2018

10 Zhu, X., "Experimental study on dynamics of rotatable bypass-valve in speed control pig in gas pipeline" 47 : 686-692, 2014

11 Xiang, C., "Energy management of a dualmode power-split hybrid electric vehicle based on velocity prediction and nonlinear model predictive control" 189 : 640-653, 2017

12 Nguyen, T. T., "Dynamic modeling and its analysis for PIG flow through curved section in natural gas pipeline" IEEE 2001

13 Tolmasquim, S. T., "Design and control of pig operations through pipelines" 62 (62): 102-110, 2008

14 Qi, X., "Deep reinforcement learning enabled self-learning control for energy efficient driving" 99 : 67-81, 2019

15 Xiangjing, L., "Control strategy of energy saving for power system in hydraulic surface drilling rig" 40 (40): 282-, 2018

16 Wu, J., "Continuous reinforcement learning of energy management with deep Q network for a power split hybrid electric bus" 222 : 799-811, 2018

17 El-Abbasy, M. S., "Artificial neural network models for predicting condition of offshore oil and gas pipelines" 45 : 50-65, 2014

18 Li, L., "Analysis of a novel energy-effi-cient system with double-actuator for hydraulic press" 47 : 77-87, 2017

19 Li, L., "An energy-saving method to solve the mismatch between installed and demanded power in hydraulic press" 139 : 636-645, 2016

20 Wu, T., "An energy-saving and velocity-tracking control design for the pipe isolation tool" 11 (11): 168781401984594-, 2019

21 Dey, P. K., "A risk-based model for inspection and maintenance of cross-country petroleum pipeline" 7 (7): 25-43, 2001