Inverse Problem of Oil Pipeline Dynamic Operation Based on Model-Free Adaptive Control Theory

Abstract

The purpose of this paper is to solve the inverse problem of the dynamic operation of dispatchers in oil pipeline system based on the real-time system process parameters. To realize online solution of system dynamic operation inverse problem, model free adaptive control theory is introduced for the first time. In particular, the frequency domain analysis method is adopted to prove that the oil pipeline system meets assumptions of the control theory. The dynamic state model suitable for solving the control process of the pipeline system is established by using the field data, and the inverse problem of the controlled components dynamic process is solved by using the model free adaptive control method. By analyzing the characteristics of pipeline flow process and comparing different linearized models, the full from dynamic linearization method is selected to model the system state. Single-input-single-output system model and Multi-Input-Single-Output system model are adopted to solve the single station and multiple stations joint operation process respectively. This method is verified with SCADA data of a real pipeline system. The results show that the proposed method can realize the online twining of valve dynamic operation process, and improve the accurate of system state parameters estimation during dynamic operation.