基于混沌时间序列的潜油柱塞泵动液面预测研究

Study of Forecasting Producing Fluid Level of Submersible Reciprocating Pump on the Basis of Chaotic Time Series

摘要: 依托嵌入式计算机系统及自动控制技术，潜油柱塞泵在油田原油生产过程中得到大范围普及应用，这极大地提高了原油的开采效率。潜油柱塞泵的核心是位于井下的潜油直线电机，地面通过计算机控制井下电机的冲次、冲程等运行状态，使其与油井油藏状态相适应，从而在原油生产过程中达到高效、节能的目的。目前潜油柱塞泵的冲次控制方法是建立在动液面时间序列预测和计算的基础上的，本文介绍一种基于混沌时间序列的动液面预测计算方法，并与线性预测方法进行对比分析，最终得到能够满足潜油柱塞泵冲次优化控制的动液面预测方法。本文的研究内容为潜油柱塞泵的计算机控制提供了理论和应用依据。

Abstract: Relying on the embedded computer system and automatic control technology, the submersible plunger pump is widely used in crude oil production, which improves the efficiency of crude oil extraction greatly. The core of the submersible plunger pump is the submersible linear motor which is located underground. The operation state, such as the impulse and the stroke, of the motor is controlled by computer on the ground, which can adapt the computer to the state of the oil wells to achieve higher efficiency and save more energy in the production of crude oil. Currently, the impulse control of the submersible plunger pump is based on the prediction and calculation of the time series of dynamic liquid surface. Our paper presents a method for predicting dynamic liquid surface on the basis of chaotic time series and analyzes its advantages compared with the linear prediction method. By this way, the dynamic liquid surface prediction method can be obtained to optimally control the submersible plunger pump impulse. Our research provides the theoretical and practical basis for the computer control of the submersible plunger pump.