论文部分内容阅读
在石油开采过程中,利用直线电机驱动潜油往复泵是一种新型举升方式。这种举升方式可使抽油机具有较好的可控性。在这一背景下,提出一种基于沉没度预测的潜油往复泵冲次优化方法。文中对在采油厂采集的沉没度数据进行时间序列的输入空间重构,基于支持向量机(support vector machines,SVM)建立沉没度预测模型。以抽油机的经济效益为目标,以预测得到的沉没度为参量,对直线电机的冲次进行优化。采用沉没度–冲次子区间匹配方法改进原优化算法,进一步降低原方法的计算量和数据量。改进的优化算法可以更好地适应现场计算机对计算量的限制。该优化方法可有效地提高往复泵的产油量,并避免直线电机长期工作于满载或过载状态。
In the process of oil exploration, the use of linear motor driven submersible reciprocating pump is a new way to lift. This lifting method can make the pumping unit has better controllability. Against this background, a suboptimal reciprocating pump stroke optimization method based on subsidence prediction is proposed. In this paper, the input space reconstruction of the time series of submerged data collected in the oil production plant is carried out. The subsidence prediction model is established based on support vector machines (SVM). The economic benefit of pumping unit is taken as the goal. Based on the predicted submergence, the stroke of linear motor is optimized. The original optimization algorithm is improved by using submergence sub-interval matching method, which further reduces the computation and data volume of the original method. The improved optimization algorithm can better adapt to the on-site computer’s calculation limit. The optimization method can effectively improve the oil production of the reciprocating pump and avoid the long-term operation of the linear motor in the full load or overload condition.