工业控制场景下5G-A终端传输时延是确定性网络能力的直接表征之一,时延预测对提高网络确定性至关重要。由于传输时延序列的不稳定性和随机性,单一模型难以准确预测。针对该问题,提出一种基于优化变分模态分解(Variational Mode Decompos...工业控制场景下5G-A终端传输时延是确定性网络能力的直接表征之一,时延预测对提高网络确定性至关重要。由于传输时延序列的不稳定性和随机性,单一模型难以准确预测。针对该问题,提出一种基于优化变分模态分解(Variational Mode Decomposition, VMD)和卷积注意力长短时记忆网络(Convolutional Attention Long Short Term Memory Network, CA-LSTM)的传输时延预测方法。为提高VMD的分解性能,利用相关系数检验法确定时延序列分解的模态数,并利用蝗虫优化寻优分解的惩罚因子和保真度系数;设计CA-LSTM网络,借助卷积滤波器以及注意力机制使得网络具备分辨时延特征重要程度的能力;将各模态预测值重建成一维时延值得到预测结果。实验研究结果表明,优化VDM能够将5G终端传输时延序列有效分解,结合CA-LSTM模型相比于经典LSTM在MSE、RMSE和MAE上分别提升了37.1%、21.3%和23.6%。展开更多
利用全球导航卫星反射信号(Global Navigation Satellite System Reflectometry,GNSS-R)进行海洋环境要素探测已成为国内外遥感领域一大热点。镜面反射点作为GNSS-R遥感探测的坐标基准,其预测的精度与速度对后续环境要素的反演有着显著...利用全球导航卫星反射信号(Global Navigation Satellite System Reflectometry,GNSS-R)进行海洋环境要素探测已成为国内外遥感领域一大热点。镜面反射点作为GNSS-R遥感探测的坐标基准,其预测的精度与速度对后续环境要素的反演有着显著影响。针对星载GNSS-R实时预测高精度镜面反射点的需求,提出了一种基于AdaGrad优化的GNSS-R镜面反射点自适应预测算法。利用空间几何关系对镜面反射点进行粗略估计,通过预处理后利用AdaGrad优化寻找镜面反射点的最优解,仿真实验证明算法的精确性、高效性。展开更多
Reduction in water injectivity would be harmful to the waterflood development of offshore sandstone oil reservoirs. In this paper the magnitude of formation damage during water injection was evaluated by analyzing the...Reduction in water injectivity would be harmful to the waterflood development of offshore sandstone oil reservoirs. In this paper the magnitude of formation damage during water injection was evaluated by analyzing the performance of water injection in the Bohai offshore oilfield, China. Two parameters, permeability reduction and rate of wellhead pressure rise, were proposed to evaluate the formation damage around injection wells. The pressure performance curve could be divided into three stages with different characteristics. Analysis of field data shows that formation damage caused by water injection was severe in some wells in the Bohai offshore oilfield, China. In the laboratory, the content of clay minerals in reservoir rock was analyzed and sensitivity tests （including sensitivity to water, ftow rate, alkali, salt and acid） were also conducted. Experimental results show that the reservoir had a strong to medium sensitivity to water （i.e. clay swelling） and a strong to medium sensitivity to flow rate, which may cause formation damage. For formation damage prevention, three injection schemes of clay stabilizer （CS） were studied, i.e. continuous injection of low concentration CS （CI）, slug injection of high concentration CS （SI）, and slug injection of high concentration CS followed by continuous injection of low concentration CS （SI-CI）. Core flooding experiments show that SI-CI is an effective scheme to prevent formation damage and is recommended for the sandstone oil reservoirs in the Bohai offshore oilfield during water injection.展开更多
文摘工业控制场景下5G-A终端传输时延是确定性网络能力的直接表征之一,时延预测对提高网络确定性至关重要。由于传输时延序列的不稳定性和随机性,单一模型难以准确预测。针对该问题,提出一种基于优化变分模态分解(Variational Mode Decomposition, VMD)和卷积注意力长短时记忆网络(Convolutional Attention Long Short Term Memory Network, CA-LSTM)的传输时延预测方法。为提高VMD的分解性能,利用相关系数检验法确定时延序列分解的模态数,并利用蝗虫优化寻优分解的惩罚因子和保真度系数;设计CA-LSTM网络,借助卷积滤波器以及注意力机制使得网络具备分辨时延特征重要程度的能力;将各模态预测值重建成一维时延值得到预测结果。实验研究结果表明,优化VDM能够将5G终端传输时延序列有效分解,结合CA-LSTM模型相比于经典LSTM在MSE、RMSE和MAE上分别提升了37.1%、21.3%和23.6%。
文摘利用全球导航卫星反射信号(Global Navigation Satellite System Reflectometry,GNSS-R)进行海洋环境要素探测已成为国内外遥感领域一大热点。镜面反射点作为GNSS-R遥感探测的坐标基准,其预测的精度与速度对后续环境要素的反演有着显著影响。针对星载GNSS-R实时预测高精度镜面反射点的需求,提出了一种基于AdaGrad优化的GNSS-R镜面反射点自适应预测算法。利用空间几何关系对镜面反射点进行粗略估计,通过预处理后利用AdaGrad优化寻找镜面反射点的最优解,仿真实验证明算法的精确性、高效性。
文摘Reduction in water injectivity would be harmful to the waterflood development of offshore sandstone oil reservoirs. In this paper the magnitude of formation damage during water injection was evaluated by analyzing the performance of water injection in the Bohai offshore oilfield, China. Two parameters, permeability reduction and rate of wellhead pressure rise, were proposed to evaluate the formation damage around injection wells. The pressure performance curve could be divided into three stages with different characteristics. Analysis of field data shows that formation damage caused by water injection was severe in some wells in the Bohai offshore oilfield, China. In the laboratory, the content of clay minerals in reservoir rock was analyzed and sensitivity tests （including sensitivity to water, ftow rate, alkali, salt and acid） were also conducted. Experimental results show that the reservoir had a strong to medium sensitivity to water （i.e. clay swelling） and a strong to medium sensitivity to flow rate, which may cause formation damage. For formation damage prevention, three injection schemes of clay stabilizer （CS） were studied, i.e. continuous injection of low concentration CS （CI）, slug injection of high concentration CS （SI）, and slug injection of high concentration CS followed by continuous injection of low concentration CS （SI-CI）. Core flooding experiments show that SI-CI is an effective scheme to prevent formation damage and is recommended for the sandstone oil reservoirs in the Bohai offshore oilfield during water injection.
