School of Computer Science, China University of Geosciences, Wuhan, 430078, China.
Sci Rep. 2023 Mar 20;13(1):4586. doi: 10.1038/s41598-023-30311-w.
Petroleum industry has started to embrace the advanced petroleum cyber-physical system (CPS) technologies. Offshore petroleum CPS is particularly hard to build, mainly due to the difficulty in detecting and preventing offshore oil leaking. During the oil exploration and transportation process, the remote multi-sensing technology is typically employed for emerging service. It can be utilized for leak detection by enabling the underwater modeling of an offshore petroleum CPS. However, such a technology suffers from insufficient remote sensing resources and expensive computational overhead. In this work, a cross-entropy based leak detection technique is proposed to detect the oil leak, which facilitates the understanding of the oil leak induced marine pollution. Furthermore, a hierarchical parallel approach is proposed on the super computer Tianhe-2 to improve the efficiency of the proposed leak detection technique. Experimental results on Penglai oil spill events demonstrate that the proposed method can effectively identify the sources of oil spilling with accuracy up to [Formula: see text].
石油行业已经开始采用先进的石油网络物理系统 (CPS) 技术。海上石油 CPS 特别难以构建，主要是因为难以检测和预防海上石油泄漏。在石油勘探和运输过程中，通常采用远程多传感器技术来提供新兴服务。它可以通过对海上石油 CPS 进行水下建模来实现泄漏检测。然而，这种技术存在远程传感资源不足和计算开销昂贵的问题。在这项工作中，提出了一种基于交叉熵的泄漏检测技术来检测石油泄漏，有助于理解石油泄漏引起的海洋污染。此外，在天河-2 超级计算机上提出了一种分层并行方法来提高所提出的泄漏检测技术的效率。蓬莱溢油事件的实验结果表明，所提出的方法可以有效地识别溢油源，准确率高达[Formula: see text]。
