This paper presents a novel miniaturized noninvasive submersible compact planar sensor to estimate petroleum contents for various industrial and remote sensing applications. The proposed sensor is based on a microstrip technique wherein a complementary split-ring resonator is used to sense samples under interrogation. The proposed compact sensor works at a resonating frequency of 8.87 GHz with quite high sensitivity. The designed sensor can accurately detect the shift in resonant frequency for corresponding change in soil samples and different level of petroleum contents. Furthermore, the sensor gives accurate permittivity extraction of various soil samples sample using an innovative automated neural network-based algorithm. The experimental results show that the proposed sensor can detect materials contents with good repeatability, which can be used to detect the presence of petroleum constituents in transportation pipelines and mining exploration.

中文翻译：

---

基于人工神经网络的紧凑型潜水射频传感器估算沥青土壤中石油含量

本文提出了一种新型微型无创潜水紧凑型平面传感器，用于估计各种工业和遥感应用的石油含量。所提出的传感器基于微带线技术，其中互补裂环谐振器用于在询问下感测样本。建议的紧凑型传感器在 8.87 GHz 的谐振频率下工作，具有相当高的灵敏度。所设计的传感器可以准确地检测到土壤样品的相应变化和不同石油含量水平的共振频率的变化。此外，该传感器使用创新的基于自动神经网络的算法对各种土壤样品样品进行准确的介电常数提取。实验结果表明，所提出的传感器能​​够以良好的重复性检测材料含量，