组合结构ECT传感器的仿真与实验研究

发布时间：2018-06-04 01:04

本文选题：电容层析成像 + 大尺度传感器　；参考：《中国科学院大学(中国科学院工程热物理研究所)》2017年硕士论文

【摘要】：多相流系统广泛存在于石油、制药、化工等工业应用领域,实现上述多相流过程的实时在线测量对理解多相流过程的机理、优化多相流过程意义重大。电容层析成像技术是一项可实现有效实时测量的多相流检测技术。该技术的诸多优点,如非侵入、非接触、无辐射、造价低、采集速度快,使其广泛应用于多相流过程的检测。但是,电容层析成像技术在大型化过程中也面临诸多挑战,如提高传感器的敏感性以及提高大尺度电容传感器的空间分辨率等。本文的目的是探索提高大尺度电容层析成像传感器空间分辨率的有效方法,文中涉及到了两种解决方法:通过改变电极激励模式提高传感器的敏感性从而来提高分辨率以及通过改变电极结构来提高空间分辨率。研究方法包括数值仿真以及实验研究。利用数值仿真方法主要研究了不同的电极结构以及不同的外层电极数对大尺度电容传感器成像效果的影响。仿真过程中设计了三种ECT传感器,即只有外层电极结构的ECT传感器(EE)、含有内外双层电极结构的ECT传感器(IEE)以及组合结构的ECT传感器(CE)。另外还改变了不同结构的传感器的外层电极数,重构图像算法用到了三种:线性反投影算法(LBP)、Landweber迭代算法以及修正的Landweber迭代算法。最后总结了外层电极数在不同的重构图像算法下对不同结构的ECT传感器最终成像效果的影响,同时依据仿真结果确定了实验过程中用到的传感器的电极结构。实验研究包括两部分:不同激励模式对大尺度IEE结构的传感器的成像效果的影响以及不同的传感器结构对大尺度ECT传感器的性能的影响。在激励模式的研究过程中设计了一个大尺度IEE结构传感器,利用该传感器研究了五种双电极激励模式和传统单电极激励模式对该传感器成像效果的影响。对比了不同激励模式下利用LBP算法获得的重构图像,同时引入了一种修正的重构方法来减小重构图像过程中噪声的干扰。最后利用重构图像的固体浓度的相对误差评价了六种激励模式的精度。在研究大尺度ECT传感器结构产生的性能影响方面,设计了上述三种结构的大尺度ECT传感器,对比了不同结构的ECT传感器的成像效果。重构图像使用的算法依然是LBP算法,但是该实验中重构图像用了两种归一化方式,对比了不同的归一化方式对不同结构的ECT传感器最终成像效果的影响,最终利用重构图像的固体浓度评估了两种归一化应用于不同结构的ECT传感器时的性能。
[Abstract]:Multiphase flow systems are widely used in petroleum, pharmaceutical, chemical and other industrial applications. It is of great significance to realize the real-time on-line measurement of the multiphase flow process in order to understand the mechanism of the multiphase flow process and optimize the multiphase flow process. Electrical capacitance tomography (ECT) is a multiphase flow detection technique which can be used for real-time measurement. The advantages of this technology, such as non-invasion, non-contact, non-radiation, low cost and fast acquisition speed, make it widely used in the detection of multiphase flow process. However, the capacitance tomography technology also faces many challenges in the large-scale process, such as enhancing the sensitivity of the sensor and improving the spatial resolution of the large-scale capacitance sensor. The purpose of this paper is to explore an effective method to improve the spatial resolution of large scale capacitance tomography sensor. Two solutions are involved in this paper: to improve the sensitivity of the sensor by changing the electrode excitation mode to improve the resolution and to improve the spatial resolution by changing the electrode structure. The research methods include numerical simulation and experimental research. The effects of different electrode structures and the number of outer electrodes on the imaging effect of large scale capacitive sensor are studied by numerical simulation. In the process of simulation, three kinds of ECT sensors are designed, namely, ECT sensors with outer electrode structure, ECT sensors with internal and external double-layer electrode structure, and ECT sensors with combined structure. In addition, the number of outer electrodes of sensors with different structures is changed, and three kinds of image reconstruction algorithms are used: the linear backprojection algorithm and the modified Landweber iterative algorithm. Finally, the effects of the number of outer electrodes on the final imaging of ECT sensors with different structures under different reconstructed image algorithms are summarized, and the electrode structure of the sensors used in the experiment is determined according to the simulation results. The experimental study consists of two parts: the influence of different excitation modes on the imaging effect of large scale IEE sensors and the influence of different sensor structures on the performance of large scale ECT sensors. A large scale IEE structure sensor is designed in the research of excitation mode. The effects of five two electrode excitation modes and the traditional single electrode excitation mode on the imaging effect of the sensor are studied. The reconstructed images obtained by using LBP algorithm under different excitation modes are compared, and a modified reconstruction method is introduced to reduce the noise interference in the process of image reconstruction. Finally, the relative error of solid concentration of reconstructed image is used to evaluate the accuracy of six excitation modes. In order to study the effect of large scale ECT sensor structure on the performance, the three kinds of large scale ECT sensors are designed, and the imaging effects of ECT sensors with different structures are compared. The algorithm used for image reconstruction is still LBP algorithm, but in this experiment, two normalization methods are used to reconstruct the image, and the effects of different normalization methods on the final imaging effect of ECT sensors with different structures are compared. Finally, the performance of two normalized ECT sensors with different structures is evaluated by using the solid concentration of reconstructed images.
【学位授予单位】：中国科学院大学(中国科学院工程热物理研究所)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP212

【参考文献】