测量原理-数据-领域知识融合ECT重建方法
Measurement Principles-data-domain Knowledge Fusion Reconstruction Method for ECT
李珍兴 1邵继续 2吴俊杰 1任婷3
作者信息
- 1. 国能河北定州发电有限责任公司
- 2. 中广核(北京)仿真技术有限公司
- 3. 中国科学院电工研究所
- 折叠
摘要
低质量的图像制约了电容层析成像在多相流参数测量中的应用.针对该问题,引入了由深度卷积神经网络预测的数据驱动先验,提出了融合测量原理、数据驱动先验和稀疏先验的成像模型;建立了新的算法实现成像模型的高效求解.评估结果证实,与其他的成像算法相比,新算法在细节重建、伪影去除和鲁棒性等方面具有显著优势.
Abstract
Low-quality images constrain the application of electrical capacitance tomography in multiphase flow parameter measurements.Aiming at this problem,the data driven prior(DDP)predicted by a deep convolutional neural network was intro-duced,and a new imaging model fusing the measurement principle,the DDP and the sparsity prior was proposed.A new algorithm was established to achieve the efficient solution of the imaging model.Evaluation results confirm that the new algorithm has significant advantages over other imaging algorithms in terms of detail reconstruction,artifact removal and robustness.
关键词
计算成像问题/数据驱动先验/深度卷积神经网络/电容层析成像/反问题/多相流测量Key words
computational imaging problem/data driven prior/deep convolutional neural network/electrical capacitance tomo-graphy/inverse problem/multiphase flow measurement引用本文复制引用
基金项目
国家自然科学基金面上项目(52276217)
出版年
2024
NETL
NSTL
维普
万方数据