【摘要】 针对低温流体液气介电常数接近1的特点,基于八电极电容层析成像（ECT）传感器,通过定量对比图像误差和相关性系数,系统分析了线性反投影算法、Tikhonov正则化算法、Landweber迭代算法、迭代Tikhonov正则化算法、代数重建技术和同步代数重建技术等传统用于室温流体的线性算法,用于LN₂-VN₂两相流的反演图像精度,并指出了各算法优缺点。通过对比水-空气的反演图像,发现LN₂-VN₂反演过程ECT线性化误差较小,从而有更好的反演结果。更多还原

【Abstract】 In view of the characteristic that the dielectric constant ratio of liquid to gas for cryogenic fluids was close to 1,the inversion image accuracy of the traditional linear algorithm for room temperature fluid for LN₂-VN₂ two-phase flow was analyzed systematically based on the eight-electrode capacitance tomography（ECT）sensor,through quantitative comparison of image error and correlation coefficient.The advantages and disadvantages of each algorithm were pointed out.These commonly-used algorithms include the linear back-projection algorithm,Tikhonov regularization algorithm,Landweber iterative algorithm,iterative Tikhonov regularization algorithm,algebraic reconstruction technology,and synchronous algebraic reconstruction technology.It is found that the ECT linearization error of cryogenic fluid is small and has better inversion results by comparing the inversion images of water-air and LN₂-VN₂.更多还原