中文摘要：

为了深入研究超声波CT温度场重建算法,总结了已有的研究成果,归纳了这一领域内的主流算法,并对现有的8种算法进行了MATLAB仿真实验。实验结果表明,使用ramp滤波器的滤波反投影算法重建精度较低,均方根误差超过38%;基于矩阵方程的重建算法效率较高,但是一般需要有较多的超声波路径;迭代算法对路径要求低,重建结果较好。由于超声波传播的复杂性和应用条件的多样性,超声CT重建算法仍然有很大的发展空间和研究价值,同时测量温度和风速的超声CT算法将是未来的发展趋势。

英文摘要：

In order to explore the temperature-field reconstruction algorithms by using ultrasonic CT, we review the former research results, summarize the main approaches in this area, and carry out the computer simulation for 8 ex- isting algorithms using MATLAB. The experiment results indicate that the filtered back projection method with ramp filter shows poor accuracy, and its root of mean square error is over 38%. The methods based on matrix equation are computationally efficient, but they usually require too many paths. The iterative methods demand fewer paths, and their reconstruction results are satisfying. As the complexity of ultrasonic propagation and the variety of ultrasonic CT application, there is much space for the research of the ultrasonic CT reconstruction. It will be a possible develop- ment trend to the ultrasonic CT research on measuring both temperature and wind velocity simultaneously.