Numerical and experimental investigations on the total-variation regularization method of temperature distribution reconstruction in acoustic tomography

Acoustic tomography (AT) is considered to be a non-intrusive and effective method of obtaining temperature distribution information. The reconstruction method is crucial to obtaining reliable results of AT measurement. This paper presents a new method of reconstructing the temperature field in a measuring domain. Numerical simulations and experimental studies are conducted in this paper to validate the performance and robustness of the proposed method. In addition, it is compared to the other acknowledged regularization methods in inverse problems. The reconstruction time is also discussed to explore the feasibility of performing the construction procedure in real time. The numerical results indicate that the proposed method has better performance than other regularization methods for different temperature distributions. Moreover, it shows better anti-noise ability in the condition that the time-of-flight is contaminated by noise. The experimental reconstruction results of these reconstruction methods are compared against thermocouple measurements, showing that the proposed method has superior reconstruction performance in most cases. Therefore it may serve as a promising reference for quality improvement of temperature distribution reconstruction in AT.