Research On Reconstruction Algorithm Of Temperature Field In Acoustic Tomography

The temperature field reconstruction technology of acoustic tomography uses multiple acoustic transceivers arranged around the tested area to form several effective acoustic paths through the tested area.By measuring the flight time of sound wave in each path,the projection data of sound velocity distribution in corresponding direction can be obtained.Then the sound velocity distribution of the measured area is reconstructed from the projection data by using the appropriate reconstruction algorithm.The temperature distribution is obtained by using the relationship between the speed of sound and the temperature.This technology is non-contact,so it will not interfere with the temperature field,and has the advantages of wide temperature measurement range,strong environmental adaptability,suitable for a wide range of space and so on.It has a good application prospect in the fields of industry,agriculture and environmental monitoring.In the reconstruction of temperature field in acoustic tomography,the reconstruction algorithm plays an important role.Least square method(LSM),algebraic iterative algorithm,(ART)and synchronous iterative algorithm,(SIRT)algorithm are commonly used reconstruction algorithms.These algorithms require that the number of grids divided by the measured area is less than the number of acoustic paths,so when the temperature distribution is complex,the reconstruction results deviate greatly from the actual distribution,and even can not correctly reconstruct the characteristics of the temperature field.SIRT algorithm is often used in the reconstruction of acoustic CT temperature field because of its good anti noise ability and stability.However,when the number of meshes divided by the reconstruction area is more than the number of acoustic paths,there are usually a lot of noise in the reconstruction of temperature field.In order to solve this problem,an improved SIRT algorithm(MSIRT)is proposed in this paper.The smoothing filter is incorporated into SIRT algorithm,and the noise in the reconstructed temperature field is suppressed by the specially designed inner and outer smoothing filter.In order to verify the effectiveness of the proposed method,three typical model temperature fields and a real single peak temperature field are reconstructed.The reconstruction results show that the reconstruction error of the improved SIRT algorithm is significantly reduced compared with the commonly used least square algorithm,ART algorithm and SIRT algorithm.The improved SIRT algorithm has better ability to reconstruct complex temperature field.In order to improve the reconstruction ability of LSM and ART algorithm,this paper proposes the improved LSM algorithm(MLSM)and improved ART algorithm.After the LSM and ART algorithm calculate the sound slowness of each pixel,the smooth filter is added to smooth the sound slowness and improve the reconstruction ability of the algorithm.In the case of noise and no noise,the reconstruction is carried out in three typical model temperature fields.The experimental results show that MLSM algorithm and MART algorithm can effectively reduce the reconstruction error of temperature field and Compared with LSM algorithm and ART algorithm,it has stronger noise resistance ability.