Research On Two-stage Reconstruction Algorithm Of Acoustic CT For Temperature Field

Acoustic CT temperature field reconstruction is based on placing multiple acoustic transceivers around the measured area to form several effective acoustic paths across the measured area.And then,by measuring the flight time of sound wave along each path,the projection data of sound velocity distribution in the corresponding direction can be obtained.Then,a suitable reconstruction algorithm is used to reconstruct the sound velocity distribution of the measured region from the projection data.According to the relationship between sound velocity and temperature,the reconstructed temperature distribution is finally obtained.This is a non-contact measurement technology,so it will not interfere with the temperature field to be measured,and has the advantages of wide temperature measurement area,strong environmental adaptability,and has broad application prospects in industry,agriculture and natural environment monitoring and other fields.Reconstruction algorithm is quite important in acoustic CT temperature field reconstruction.Traditional reconstruction algorithms mainly include least square method,ART iteration method,SIRT iteration method and so on.The number of grids divided by these algorithms is less than the number of acoustic paths.Therefore,these algorithms not only have outstanding advantages of good stability,but also lack of edge temperature information.Two two-stage reconstruction algorithms based on traditional algorithms are proposed in this paper.In this paper,a two-stage reconstruction algorithm based on traditional algorithm and radial basis function is proposed.Firstly,the temperature distribution described by coarse grid is reconstructed by traditional algorithm.Then the radial basis function with linear extension is used to obtain the complete and detailed temperature distribution.Then a two-stage algorithm based on traditional algorithm and Gauss process regression model is proposed.Firstly,the temperature distribution described by coarse grid is reconstructed by traditional algorithm.Gaussian process regression model was used to obtain the complete and detailed temperature distribution.The traditional algorithms in the two-stage algorithm adopt the least square method,ART iteration method and SIRT iteration method respectively.By simulating the typical temperature field,the performance of the traditional algorithm and the two-stage reconstruction algorithm proposed in this paper are compared.The results show that the proposed two-stage algorithm has high reconstruction accuracy and strong anti-interference ability.It also expands the range of measurements for the area being measured.In this paper,a two-dimensional temperature field reconstruction software is designed and integrated based on Matlab.The software can easily,quickly and accurately reconstruct the two-dimensional temperature field and refine the internal and extrapolation.