Research On Reconstruction Algorithms Of Acoustic CT For Temperature Field Measurement

Temperature field detection technology is of great significance to the atmospheric environment,marine environment,graim storage,industrial combustion and other fields.Acoustic CT temperature field detection technology belongs to the non-contact measurement method.It has the advantages of non-interference in the measured temperature field,easy installation of measurnng devices,strong environmental adaptability,large-scale real-time measurement of temperature and space,and is suitable for applications in the fields of atmosphere,ocean,grain storage,industrial furnaces and so on.Reconstruction algorithm plays an important role in acoustic CT temperature field detection.In this paper,the reconstruction algorithm of acoustic CT temperature field is studied.The main work is as follows.Firstly,the basic principle of acoustic temperature measurement and the reconstruction principle of acoustic CT temperature field are introduced.The factors affecting the reconstruction quality of temperature field are analyzed,and four indexes for evaluating the reconstruction quality of temperature field are introduced.Then two representative reconstruction algorithms of temperature field are introduced,namely least squares method(LSM algorithm)and Markov radial basis function Tikhonov regularization method(MTR algorithm).In order to improve the reconstruction ability of complex temperature field in acoustic CT,this paper presents a temperature field reconstruction algorithm based on Markov radial basis function approximation and Kalman filter,MKR algorithm.Firstly,the linear combination of Markov radial basis function is used to approximate the sound velocity distribution in the measured area,and the system state observation equation corresponding to acoustic CT is established.Then,the Kalman filter is used to reconstruct the acoustic distribution from the multipath time-of-flight data.Then the temperature distribution is obtained by using the relationship between sound velocity and temperature.In this paper,the parameter selection of MKR algorithm is studied through four typical three-dimensional model temperature field simulation and reconstruction experiments.The experimental results show that when the reconstruction results of MTR algorithm are taken as the initial value of the state estimator,the estimation error covariance matix is 10×I and the Kalman filter iteration number is 50 times,the reconstruction results of MKR algorithm are the best.The influence of the layout of the acoustic transceiver and the meshing method of the measured area on the reconstruction results of MKR algorithm is also studied.The experimental results show that the reconstruction effect of 32 acoustic transceivers is the best when they are divided into four layers and the measured area is divided into 10×10×10=1000 meshes.In this paper,MKR,MTR and LSM algorithms are used to reconstruct the noiseless and noiseless simulation data of four typical three-dimensional models.The reconstruction results show that the reconstruction quality of MKR and MTR is better than LSM algorithm,and the reconstruction quality of MKR is the best.Then,the MKR and MTR algorithms are further compared by using 1000 single-peak model temperature fields with different hot spot locations.The reconstruction results prove that MKR is better than MTR in reconstruction ability,especially in the case of noise.Therefore,the proposed MKR algorithm has better ability to reconstruct complex temperature field.