Acoustic Temperature Measurement Technique And Temperature Field Reconstruction Algorithm

The ultimate goal of temperature field measurement is to visualize the measured temperature field, and thus detect and control the temperature state. Acoustic temperature measurement technology has many advantages such as non-contact, wide temperature range, large space of measuring object, high-precision, real-time, continuous measurement, easy maintenance and so on. For industrial boiler temperature field measurement, it provides an effective means. Besides, it also can be used to monitor and adjust the temperature of a theater, sense the temperature distribution of atmosphere and ocean. However, its application in monitoring the temperature distribution in grain storage is entire just a new exploration field.In this paper, the basic principle of acoustic measurement technique and its application are expatiated. The development status of temperature field measurement by acoustic tomography at home and abroad is introduced. The measurement methods of temperature field are outlined. Temperature field reconstruction algorithms are focused on. A new acoustic temperature reconstruction algorithm, based on radial basis function and the truncated singular value decomposition, is proposed. First expand the temperature-dependent sound slowness function over a finite set of radial basic functions, that is, establish the forward problem model of acoustic temperature field measurement. Then acquire a stable inverse problem solution by using matrix truncated singular value decomposition, and reconstruct the temperature field.The algorithms are investigated using simulation data calculated by MATLAB. With 6 or 8 acoustic sound emitters/receivers, four industry furnace temperature distributions, i.e. symmetrical single temperature peak model, asymmetrical single temperature peak model, symmetrical dual temperature peaks model, and asymmetrical dual temperature peaks model, are reconstructed by the new algorithm. With 16 acoustic sound emitters/receivers, eight stored grain temperature distributions, i.e. symmetrical single temperature peak model, asymmetrical single temperature peak model, symmetrical dual temperature peaks model, three peaks, four peaks, four contour peaks, asymmetrical four high-low peaks and five peaks, are reconstructed by the new algorithm. Reconstruction results indicate that the new algorithm has high accuracy, fast speed and good antinoise ability, and suitable for the temperature field monitoring of industrial furnace and stored grain. In this paper, the reconstruction of the model temperature fields with 400 single-hotspots confirmed that the new algorithm has good detection ability for the single-hotspots in stored grain.With 8 acoustic sound emitters/receivers, two-dimension temperature field reconstruction experiment is implemented in Lab by the new algorithm. The temperature distribution reconstructed accords with the real temperature field.