Study On Measurement System Of Temperature Field Based On Acoustic Sensors

Combustion temperature testing technique is one of the important foundations in combustion theory and technique developing. The distribution and visualization of temperature field in the furnace is important to its controlling and diagnosing combustion. However, the processing of combustion in furnace is instantaneous, random and in hard environment, and the scale of furnace is large, and makes it hard for the on line measurement of actual furnace temperature field. As an effective non-contact measurement method for the temperature measurement in the boiler, the acoustic method is much faster and with a high precision as well as a broader measuring range. It can satisfy the requirements of measuring accurately and controlling on-line on the temperature field in industrial productions and scientific investigations. It has important instructional meaning for optimizing combustion, reducing pollution, equipment circulating safely and designing the hearth.This paper improves the acoustic temperature measurement system on questions of based on the predecessor work. On the basis of acoustic temperature measurement fundamental, the temperature field of the target area was reconstructed by the measurement of flight time of acoustic signals in the effective paths. The analysis has been done with the factors which affect the veracity of measurement. The main work is as following:(1)According to the theory of acoustic pyrometer, improvements have been made in the analog circuit based on former students. With Multi-channel switches of the added CMOS analog multiplexers, the on-line and real-time measurements of flight time of acoustic signals in the effective paths are taken in the system. Experiments of different type sensors have been carried out, include Pressure-loop Sound Convertor、Electret Sound Convertor、Piezoelectric Ceramic Sound Convertor, and finally determine to use the Pressure-loop Sound Convertor, dual-channel microphone scheme;(2)Based on dual-channel measurement system, the time delay estimation algorithm based on cross correlation is given. The generalized cepstrum correlation algorithm and the generalized cross power spectrum algorithm are proposed to overcome the convolution interference, which may exist in the detection process. The generalized cepstrum correlation algorithm can transform the product and convolution homomorphism system to linear system by using cepstrum operation, thereby, this algorithm can separate product signals or convolution signals, filter the product and convolution interferences. The generalized cross power spectrum algorithm whitens the sample signals, adjusts the weight of cross power spectrum with SNR, and improves the performance of noise proof and reverberation suppression. The experimental result and theory proved that, compared to the traditional correlation function analytical algorithm, the new algorithms can overcome the convolution interferences from reverberation, and sharpen peak value, thereby estimate the time delays of signals accurately;(3) Based on the measurement of flight time of acoustic signals travelling between acoustic sensors and by using the least square root field reconstruction algorithm, the temperature fields of the Single-peak symmetry temperature field, the Single-peak asymmetry temperature field and the Double-peak temperature field were reconstructed, and a good many of elements which may affect the precision of temperature field reconstruction were summarized. Correlative improvements were attached;(4)Improvements and accomplishments have been made in the software platform of the system of acoustic for measuring temperature. Based on the measurement of flight time of acoustic signals in the effective paths, the on-line and real-time temperature reconstructed is taken by calling the MATLAB DDL (dynamic link library) method.