Research On Reconstruction Algorithm Based On Acoustic Temperature Measurement Techniques In Boilers

The combustion of pulverized-coal in power plant boiler is a complexly reaction, which is occurred in larger space with instantaneous and random character, monitoring the combustion status is of great importance for boiler’s safety, energy conservation and emission reduction. And the temperature distribution in power plant boiler is an important parameter which can reflect the combustion directly, so the monitoring of the furnace temperature field is of great meaning. As a non-contact measurement, acoustic temperature measurement technique have developed rapidly for its wide measuring range, high-precision, real-time, continuous measurement and appliance in high-temperature and high-corrosiveness environment. In this paper, much studies about reconstruction algorithm and interpolation method of acoustic temperature measurement technique had been done based on acoustic theory. This paper has mainly completed the following works:First build a three-dimensional temperature field model, put forward an acoustic measurement scheme based on the model. Then single-peak symmetry temperature field, single-peak asymmetry temperature field and double-peak symmetry temperature field were reconstructed by severally use the least square method, steepest descent method and Tikhonov regularization method with spline interpolation method. The results and errors show that the three algorithms can all achieve reconstruction of temperature field, different algorithm may cause different error, but the difference is small. And the errors of simple temperature field reconstruction were bigger than that of complex ones.To study the influence of different interpolation for the reconstruction, single-peak symmetry temperature field, single-peak asymmetry temperature field and double-peak symmetry temperature field were reconstructed by severally use the spline interpolation method, inverse distance squares method and Kriging method with Tikhonov regularization method. The results and errors show that the reconstructions of temperature field by using different interpolation methods are much different, the result of Kriging method is much better than the results of the other two interpolation methods.