The Research on Method of Detection for Three-Dimensional Temperature of the Furnace Based on Support Vector Machine

SVM is a rising in common use learning technology base on Statistical Learning Theory. In view of the variety of the changes of burning flame in large-scale boiler, in order to master the status of the furnace flame in real time. In this paper, it introduces the SVM network absorb the basic method of flame-picture’s temperature, to study thoroughly discussed about the SVM network model, and puts forward an improved SVM network method, as for the method to calculate three-dimensional temperature field, which has many advantages to the traditional optimization method. Taking the SVM network using to test the three-dimensional temperature field in the furnace is feasible by the simulation studies and the experimental results.