Steady-state detection method of burning flame temperature based on wavelet transform and least squares fitting

The steady-state detection of burning flame temperature plays an important role in the modelling, state identification and optimization control of the cement clinker burning process. In this paper, the steady-state detection method of burning flame temperature based on wavelet transform and least squares method is studied. First, the burning flame temperature data were detected accurately using a video detection device. Then, the temperature signal was decomposed into the high-frequency and low-frequency components based on the wavelet transform method, and the wavelet basis function and the decomposition layer were determined by least squares fitting error. Thus, the signal trend item can be obtained by removing the high-frequency component that represents the signal noise, and reconstructing the low-frequency component that reflects the basic trend of the signal. On this basis, the first derivative of the trend was further obtained, and the steady-state detection threshold was set to achieve steady state-detection of the burning flame temperature. The results showed that the method proposed in this paper can accurately extract the burning flame temperature trend and realize steady-state detection. This paper provides a feasible method for the steady-state detection of burning temperature.