Functional Development And Practical Application Of The Flame Detection System In Boilers

The physical and chemical mechanisms of flame combustion are complex and therefore the flame stability is difficult to evaluate accurately. In this paper, a comprehensive assessment method for the stability of flame combustion using comprehensive fuzzy judgment is proposed. Flame radiation luminance fluctuating digital signals are acquired on a low-cost flame detection device. On the basis of using wavelet transform for noise reduction, power spectrum of frequency domain and bi-spectrum information of phase domain are derived from time domain signals respectively based on fast fourier transform and higher order statistics. The features of the time-frequency domain and bi-spectrum under various combustion conditions are qualitatively analysed, and the features including fluctuation variance, flicker frequency and bi-spectrum phase which reflect time-frequency domain and phase domain are respectively extracted. A mathematical model for comprehensive fuzzy assessment has been constructed and a stability index system for the assessment of flame combustion established. Results obtained demonstrate that this method can effectively reveal the stability condition of the flame radiation, which provides to the operation staff with reliable reference information in judging the combustion stability.At present, the flameout early-warning function has not been exploited by all of the flame detectors in the power plant boilers. The tap water is used as fire-extingushing media to simulate the flameout process. The time-frequency variation information is used to analysis the flame stability and the forewarning flameout. The typical characteristic information of the flameout early-warning is summarized. The threshold value of the fluctuation variance and the flicker frequency are setted to accomplish the millisecond flameout early-warning function and to establish the criterion of the forewarning flameout.