Research On Detecting The State Of Flame Combustion On-line

Whether burnable flame is stability or not can reflect the combustion state most directly. If the combustion is unsteady or unsuitable, it will cause lower efficiency, more pollution and more noise. Furthermore, it may cause serious trouble, such as personnel casualty and properties loss. So, in order to improve the burnable efficiency and decline the latent danger, the flame burnable state detecting is needed.This paper builds up a model to recognize the flame combustion images, by processing the collected flame images, combine the other sensor information, thereby gain the burnable state characteristic parameter, make use of the detected information of the flame's shape and position, and the radiation intensity to define the flame burnable state synthetical parameter, using the artificial neural network technique to analyse and process the parameter, thus pinpoint the parameter of the flame combustion burnable state, and pinpoint the present flame burnable state lastly. It's for workers to supervise and control the flame combustion, improving the fuel's burnable efficiency, saving energy and protecting the environment have much meaning.