Intelligent control of the torch system

In the refinery and petrochemical industries, flare control is carried out manually, which has many environmental consequences. In order to better burn gases from process units, steam is used by the operator to control it, which results in a reduction or increase in the amount of steam, resulting from the emission of gases or flame shutdown. By the image processing control system, you can detect the amount of flame and set it to the required level, even if the steam is set. In this project, flame pixels are first detected using color information and flame pixels are analyzed in successive frames. Measurement of a flame must be accurate and fast, as flame control is controlled simultaneously with flame changes. Image processing begins with image acquisition by the digital camera, and after storing images, image segmentation, filtering, edge detection, noise reduction, and color-to-gray conversion. The results of the processed image analysis, the control by the fuzzy control system are optimized, and by the digital-to-analogue converter, a steam valve control signal is issued.

Keywords:

Intelligent control , Fuzzy control , Torch , Flair , Image processing