Research On Methods For Extracting The Features Of The Air Flow Field For The Single-Level Burner Of A Tangentially Fired Boiler

The distribution of air flow field in a tangentially fired boiler is of vital importance for safe and economic operation and pollutants control. However, the only effective approach for improving the distribution of overall air flow-field in furnaces seems to be adjustment of the air flow field at the burners'outlet. Therefore, it's significant to monitor and study the air flow field at the burners'outlet.Based on the images of single-angle jet flow at a burner's outlet and the single-level jet flow obtained from image processing tools, a software is developed for processing the single-level jet air flow images. By experimental comparisons, algorithms are designed and optimized for analyzing the single-level outlet jet air flow and extracting parameters. Finally, the boundary contours of jet flow and corresponding geometric parameters are extracted successfully. The procedure of algorithms steps as follows.①noises of the discrete-points in jet flow images are removed by the weighted mean filter algorithm based on property of the noises;②the target image is separated from background by using image segmentation algorithm based on region growing;③the edge pixels are detected based on adaptive edge detection algorithm;④the detected edge pixels are fitted based on least square method;⑤parameters for centerline position of jet flow and other geometric parameters such as the angle between two linear edges and angle information of jet geometric centerline are extracted. During the design of algorithm, the size of sliding window for denoising is 10×10, and the growing rule is decided by setting the 5% of lightest pixels in the jet flow images as error bounds. The experimental results indicate:the relationship between time-consuming and data precision is optimized; the time-consuming is approximately 3 seconds to complete all images processing procedure by a personal computer equipping an AMD 2500 + CPU and a 768MB memory. The software can process data rapidly and accurately, and the algorithm is stable and reliable, which makes the approach have a potential meaning for real-time analysis of air flow of burner's outletsImage processing techniques provide sufficient data to support computer aided analysis for air flow field of the burners'outlets in power stations. It is the basis for visualization and makes be possible for the real-time analysis of air flow field. visualization method based on Image processing techniques not only reflects the properties of air flow field from outside contours, but also quantifies the properties of air flow field inside the burners by using the geometric parameters of jet flow.