基于主动双目立体视觉的干排渣机渣量检测方法研究

Research on Detection Method of Slag Quantity in Dry Slag Extractor Based on Active Binocular Stereo Vision

实现燃煤电站锅炉底部渣量在线检测,进而实现干排渣机冷却风量有效调整,对提高干排渣机运行的可靠性和经济性具有重要意义。提出一种基于主动双目视觉的炉渣体积流量在线检测方法,通过线型激光照射干排渣机输送带上炉渣表面,有效地增强了图像的亮度、纹理等特征,进一步利用改进的自适应Census算法实现了双目视觉系统准确匹配。在此基础上,开发了炉渣体积流量检测系统,开展了评价研究,并在某660 MW超临界机组上进行了现场应用。结果表明:该系统能够准确可靠的获得灰渣的深度信息,误差小于8%,整体精度较高。660 MW超临界机组现场应用结果进一步验证了渣量在线检测系统的可靠性和准确性,满足工业现场应用要求。

To achieve the on-line measurement of the slag amount at the bottom of the coal-fired boiler,thereby to effectively adjust the cooling air volumetric flowrate,is of great significance for improving the reliability and economy of the dry slag extractor.An active binocular vision based-measurement method for the slag volumetric flowrate is presented in this paper.By irradiating the slag surface on the conveyor belt of the slag extractor with a linear laser,the brightness and texture of the slag image are effectively enhanced.Furthermore,an improved adaptive Census algorithm is used to realize the accurate matching of binocular vision system.On the basis,an online measurement system for slag volumetric flowrate is developed and experimentally evaluated in laboratory,and then is applied in a 660 MW supercritical unit.Results show that the active binocular vision system can obtain the depth information of slag accurately and reliably,the relative error is less than 8%,and the overall accuracy relatively high.Field applications in a 660 MW supercritical unit further verified the reliability and accuracy of the on-line slag measurement system,which meets the requirements of industrial field.