液体连续相撞击流PIV测量系统中坐标转换方法研究  

Method of coordinates mapping in PIV measurement system of a liquid-continuous impinging streams

作　　者：李福宝[1] 伍沅[2] 徐成海[1] 李勤[3] 

机构地区：[1]东北大学机械工程与自动化学院,沈阳110004 [2]武汉工程大学化工与制药学院,武汉430073 [3]沈阳工业大学石油化工学院,辽阳111003

出　　处：《仪器仪表学报》2008年第9期1827-1830,共4页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(29276260;20176043)资助项目

摘　　要：在液体连续相撞击流反应器圆柱壳体内流场PIV测量系统中,为了实现由序列图象解算真实流场数据,设计了应用人工神经网络的坐标转换方法,该方法将一画有标准间距网格的平板放置在透明的液体连续相撞击流反应器撞击区域内,网格平面通过反应器圆筒轴线,反应器内充满水,数字摄像机固定在容器外,光轴垂直于网格平面,借助软件工具从网格图象中得到用于训练人工神经网络的训练矢量集合,完成训练的网络可以液体内被测平面的象素坐标正确地转换为真实坐标。In order to acquire flowing field parameters in PIV measurement system of a liquid-continuous impinging streams, a special coordinates mapping method from image space to real space is put forward. The method based on neural net work includes following steps. Firstly we put a board with grid pattern on it into a transparent cylinder container and fill the container with water. Secondly we adjust the board to ensure the axis of the container right goes through the grid plane. Thirdly a camera facing to the board was fixed outside the container. By taking picture of the gird and processing the picture with software tool, the training vector set was obtained, and after training the neural net work can mapping coordinates from image space to real word space correctly.

关 键 词：撞击流 粒子图象 图象处理 人工神经网络 坐标变换 

分 类 号：TQ560.6[化学工程—炸药化工] P208[天文地球—地图制图学与地理信息工程]