基于SOM网络的智能化粒子测速算法

Intelligentized method for particle image velocimetry based on SOM network

提出了基于自组织映射(self-organized mapping,SOM)神经网络的粒子图像测速算法。该方法使用SOM神经网络对粒子测速技术中运动追踪方法进行了改进,并在匹配过程中根据兴趣区域的粒子密度对粒子追踪算法与粒子相关算法进行了选择处理。经SOM网络改进的测速算法首先利用相关后的结果进行网络构建,然后使用追踪技术对候选匹配点进行筛选。该算法不仅消除了粒子密度与灰度分布的敏感性,而且也降低了相关时对分析窗口尺寸的敏感。最后,使用人工合成的粒子图进行了算法验证及误差分析。结果表明:所提算法在分析精度方面有很大的提高并且具有很强的鲁棒性。

A modified particle image velocimetry （PIV） method based on Kohonen self-organized mapping （SOM） neural network is presented. In the proposed algorithm, SOM network is used to improve the motional tracking capability; the particle tracking algorithm and the particle correlation algorithm are combined to process images according to particle density. Firstly, the results of cross-correlation are used to build network. Secondly, the tracking method is used to select matching points. The new PIV algorithm based on SOM network can reduce the dependence on particle density, intensity distribution and interrogation window size. At last, synthetic particle images are tested and the errors are analyzed. The experimental results show that the modified method is a robust algorithm for measuring the movement of particles and the velocity field can be obtained with high precision.