一种改进的粒子图像测速混合算法研究  

An Improved Hybrid Algorithm for Particle Image Velocimetry

作　　者：王甜[1] 房红兵[1] 黄海龙[2] 王驰[2] 

机构地区：[1]南京理工大学电子工程与光电技术学院,南京210094 [2]南京水利科学研究院,南京210024

出　　处：《长江科学院院报》2017年第7期144-148,共5页Journal of Changjiang River Scientific Research Institute

基　　金：中央级公益性科研院所基本科研业务费专项基金(2011YQ070055)

摘　　要：粒子图像测速技术作为一种新的流场测速方法能够在不干扰流场的情况下获得整个流场的速度信息。粒子图像测速技术最关键的步骤在于粒子匹配。针对粒子密度分布不均匀、流场不同等实际情况,提出了混合算法,即结合互相关和松弛算法能够更准确地搜索粒子,进而对粒子进行匹配。对3种匹配算法的匹配概率进行比较分析,发现混合算法能更准确地分析粒子的运动状态,减少错误矢量的产生;另外,对松弛算法进行改进,通过优化筛选加权因子发现改进的松弛算法在运行速度上相比原始算法有了较大提高,匹配率与原始算法基本一致。As a new method of flow velocity measurement, particle image velocimetry （PIV） could obtain velocity information of the whole flow field without disturbing the flow field. The most critical step in PIV is particle matc-hing. A hybrid algorithm combining cross-correlation algorithm and relaxation algorithm is proposed in view of the actual conditions of uneven distribution of particle density and different flow fields. The hybrid algorithm could search the particles more accurately so as to match the particles. The matching probabilities of three matching algo-rithms are compared and results suggest that the hybrid algorithm can analyze the motion state of particles more ac-curately and reduce the generation of error vectors. In addition, the relaxation algorithm is improved in this paper. By optimizing weighting factor, the running speed of the improved relaxation algorithm has greatly improved com-pared with the original algorithm, while the matching rate is basically consistent with the original algorithm.

关 键 词：流场速度 粒子图像测速 混合算法 超松弛迭代粒子追踪 粒子匹配 匹配概率 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]