基于粒子群优化Hopfield网络匹配的稀相颗粒速度测量  

Measurement of particle velocity in dilute phase based on PSO and Hopfield network matching

作　　者：周云龙[1] 宋连壮[2] 周红娟[2] 

机构地区：[1]东北电力大学能源与机械工程学院 [2]东北电力大学自动化工程学院,吉林省吉林市132012

出　　处：《化工学报》2011年第2期348-354,共7页CIESC Journal

摘　　要：提出了一种基于粒子群算法(PSO)和Hopfield神经网络相结合的粒子跟踪测速算法。该方法采用高速摄影系统拍摄气固两相流的稀相颗粒运动图像,经图像处理后,提取形心参数。将粒子匹配问题转化为优化问题,采用粒子群优化算法与Hopfield神经网络相结合的方法进行优化,求出最优解来实现颗粒的正确匹配,然后计算出颗粒的速度矢量,并与互相关法求出的速度进行对比,实验结果表明,该方法能准确地跟踪稀相颗粒,是一种有效的稀相流场速度测量方法。A particle tracking velocimetry algorithm was proposed based on particle swarm optimization（PSO） and Hopfield neural network.Dilute-phase particle images were captured by digital high-speed video system in gas-solid flow,and the image centroid parameters were extracted by using image processing techniques.The particle matching was changed to a optimization problem,which was realized by using particle swarm optimization algorithm and Hopfield neural network,and the correct matching of particles was achieved by getting the optimal solution.Then the velocity vector of particles was calculated with PSO and Hopfield network matching and compared with the one obtained by using the cross-correlation algorithm.The results showed that this method could accurately track dilute-phase particles and could be an effective way to measure the flow velocity of dilute-phase.

关 键 词：稀相输送 图像处理 粒子群优化 HOPFIELD网络 粒子跟踪测速技术 

分 类 号：O359.1[理学—流体力学]