浮选过程气泡流场体视化测试方法与运动机理研究

浮选（Flotation）是一种以多相流流体力学理论为基础的矿物分选技术，气泡作为浮选过程的关键媒介直接影响浮选效率。测量气泡流场的时空特性，对于指导工业生产设计、推动浮选机理发展具有重要意义。本课题将实验测量与人工智能相结合，旨在实现浮选气泡三维流场的时空特性智能测量和流动机理准确描述。主要研究内容有：（1）研究用于三维高速动态显微图像测试(3D-SDMI）技术的三维图像测量方法，建立时/空多尺度智能预测模型，实现气泡参数的智能测量。（2）基于三维粒子图像测速（3D-PIV）技术，开发深度学习体视重构算法，实现气泡流场的体视化，分析气泡的三维时空特性。（3）建立基于智能预测与运动机理的混联模型，提高特征参数预测精度，获得气泡运动机理描述的新方法。

Flotation is a mineral separation technique, which is based on the multi-phase flow mechanics. As a key medium of the flotation process, bubbles affect the efficiency of flotation directly. Researching on the temporal-spatial characteristics of bubble flow is very important to improve the industrial design production and promote the development of flotation theory. In order to realize the intelligent measurement of the spatial-temporal characteristics and flow mechanism of the three-dimensional bubble flow field in flotation, experimental measurement and artificial intelligence technique are combined in this project. The main contents are as follows: (1) Study the three-dimensional image measuring method for the Three Dimensional measurement with Speedy-Dynamic-Microscopic Image (3D-SDMI) technique. The intelligent prediction models based on spatial and temporal scale are established to predict the bubble parameters. (2) Develop the depth learning stereoscopic reconstruction algorithm for the Three Dimensional Particle Image Velocimetry (3D-PIV) technique. Then, get the 3D spatial-temporal characteristics of the bubble flow based on the reconstructed space. (3) Establish the series-parallel fusion models of intelligent prediction and dynamic mechanism, which could improve the prediction accuracy of parameter and develop the description of bubble flow mechanism.