应用神经网络预测蛇形微通道中流体的流动特性

Prediction of Flow Characteristics of Fluid in Serpentine Microchannel Using Neural Network

基于伯努利方程建立惯性力作用下流体在蛇形微通道的流动方程。为预测惯性力作用下微流体的流动特性,提取了方程黏性力项中的阻尼特征系数,采用三层神经网络,建立微通道的结构特征参数与阻尼特征系数的映射关系。以数值仿真作为训练样本,通过神经网络提取其阻尼特征系数。制作了蛇形微通道的样机,通过离心实验验证理论流动特性。神经网络模型所预测的阻尼特征系数和实验得出的阻尼特征系数相差9.1%,证明神经网络模型能够较好地提取蛇形微通道的阻尼特征系数并预测微通道中流体的流动特性。

The flow equation of fluid in serpentine microchannel under inertia force was established based on the Bernoulli equation.In order to predict the flow characteristics of the microfluid under inertia force,the damping characteristic coefficient in the viscous force term of the equation was extracted.A three-layer neural network was used to establish the mapping relationship between structural characteristic parameters and damping characteristic coefficient of the microchannel.Numerical simulation was used as a training sample,and the damping characteristic coefficient was extracted by the neural network.A prototype of the serpentine microchannel was fabricated,the theoretical flow characteristics were verified by centrifugation experiments.The difference between the predicted damping characteristic coefficient by the neural network model and the obtained damping characteristic coefficient by the experiment is 9.1%,which proves that the neural network model can extract the damping characteristic coefficient of the serpentine microchannel and predict the flow characteristics of the fluid in the serpentine microchannel.