Subjects: Physics >> General Physics: Statistical and Quantum Mechanics, Quantum Information, etc. submitted time 2018-01-05 Cooperative journals: 《工程热物理学报》
Abstract：本文通过人工神经网络预测方法对非球形颗粒气固曳力系数进行了预测及分析。首先比较了BP（Backpropagation）神经网络模型和RBF（Radical Basis Function）基神经网络模型对Pettyjohn 和Christiansen 等人实验工况中的结果进行了预测。结果表明，采用RBF 方法预测非球形颗粒气固曳力系数误差较小，计算效率较高。同时，应用RBF 基神经网络模型，对不同形状因子下的气固曳力系数进行了预测和分析。研究结果表明，人工神经网络可以用于非球形颗粒气固曳力系数的预测研究，本文研究结果为复杂形状颗粒气固曳力系数的预测提供了一种有效的手段。
Subjects: Dynamic and Electric Engineering >> Engineering Thermophysics submitted time 2017-11-07 Cooperative journals: 《工程热物理学报》
Abstract： In this paper, the prediction and investigation on drag coefficient of non-spherical particles is presented applying artificial neural network. The performance between BP (Backpropagation) model and RBF (Radical Basis Function) model is compared. The RBF model is employed to predict the drag coefficient of non-spherical particles with higher efficiency and less error. The simulation results are compared with the experimental results in the literature by using RBF model. It reveals that artificial neural network can be applied to the prediction on the drag coefficient of non-spherical particles.
Hits
YES
Hits
Downloads
Comment