利用神经网络外推预测干燥过程降水率

利用神经网络所具有的捕捉过程输入－输出之间的非线性关系的能力和强大的学习推理能力，给出了一种基于神经网络的外推预测垂直对撞流干燥过程降水率的方法．为提高预测的快速性和准确性，针对ＢＰ算法学习参数难以确定的缺点，给出了一种基于目标函数的一阶和二阶导数同时优化学习率和确定动量系数的方法，并将此法应用于外推预测物料降水率的过程之中．仿真结果表明，对于运动规律十分复杂、目前仍无法从其内部的运动机理和传热传质特性出发建立干燥特性预测模型的一类高强度的干燥方式而言，文中所提出的神经网络模型能够较正确、快速地预测干燥过程中物料降水率的变化．

Predicting Water Content of Drying Process Using Neural Network Model

Proper modelling of a dryer is an important basis for designing a dryer control system and developing the related analysis,prediction and computer simulation.A method to simultaneously determine the optimal learning rate and momentum coefficient based on the first and second derivatives of the objective function with respect to the learning rate was proposed and applied to simulate the drying process of a vertical impingement stream dryer.The simulation results show that for this kind of new type high efficient dryers,which can not be modelled based on the motion as well as heat and mass transfer of products during drying,the neural network model given in this paper is capable of predicting the behaviour of the dryers exactly and rapidly.