基于BP神经网络的水力旋流器建模仿真研究.pdf

29 9 煤 矿 机 械 Vol29No9 2008 9 Coal Mine Machinery Sep. 20 08 基于BP 神经网络的水力旋流器建模仿真研究 , , , , ( , 266510) : 通过对某一固定的水力旋流器自身工艺的研究, 确定水力旋流器的输入与输出分 2 种情况对有相互 响的输入参数分别建立BP 神经网络的模型和利用Matlab 工具箱中的BP 神经 网络建立水力旋流器的三层神经网络模型通过对收集的该设备的实例数据进行仿真训练, 结果 表明不仅是旋流器本身参数, 调浆槽的液位等也会对旋流器效率产生 响 : 水力旋流器; BP 神经网络; 模型; 样本训练 : TD462; TP391 : A : 2008) Research onModel Simulationof HydrocycloneBasedonBP Neural Network ZHANG Jing, XU en- shang, GAI en- dong, YU Qing- ming, YUZhen- bo ( College of nformation and Electrical Engineering, Shandong University of Science and T echnology, Qingdao 266510, China) Abstract: Based on the research of a f ixed hydrocyclone technics, definite the inputs and outputs of hydrocy clone. On the foundation of the input parameters with mutual influence divide them into two different situations to establish separately the model of BP NN( Neural Network) and use the BP neural network toolbox in Matlab to establish hydrocyclone three neural network model. Carrying on the simulation training of the equipments collected example data, the result can indicate that not only the hydrocyclone parameters, but also the level of tank have the influence to the efficiency of hydrocyclone. Keywords: hydrocyclone; BP neural network; model; sample training 0 , 4 4 : [ 1] , . [ J] . , 1994, 30( 4) : 58- 63.