欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
提出基于PSO混合算法对BP网络进行优化,改善了BP网络权阈值求解所面临的局部极值与收敛速度慢的问题。通过赫兹模拟试验方法,在JD-1轮轨模拟试验机上获得BP神经网络训练数据,运用基于PSO混合算法优化BP网络对钢轨磨损量进行拟合。结果表明：基于PSO混合算法优化BP网络对钢轨磨损量数据能够进行极好的拟合,为钢轨磨损量预测提供一种更有效的数据分析途径。
英文摘要：
Hybrid algorithm based on PSO algorithm is pro-posed in this paper,which makes the local extremum and slows convergence velocity of BP ANN faster.With the Hertz testing simulation method,the experiment investigation on rolling wear of railwheel is performed on JD-1 wheel / rail simulation facilities obtain wear volume as ANN’s training data.Then,PSO hybrid algorithm is applied in the fitting of wear volume data of rail steel.The results show that PSO hybrid algorithm can be fitting for wear volume data of rail steel perfectly,which provides a more effective data analysis approach for prediction rail steel wear volume.
同期刊论文项目
同项目期刊论文
Real time observation and friction vibration analysis of torsional fretting wear of polymethylmethac
An experimental study on bending fretting fatigue characteristics of 316L austenitic stainless steel
Torsional wear behaviors of UHMWPE against Ti6Al4V alloy and alumina femoral head ball in dry condit
Comparison on the torsional fretting wear of nitrogen ion implanted pure titanium and Ti6Al7Nb alloy
The Torsional Fretting Wear Behavior of Nitrogen Ion Implantation Layer of Ti6Al7Nb Ally Under Bovin
期刊信息
位置：