基于径向基函数神经网络的高聚物粘结炸药切削表面粗糙度预测研究  

Prediction Research on Cutting Surface Roughness of PBX Based on RBF Neural Network

作　　者：唐贤进[1] 张丘[1] 邹刚[1] 吴松[1] 刘维[1] 尹锐[1] 

机构地区：[1]中国工程物理研究院化工材料研究所,四川绵阳621900

出　　处：《兵工学报》2014年第2期200-206,共7页Acta Armamentarii

基　　金：中国工程物理研究院基金项目(2009A0203010);中国工程物理研究院化工材料研究所所长基金项目(626010939)

摘　　要：高聚物粘结炸药(PBX)的切削表面质量是影响炸药部件甚至是武器性能的重要因素。通过对PBX炸药切削表面三维轮廓的观察和分析,发现其切削表面形貌由于材料、工艺和工况等多因素作用而产生不确定崩落现象,导致对切削表面进行二维轮廓算术平均偏差与三维轮廓算术平均偏差的计算值最大相差32%.基于此,综合考虑多因素影响,利用RBF人工神经网络构建了炸药切削表面粗糙度预测模型。通过网络训练和验证表明,该模型基本能够反映PBX炸药切削表面成形的基本规律,并且预测值与实际值误差不超过3%.The surface quality of polymer-bonded-explosive（PBX） is a key factor to influence the explosive components and the weapons. An avalanche phenomenon, which is created by material, process and operating condition, on the cutting surface of PBX is observed by analyzing the cutting surface 3 D-contour of PBX, which causes the difference between 2D arithmetical mean deviation of the profile and 3D arith- metical mean deviation of the profile to be 32%. Hence, a prediction model which consideres the multi- ple factors is established with the RBF neural network. The training and test of the prediction model illustrates that the model could reflect the regularity of cutting process, and the predicted error is within 2%.

关 键 词：机械制造工艺与设备 高聚物粘结炸药 表面形貌 粗糙度 神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]