基于RBF神经网络的机械加工误差质量模型  

The Quality Model of Mechanic Processing Error Based on RBF Network

作　　者：何涛[1] 龚立雄[1] 钟飞[1] 

机构地区：[1]湖北工业大学机械工程学院,湖北武汉430068

出　　处：《湖北工业大学学报》2007年第1期57-59,共3页Journal of Hubei University of Technology

基　　金：湖北省现代制造质量工程重点实验室开放基金(LMQ2005A02)

摘　　要：针对机械加工过程中产生误差的复杂性,用传统的数学建模方法难以实现加工误差和工艺参数、系统刚度和加工过程变量之间的关系模型.采用RBF神经网络建立加工误差的质量模型,并通过实验仿真,取得了较好的仿真效果.证明了RBF网络模型具有很强的逼近能力,可用于各种质量建模,达到减少加工误差、指导生产的目的.It is complicated and impossible to construct relational model of the processing error and the craft parameter, the system rigidity and the process variable by using conventionally mathematics methods. This article establishes the quality model of Mechanic Processing Error by using the RBF network, and has obtained the better simulation effect. It is proved the RBF network model has a very strongly approximate ability, which can be applied to all kinds of quality modeling, which can achieve the goal of reducing processing error and instructing oroduction.

关 键 词：径向基函数 神经网络 MATLAB 加工误差 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]