基于RBF的AR模型在数控振动仿真中的应用  

The Application of RBF Networks-based Autoregressive Model in Vibration Simulation of NC Machine Tool

作　　者：朱英杰[1] 王国锋[2] 王太勇[2] 

机构地区：[1]天津商学院机械工程学院,天津300134 [2]天津大学机械工程学院,天津300072

出　　处：《中国制造业信息化（学术版）》2006年第4期32-35,共4页

摘　　要：为了更加有效地提高生产效率,数控加工仿真系统在进行几何仿真的同时,还要对加工系统的动态性能进行分析。提出了一个对数控加工过程中振动信号仿真的新方法,可以实现加工状态下的车削振动信号的仿真。该方法应用RBF神经网络预测AR模型的参数,从而建立不同切削条件下的振动仿真模型。实验表明,仿真信号和测量信号可以很好的吻合。To increase productivity in manufacturing, it uses the accurate cutting simulation systems to study the performance of machining process. It presents a new method which can simulate the vibration in NC machine tools under working condition. According to the features of vibration signal, it builds autoregressive model, in which the model parameters is obtained by RBF（radius basis function）network so as the system can simulate the vibration under different working conditions. There is a good agreement between the simulated and measured signal.

关 键 词：RBF 神经网络 AR模型 振动仿真 NC车削 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]