遗传算法优化的模糊神经网络在刀具磨损诊断中的应用  

Application of GA-fuzzy-neural Networks in Tool Wear Diagnosis

作　　者：曹伟青[1] 傅攀[1] 张尔卿 

机构地区：[1]西南交通大学机械工程学院,成都610031

出　　处：《机械科学与技术》2014年第11期1682-1687,共6页Mechanical Science and Technology for Aerospace Engineering

基　　金：中央高校基本科研业务费专项资金项目(SWJTU12CX039);国家科技支撑计划项目(2009BAG12A01-E04)资助

摘　　要：针对刀具磨损监测时,采集的振动信号含有强烈的背景噪声,难以提取故障频率的问题,提出采用形态滤波消噪后进行经验模态分解来提取故障频率;同时,为了准确监测刀具的磨损状态,将提取的故障特征输入到遗传算法优化的模糊神经网络对刀具的磨损进行识别,模糊神经网络的基函数采用B样条基函数。传统的网络学习算法采用梯度下降法,这在学习过程中容易陷入局部最小,论文采用遗传算法寻求全局的最优解。实验表明,该方法能有效地应用于强噪声背景下的刀具磨损识别。In view of the strong background noise involved in the fault signal of tool wears and the difficulty to obtain fault feature frequencies, in this paper, a fault feature extraction method was proposed based on morphological filters and combining with empirical mode decomposition. At the same time,tool wears were identified by genetic algorithm（ GA）-fuzzy-neural networks with B-spline membership functions. Fuzzy neural networks are traditionally trained by using gradient-based methods,and may fall into local minimum during the learning process. So,the genetic algorithm was adopted for global optimization in this study. The experimental results show that the diagnosis approach put forward in this paper can effectively identify tool wears in strong background noise.

关 键 词：经验模态分解 形态滤波 B样条基函数 模糊神经网络 遗传算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TH165.3[自动化与计算机技术—控制科学与工程]