基于盲源信号分离的加工误差分离方法研究  

Independent Component Analysis Based on Machining Error Separation

作　　者：张发平[1] 吴迪[1] 张体广[1] 张凌雲 杨吉彬 ZHANG Fa-ping WU Di ZHANG Ti-guang ZHANG Ling-yun YANG Ji-bin(School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China)

机构地区：[1]北京理工大学机械与车辆学院,北京100081

出　　处：《兵工学报》2016年第9期1692-1699,共8页Acta Armamentarii

基　　金：国家自然科学基金项目(51275049);国防"973"计划项目(613243)

摘　　要：针对当前加工误差分离方法无法分离相近尺度系统误差的缺点,提出了一种基于独立成分分析的加工误差分离方法。建立了各误差源所导致的系统误差和总的系统误差之间的传递模型;根据盲源信号分离相关理论,建立了基于负熵固定点算法的加工误差分离模型,实现了相近尺度的系统加工误差分离,并利用主成分分析法给出了误差源数量确定的方法。以某陀螺仪的某加工面加工误差分离为例对所提出的方法进行了验证,结果表明所提出的误差分离方法可有效实现相近尺度的加工误差分离。For the unsuccessful separation of the multiple systematic errors with similar scales by current machining error separation method, a new method to separate the machining errors is proposed based on independent component analysis. An error transfer model is built to describe the relationship between the systematic error caused by individual error source and the final systematic error measured from the machi- ning surface. Then according to the theory of blind signal separation, an optimization model is used for the machining error separation where the negative entropy of the estimated error is used as the optimization objective function, and the fixed point algorithm is used as the optimization method. A method to determine the number of machining error sources is also given by means of the principal component analysis. A study case of a certain gyroscope surface is tested to verify the efficiency of the error separation method. The proposed method provides a new way for separation of machining errors and tracing of error sources.

关 键 词：机械制造工艺与设备 误差源 误差传递模型 独立成分分析 主成分分析 误差分离 

分 类 号：TH165.3[机械工程—机械制造及自动化]