基于信息融合的精密磨削砂轮磨损状态在线识别方法研究

ON-LINE IDENTIFICATION OF WHEEL WEAR CONDITION IN PRECISION GRINDING BASED ON INFORMATION FUSION

高质量非球面光学元件批量制造是目前精密磨削技术力求实现的目标。为了提高非球面光学元件精密磨削的加工效率,必须在加工过程中动态识别砂轮磨损状态,在砂轮接近或达到寿命周期时对其进行修整。寻求一种经济可行的方式,实现砂轮寿命周期在线评估,利用声发射、砂轮振动、磨削力等多种类型加工过程信号,提取和选择能够全面、灵敏反应砂轮磨损状态的特征,基于Dempster-Shafer证据理论,进行多源信息融合,实现精密磨削砂轮磨损状态在线识别。

Grinding wheel should be addressed when it just reaches its expectancy in order to achieve high machining efficiency of aspheric optical lens. The realization of on-line estimating wheel life is based on the automatic identification of the wheel wear condition. Characters of dynamical process signals change accompanied with the lapse of the wheel life. Therefore, process signals can be used to monitor and estimate the wheel condition. Acoustic emission （AE） , wheel vibration and grinding force are picked up to abstract three kinds of representative monitoring parameters. They are the skewness of the AE power spectrum, the complexity degree of the wheel vibration and the ratio of normal component and tangential component of the grinding force. These monitoring parameters are sensitive to different macro-and micro-wear of grinding wheel. The Dempster-Shafer evidence theory, which is one of the decision-level information fusion technologies, is employed to acquire a reliable decision about the status of the grinding wheel.