基于小波网络的砂轮状态监测方法  

ON-LINE INTELLIGENT MONITORING OF THE GRINDING WHEEL STATE BY WAVELET-NETWORK-BASED METHOD

作　　者：刘贵杰[1] 巩亚东[2] 王宛山[2] 

机构地区：[1]山东轻工业学院机电工程系,济南250100 [2]东北大学机械工程及自动化学院,沈阳110004

出　　处：《金刚石与磨料磨具工程》2003年第3期21-24,共4页Diamond & Abrasives Engineering

基　　金：教育部科学技术研究重点资助项目 (2 0 0 0 32 )

摘　　要：提出了一种基于小波神经网络和模糊控制的砂轮状态在线监测方法 ,该方法通过提取磨削加工过程中有关砂轮状态的声发射信号和功率信号 ,利用小波神经网络 ,实现砂轮状态的在线智能化监测 ;针对多输入输出带来的网络规模增大、收敛速度缓慢等问题 ,提出采用尺度参数的自适应调整法及平移参数的寻优搜索法 ,寻找最优小波基元 ,同时采用模糊自适应BP算法对训练速率系数 η和惯性系数α进行在线调整 ,从而简化了小波网络 ,减少了学习次数 ,加快了网络的收敛速度 。A wavelet neural network & fuzzy BP algorithm monitor method was developed to realize the on line intelligent monitoring of the grinding wheel state. The method works by collecting the acoustic emission (AE) signals and motor power signals about the grinding wheel state during the grinding process. In order to solve the problem of the'dimension disaster' and the slow learning speed caused by multi input & output, the wavelet network was improved by optimum algorithm that could adjust and search for the wavelet parameter adaptively, at the same time the fuzzy adaptive BP algorithm was used to on line adjust the training speed coefficient η and inertia coefficient α. Finally, a simplified structure and quickly convergent velocity of the new algorithm was demonstrated by the simulation results.

关 键 词：砂轮 状态监测 小波网络 参数寻优搜索 模糊自适应BP算法 磨削加工 

分 类 号：TG743[金属学及工艺—刀具与模具]