中文摘要：

为实现高速加工时刀具渐变磨损状态的在线准确识别，提出了一种集合多种智能的间接检测刀具磨损状态方法的模糊数据融合方法。尽管这些方法具有算法实现较为简单、处理速度较快的优点，但单一的信号检测及单一的智能建模方法难以获得全面的加工状态信息和准确的识别结果。为此，利用F推理技术对上述方法的冗余和互补信息进行数据融合，应用Makino—Fanuc 74-A20型加工中心的测试数据验证了该方案的可行性，并将刀具后刀面磨损的预测值与基于机器视觉检测的实测值进行比较。实验结果分析表明，多参数模糊融合识别方法能快速获得切削刀具磨损状态更加准确的预测值。

英文摘要：

For detecting gradual tool wear state on line in high cutting speed,the methods of wave- let fuzzy neural network9 regression neural network and sample classification fuzzy neural network by detecting cutting force, motor power of machine tool and AE signal respectively are presented. A1- though these methods are not difficult to come true and processed accurately and rapidly,it is difficult to obtain comprehensive information of machining and exact value of tool wear when using single method of intelligent modeling and single signal detecting. For this purpose, fuzzy inference technique is adopted to fuse the recognized data. Emu-lation experiment is carried out by using Matlab software platform and this method is verified to be feasible. Experimental result indicates that by applying fuzzy data fusion, we can get an exact tool wear forecast rapidly.