刀具磨损的在线监测

On-line Tool Wear Sensing

采用模式识别技术，对反映刀具磨损状态的几个特征量构成的状态模式进行分析。针对状 态模式［６_(ＴＲ)～２ρ１〕Ｔ和［σ＿（ｘ）～２ρ＿１］Ｔ的分布统计特性，构造了具有自适应性的“与 门”式刀具磨损状态的综合判据－无人管理学习分类器。结果证明，由状态模式［６_(Ｔｒ)~2ρ_1] 构成的刀具磨损状态综合判据在识别精度、通用性、实时性及灵敏度方面都令人满意。

The state patterns based on the charcteristics related on tool wear are analysed by the pattern recognition technigue. The recognizing critera being self-adaptive in reterence to the state patterns [σ_(TR)~2ρ_1]T and [σ_x^2ρ_1]T are given .Experimental results show that the criterion refering to [σ_(TR)~2ρ_1] T proves to be satisfactory in recognizing accuracy, flexibility, practicability and sensitivity.