摘要
从实验稳定阈图出发，将切削参数n，b，f与颤振状态贴近度I一起构成模式向量，通过聚类分析得到不稳定区和条件稳定区模式中心，利用颤振状态贴近度I置补的方式，以欧氏距离最小为准则建立一求无颤振切削参数值的神经网络方法，实验证明了该方法的可行性与有效性。
From a testing stability chart, pattern vector is constructed with thecutting parameters (n.b.f) and the chatter state closing degree I. The pattern center of instability and stability region is obtained through clustering analysis. A neural network searching method for free－chatter cutting parameters using (1 －I) andin accordance with the criterion of minimum Euclidean distances is developed. Experimental results proved the feasibility and effectiveness of this system.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
1998年第4期335-338,共4页
Acta Armamentarii
