摘要
为了研究电火花微小孔加工工艺过程,通过对一些智能算法的比较分析,采用线性规划(LP)形式的ε-支持向量回归机(ε-SVR)这种算法,对电火花微小孔加工进行了加工工艺的仿真。其中结合正交试验获得所需样本,并将样本随机分配为训练样本和预测样本,利用训练样本建立微小孔加工速度的仿真模型,再利用预测样本对仿真模型进行验证。结果表明:预测结果与试验结果较为一致,所建立的电火花微小孔加工速度模型可用,本方法可行性较好。
In order to study the EDM machining process for the micro-hole processing technology,it analyzed and compared several intelligent algorithms,finding out a way to simulate the EMD machining process for the micro hole processing technology by using the ε-support vector regression of linear programming （LP）,which combined the training samples with the prediction samples that had gotten from the orthogonal experiment,then established the simulation model of micro hole processing speed by the training samples.Final verification of the model by the prediction samples,it turned out that this method is feasible becaue of and the EMD machining speed modeling is available and the prediction results in conformity with the test results.
出处
《机械设计与制造》
北大核心
2013年第11期251-252,255,共3页
Machinery Design & Manufacture
