Abstract

Drilling trials were conducted on E glass woven composites and machine learning models were employed for correlating drilling variables namely point angle, feed rate and spindle speed with response measures. The thrust force, surface roughness (Ra) and burr height were considered as performance characteristics in this study. Using self-organizing map (SOM) method, Neighboring weight distances in SOM with 10 × 10 neurons, Hit map corresponding to the neighboring weight distances, Wight planes showing weight distributions were developed. The validation tests have also been conducted to verify the results obtained by ANN technique. The predictions of the artificial neural network (ANN) model result were in good agreement with experimental results.

摘要

在 E 玻璃编织复合材料上进行了钻孔试验，并采用机器学习模型将钻孔变量（即点角、进给率和主轴速度）与响应措施相关联。在这项研究中，推力、表面粗糙度 (Ra) 和毛刺高度被视为性能特征。使用自组织图 (SOM) 方法，开发了具有 10 × 10 神经元的 SOM 中的相邻权重距离、对应于相邻权重距离的命中图、显示权重分布的 Wight 平面。还进行了验证测试以验证人工神经网络技术获得的结果。人工神经网络（ANN）模型结果的预测与实验结果非常吻合。