基于核极限学习机的碳化钨涂层砂带磨削表面粗糙度的研究

Research on Surface Roughness of Tungsten Carbide Coating Belt Grinding Based on Kernel Extreme Learning Machine

采用单因素和正交试验方法研究砂带磨削碳化钨涂层过程中各磨削参数与表面粗糙度的关系,发现砂带磨削碳化钨涂层的表面粗糙度的变化规律:砂带粒度对表面粗糙度的影响最大;磨削时磨削压力对表面粗糙度的影响较大;砂带电动机转速频率和工件主轴旋转速度对表面粗糙度的影响较小,其中工件旋转速度影响最小。同时,应用人工智能算法,分别采用极限学习机(ELM)和核极限学习机(KELM)算法建立碳化钨涂层表面粗糙度的预测模型,进行了相关对比试验验证,KELM具有较好的预测效果。

This paper uses single factor and orthogonal test methods to study the relationship between various grinding parameters and surface roughness in the process of abrasive belt grinding of tungsten carbide coatings.It is found that the surface roughness of the tungsten carbide coating changes during the abrasive belt grinding process:the particle size of the abrasive belt has the greatest influence on the surface roughness;the grinding pressure has a greater influence on the surface roughness during grinding;the abrasive belt motor rotation speed and the workpiece spindle rotation speed have small effect on the surface roughness,and the rotation speed of the workpiece has the least effect.The artificial intelligence algorithm is applied,and the Extreme Learning Machine(ELM)and Kernel Based Extreme Learning Machine(KELM)algorithms are used to establish the prediction model of the surface roughness of the tungsten carbide coating,and the relevant comparative experiments are carried out to verify that KELM has a good prediction effect.