中文摘要：

电化学砂带超精加工（ECBS）轴承滚道比传统油石超精法在加工质量及加工效率上具有优势,但其加工效果受很多因素和复杂的动态关系影响.为预测加工质量和选择加工参数,基于最小二乘支持向量机（LS-SVM）理论,在ECBS实验基础上建立了加工参数间多元回归非线性智能化预测模型.结合正交实验,评价了光整加工参数对表面粗糙度的影响,并获得训练预测模型所需的数据样本.研究结果表明：预测值和实验值具有较好的一致性,表面粗糙度预测值的平均绝对百分误差为3.33%,电流密度预测值的平均绝对百分误差为2.52%.

英文摘要：

Electrochemical belt superfinish （ECBS） technology applied to bearing raceway finishing has the advantages of high surface quality and processing efficiency,and it is superior to traditional oilstone superfinish. However,the finishing effect of ECBS is dominated by many factors and complicated dynamic behavior of the factors. Therefore,it is difficult to predict the finishing results and select the suitable processing parameters in ECBS. To solve this problem an intelligent multiple regression predictive model thinking about the non-linear relationship between processing parameters and the machining effect was established based on least squares support vector machines （LS-SVM）. Taguchi method was introduced to assess the effect of finishing parameters on surface roughness,and the training data was also obtained through experiments. The comparison between the predicted values and the experimental values under the same condition is carried out,and the results show that the predicted values are approximately consistent with the experimental ones. The mean absolute percent error is 3.33% for surface roughness and 2.52% for current density.