基于多元非线性回归方法的脉冲电化学光整加工表面质量预测研究
Surface Quality Prediction of Pulsed Electrochemical Finishing Based on Multiple Nonlinear Regression Method

本文对脉冲电化学光整加工零件的材料去除率以及表面粗糙度进行预测。首先对实验所得的数据进行分析，找出其中最主要的影响因素，然后从搜集到的数据中取出21组数据建立了数学回归模型。求解模型，对材料的去除率以及表面粗糙度进行了预测，并对模型进行了验证。根据验证结果可以看出，我们所得到的模型可以比较准确的预测出材料去除率以及表面粗糙度。最后我们依据以上结论对模型进行评估和改进，从而得出在其他条件一定时：粗糙度与电解液温度成反比，材料去除率与电解液温度成正比；粗糙度和材料去除率均与电流密度成正比；粗糙度受电解液温度影响更大，材料去除率受电流密度影响更大等结论。
In this paper, the material removal rate and surface roughness of parts processed by pulse electrochemical finishing are predicted. First, we analyze the data obtained from the experiment, find out the most important influencing factors, and then take 21 groups of data from the collected data to establish a mathematical regression model. The mate-rial removal rate and surface roughness were predicted by solving the model, and the model was verified. According to the verification results, the model we obtained can accurately predict the material removal rate and surface roughness. At last, we evaluate and improve the model according to the above conclusions, so that when other conditions are constant, roughness is inversely propor-tional to electrolyte temperature, and material removal rate is proportional to electrolyte temper-ature; roughness and material removal rate are directly proportional to current density; the roughness is more affected by electrolyte temperature, and the material removal rate is more af-fected by current density.