It has been seen that there is a growing demand for the production of micron size grooves with a uniform flat surface, minimal heat-affected zone (HAZ) thickness, debris deposition, microcrack, microcavity, and recast layer thickness on orthopedic implants. In the present study, a multi-response optimization approach is adopted for fiber laser-based micro engraving to enhance groove characteristics and implant functionality during fixation. Grey relational analysis (GRA) is applied to analyze the experimental data sets and to decide key input factors' (pulse frequency, scanning speed, number of passes, laser power) level setting. These optimized data sets maintain groove quality (groove width, HAZ, debris deposition, recast layer thickness). Laser power and pulse frequency have been identified as the most significant factors for controlling groove quality. A neural network model has been developed and trained through experimental data sets. During analyzing the model, it has been recognized that the regression analysis score is very nearer to 1, and model performance is 2.45e^-18. It represents its adaptability for determining the response quality characteristics when factor level settings are out of defined boundary. Improvements in groove quality are noticed in terms of minimal HAZ thickness, debris deposition, recast layer thickness, microvoids, and microcavity under optimum engraving conditions.

已经看到，越来越需要在整形外科植入物上生产具有均匀的平坦表面，最小的热影响区（HAZ）厚度，碎屑沉积，微裂纹，微腔以及重铸层厚度的微米级凹槽。在本研究中，基于光纤激光的微雕刻采用了多响应优化方法，以增强固定过程中的凹槽特性和植入功能。应用灰色关联分析（GRA）来分析实验数据集并确定关键输入因素（脉冲频率，扫描速度，通过次数，激光功率）水平设置。这些优化的数据集可保持凹槽质量（凹槽宽度，HAZ，碎屑沉积，重铸层厚度）。激光功率和脉冲频率已被确定为控制凹槽质量的最重要因素。通过实验数据集已经开发并训练了神经网络模型。在分析模型期间，已经认识到回归分析得分非常接近1，模型性能为2.45e^-18。它表示当因子级别设置超出定义的边界时，用于确定响应质量特征的适应性。在最小的HAZ厚度，碎屑沉积，重铸层厚度，微孔和最佳雕刻条件下的微腔方面，凹槽质量得到了改善。