Dataset and Code for "Optimization of selective laser melting process parameters for 316L stainless steel based on machine learning"

This repository contains all experimental and simulation data, as well as complete machine learning code, used in the study on optimizing SLM process parameters for 316L stainless steel based on machine learning. The contents include: (1) Raw experimental data: SLM process parameters (laser power, scanning speed, substrate preheating temperature, etc.) and corresponding maximum residual stress test results; (2) Preprocessed data: Z-score normalized datasets for machine learning modeling, divided into training/validation/test sets at an 8:1:1 ratio; (3) Machine learning code: MATLAB scripts for nine regression algorithms (MLR, LSBoost, GPR, GKR, GAM, DTR, RF, BP, LSTM), including model training, cross-validation, and performance evaluation (MAE, MAPE, MSE, RMSE, R²); (4) Simulation files: Ansys 2022 Workbench finite element model parameters for SLM temperature and stress field simulation; (5) Documentation: Detailed descriptions of data fields, code running environments, and simulation setup steps to ensure full reproducibility of the research results. All data and code support the findings on laser power as the dominant factor of residual stress and the optimal SLM process window (200 W laser power with 950/1050 mm/s scanning speed) for 316L stainless steel.