Synthetic Data and Models for the paper "Learning-Based Control Barrier Functions for Safety-Critical Multi-Objective Optimization in Human-Robot Collaborative Manufacturing Systems"

<p dir="ltr">This dataset contains all training data, pre-trained models, and experimental results to reproduce the findings from "Learning-Based Control Barrier Functions for Safety-Critical Multi-Objective Optimization in Human-Robot Collaborative Manufacturing Systems" (Mathematics, MDPI, 2026).<br><br>Dataset includes: <br>- 50,000 labeled training samples for CBF neural network. <br>- Pre-trained Lipschitz-constrained neural network (98.3% validation accuracy) - Gaussian Process human motion predictor. <br>- Complete experimental data from 720 runs (4 scenarios × 6 methods × 30 runs) - CMU MoCap-based human motion sequences. <br>- Statistical analysis with Friedman and Wilcoxon tests.<br><br>All data is validated and consistent with paper results (Tables 2-5). <br><br>Source code: https://github.com/ClaudioUrrea/ur5-human_CoppeliaSim_EDU</p><p dir="ltr"><br></p>