Simulation Data for Hazard Prediction in Artificial Pancreas Systems

该数据集是由模拟人工胰腺系统（APS）在注入不同故障情景下使用患者模拟器生成的数据。它包含了患者特征、初始血糖值以及故障注入情景的多种变化，这些变化为训练和评估安全监控器的鲁棒性提供了丰富的数据。该数据集规模达到2,646,000个模拟样本，旨在应对医疗信息物理系统中的危险预测与安全监测任务。

This dataset is generated by an artificial pancreas system (APS) using a patient simulator under various injected fault scenarios. It includes patient characteristics, initial blood glucose levels, and diverse variations of fault injection scenarios, which provide rich data for training and evaluating the robustness of safety monitors. With a total of 2,646,000 simulated samples, this dataset is designed to address the tasks of hazard prediction and safety monitoring in medical cyber-physical systems.