Tennessee Eastman Reference Data for Fault-Detection and Decision Support Systems

Extended Tennessee Eastman Reference Data as presented in Reinartz et al. (2021). <br><br>The data contains simulations of 28 process faults for 6 different operating modes. Each fault is simulated 500 times with different seeds for the random number generator. <br>Additional simulations feature setpoint changes and mode transitions between the 6 operating modes. <br>All simulations have a duration of 100 hours with a sampling rate of 3 minutes. <br>Please refer to the Readme and publication for detailed information. <br>

如Reinartz等人（2021年）的研究中所提出的扩展田纳西东蒙基准数据集（Extended Tennessee Eastman Reference Data）。<br><br>该数据集包含针对6种不同运行模式下的28种过程故障的仿真数据。每种故障均采用不同的随机数生成器种子，重复开展500次仿真实验。额外仿真实验涵盖了6种运行模式间的设定值变更与模式切换场景。所有仿真实验的时长均为100小时，采样率为3分钟。如需获取详细信息，请参阅数据集说明文档（Readme）及相关发表文献。