FDM Environmental sensor data

This is a real manufacturing dataset that contains environmental sensor values of the additive manufacturing process of a fused deposition modeling (FDM) system. The sensor values in this dataset were used to monitor, analyze and document the quality of the 3D printing process and the 3D printed parts. The generated dataset consists of individual data files with the recorded environmental sensor values for several FDM print jobs.

The dataset consists of two folders and two CSV files:
The Environmental_sensor_dataset folder contains sensor values of different 3D printing conditions of the FDM process. The folder is divided into subgroups, each containing 40 sequenced environmental sensor data files from one print job with exactly 500 consecutive pairs of values and a title line for the column names. The Environmental_sensor_dataset.csv file is a combined and unbalanced version of all 3D printing conditions and environmental sensor data sequences of the environmental_sensor_dataset folder for machine learning (ML) analyzes with different algorithms.

The Environmental_sensor_ablation folder contains also sensor values of different 3D printing conditions of the FDM process. However, this folder contains far fewer sensor values than the Environmental_sensor_dataset folder. The folder is also divided into subgroups, each of which contains 24 sequenced environmental sensor data files from one print job with exactly 500 consecutive pairs of values, but only for one print job of the respective 3D printing condition. The Environmental_sensor_ablation.csv file is a combined and balanced version of all 3D printing conditions and environmental sensor data sequences in the environmental_sensor_ablation folder for an ablation study that aims to investigate the influence of certain analysis conditions (in this case the size of the data set and the distribution of the data) on changes in the performance of the ML algorithms.

A more detailed description of the dataset can be found in the associated article:

https://doi.org/10.1016/j.addma.2021.102535

本数据集为真实制造类数据集，收录了熔融沉积成型（Fused Deposition Modeling, FDM）系统增材制造过程中的环境传感器数据。本数据集收录的传感器数据可用于监测、分析并记录3D打印过程及打印成品的质量。本数据集包含多组FDM打印任务对应的环境传感器记录数据，以独立数据文件形式存储。

本数据集包含两个文件夹与两个CSV文件：
Environmental_sensor_dataset文件夹收录了FDM工艺下不同3D打印工况对应的传感器数据。该文件夹下设多个子分组，每个子分组包含来自单次打印任务的40个按序排列的环境传感器数据文件，每个文件均包含恰好500组连续数值对，以及一列用于标注列名的标题行。Environmental_sensor_dataset.csv文件为该文件夹下所有3D打印工况与环境传感器数据序列的合并版本，且数据分布存在不平衡性，可用于搭载不同算法的机器学习（Machine Learning, ML）分析任务。

Environmental_sensor_ablation文件夹同样收录了FDM工艺下不同3D打印工况对应的传感器数据，但该文件夹包含的数据量远少于Environmental_sensor_dataset文件夹。该文件夹同样下设多个子分组，每个子分组仅包含对应单一3D打印工况单次打印任务的24个按序排列的环境传感器数据文件，每个文件同样包含恰好500组连续数值对。Environmental_sensor_ablation.csv文件为该文件夹下所有3D打印工况与环境传感器数据序列的合并版本，且数据分布经过均衡处理，可用于消融实验。该实验旨在探究特定分析条件——即数据集规模与数据分布——对机器学习算法性能变化的影响。

本数据集的详细说明可参阅关联论文：
https://doi.org/10.1016/j.addma.2021.102535