当前位置: X-MOL 学术Comput. Chem. Eng. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
INDUSTRIAL DATASETS AND A TOOL FOR SISO DATA VISUALIZATION AND ANALYSIS
Computers & Chemical Engineering ( IF 3.9 ) Pub Date : 2020-12-09 , DOI: 10.1016/j.compchemeng.2020.107198
Jônathan W. V. Dambros , Jorge O. Trierweiler , Marcelo Farenzena

The demand for high performance in industries increases as competitive pressure grows. This opens opportunities for control loop performance monitoring (CPM) techniques, which have been widely proposed over the last three decades. The large number of techniques is valuable, but it brings problems in selecting the right technique for a given application. Industrial data are not easily available due to confidential reasons. Thus, new techniques are commonly tested in a few examples that cover a limited number of scenarios. Also, the tests are not standardized and are not performed using the same dataset. This complicates the comparison of the techniques. This work shares three datasets with different features found on industries, which go from long-range raw data to selected fragments with adequate sampling time. Also, an open-source tool for SISO data visualization and analysis is provided. This tool allows for easy visualization and testing of new techniques.



中文翻译:

工业数据集和用于SISO数据可视化和分析的工具

随着竞争压力的增加,行业对高性能的需求也在增加。这为控制环性能监视(CPM)技术打开了机遇,在过去的三十年中,该技术已被广泛提出。大量的技术很有价值,但是在为给定的应用选择正确的技术时会带来问题。由于机密原因,不容易获得工业数据。因此,通常在覆盖有限数量场景的几个示例中测试新技术。同样,测试不是标准化的,并且不会使用同一数据集执行。这使技术的比较复杂。这项工作共享了三个具有不同行业特征的数据集,这些数据集从远程原始数据到具有足够采样时间的选定片段。也,提供了用于SISO数据可视化和分析的开源工具。该工具可以轻松地可视化和测试新技术。

更新日期:2020-12-09
down
wechat
bug