An AIoT-based operation guidance system for constrained emissions control and economic optimization of an operational commercial-scale chemical process
Journal
Journal of Industrial and Engineering Chemistry
ISSN
1226086X
Date Issued
2026
Author(s)
Liu, Hsiao-Te
Fang, Ming-Chun
Lee, Hao-Yeh
Hsieh, Cheng-Ting
Hua, Tzu-Chieh
Lin, Shih-Chieh
Lee, Chih-Lung
Huang, Tzu-Hsien
Chou, Wei-Ti
Abstract
As environmental regulations become increasingly stringent, chemical companies strive to remain profitable while achieving regulatory compliance. Here, an AIoT-based guidance system is developed and applied to an existing chemical plant where light oil (benzene, toluene and xylene (BTX)) is recovered from coke oven gas (COG). The objective is to maximize profit while keeping the BTX concentration in the treated gas under the constraint. At each processing time step, current values of measured, uncontrollable model inputs are retrieved from the DCS. Controllable inputs are discretized and the GRU-based AI model is used to predict outputs for all possible combinations of controllable model inputs. Combinations of input variables for which the model predicts a violation of the emission regulation are discarded and an economic measure is used to identify the best combination of input variables from those that remain. Compared to historical operations, the system improves economic performance by 28.52%, reduces outlet BTX concentration by 79.45%, and lowers carbon emissions by 15.96% on average. These results demonstrate the system’s capability to achieve both profitability and regulatory compliance.
Subjects
Artificial intelligence of things (AIoT)
Decision-support system
Emission strategies
Industry 4.0
Smart manufacturing
Publisher
Korean Society of Industrial Engineering Chemistry
Type
journal article