Neural Network-Based Control Optimization for NH3 Leakage and NOx Emissions in SCR Systems
Open Access
Abstract
This study proposes a data-driven optimization framework to enhance emission control performance in diesel engine selective catalytic reduction (SCR) systems under transient operating conditions. A one-dimensional SCR model was constructed in GT-Power, and simulation datasets were generated using experimentally measured inputs from the World Harmonized Transient Cycle (WHTC), with representative emission responses obtained by varying fixed ammonia-to-NOx (A/N) ratios. Building on these datasets, a hybrid prediction model combining Long Short-Term Memory (LSTM) networks and multi-head attention mechanisms was developed to accurately forecast SCR outlet NH3 leakage and NOx emissions. The model exhibited high predictive accuracy, achieving R2 values exceeding 0.977 and low RMSE across training, validation, and test sets. Based on the model predictions, a constrained dynamic multi-objective optimization strategy was implemented to adaptively adjust ammonia dosing, aiming to simultaneously minimize NH3 leakage and NOx emissions. The optimized NH3 injection profiles were validated through reapplication in the GT-Power simulation environment. Compared to the baseline fixed-ratio control strategy, the proposed approach reduced NH3 leakage and NOx emissions by 34.40% and 11.15%, respectively, as determined for the transient segment of the WHTC cycle. These results demonstrate the effectiveness of integrating physics-based simulation, deep learning prediction, and dynamic optimization for improving aftertreatment adaptability and emission compliance in real-world diesel engine applications. All reported values are based on a single simulated WHTC cycle without statistical uncertainty analysis.Keywords
Funding Information
- Young Innovative Talent Program (2022A 135-G, 2021Z036, 2022Z151, 2022YFE0209000)
- Ningbo Major Research and Development Plan Project (2022A 135-G, 2021Z036, 2022Z151, 2022YFE0209000)
- National Key R&D Program of China (2022A 135-G, 2021Z036, 2022Z151, 2022YFE0209000)
This publication has 30 references indexed in Scilit:
- An artificial intelligence strategy for multi-objective optimization of Urea-SCR for vehicle diesel engine by RSM-VIKOREnergy, 2025
- Machine learning-based design of target property-oriented fuels using explainable artificial intelligenceEnergy, 2024
- Multi-objective optimization of Fe-based SCR catalyst on the NOx conversion efficiency for a diesel engine based on FGRA-ANN/RFEnergy, 2024
- Model-based NH3 Adaptation Method for SCR ControlInternational Journal of Automotive Technology, 2022
- Performance evaluation of neural networks in modeling exhaust gas aftertreatment reactorsChemical Engineering Journal, 2022
- Experimental investigation of urea injection strategy for close-coupled SCR aftertreatment system to meet ultra-low NO emission regulationApplied Thermal Engineering, 2022
- Reinforcement learning for online adaptation of model predictive controllers: Application to a selective catalytic reduction unitComputers & Chemical Engineering, 2022
- Global kinetic modeling of NH3-SCR with two sites of NH3 storage on Cu-SSZ-13Chemical Engineering Journal, 2021
- Review of state of the art technologies of selective catalytic reduction of NOx from diesel engine exhaustApplied Thermal Engineering, 2014
- Detailed kinetic modeling of the NH3–NO/NO2 SCR reactions over a commercial Cu-zeolite catalyst for Diesel exhausts after treatmentCatalysis Today, 2012