Real-time nonlinear model predictive control for optimizing the ammonia coverage ratio of an SCR system based on TC-QPSO
Jinghua Zhao,Guogang Chang,Yao Sun,Yunfeng Hu,Hong Chen,Jianhua Yu,Huan Chen,Xun Gong
DOI: https://doi.org/10.1016/j.conengprac.2022.105409
IF: 4.057
2023-01-05
Control Engineering Practice
Abstract:SCR systems with complex chemical reaction dynamics are important components of emission aftertreatment systems. Their contradictory emission requirements of higher NOx conversion efficiency and lower NH3 slip can be simultaneously achieved by optimizing the ammonia coverage ratio, which can be accomplished with a nonlinear model predictive control (NMPC) framework. However, how to find a reasonable NMPC approach for optimizing the ammonia coverage ratio and utilize an appropriate method to solve it remain a formidable challenge. In this paper, a two-layer hierarchical control framework for an SCR system is developed to decouple the optimization and tracking control of the ammonia coverage ratio. In the upper layer, an NMPC controller with a long prediction horizon is designed to account for the emission objectives and constraints and generate optimal trajectories for the ammonia coverage ratio. In the lower layer, a robust observing and tracking controller with a short horizon is employed which takes the planned trajectories as the tracking target. Moreover, an improved thermonuclear-initiation and cross-boundary quantum-behaved particle swarm optimization (TC-QPSO) method for optimizing the ammonia coverage ratio based on its dynamic characteristics is proposed to solve the NMPC problem. The simulation results show that the proposed TC-QPSO emission optimization effects of NOx and NH3 are closest to those of the SQP method (the benchmark controller) with a slight computation time penalty. The experimental results under the three transient cycles show that the average NOx conversion efficiency of the proposed two-layer hierarchical control framework with the TC-QPSO algorithm can reach more than 98% with an average NH3 slip of less than 11 ppm, which is able to satisfy emission requirements, and that compared to the traditional QPSO algorithm, the average NH3 slip of the proposed TC-QPSO algorithm is improved by 45.6% with an average NOx conversion efficiency penalty of 1.09% and its emission optimization effects have better robustness for uncertainties in transient cycles.
automation & control systems,engineering, electrical & electronic