基于数据的中重型柴油机选择性还原系统闭环控制理论研究

Selective catalytic reduction (SCR) systems have been widely regarded as the most promising technique for the Diesel engine NOx reduction and the stringent NOx emission regulation. However, the SCR system may also lead to the ammonia slip in the tailpipe. Since the gaseous ammonia is one kind of greenhouse gases and harmful for the health of human being, the ammonia slip is undesired and should be constrained. Due to the facts that the NOx emission regulation is becoming more and more stringent and the strict tailpipe ammonia slip constraint is required, the existing single-cell SCR system cannot simultaneously achieve a high NOx conversion efficiency and a low ammonia slip constraint...Motivated by this point, this proposal aims to study the series SCR system and theoretically analyze the possibility of simultaneously achieving a high NOx conversion efficiency and a low ammonia slip constraint. Specifically, the objectives of this proposal are four-fold: to derive precise dynamics model of Diesel engine series SCR systems and reveal the main principle of the systems; to propose globally optimized algorithm for stiff nonlinear systems with high dimensions and develop the matching theory for the series SCR systems; to analyze the observability of delayed nonlinear systems and develop the best physical sensor placement theory; and to build an online model of dynamic optimal ammonia coverage ratios and propose an optimal closed-loop control strategy. The results of this proposal would not only prompt the development of control theory for delayed nonlinear systems, but also prompt the development of future SCR aftertreatment systems in China.

SCR系统是公认的降低柴油机氮氧化物排放、满足苛刻排放法规最有前景的技术，但在催化还原氮氧化物时也可能产生氨气逃逸，导致新的污染。随着氮氧化物排放法规的日益严格和氨气逃逸量的限制，现有的单催化剂罐SCR系统很难同时满足高氮氧化物转化率和低氨气逃逸量要求。拟开展基于数据的SCR系统的理论研究，从理论层面上分析基于数据的SCR系统同时满足高氮氧化物转化率和低氨气逃逸量的可行性及系统匹配与控制策略。围绕该目标，建立基于数据的柴油机SCR系统动力学模型，揭示SCR系统的内部机理；提出适用于刚性非线性高维系统的全局最优优化算法，建立SCR系统的最佳匹配理论；建立基于数据的时滞非线性系统的可观性判据，确定串联SCR系统物理传感器最佳分布原则；实现SCR系统氨气覆盖率动态最优跟踪目标的在线生成，提出基于数据的最优闭环控制策略。研究成果能够推动时滞非线性系统基于数据的控制理论和我国未来SCR系统的发展。

SCR系统可以有效降低中重型柴油机尾气中的NOx，但在催化还原氮氧化物时也可能产生氨气逃逸，导致新的污染。随着氮氧化物排放法规的日益严格和氨气逃逸量的限制，现有的单催化剂罐SCR系统很难同时满足高氮氧化物转化率和低氨气逃逸量要求。本项目展开了基于数据的SCR系统的理论研究，从理论层面上分析基于数据的SCR系统同时满足高氮氧化物转化率和低氨气逃逸量的可行性及系统匹配与控制策略。围绕该目标，建立基于数据的柴油机SCR系统动力学模型，揭示SCR系统的内部机理；建立基于数据的时滞非线性系统的可观性判据，并对SCR系统内部的不可直接测量的重要状态量设计了观测器，实现了SCR系统氨气覆盖率动态最优跟踪目标的在线生成。并提出了基于数据的最优闭环控制策略，该策略基于模糊控制算法。研究结果表明，该控制算法能够实现优秀的NOx控制效果，在保证较高NOx转化的同时，还能保持较低的氨泄露。