Route-dependent optimal control of the after-treatment system of diesel engines

Abstract

In this article, the optimal control problem for nitrogen oxide emission reduction is investigated for diesel engines with a lean nitrogen oxide trap. First, a control-oriented model is developed based on conservation laws. Then, the optimal control problem is formulated as a multistage decision problem and solved using a dynamic programming algorithm under dynamical model constraints. A trade-off between fuel economy and nitrogen oxide emission is considered in the cost function of optimization. To demonstrate the obtained optimal control scheme, the parameters of the lean nitrogen oxide trap model are identified with data obtained from a GT-power-based diesel engine simulator. The numerical simulation results for two standard driving cycles and a stochastically generated driving cycle in comparison to a conventional logic-based control scheme are provided using the identified model in the MATLAB/Simulink platform.

Ammonia injection optimization for selective catalytic reduction aftertreatment systems

“…Advances in computation capabilities made optimal control (OC) theory attractive to optimize a complex dynamic system. 24,25 The OC theory was introduced by Lev Pontryagin et al 26 and Richard Bellman, 27 and a more detailed explanation of the subject can be found in their work. The OC theory is a model-based control approach where different mathematical methods are used to calculate the controls that must be applied to a system to minimize a predefined cost index.…”

MPC-Based Optimal Control for Diesel Engine Coupled with Lean NOx Trap System

“…Knowledge of the dynamic process of NOx reduction is critical to capture LNT NOx emission. In this paper, a two-order LNT model is used for optimal control scheme design, which is based on energy and mass conservation [20]. The LNT temperature T l and the fill ratio F r are modeled as follows:…”

“…where β, a k , b k , c k , and d k are determined through an identification method. It is noted that the order of the air-fuel ratio λ is selected as three for better identification performance, which is written in previous work [20]. Actual air-fuel ratio with consideration of additional fuel injection is calculated as:…”

Deep reinforcement learning to discover multi-fuel injection strategies for compression ignition engines

“…Many researchers in the literature have used machine learning (ML) techniques for various aspects of engine research such as simulation, 19 modeling, [20][21][22][23][24][25] optimization, [26][27][28][29][30][31][32] and control. [33][34][35][36][37][38] Traditional reinforcement learn-ing (RL) has been used for hard-coal combustion processes in a power plant, 39,40 for spark ignition and injection timing, 41,42 for energy management strategies for hybrid-electric vehicles, [43][44][45] and for the control of the air-fuel ratio 46 and spark engine exhaust gas recirculation (EGR) operation. 47 Deep reinforcement learning (DRL) leverages the expressiveness of deep neural networks to tackle complex control challenges.…”