Fuel and Emissions Optimization for Connected Diesel Engine Vehicles With Hierarchical Model Predictive Control

This paper studies a hierarchical model predictive control (MPC) strategy to optimize the powertrain and aftertreatment systems for connected diesel-powered vehicles. With the development of vehicle connectivity and autonomy, it is convenient to acquire vehicle speed prediction and future geographic information for improving driving safety and fuel economy. Inspired by such achievements, preview speed and geographic information is utilized to enhance the control performance of diesel engines and urea-based selective catalytic reduction (SCR) systems simultaneously in this work. With the short-term prediction of vehicle speed and road grade, the upper-level controller for the diesel engine could respond in advance and hence reduce fuel consumption by avoiding sudden braking and acceleration. Similarly, according to the engine-out NO
$_{x}$
emissions predicted through upper-level control actions, the lower-level dosing controller of SCR system could remove the NO
$_{x}$
emissions more efficiently as well. Finally, to explore the effectiveness of the designed predictive control strategy, several simulations are implemented based on the real experimental data. The comparison results demonstrate the remarkable improvements of our proposed approach.