Byounggul Oh Advanced Combustion & Engine Technoiogy Team, Institute ot Tectinoiogy, Docsan Intracore Co., Ltd., 39-3, Sungbok-Dong, Suji-Gu, Yongin-Si, Gyecnggi-Do 448-795, Korea Minkwang Lee Yeongseop Park Jeongwon Sohn Department of Automotive Engineering, Hanyang tJniversity, 222 Wangsimni-ro, Seongdong-gu, Seoul 133-791, Korea Jongseob Won Department ot Mectianicai and Automctive Engineering, Jeonjj university, 303 Cheonjam-ro, Wansan-gu, Jeonju 560-759, Kcrea Myoungho Sunwoo^ Department ot Automotive Engineering, iHanyang University, 222 Wangsimni-ro, Secngdong-gu, Seoui 133-791, Korea e-mail: msunwoo@tianyang.ac.i<r VGT and EGR Control of Common-Rail Diesel Engines Using an Artificial Neural Network In diesel engines, variable geometry turbocharger (VGT) and exhaust gas recirculation (EGR) systems are used to increase engine specific power and reduce NO^ emissions, respectively. Because the dynamics of both the VGT and EGR ar'e highly nonlinear and coupled to each other, better perfor~mance may be attained by substituting nonlinear mul- tiple input, multiple output (MIMO) controllers for the existing conventional lookup table-based linear contrvllers. This paper presents a coordinated VGT/EGR control sys- tem for common-rail direct injection diesel engines. The objective of the control system is to track target mass air flow and target intake manifold pressure by adjusting the EGR and VGT actuator positions. We designed a nonlinear MIMO control system using a neu- r-al control scheme that adopts an indirect adaptive control approach. The neural contrvl system is comprised of a neural network identifier, which mimics the target air system, and a neural network controller, which calculates the actuator positions. The proposed control system has been validated with engine experiments under transient operating conditions. It was demonstrated from experimental results that the proposed control system shows improved target value tr-acking performance over conventional VGT/EGR control system. [DOI: 10.1115/1.4007541] Keywords: control, diesel engine, variable geometry turbocharger, exhaust gas recircu- lation, mass airflow, neural network, indirect adaptive control 1 Introduction Diesel engines are used extensively as a power source for vehicles because of their higher fuel efficiency than gasoline engines. However, they have lower specific power and higher nitrogen oxide (NO^) and particulate matter (PM) emissions, because of their lean combustion and compression ignition. In order to reduce these harmful emissions while still maintaining fuel efficiency, various electronic control technologies have been applied, the most common of which are common-rail direct injec- tion, variable geometry turbocharger (VGT) and exhaust gas recir- culation (EGR) systems [ 1 ^ ] . VGT systems are used to increase the engine's specific power in various operating conditions by boosting the intake air charge. EGR systems reduce NOx emissions by recirculating exhaust gas back into the air intake [1,5-7]. The boost pressure and amount of EGR gas are regulated by the VGT vane position and EGR valve lift, respectively. An appropri- ate level of exhaust gas feed into the cylinder air charge is effec- tive in NOx emissions reduction, but an excessive EGR rate generates increased PM emissions, as shown in Fig. 1. For this reason, it is important to accurately control the EGR in diesel engines. Corresponding author. Contributed by the IC Engine Division of ASME for publication in the JOURNAL OF ENGINEERING FOR GAS TURBINES AND POWER. Manuscript received February 29, 2012; final manuscript received August 6. 2012; published online November 21. 2012. Assoc. Editor: Christopher J. Rutland. From a control point of view, the diesel engine air system is classified as a nonlinear multiple input multiple output (MIMO) system. The inputs to the air system are the VGT vane position and the EGR valve lift; outputs are the manifold absolute pressure (MAP) and mass air flow (MAF) at the compressor inlet. Both MAP and MAF are regulated by both VGT vane position and EGR valve lift. This coupling makes a MIMO problem. Diesel engine air systems show highly nonlinear dynamics due to the nonlinearity and interactions between the VGT and EGR systems [8,9]. For these reasons, a nonlinear control strategy that considers the interaction of the VGT and EGR systems is required to control the diesel engine air system. However, most conventional engine management systems have adopted lookup table-based gain scheduling n controllers to track PM [g/kWh]' NOx [g/kWh] Fig. 1 Trade-off between particulate matter (PM) emission and nitrogen oxide (NO,) emission with respect to the EGR rate Journal of Engineering for Gas Turbines and Power Copyright © 2013 by ASME JANUARY2013, Vol. 135 / 012801-1