IFAC-PapersOnLine 49-11 (2016) 490–496 ScienceDirect Available online at www.sciencedirect.com 2405-8963 © 2016, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. Peer review under responsibility of International Federation of Automatic Control. 10.1016/j.ifacol.2016.08.072 © 2016, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved. Keywords: Fuel injection modelling, Compressed natural gas, Common rail injection system, Injection pressure control, Dynamic modelling. 1. INTRODUCTION Automotive engines employing compressed natural gas (CNG) can lower the impact of harmful emissions and adhere to more and more strict regulations. Even if the combustion process in Diesel engines greatly reduced the emissions of gas and particulate matter, the interest in CNG is because of its wide availability and simple de- livery by well spread infrastructure (International Gas Union, 2005; Geok et al., 2009). Moreover, reducing fuel consumption is a shared objective by manufacturers and users. At the same time, certain levels of performance must be maintained or increased. These objectives can be reached by accurately metering the air-fuel mixture (Baratta et al., 2015), which can be obtained in two ways in the common rail injection systems. One is to precisely control the opening/closing time intervals of the injectors. Another way is to control the injection pressure, even if an accurate and efficient regulation is difficult mainly because of gas compressibility. Namely, the speed variations, the load changes and the power requirements may determine substantial (and frequent) changes of the injection system working points that are necessary to inject the proper fuel amount. Moreover, parametric variations that affect the system performance and disturbances must be compen- sated (Baratta et al., 2015). In this context, the aim is to improve the robustness and performance of the controllers by advanced schemes and by model-based design approaches. Namely, accurate modeling the CNG injection system is useful both for behavior analysis and prediction and for designing con- trollers that better regulate the rail pressure. The model must be simple enough for control design. However, even if many control-oriented models have been developed for diesel and gasoline injection systems, few regard CNG injection systems (Lino et al., 2008; Lino and Maione, 2013a,b, 2014). Instead, existing accurate models that are able to replicate the real system behavior are not suitable for model-based control design. These models are useful to improve the knowledge of the injection process, to evaluate and validate the effects of operational conditions, and to explore different configurations and alternative functional designs (see Lino and Maione (2007); Dellino et al. (2009); Misul et al. (2014); Baratta et al. (2015), and references therein). The aim of this work is then to develop an accurate model of an electronically controlled CNG injection system, by trading off between accuracy, simplicity and suitability for control development. The approach is based on the physical equations underlying the involved processes, and requires the tuning of a minimal set of parameters. The starting point is the simple second-order nonlinear model presented in Lino et al. (2008), which includes many simplifying assumptions. Then, to improve the prediction accuracy, a new higher-order nonlinear model is developed by properly including the electro-magnetic forces, inertia of moving parts, load losses, propagation delays, and volume changes due to motion of mechanical elements. The paper is organized as follows. Section 2 presents the CNG injection system and Section 3 describes the mod- eling approach. Section 4 reports on a validation analysis performed by comparing simulation and experimental data to evaluate the prediction capabilities, either in time and frequency domains. Finally, conclusions are presented in Section 5. 2. THE METHANE INJECTION SYSTEM This paper considers an innovative CNG fuel feeding sys- tem developed by the FIAT Research Center, Valenzano branch, Italy (Amorese et al., 2004; Lino et al., 2008) shown in Fig. 1. The elements that can be distinguished Dept. of Electrical and Inform. Eng., Politecnico di Bari, Bari, Italy (e-mail: {paolo.lino,guido.maione}@poliba.it) Abstract: Compressed natural gas engines provide a non-standard solution for reducing emission of polluting gases and particulate matter. However, the gas compressibility and the complex nonlinear fluid-dynamics make the task of metering the air-gas mixture very difficult. To address this issue and to synthesize effective strategies for controlling the common rail pressure, it is necessary to develop an accurate control-oriented model that describes the relevant components, processes and dynamics. This paper presents a model-based approach that can be used for prediction, analysis of performance and control design. The developed model is validated by comparison of simulation results with experimental data from a real injection system. Paolo Lino, Guido Maione Accurate dynamic modeling of an electronically controlled CNG injection system