Introduction

In recent years, serious air pollutions have affected people’s daily life. According to the survey of the environment protection department in 2015, one of the main emission resources of NOx and C is from the transportation section, in which the emissions of diesel engine occupy the most important part. Thus improvement of the emission performance of diesel engine becomes a rigorous and important task. High Pressure Common Rail Injection System (HPCRIS) is a key technology to reduce the emission of diesel engine. Its function determines not only the injection-fuel pressure, but also the emission measurement of fuel-injection. High and steady common-rail pressure is important to diesel engine.

To study the control of common-rail pressure, a model that can describe the dynamics of the common-rail pressure is required firstly. Many researchers have built different common-rail pressure models, like an empirical common-rail pressure model in [1], a hybrid common-rail injection system model in [2], and a controloriented model in [12] which is adopted by
most researchers.

In [13], Lino et al based on the referred control-oriented model proposed a sliding mode rail pressure controller, while many other researchers based on classical PID controller, propose derived PID controllers for common rail pressure control, such like RBF neural network adaptive PID control in [10], feed forward fuzzy PID control in [6], and genetic algorithm nonlinear PID control in [20], etc.

For the pressure wave in the common rail caused by the discontinuous inlet fuel flow and the discontinuous outlet fuel flow, some different control methods have been applied to the rail pressure control, such as the QFT (quantitative feedback theory) method is introduced to regulate and control rail pressure [3], and a
coordinated control strategy for the rail pressure using a metering unit and a pressure control valve in proposed in [9].

More recently, to improve the control performance of the considered common rail system, particularly with the consideration of the uncertainties, and external disturbance such as measurement noise, etc in system dynamics, a novel disturbance estimation and rejection control is designed for rail pressure control in [16]. From the above overview of the rail pressure control, the requirements for the rail pressure controller are: simple parameters tuning, good rejection to external disturbances, and easy adaptation to variable working conditions.

Thus in this paper, an Extended State Observer – based intelligent Proportional Integral (ESObased iPI) controller is designed for common rail pressure of HPCRIS. This proposed ESObased iPI control is based on new results from intelligent PID Control which employs an ultralocal model and algebraic based derivatives estimation of systems output signals [4-5]. While for the referred algebraic based derivatives estimation, it needs complex calculation steps and easily perturbed by external measurement noises. To improve the entire method performance, an extended state observer (ESO) which are derived from [8, 14] is proposed via the knowledge of the control and output signals [6].

The following paper is organized as follows: a detailed nonlinear model of HPCRIS is proposed in section 2, which is followed in section 3, the referred mathematical model of HPCRIS is validated both by MATLAB/Simulink and AMESim software. Then in Section 4, for the considered HPCRIS, a new ESO-based iPI controller is designed. Moreover to validate the proposed entire control, its corresponding simulation results which are implemented in both Matlab/Simulink and AEMSim environment, and compared with classic PID controller and ADRC are illustrated in section 5. Finally, some conclusions and future work comments are summarized in section 6.