Online Adaptive Residual Mass Estimation in a Multicylinder Recompression Hcci Engine

This work presents two advances to the estimation of homogeneous charge compression ignition (HCCI) dynamics. Combustion phasing prediction in control-oriented models has been achieved by modeling the in-cylinder temperature and composition dynamics, which are dictated by the large mass of residuals trapped between cycles. As such, an accurate prediction of the residual gas fraction as a function of the variable valve timing is desired. Energy and mass conservation laws applied during the exhaust valve opening period are complemented with online in-cylinder pressure measurements to predict the trapped residual mass in real time. In addition, an adaptive parameter estimation scheme uses measured combustion phasing to adjust the residual mass prediction. Experimental results on a multicylinder gasoline HCCI engine demonstrate the closed loop residual estimation’s ability to compensate for modeling errors, cylinder to cylinder variations, and engine wear. Additionally it is shown that using the adaptive parameter estimation reduces the model parameterization effort for a multicylinder engine. INTRODUCTION Recompression homogeneous charge compression ignition (HCCI) is a promising combustion strategy that can achieve high thermal efficiency with low engine-out emissions. It is characterized by compression-driven near simultaneous auto-ignition events at multiple sites throughout a homogeneous mixture. Autoignition timing control in HCCI combustion requires careful regulation of the temperature, pressure, and composition of the precombustion cylinder charge. This regulation of charge properties is carried out in recompression HCCI by retaining a large frac∗Address all correspondence to this author. tion of the post-combustion residual gases before they can be exhausted [1, 2]. Since neither the temperature nor the mass of the trapped residuals can be measured directly, model-based control of recompression HCCI requires the development of accurate controloriented models that can be run in real-time on embedded control hardware. Examples of HCCI control-oriented models can be found in literature [3–9]. The combustion phasing prediction accuracy of these models is especially crucial when they are used in model-based predictive control strategies, for example [9–12]. This is because the phasing, or timing, of HCCI combustion must be maintained within a narrow acceptable range to satisfy stability and mechanical constraints. The models must be robust to engine aging, parameter drift and changes in environmental conditions. The current work advances the state of art in two important ways. First, the authors propose a novel, physics-based method of calculating the residual gas mass in real-time. Second, an adaptive parameter estimation scheme is implemented in a previously developed HCCI model [8, 9], and is shown in experiments to improve prediction performance and robustness. Accurate modeling of the residual gas fraction is important for a control oriented model due to HCCI’s high sensitivity to the thermal energy associated with the residual gases. If too much residual mass is trapped the combustion can occur very early causing potential engine damage and a loss in efficiency. If too little mass is trapped the combustion can become highly oscillatory [13, 14], and misfires may occur. While there have been methods described in literature for estimating the residual mass [15–18], none have presented an algorithm capable of online implementation without the use of a steady-state assumption. This work aims to provide a solution to this problem. In addition, adaptive parameter estimation is used to increase 1 Copyright © 2013 by ASME Proceedings of the ASME 2013 Dynamic Systems and Control Conference DSCC2013 October 21-23, 2013, Palo Alto, California, USA