0
Research Papers

Model-Based Control of the Air Fuel Ratio for Gasoline Direct Injection Engines via Advanced Co-Simulation: An Approach to Reduce the Development Cycle of Engine Control Systems

[+] Author and Article Information
Alessandro di Gaeta1

 Istituto Motori National Research Council, Naples, Italy 80125a.digaeta@im.cnr.it

Umberto Montanaro

 Istituto Motori National Research Council, Naples, Italy 80125u.montanaro@im.cnr.it

Veniero Giglio

 Istituto Motori National Research Council, Naples, Italy 80125v.giglio@im.cnr.it

1

Corresponding author.

J. Dyn. Sys., Meas., Control 133(6), 061006 (Sep 29, 2011) (17 pages) doi:10.1115/1.4004067 History: Received March 16, 2010; Revised February 24, 2011; Published September 29, 2011; Online September 29, 2011

Nowadays, the precise control of the air fuel ratio (AFR) in spark ignition (SI) engines plays a crucial role in meeting the more and more restrictive standard emissions for the passenger cars and the fuel economy required by the automotive market as well. To attain this demanding goal, the development of an advanced AFR control strategy embedding highly predictive models becomes mandatory for the next generation of electronic control unit (ECU). Conversely, the adoption of more complex control strategies affects the development time of the ECU increasing the time-to-market of new engine models. In this paper to solve the AFR control problem for gasoline direct injection (GDI) and to speed up the design of the entire control system, a gain scheduling PI model-based control strategy is proposed. To this aim, AFR dynamics are modeled via a first order time delay system whose parameters vary strongly with the fresh air mass entering the cylinders. Nonlinear relations have been found to describe the behavior of model parameters in function of air mass. Closed loop performances, when this novel controller is nested in the control loop, are compared to those provided by the classical PI Ziegler–Nichols control action with respect to different cost functions. Model validation as well as the effectiveness of the control design are carried out by means of ECU-1D engine co-simulation environment for a wide range of engine working conditions. The combination in one integrated designing environment of control systems and virtual engine, simulated through high predictive commercial one dimensional code, becomes a high predictive tool for automotive control engineers and enables fast prototyping.

Copyright © 2011 by American Society of Mechanical Engineers
Topics: Engines , Cylinders , Fuels
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 1

AFR operational modes of the JTS 2.0 GDI Alfa Romeo engine (taken from D. Andriesse., F. Guarnaccia., M. Guazzaroni and A. Oreggioni, 2001. “The New Alfa Romeo 2 litre JTS Engine with Direct Gasoline Injection,” Proceedings of 10th Colloquium on Vehicle and Engine Technology, Aachen,Germany.)

Grahic Jump Location
Figure 2

Co-simulation framework

Grahic Jump Location
Figure 3

Schematization of the AFR path in a GDI engine taken from Ref. [22]

Grahic Jump Location
Figure 4

Comparison between modeled (solid line) and estimated (circle) model parameters in function of the breathed air-mass: (a) open-loop delay θD(mac); (b) mixing time constant θM(mac)

Grahic Jump Location
Figure 5

Identification results for tests belonging to the ϒ set: (a) STR step at low loads m¯ac=0.2962 (g/cc); (b) STL step at high loads mac=2.2847 (g/cc); (c) STL at medium loads mac=0.8711 (g/cc); (d) STR at medium loads mac=1.8083 (g/cc)

Grahic Jump Location
Figure 6

Validation results for tests not belonging to the ϒ set: (a) STL step at low loads m¯ac=0.3160 (g/cc); (b) STR step at medium loads mac=0.9195 (g/cc); (c) STR at medium loads mac=1.4051 (g/cc); (d) STR at high loads mac=2.1435 (g/cc)

Grahic Jump Location
Figure 7

D plots of the quasi-steady speed-density model at standard temperature T0 : (a) air mass breathed by the engine mac(P,N) (g/cc); (b) air mass flow rate mac(P,N)(3600·N)/(120·1000) (g/cc)

Grahic Jump Location
Figure 8

Comparison between the in-cylinder air mass measured on the 1D-engine and ones provided by regressor (8) for different engine working points

Grahic Jump Location
Figure 9

Validation of the quasi-steady speed-density model during a high load transients at low speed [ 1500 (rpm)]: (a) throttle position (above) and manifold pressure (bottom); (b) MAF (1D-ICE) is the air incoming into the manifold (blue dotted line), IN-CYLS. (1D-ICE) is the sum of the air mass flowing through the intake runners (red dash-dot line), MODEL is the in-cylinders air mass predicted by the speed-density model (black solid line).

Grahic Jump Location
Figure 10

Validation of the quasi-steady speed-density model during a high load transients at high speed [ 4000 (rpm)]: (a) throttle position (above) and manifold pressure (bottom); (b) MAF (1D-ICE) is the air incoming into the manifold (blue dotted line), IN-CYLS. (1D-ICE) is the sum of the air mass flowing through the intake runners (red dash-dot line), MODEL is the in-cylinders air mass predicted by the speed-density model (black solid line).

Grahic Jump Location
Figure 11

First model validation: (a) steady air and fuel transient; (b) comparison between the equivalence AFR 1D engine data (black dotted line) and model prediction (blue solid line)

Grahic Jump Location
Figure 12

Second model validation: comparison between the equivalence AFR 1D engine data (black dotted line) and model predictions (blue solid line) when the in-cylinder air mass is a sequence of steps and the fuel injected is kept constant at (a)–(d)

Grahic Jump Location
Figure 16

Model-based AFR control results in the case of medium engine speed [N=3250 (rpm)] obtained via 1D-engine and control co-simulation approach: (a) estimated air breathed by the engine (disturbance); (b) injected fuel mass; (c) equivalence ratio

Grahic Jump Location
Figure 17

Model-based AFR control results in the case of high engine speed [N=5200 (rpm)] obtained via 1D-engine and control co-simulation approach: (a) estimated air breathed by the engine (disturbance); (b) injected fuel mass; (c) equivalence ratio

Grahic Jump Location
Figure 18

Model-based AFR control results in the case of low engine speed [N=1500 (rpm)] obtained via 1D-engine and control co-simulation approach: (a) estimated air breathed by the engine (disturbance); (b) injected fuel mass; (c) equivalence ratio

Grahic Jump Location
Figure 19

Graphical meaning of the performance indexes

Grahic Jump Location
Figure 20

Cost indexes for tests in Table 5 in the case of Ziegler-Nichols PI controller (yellow bars) and PI Model-Based with Smith Predictor (red bars); (a) peak-peak overshoots ξpp ; (b) recovery cycles Cr ; (c) subtended area As ; (d) cost index Q

Grahic Jump Location
Figure 15

Model-based AFR control results in the case of low engine speed (N=2600 [rpm]) obtained via 1D-engine and control co-simulation approach: (a) estimated air breathed by the engine (disturbance); (b) injected fuel mass; (c) equivalence ratio

Grahic Jump Location
Figure 14

Three level gain scheduling model-based control scheme

Grahic Jump Location
Figure 13

Third model validation: comparison between the equivalence AFR 1D engine data (black dotted line) and model predictions (blue solid line) when both in-cylinder air mass and fuel injected change

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In