Christopher H. 0nder
`
`Lino Guzzella
`
`1
`
`PAICE 2028
`BMW v. Paice
`|PR2020-01386
`
`1
`
`PAICE 2028
`BMW v. Paice
`IPR2020-01386
`
`

`

`Introduction to Modeling and Control of Internal
`Combustion Engine Systems
`
`2
`
`

`

`Lino Guzzella and Christopher H. Onder
`
`Introduction to Modeling
`and Control of Internal
`Combustion Engine
`Systems
`
`ABC
`
`3
`
`

`

`Prof. Dr. Lino Guzzella
`ETH Zürich
`Institute for Dynamic Systems & Control
`Sonneggstr. 3
`8092 Zürich
`ETH-Zentrum
`Switzerland
`E-mail: lguzzella@ethz.ch
`
`Dr. Christopher H. Onder
`ETH Zürich
`Institute for Dynamic Systems & Control
`Sonneggstr. 3
`8092 Zürich
`ETH-Zentrum
`Switzerland
`E-mail: onder@ethz.ch
`
`ISBN 978-3-642-10774-0
`
`e-ISBN 978-3-642-10775-7
`
`DOI 10.1007/978-3-642-10775-7
`
`Library of Congress Control Number: 2009940323
`c(cid:2) 2010 Springer-Verlag Berlin Heidelberg
`
`This work is subject to copyright. All rights are reserved, whether the whole or part of the mate-
`rial is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
`broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Dupli-
`cation of this publication or parts thereof is permitted only under the provisions of the German
`Copyright Law of September 9, 1965, in its current version, and permission for use must always
`be obtained from Springer. Violations are liable to prosecution under the German Copyright Law.
`
`The use of general descriptive names, registered names, trademarks, etc. in this publication does
`not imply, even in the absence of a specific statement, that such names are exempt from the relevant
`protective laws and regulations and therefore free for general use.
`
`Typesetting: Data supplied by the authors
`
`Production: Scientific Publishing Services Pvt. Ltd., Chennai, India
`
`Cover Design: WMX Design, Heidelberg, Germany
`
`Printed in acid-free paper
`
`9 8 7 6 5 4 3 2 1
`
`springer.com
`
`4
`
`

`

`Contents
`
`1
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`1
`1
`1.1 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4
`1.2 Control Systems for IC Engines . . . . . . . . . . . . . . . . . . . . . . . . . . .
`4
`1.2.1 Relevance of Engine Control Systems . . . . . . . . . . . . . . . .
`5
`1.2.2 Electronic Engine Control Hardware and Software . . . . .
`6
`1.3 Overview of SI Engine Control Problems . . . . . . . . . . . . . . . . . . .
`6
`1.3.1 General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
`8
`1.3.2 Main Control Loops in SI Engines . . . . . . . . . . . . . . . . . . .
`1.3.3 Future Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
`1.4 Overview of Control Problems in CI Engines . . . . . . . . . . . . . . . . 11
`1.4.1 General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
`1.4.2 Main Control Loops in Diesel Engines . . . . . . . . . . . . . . . . 14
`1.4.3 Future Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
`1.5 Structure of the Text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
`
`2 Mean-Value Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
`2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
`2.2 Cause and Effect Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
`2.2.1 Spark-Ignited Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
`2.2.2 Diesel Engines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
`2.3 Air System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
`2.3.1 Receivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
`2.3.2 Valve Mass Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
`2.3.3 Engine Mass Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
`2.3.4 Exhaust Gas Recirculation . . . . . . . . . . . . . . . . . . . . . . . . . 37
`2.3.5 Supercharger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
`2.4 Fuel System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
`2.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
`2.4.2 Wall-Wetting Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
`2.4.3 Gas Mixing and Transport Delays . . . . . . . . . . . . . . . . . . . 63
`2.5 Mechanical System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
`
`5
`
`

`

`X
`
`Contents
`
`2.5.1 Torque Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
`2.5.2 Engine Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
`2.5.3 Rotational Vibration Dampers . . . . . . . . . . . . . . . . . . . . . . 81
`2.6 Thermal Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
`2.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
`2.6.2 Engine Exhaust Gas Enthalpy . . . . . . . . . . . . . . . . . . . . . . 86
`2.6.3 Thermal Model of the Exhaust Manifold . . . . . . . . . . . . . 88
`2.6.4 Simplified Thermal Model . . . . . . . . . . . . . . . . . . . . . . . . . . 89
`2.6.5 Detailed Thermal Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
`2.7 Pollutant Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
`2.7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
`2.7.2 Stoichiometric Combustion . . . . . . . . . . . . . . . . . . . . . . . . . 98
`2.7.3 Non-Stoichiometric Combustion . . . . . . . . . . . . . . . . . . . . . 100
`2.7.4 Pollutant Formation in SI Engines . . . . . . . . . . . . . . . . . . . 102
`2.7.5 Pollutant Formation in Diesel Engines . . . . . . . . . . . . . . . 108
`2.7.6 Control-Oriented N O Model . . . . . . . . . . . . . . . . . . . . . . . . 110
`2.8 Pollutant Abatement Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
`2.8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
`2.8.2 Three-Way Catalytic Converters, Basic Principles . . . . . 114
`2.8.3 Modeling Three-Way Catalytic Converters . . . . . . . . . . . . 117
`2.9 Pollution Abatement Systems for Diesel Engines . . . . . . . . . . . . . 137
`
`3 Discrete-Event Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
`3.1 Introduction to DEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
`3.1.1 When are DEM Required? . . . . . . . . . . . . . . . . . . . . . . . . . . 148
`3.1.2 Discrete-Time Effects of the Combustion . . . . . . . . . . . . . 148
`3.1.3 Discrete Action of the ECU . . . . . . . . . . . . . . . . . . . . . . . . . 150
`3.1.4 DEM for Injection and Ignition . . . . . . . . . . . . . . . . . . . . . 153
`3.2 The Most Important DEM in Engine Systems . . . . . . . . . . . . . . . 156
`3.2.1 DEM of the Mean Torque Production . . . . . . . . . . . . . . . . 156
`3.2.2 DEM of the Air Flow Dynamics . . . . . . . . . . . . . . . . . . . . . 161
`3.2.3 DEM of the Fuel-Flow Dynamics . . . . . . . . . . . . . . . . . . . . 164
`3.2.4 DEM of the Back-Flow Dynamics of CNG Engines . . . . 173
`3.2.5 DEM of the Residual Gas Dynamics . . . . . . . . . . . . . . . . . 175
`3.2.6 DEM of the Exhaust System . . . . . . . . . . . . . . . . . . . . . . . . 178
`3.3 DEM Based on Cylinder Pressure Information . . . . . . . . . . . . . . 180
`3.3.1 General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
`3.3.2 Estimation of Burned-Mass Fraction . . . . . . . . . . . . . . . . . 181
`3.3.3 Cylinder Charge Estimation . . . . . . . . . . . . . . . . . . . . . . . . 183
`3.3.4 Torque Variations Due to Pressure Pulsations . . . . . . . . . 188
`
`6
`
`

`

`Contents
`
`XI
`
`4 Control of Engine Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
`4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
`4.1.1 General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
`4.1.2 Software Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
`4.1.3 Engine Operating Point . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
`4.1.4 Engine Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
`4.2 Engine Knock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
`4.2.1 Autoignition Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
`4.2.2 Knock Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
`4.2.3 Knock Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
`4.2.4 Knock Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
`4.3 Air/Fuel-Ratio Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
`4.3.1 Feedforward Control System . . . . . . . . . . . . . . . . . . . . . . . . 210
`4.3.2 Feedback Control: Conventional Approach . . . . . . . . . . . . 215
`4.3.3 Feedback Control: H∞ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
`4.3.4 Feedback Control: Internal-Model Control . . . . . . . . . . . . 229
`4.3.5 Multivariable Control of Air/Fuel Ratio and Engine
`Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
`4.4 Control of an SCR System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
`4.5 Engine Thermomanagement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
`4.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
`4.5.2 Control Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . 250
`4.5.3 Feedforward Control System . . . . . . . . . . . . . . . . . . . . . . . . 252
`4.5.4 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
`
`A Basics of Modeling and Control-Systems Theory . . . . . . . . . . . 261
`A.1 Modeling of Dynamic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
`A.2 System Description and System Properties . . . . . . . . . . . . . . . . . . 270
`A.3 Model Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
`A.4 Control-System Design for Nominal Plants . . . . . . . . . . . . . . . . . . 279
`A.5 Control System Design for Uncertain Plants . . . . . . . . . . . . . . . . 288
`A.6 Controller Discretization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
`A.7 Controller Realization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
`A.7.1 Gain Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
`A.7.2 Anti-Reset Windup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
`A.8 Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
`
`B Case Study: Idle Speed Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
`B.1 Modeling of the Idle Speed System . . . . . . . . . . . . . . . . . . . . . . . . 306
`B.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
`B.1.2 System Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
`B.1.3 Description of Subsystems . . . . . . . . . . . . . . . . . . . . . . . . . . 308
`B.2 Parameter Identification and Model Validation . . . . . . . . . . . . . . 315
`B.2.1 Static Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
`B.2.2 Dynamic Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
`
`7
`
`

`

`XII
`
`Contents
`
`B.2.3 Numerical Values of the Model Parameters . . . . . . . . . . . 321
`B.3 Description of Linear System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
`B.4 Control System Design and Implementation . . . . . . . . . . . . . . . . . 326
`
`C Combustion and Thermodynamic Cycle Calculation of
`ICEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
`C.1 Fuels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
`C.2 Thermodynamic Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
`C.2.1 Real Engine-Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
`C.2.2 Approximations for the Heat Release . . . . . . . . . . . . . . . . 337
`C.2.3 Csallner Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
`
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
`
`8
`
`

`

`1 I
`
`ntroduction
`
`In this chapter, first the notation used throughout this text is defined. It fur-
`ther contains some general remarks on electronic engine control systems and
`introduces the most common control problems encountered in spark ignition
`(Otto or gasoline) and compression ignition (Diesel) engine systems. The in-
`tention is to show the general motivation for using control systems and to
`give the reader an idea of the problems that can be tackled by feedforward and
`feedback control systems for both SI and CI engines.
`The emphasis in this chapter is on qualitative arguments. The mathemati-
`cally precise formulation is deferred to subsequent chapters. Those readers not
`familiar with modern electronic sensors, actuators, and control hardware for
`automotive applications may want to consult either [7], [108], or [125].
`
`1.1 Notation
`
`The notation used in this text is fairly standard. The derivative of a variable
`x(t), with respect to its independent variable t, is denoted by
`
`x(t)
`
`d d
`
`t
`
`˙x(t)
`
`while the notation
`
`is used to indicate a flow of mass, energy, etc. Both variables d
`dt x(t) and ˙x(t)
`have the same units, but they are different objects. No special distinction is
`made between scalars, vectors and matrices. The dimensions of a variable, if
`not a scalar, are explicitly defined in the context. Input signals are usually
`denoted by u... and output signals by y..., whereas the index . . . specifies what
`physical quantity is actuated or measured.
`Concentrations of chemical species C are denoted by [C], with units
`mol/mol, with respect to the reference substance. The concentrations are
`
`9
`
`

`

`2
`
`1 Introduction
`
`therefore limited to the interval [0, 1]. The concentration of pollutant species
`are often shown in plots or tables using ppm units (part per million), i.e., by
`using a amplification factor of 106. For mass storage and transportation mod-
`els it is advantageous to use mass fractions, which are denoted by ξ having
`units [kg/kg].
`In general, all variables are defined at that place in the text where they
`are used for the first time. To facilitate the reading, some symbols have been
`reserved for special physical quantities:
`
`[ W
`α
`m2K ]
`A [m2]
`[ J
`kgK , -]
`cx
`ε
`[-]
`η
`[-]
`◦
`φ
`[
`, rad]
`γ
`[-]
`H [J]
`κ
`[-]
`λ
`[-]
`m [kg]
`M [ kg
`mol ]
`N [-]
`
`heat-transfer coefficient
`area
`specific heat capacities (x = p, v), concentration of x
`compression ratio, volume fraction
`efficiency
`crank angle
`gear ratio
`enthalpy
`ratio of specific heats
`air-to-fuel ratio, volumetric efficiency, Lagrange multiplier
`mass
`molar mass
`number of engine revolutions per cycle
`(1 for two-stroke, 2 for four-stroke engines)
`stoichiometric coefficient
`[-]
`ν
`[P a, bar] pressure
`p
`[W ]
`power
`P
`Π [-]
`pressure ratio
`Q [J]
`heat
`[m, mol
`r
`s ] radius, reaction rate
`[ kg
`m3 ]
`density
`ρ
`R [ J
`kgK ]
`specific gas constant
`R [
`J
`molK ]
`universal gas constant
`σ0
`[-]
`stoichiometric air-to-fuel ratio
`t
`[s]
`time (independent variable)
`τ
`[s]
`time (interval or constant)
`◦
`ϑ
`[K,
`temperature
`θ
`[-]
`occupancy
`T
`[N m]
`torque
`Θ [m2kg]
`rotational inertia
`u, y [-]
`control input, system output (both normalized)
`[m3, l]
`V
`volume
`◦
`ζ
`[
`]
`ignition angle
`[ rad
`s ]
`rotational speed or angular frequency
`ω
`ξ
`[-]
`mass fraction
`
`C]
`
`10
`
`

`

`Similarly, some indices have been reserved for special use. The following list
`shows what each of them stands for:
`
`1.1 Notation
`
`3
`
`α, a, β ambient air
`c
`compressor or cylinder
`e
`engine
`eg
`exhaust gas
`egr, ε
`exhaust-gas recirculation
`f, ϕ, ψ fuel
`γ
`engine outlet
`l
`load
`m
`manifold or mean value
`seg
`segment
`t
`turbine
`ξ
`combustion
`ζ
`timing (e.g. of ignition or injection)
`
`In a turbocharged engine system, the four most important locations are des-
`ignated by the indices 1 for “before compressor,” 2 for “after compressor,” 3
`for “after engine,” and 4 for “after turbine.”
`In general, all numerical values listed in this text are shown in SI units. A
`few exceptions are made where non-SI units are widely accepted. These few
`cases are explicitly mentioned in the text.
`The most commonly used acronyms are:
`
`BDC (TDC)
`
`BMEP or pme
`bsfc
`CA
`CI
`CNG
`COM
`DEM
`DPF
`ECU
`IEG
`IPS
`IVC (IVO)
`EVC (EVO)
`MBT
`OC
`ODE
`ON
`PDE
`PM
`
`bottom (top) dead center (piston at lowest (topmost)
`position)
`(brake) mean-effective pressure
`brake specific fuel-consumption
`crank angle
`compression ignition (in Diesel engines)
`compressed natural gas
`control-oriented model
`discrete-event model
`Diesel particulate-filter
`electronic (or engine) control unit
`induction-to-exhaust delay
`induction-to-powerstroke delay
`inlet-valve closing (opening)
`exhaust-valve closing (opening)
`maximum brake torque (ignition or injection timing)
`oxidation catalyst
`ordinary differential equation
`octane number
`partial differential equation
`particulate matter
`
`11
`
`

`

`4
`
`1 Introduction
`
`SCR
`SI
`TPU
`TWC
`VNT
`WOT
`
`selective catalytic reduction
`spark ignition (in Otto/gasoline/gas engines)
`time-processing unit
`three-way catalytic converter
`variable-nozzle turbine
`wide-open throttle
`
`1.2 Control Systems for IC Engines
`
`1.2.1 Relevance of Engine Control Systems
`
`Future cars are expected to incorporate approximately one third of their parts
`value in electric and electronic components. These devices help to reduce the
`fuel consumption and the emission of pollutant species, to increase safety,
`and to improve the drivability and comfort of passenger cars. As the elec-
`tronic control systems become more complex and powerful, an ever increasing
`number of mechanical functions are being replaced by electric and electronic
`devices. An example of such an advanced vehicle is shown in Fig. 1.1.
`
`Fig. 1.1. Wiring harness of a modern vehicle (Maybach), reprinted with the per-
`mission of Daimler AG.
`
`In such a system, the engine is only one part within a larger structure.
`Its main input and output signals are the commands issued by the electronic
`
`12
`
`

`

`1.2 Control Systems for IC Engines
`
`5
`
`control unit (ECU) or directly by the driver, and the load torque transmitted
`through the clutch onto the engine’s flywheel. Figure 1.2 shows a possible
`substructure of the vehicle control system. In this text, only the “ICE” (i.e.,
`the engine and the corresponding hardware and software needed to control
`the engine) will be discussed.
`Control systems were introduced in ICE on a larger scale with the advent of
`three-way catalytic converters for the pollutant reduction of SI engines. Good
`experiences with these systems and substantial progress in microelectronic
`components (performance and cost) have opened up a path for the application
`of electronic control systems in many other ICE problem areas. It is clear
`that the realization of the future, more complex, engine systems, e.g., hybrid
`power trains or homogeneous charge compression ignition engines, will not be
`possible without sophisticated control systems.
`
`Fig. 1.2. Substructure of a complete vehicle control system.
`
`1.2.2 Electronic Engine Control Hardware and Software
`
`Typically, an electronic engine control unit (ECU) includes standard micro-
`controller hardware (process interfaces, RAM/ROM, CPU, etc.) and at least
`one additional piece of hardware, which is often designated as a time pro-
`cessing unit (TPU), see Fig. 1.3. This TPU synchronizes the engine control
`commands with the reciprocating action of the engine. The synchronization of
`
`13
`
`

`

`6
`
`1 Introduction
`
`the ECU with the engine is analyzed in more detail in Sec. 3.1.3.1 Notice also
`that clock rates of ECU microprocessors are typically much lower than those
`of desktop computers due to electromagnetic compatibility considerations.
`ECU software has typically been written in assembler code, with propri-
`etary real-time kernels. In the last few years there has been a strong ten-
`dency towards standardized high-level programming interfaces. Interestingly,
`the software is structured to reflect the primary physical connections of the
`plant to be controlled [70].
`
`command
`signals to
`engine
`
`amplifier, relays, etc.
`
`DA converter, digital output
`
`event
`controller
`(TPU)
`
`microcontroller
`
`AD converter, digital input
`
`input
`signals
`from
`engine
`
`crank−angle pulses
`
`Fig. 1.3. Internal structure of an electronic engine control unit.
`
`1.3 Overview of SI Engine Control Problems
`
`1.3.1 General Remarks
`
`The majority of modern passenger cars are still equipped with port (indi-
`rect) injection spark-ignited gasoline engines. The premixed and stoichiomet-
`ric combustion of the Otto process permits an extremely efficient exhaust gas
`purification with three-way catalytic converters and produces very little par-
`ticulate matter (PM). A standard configuration of such an engine is shown in
`Fig. 1.4.
`The torque of a stoichiometric SI engine is controlled by the quantity
`of air/fuel mixture in the cylinder during each stroke (the quality, i.e., the
`air/fuel ratio, remains constant). Typically, this quantity is varied by chang-
`ing the intake pressure and, hence, the density of the air/fuel mixture. Thus,
`a throttle plate is used upstream of the combustion process in the intake
`system. This solution is relatively simple and reliable, but it produces sub-
`stantial “pumping losses” that negatively affect the part-load efficiency of the
`
`1 The reciprocating or event-based behavior of all ICE also has important conse-
`quences for the controller design process. These problems will be addressed in
`Chapters 3 and 4.
`
`14
`
`

`

`1.3 Overview of SI Engine Control Problems
`
`7
`
`engine. Novel approaches, such as electronic throttle control, variable valve
`timing, etc., which offer improved fuel economy and pollutant emission, will
`be discussed below.
`
`ECU
`
`MA
`
`ET
`
`PM
`
`TA
`
`VE
`
`FP
`
`CP
`
`IC
`
`λ
`1
`
`λ
`2
`
`DP
`
`TWC
`
`TWC
`
`SA
`
`CCV
`
`CC
`
`Tank
`
`CCV
`
`AK
`
`TE
`
`SE
`
`AK
`CP
`IC
`MA
`SE
`FP
`
`knock sensor
`camshaft sensor
`ignition command
`air mass-flow sensor
`engine speed sensor
`fuel pressure control
`
`PM
`ET
`TA
`TE
`CC
`λ
`
`1,2
`
`manifold pressure sensor
`electronic throttle
`intake air temperature sensor
`cooling water temperature sensor
`active carbon canister
`air/fuel ratio sensors
`
`VE
`SA
`TWC
`ECU
`CCV
`DP
`
`EGR valve
`secondary air valve
`3-way catalyst
`controller
`CC control valves
`driver pedal
`
`Fig. 1.4. Overview over a typical SI engine system structure.
`
`A simplified control-oriented substructure of an SI engine is shown in
`Fig. 1.5. The main blocks are the fuel path Pϕ and the air path Pα, which
`define the mixture entering the cylinder, and the combustion block Pχ that
`determines the amount of torque produced by the engine.
`Other engine outputs are the knock signal yζ (as measured by a knock
`sensor Pζ) and the engine-out air/fuel ratio yλ (as measured by a λ sensor Pλ
`mounted as close as possible to the exhaust valves). The engine speed ωe is
`the output of the block PΘ, taking into account the rotational inertia of the
`engine, whose inputs are the engine torque Te and the load torque Tl.
`The four most important control loops are indicated in Fig. 1.5 as well:
`the fuel-injection feedforward loop;
`the air/fuel ratio feedback loop;
`the ignition angle feedforward2 loop; and
`the knock feedback loop.
`In addition, the following feedforward or feedback loops are present in
`many engine systems:3
`
`•
`•
`•
`•
`
`2 Closed-loop control has been proposed in [60] using the spark plug as an ion
`current sensor.
`3 Modern SI engines can include several other control loops.
`
`15
`
`

`

`8
`•
`•
`•
`•
`
`1 Introduction
`
`idle and cruise speed control;
`exhaust gas recirculation (for reducing emission during cold-start or for
`lean-burn engines);
`secondary air injection (for faster catalyst light-off); and
`canister purge management (to avoid hydrocarbon evaporation).
`
`Fig. 1.5. Basic SI engine control substructure.
`
`1.3.2 Main Control Loops in SI Engines
`
`Air/Fuel Ratio Control
`
`The air/fuel ratio control problem has been instrumental in paving the road
`for the introduction of several sophisticated automotive control systems. For
`this reason, it is described in some detail.
`The pollutant emissions of SI engines (mainly hydrocarbon (HC), carbon
`monoxide (CO), and nitrogen oxide (N Ox)) greatly exceed the limits imposed
`by most regulatory boards, and future emission legislation will require sub-
`stantial additional reductions of pollutant emission levels. These requirements
`can only be satisfied if appropriate exhaust gas after-treatment systems are
`used.
`The key to clean SI engines is a three-way catalytic converter (TWC)
`system whose stationary conversion efficiency is depicted in Fig. 1.6. Only for
`a very narrow air/fuel ratio “window,” whose mean value is slightly below
`the stoichiometric level, can all three pollutant species present in the exhaust
`
`16
`
`

`

`1.3 Overview of SI Engine Control Problems
`
`9
`
`Fig. 1.6. Conversion efficiency of a TWC (after light-off, stationary behavior).
`
`gas be almost completely converted to the innocuous components water and
`carbon dioxide. In particular, when the engine runs under lean conditions,
`the reduction of nitrogen oxide stops almost completely, because the now
`abundant free oxygen in the exhaust gas is used to oxidize the unburned
`hydrocarbon and the carbon monoxide. Only when the engine runs under
`rich conditions do the unburned hydrocarbon (HC) and the carbon monoxide
`(CO) act as agents reducing the nitrogen oxide on the catalyst, thereby causing
`the desired TWC behavior.
`The mean air/fuel ratio can be kept within this narrow band only if elec-
`tronic control systems and appropriate sensors and actuators are used. The
`air/fuel ratio sensor (λ sensor) is a very important component in this loop. A
`precise fuel injection system also is necessary. This is currently realized using
`“sequential multiport injectors.” Each intake port has its own injector, which
`injects fuel sequentially, i.e., only when the corresponding intake valves are
`closed.
`Finally, appropriate control algorithms have to be implemented in the
`ECU. The fuel-injection feedforward controller Fϕ tries to quickly realize a
`suitable injection timing based only on the measured air-path input informa-
`tion (either intake air mass flow, intake manifold pressure, or throttle plate
`angle and engine speed). The air/fuel ratio feedback control system Cλ com-
`pensates the unavoidable errors in the feedforward loop. While it guarantees
`the mean value of the air/fuel ratio to be at the stoichiometric level, it cannot
`prevent transient excursions in the air/fuel ratio.
`
`Ignition Control
`
`Another important example of a control system in SI engines is the spark angle
`control system. This example shows how control systems can help improve fuel
`economy as well.
`
`17
`
`

`

`10
`
`1 Introduction
`
`In fact, the efficiency of SI engines is limited, among other factors, by
`the knock phenomenon. Knock (although still not fully understood) results
`from an unwanted self-ignition process that leads to locally very high pressure
`peaks that can destroy the rim of the piston and other parts in the cylinder.
`In order to prevent knocking, the compression ratio must be kept below a safe
`value and ignition timing must be optimized off-line and on-line.
`A first optimization takes place during the calibration phase (experiments
`on engine or chassis dynamometers) of the engine development process. The
`nominal spark timing data obtained are stored in the ECU. An on-line spark
`timing control system is required to handle changing fuel qualities and engine
`characteristics. The key to this component is a knock sensor and the corre-
`sponding signal processing unit that monitors the combustion process and
`signals the onset of knocking.
`The feedforward controller Fζ, introduced in Fig. 1.5, computes the nomi-
`nal ignition angles (realizing maximum brake torque while avoiding knock and
`excessive engine-out pollution levels) depending on the engine speed and load
`(as measured by manifold pressure or other related signals). This correlation
`is static and is only optimal for that engine from which the ignition data was
`obtained during the calibration of the ECU. The feedback control system Cζ
`utilizes the output of the knock detection system to adapt the ignition angle to
`a safe and fuel efficient value despite variations in environmental conditions,
`fuel quality, etc.
`
`1.3.3 Future Developments
`
`Pollutant emission levels of stoichiometric SI engines are or soon will be a
`“problem solved” such that the focus of research and development efforts can
`be redirected towards the improvement of the fuel economy. The most severe
`drawbacks of current SI engines are evident in part-load operating conditions.
`As Fig. 1.7 shows, the average efficiency even of modern SI engines remains
`substantially below their best bsfc4 values. This is a problem because most
`passenger cars on the average (and also on the governmental test cycles) utilize
`less than 10% of the maximum engine power.5 Not surprisingly, cycle-averaged
`“tank-to-wheel” efficiency data of actual passenger cars are between 12% and
`18% only. The next step in the development of SI engines therefore will be a
`substantial improvement of their part-load efficiency.
`Several ideas have been proposed to improve the fuel efficiency of SI en-
`gines, all of which include some control actions, e.g.,
`• variable valve timing systems (electromagnetic or electrohydraulic);
`• downsizing and supercharging systems;
`• homogeneous and stratified lean combustion SI engines;
`4 Brake-specific fuel consumption (usually in g fuel/kWh mechanical work).
`5 Maximum engine power is mainly determined by the customer’s expectation of
`acceleration performance and is, therefore, very much dependent on vehicle mass.
`
`18
`
`

`

`1.4 Overview of Control Problems in CI Engines
`
`11
`
`p [bar]
`me
`
`10
`
`8
`
`6
`
`4
`
`2
`
`0.36
`
`0.33
`
`0.35
`
`0.3
`
`η=0.25
`
`0.2
`
`0.1
`
`4
`
`6
`
`8
`
`10
`
`12
`
`14
`
`16
`
`cm
`
`[m/s]
`
`Fig. 1.7. Engine map (mean effective pressure versus mean piston speed) of a
`modern SI engine, gray area: part-load zone, η =const: iso-efficiency curves. For the
`definition of pme and cm see Sect. 2.5.1.
`
`• variable compression ratio engines; and
`•
`engines with improved thermal management.
`These systems reduce the pumping work required in the gas exchange part
`of the Otto cycle, reduce mechanical friction, or improve the thermodynamic
`efficiency in part-load conditions.
`Another approach to improving part-load efficiency is to include novel
`power train components, such as starter-generator6 devices, CVTs7, etc. As
`mentioned in the Introduction, these approaches will not be analyzed in this
`text. Interested readers are referred to the textbook [81].
`
`1.4 Overview of Control Problems in CI Engines
`
`1.4.1 General Remarks
`
`Diesel engines are inherently more fuel-efficient than gasoline engines (see
`Appendix C), but they cannot use the pollutant abatement systems that have
`proved to be so successful in gasoline engines. In fact, the torque output of
`Diesel engines is controlled by changing the air/fuel ratio in the combustion
`
`6 These advanced electric motors and generators typically have around 5 kW me-
`chanical power and permit several improvements like idle-load shut-off strategies
`or even “mild hybrid” concepts.
`7 Continuously variable transmissions allow for the operation of the engine at the
`lowest possible speed and highest possible load, thus partially avoiding the low
`efficiency points in the engine map.
`
`19
`
`

`

`12
`
`1 Introduction
`
`chamber. This approach is not compatible with the TWC working principle
`introduced above.
`In naturally aspirated Diesel engines, the amount of air available is appro-
`ximately the same for all loads, and only the amount of fuel injected changes
`in accordance with the driver’s torque request. In modern CI engines the situa-
`tion is more complex since almost all engines are turbocharged. Turbochargers
`introduce additional feedback paths, considerably complicating the dynamic
`behavior of the entire engine system. Additionally, pre-chamber injection has
`been replaced by direct-injection systems.