`
`SA E TECHNICAL
`PAPER SERIES
`
`Development of Model-Following Idle Speed
`Control System Incorporating Engine
`Torque Models
`
`Satoru Watanabe, Masamichi Imamura, and Naoki Tomisawa
`Japan Electronic Control Systems Co., Ltd.
`
`Hai-jio Guo, Mitsuo Satoh, and Hiroshi Takeda
`Tohoku Univ.
`
`The Engineering Society
`eA =For
`Advancing Mobility
` and sea ~ i r
`and Spacem
`
`International Congress & Exposition
`Detroit, Michigan
`February 24-28,1992
`
`4 0 0 C O M M O N W E A L T H D R I V E , W A R R E N D A L E , P A 1 5 0 9 6 - 0 0 0 1 U.S.A.
`
`PAICE 2027
`BMW v. Paice
`IPR2020-00994
`
`1
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`The appearance of the ISSN code at the bottom of this page indicates SAE's consent
`that copies of the paper may be made for personal or internal use of specific clients.
`Thisconsent is given on thecondition, however, that thecopier pay a $5.00 per article
`copy fee through the Copyright Clearance Center, Inc. Operations Center, 27
`Congress St., Salem, MA 01970 for copying beyond that permitted by Sections 107
`or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of
`copying such as copying for general distribution, for advertising or promotional
`purposes, for creating new collective works, or for resale.
`
`SAE routinely stocks printed papers for a period of three years following date of
`publication. Direct your orders to SAE Customer Service Department.
`
`To obtain quantity reprint rates, permission to reprint a technical paper or permission
`to use copyrighted SAE publications in other works, contact the SAE Publications
`Group.
`
`All SAE papers, standards, and selected
`books are abstracted and indexed in the
`SAE Global Mobility Database.
`
`No part of this publication may by reproduced in any form, in an electronic retrieval
`system or otherwise, without the prior written permission of the publisher.
`
`ISSN 0148-7191
`Copyright 1992 Society of Automotive Engineers, Inc.
`
`Positions and opinions advanced in this paper are those of the author(s) and not
`necessarily those of SAE. The author is solely responsible for the content of the
`paper. A process is available by which discussions will be printed with the paper if
`it is published in SAE transactions. For permission to publish this paper in full or in
`part, contact the SAE Publications Division.
`
`Persons wishing to submit papers to be considered for presentation or publication
`through SAE should send the manuscript or a 300 word abstract of a proposed
`manuscript to: Secretary, Engineering Activity Board, SAE.
`
`Printed in USA
`
`2
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`Development of Model-Following ldle Speed
`Control System Incorporating Engine
`-
`Torque Models
`Satoru Watanabe, Masamichi Imamura, and Naoki Tomisawa
`Japan Electronic Control Systems Co., Ltd.
`
`Hai-jio Guo, Mitsuo Satoh, and Hiroshi Takeda
`Tohoku Univ.
`
`1. Abstract
`Control of internal combustion engines depends
`on reduced idle speed and stabilized idling to
`improve fuel consumption.. Using a valve (ISCIV) to
`adjust air intake at idling, various idle speed control
`techniques have been proposed to improve response
`to variations in the target idle speed due to
`disturbance and to enhance speed control stability.
`
`Many conventional idle speed control systems
`utilize P and I control. They cannot compensate for
`the phase difference caused by the intake air volume
`between the throttle and intake valves. This leads to
`poor response to variations in the target idle speed
`and unsatisfactory control accuracy, that is, problems
`in preventing engine stalling and maintaining idle
`speed stability.
`
`To solve these problems, an idle speed control
`system based on fuzzy control(l)'theory is proposed.
`However, it is impossible to determine the control
`constant (membership function) theoretically. This
`must be determined on an experimental basis and it is
`very labor-intensive.
`
`This paper describes a control system with the
`following features:
`
`(1) As idle speed control using engine torque
`signals offers an effective
`response to
`disturbance, a model was prepared that
`provides the desired control characteristics of
`both engine speed and torque.
`
`(2) A feedback control system with rapid following
`to model output was built.
`
`(3) Phase difference caused by intake air volume is
`compensated.
`The model following idle speed control system
`with these features was installed on an engine
`control unit and its evaluated performance was
`far superior to conventional control systems.
`
`.
`.
`2. m e c t ~ v e s of ldle Speed C o n t r a
`
`ldle speed control recovers and stabilizes the
`actual engine speed reduced by load variations due
`to known and unknown disturbances to the target
`speed.
`
`One method provides extra intake air and fuel to
`generate the torque required for the given load and to
`control ignition timing. This method is unfavorable
`because fuel cost regulations are expected to be
`stricter
`in the
`future, and fuel consumption
`deteriorates when idling with this method. For this
`reason, we selected a method that reduces intake air
`as much as possible and controls air intake to
`generate the required torque.
`
`3 . Dbiectives o f AuxlUarv Air Control
`
`Figure 1 shows a typical intake system which is
`controlled electronically.
`
`The opening of the air passage that bypasses
`the throttle valve is controlled for variations in engine
`speed. However, the size of the intake system
`extending from the throttle valve to the intake valve is
`far greater than the volume of bypass air. This makes
`response to load changes slow and increases
`variations in engine speed. The method we now use
`is designed to control air intake by the difference
`between the target idle speed and actual engine
`speed. The situation is further aggravated by the
`engine speed changing energy or the delay in the
`transmission system, that is, the torque-speed
`converter.
`
`To solve these problems, we studied how to
`compensate for the delay in the intake system and to
`estimate actual torque.
`
`3
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`Idle speed control valve
`
`- Fig. 1 Typical Intake System
`
`Figure 2 shows a model of a simple intake
`system.
`
`INTIMANI ENGINE
`
`Fig. 2 A Simple intake System
`
`To compensate for the delay in the intake
`
`system, we need only c~mpensate'~) for the delay in
`the collector filling which has the largest size in the
`intake system. The volume of air passing through an
`intake value, Q, is shown by equation 1.
`
`Qt = Qc + p-Vm0A(Pmflm)/At .......................... (1
`where p: air density.
`
`Assuming that the change in pressure and
`temperature in the collector is the same as that
`immediately before the intake valve, we obtain the
`following equation:
`
`A(Pmflm)=A(Pcflc) = ACQd(vcep*~)) ........... (2)
`where E: ratio of fresh air.
`
`From equations 1 and 2, we obtain:
`
`Qt = Qc+Vm*A(Qd(Vc-&))/At
`.................... (3)
`= Qc+Vme(QoQc-1 )/(V~*E*A~)
`where Ne: engine speed,
`q: filling efficiency
`
`Total volume of intake air in the cylinder is given
`by equation 4.
`Qc = Vc*Ne*p*q ..................................................
`where Ne: engine speed,
`q: filling efficiency
`
`(4)
`
`Equations 3 and 4 lead us to:
`............... (5)
`= ~e*p*(ve*rl+vm*(q-q-i/(~*~t))
`
`Compensation for delay in the collector filling is
`determined by equation 5, so we need only to
`introduce the result in the control system.
`
`Torque generated in a normal operating engine
`is determined by total inertia moment I, coefficient of
`viscosity of engine C, and pumping loss Tpump, and
`is given by equation 6.
`
`Pumping loss is determined uniquely by the
`piston displacement of an engine, and hence may be
`omitted when considering a feedback control system.
`
`Target torque Tset at no-load is given by
`equation 8.
`
`Tset = cmset-k ..................................................... (8)
`where oset: target speed,
`k: load torque.
`
`Torque = Tset must be satisfied if actual torque
`is to coincide with the target torque. From equations 7
`and 8, we obtain
`K = -Ih+C(oset-o) ..................................... (9)
`
`4
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`This means that torque can be estimated using
`an engine's angular velocity and angular
`acceleration.
`
`6 . Seletlon of Control S v s t e q
`
`Conventional idle speed controls use classical
`control formats such as PID control shown in Figure 3.
`However, since engine characteristics are non-linear
`and include secondary delay, it is necessary to assign
`PID control constants to each condition for various
`types of disturbance. This requires a huge amount of
`time and skill for proper vehicle fitting.
`
`PII) c o n t r o l
`I-1'1) c o , , t r . o l
`P r 0 g r ; l l n control
`O t l ~ c r c o l ~ t r o l
`
`Inputs are made to a control model with the
`same characteristics as the control subject. Outputs
`from both controller and control subject are detected
`and the difference compared to perform necessary
`compensation. Outputs of the control subject always
`follow those of the control model. Figures 5 and 6
`show block diagrams of idle speed control that uses
`model following control.
`
`Target speed is input as a control input to the
`engine model. The torque and speed of the model
`and those of the actual engine are compared. Torque
`error calculated from the comparison is compensated
`by load and no-load torque estimating circuits and an
`integrator followed by compensation of delay in the
`collector filling using a collector model.
`
`L<,r,l"" (.,,a
`"I , L ,m , l , , ,
`
`,,,,,
`colllrol of
`
`Torque c . l l c s ~ l ~ l ~ u s
`
`I
`
`Stability of nlsp<,sre
`lay cc>l.ge,,a;,ling
`'I<.l.ly collector rrlli,,~
`
`Engine rpcccl
`llnginc lorqlla
`
`Ellgina
`
`L--1 ~ o n t r o l o f - T j
`
`J
`
`Flg. 5 Application to Idle Speed Control
`
`Fig. 3 Position of Model Following Control
`i n Control Theory
`
`Since the characteristics of the control subject
`can be clearly expressed by modeling the delay of the
`collector filling and estimating the torque required for
`an engine, we studied application of modern control
`theory using these characteristics.
`
`MRACS, which has a reference model and an
`application system for the model, is considered the
`most suitable control because of its adaptability and
`ability to withstand environmental changes. But,
`MRACS requires complex algorithms and is difficult to
`mount on an actual engine control unit. For these
`reasons, we decided to adopt model following control
`that compensates for errors in the model by integral
`element.
`
`7. p ~ ~ l i c a t i o n o f Model Followina Control
`
`The outline of model following control is shown
`in Figure 4.
`
`D i s t u r b a n c e
`
`Fig. 6
`
`Idle Speed Control Block Diagram
`
`The load torque estimating circuit cancels the
`angular velocity of rotation caused by load torque and
`maintains constant speed as shown in Figure 7. The
`no-load torque estimating circuit and the integrator
`cancel the rotational deviation from the target and
`converge the speed to the target speed as shown in
`Figure 8.
`
`Flg. 4 Concept of Model Followlng Control
`
`Fig. 7 Load torque estimating circuit
`
`5
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`Tmodel* Teng
`
`Nmodel
`
`Model input
`T P ~ P ~
`
`Fig. 8 No-load torque estimating circuit
`
`Figures 9 through 11 are examples of outputs
`from torque estimating circuits when a step load is
`input. Examples of the final control torque output, are
`shown in Figure 12.
`
`~ m o d a
`
`Ne
`
`Model input
`
`Tmodel
`
`Engine input
`
`Teng
`
`Nmodel
`
`NeF
`
`Model output
`
`Engine output
`
`Tmodel
`
`Ten-
`
`.
`
`Load input torque
`Model output torque - engine output torque
`Load output torque
`Model input torque -
`engine input torque n
`Final output torque
`Load input torque -
`load output torque
`
`Engine output
`
`TPmPe L
`I
`
`Final output torque
`
`NIL output torque - NIL input torque
`
`Fig. 10 Output of No-ioad Torque
`Estimating Circuit
`(Tpump)
`
`Model input
`
`Model output
`
`- Nmodel
`
`Nmodel
`
`Engine input
`
`Engine output
`
`Fig. 9 Output of Load Torque Estimating
`Circuit
`(Tmodel and Teng)
`
`Final output torque
`
`Model ouput speed - engine
`output speed
`
`i
`
`Fig. 11 Output of Deviation integrating
`Torque (Tintg)
`
`6
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`Good controllability was also observed when
`the target control speed of the engine was decreased
`150 rpm, as shown in Figure 18.
`
`/ Nmodel --Q-
`
`ISCN flow rate resolution: 4 ~ I n ~ i ~ i l s t e ~
`1l;~rdwnrs targcl opclling
`
`I
`
`kzIIertIw;bro opcr;ltion spard
`
`, ---- l--:- ---------------
`i
`/,/
`I ,/
`t
`
`Tmodel
`
`Tintg
`
`-
`
`'/1
`i
`;j
`
`100
`
`0
`
`7 2 5 -
`700-
`
`,!
`, I I
`I
`I
`
`\
`/I\IF-ISC c ; ~ l e ~ l l a t i ~ n Lilrget o1,ening
`
`Lo;,cl
`0 N
`
`f'
`
`---- MP-ISC colltro~
`----
`Actual l~nrdwnro
`cnp:tI~ility
`
`GOO -
`2 5 5 0
`5011 -
`415.
`0 00
`
`Time [SCCI
`
`2.55
`
`3.110
`
`Fig. 13 Comparison of MF-ISC and Actual
`Hardware Capability
`
`Idle s ~ e e d control P
`-
`I
`air volume
`
`L
`
`
`
`Fig. 12 Final Control Torque Output
`
`8 . Controllablllty
`
`Our simulation used the control block diagram
`shown in Figure 6 to confirm the effect of model
`following control.
`First, we evaluated control
`performance in relation to the marginal performance
`of the hardware. We defined the hardware limit to
`the same
`have an excess ISCIV target opening at
`timing as load input, followed up by the marginal
`operation speed of the ISCIV. The results were
`compared with the model following control. As shown
`in Figure 13, its controllability is equivalent to the
`margianl performance of the hardware for the
`decreased speed due to load input. As the results in
`Figure 14 show, with step load input we achieved an
`engine speed reduction of 50 rpm or better than our
`conventional PI control.
`Verification tests on an
`actual vehicle have produced an effect equivalent to
`the results of our simulation.
`
`We also confirmed the controllability of load
`inputs such as air conditioner, power steering and
`headlights using an actual vehicle. The results are
`shown in Figure 15. To evaluate idling controllability
`in an actual vehicle, we used two elements: speed
`variation ratio and target speed reset time. The
`results are plotted in Figure 16, where the two
`elements are assigned to vertical and horizontal axes.
`Distance from the origin was defined as the idling
`index, and the size of the index was evaluated. The
`results were
`improved controllability over
`conventional controls, as shown in Figure 17.
`
`1 Results of c x p e r i ~ a c n t
`
`llcs~rlts of s i n ~ u l ; ~ t i o n
`
`700r1~111r
`
`675rpt11 -
`
`G50rpm .
`
`625r1)n1-
`
`575rp111
`
`- 1'1 control 4 ,/
`
`V
`Resuits of sin~ulntion an11 c x p e ~ . i ~ ~ ~ e n l
`Spcecl loss retlr~cc[l 50 rprn o r nlorc in
`
`Fig. 14 Correlation between Actual
`Machine and Simulation
`
`I.oitcl: A'I'IN
`
`-+ 1) select
`
`: I.o:sd corn1~ons;ttion ~tvnil;ll,le
`-MF-ISC
`------ Cun"~h1 control:
`KISCP (5kl
`KISCI' 9CII
`
`"
`
`"
`
`Fig. 15 Comparison s f Current Control vs.
`MF-ISC Control
`
`7
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`I.o:~cl: AT/N --i I) select
`
`000-
`
`50(1 -
`
`to
`
`4 0 0 -
`
`.g
`3 'g 300-
`B
`f 200-
`
`W ~ f i l O lSClV Sperifi<:;utionb
`0 Current central. I.a;td conlpenr;ttion not i~v;~il.tble
`1CISCI' DCII
`X Cntrranl eantt.ol: 1.0;t'l cvnlllens;ttia,n not nvi~il:~ble
`KISCI'+SII
`A Current control: I.uilcl contpenr;stion itvnil.tble
`I<ISCI' BCll
`O Current control; I.o;xl compens;mlian ;~v;tilslrlo
`
`blli-1SC: I.o;ncl conbprnsntiun avi~iI.tblc
`o hlli-ISC: I-I,ZIII ~ ~ ~ z ~ ~ p e n ~ i l l i o n
`not i $ ~ a i l n l d ~
`
`: p x n IOSCI'PSII
`,
`
`A
`
`I
`
`@
`
`Fig. 16 Comparison of Current ISC vs. MF-
`ISC Control
`
`Fig. 17 Effect of MIF-ISC on lndivldual
`Loads
`
`Lo;bd: PIS (N r;r~igc)
`
`0 N
`
`LOAD
`
`OFF
`
`OFF
`
`MF-ISC
`increase 7%
`
`1
`
`I
`
`PI control
`
`increase 28%
`
`I
`
`RAM
`
`I
`
`I
`
`I
`
`ROM
`
`Fig. 19 Comparison of Memory
`Requirement
`
`As shown in Figure 14, model following idle
`speed control correlate highly with actual engine
`characteristics. This proves the accuracy the physical
`models of the torque estimating circuits and collector
`compensator.
`
`We developed a matching support tool using
`simulation as the core. The matching support tool is
`capable of calculating optimum control constants
`including optimum design values for the control
`device (ISCIV). The effect of the matching support
`tool is equal to a 50 Oh
`improvement in development
`efficiency over conventional PI controls, as shown in
`Figure 20.
`
`- 0
`
`2
`1
`Time (soc)
`
`3
`
`
`
`I
`
`gj '5
`E er '5 a
`@ <
`0 0
`0 0
`
`Conventional control
`
`50% reduction
`
`20%
`30%
`
`Evaluation of
`Evaluation of
`simulation
`
`&IF-ISC
`
`I
`design man-hours 2
`
`Fig. 18 MF-ISC Control Evaluatlon of Step
`Load Speed Variation
`
`9. Qn-Board Control
`
`Although it is difficult to build control system
`incorporating internal models in an engine control unit
`that has many intrinsic restrictions on memory
`capacity and processing speed, we have achieved an
`on-board version with minimal increase in memory
`capacity as shown in Figure 19. We simplified the
`model based on the fact that the control ragne is
`limited to the idling area an by compensating errors of
`the model with integrating elements.
`
`-Control
`
`Effect
`
`Time
`
`Actual vehicle
`matching
`
`50% reduction
`
`Evaluation of
`simulation
`
`20%
`
`MF-ISC
`
`I C o n f o r r n i
`
`t y man-hours A
`Fig. 20 Effect of Introducing Matching
`Support Tool
`
`8
`
`
`
`Downloaded from SAE International by University of Alberta Libraries, Monday, January 18, 2021
`
`11. C o n c l u s i o n
`
`We built a torque estimating model and a
`collector filling model and incorporated them in a
`model following idle speed control system. With this
`system we have achieved:
`
`to
`(1) Equivalent controllability
`performance of the hardware.
`
`the marginal
`
`(2)
`
`Improved controllability 55% over conventional
`controls.
`
`(3) On-board version of model following control with
`minimal increase in memory.
`
`(4) A matching support tool incorporating simulators
`of torque estimating and collector filling models.
`This has enabled matching, optimum design of
`the control device and efficiency improved 50%
`over conventional methods.
`
`'1
`
`Y. Nishimura, K. Ishii, Engine Idle Stability
`Analysis and Control, SAE paper 860415, 1986.
`JSAE Handbook, vol. 1, 1989
`*2
`*3 Y. Ohoshima, S. Inaba, Introduction of Automatic
`Control, Maruzen Co., 1979
`
`9
`
`