CN111648872A - Method for calibrating volumetric efficiency map of automobile engine - Google Patents

Method for calibrating volumetric efficiency map of automobile engine Download PDF

Info

Publication number
CN111648872A
CN111648872A CN202010465311.0A CN202010465311A CN111648872A CN 111648872 A CN111648872 A CN 111648872A CN 202010465311 A CN202010465311 A CN 202010465311A CN 111648872 A CN111648872 A CN 111648872A
Authority
CN
China
Prior art keywords
map
volumetric efficiency
engine
intake manifold
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010465311.0A
Other languages
Chinese (zh)
Inventor
王昌辉
梁美
柴永生
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yantai University
Original Assignee
Yantai University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yantai University filed Critical Yantai University
Priority to CN202010465311.0A priority Critical patent/CN111648872A/en
Publication of CN111648872A publication Critical patent/CN111648872A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D41/00—Electrical control of supply of combustible mixture or its constituents
    • F02D41/02—Circuit arrangements for generating control signals
    • F02D41/14—Introducing closed-loop corrections
    • F02D41/1438—Introducing closed-loop corrections using means for determining characteristics of the combustion gases; Sensors therefor
    • F02D41/1444—Introducing closed-loop corrections using means for determining characteristics of the combustion gases; Sensors therefor characterised by the characteristics of the combustion gases
    • F02D41/1454—Introducing closed-loop corrections using means for determining characteristics of the combustion gases; Sensors therefor characterised by the characteristics of the combustion gases the characteristics being an oxygen content or concentration or the air-fuel ratio
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D2200/00—Input parameters for engine control
    • F02D2200/02—Input parameters for engine control the parameters being related to the engine
    • F02D2200/04—Engine intake system parameters
    • F02D2200/0406—Intake manifold pressure
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D2200/00—Input parameters for engine control
    • F02D2200/02—Input parameters for engine control the parameters being related to the engine
    • F02D2200/10—Parameters related to the engine output, e.g. engine torque or engine speed
    • F02D2200/101—Engine speed
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
    • F02D—CONTROLLING COMBUSTION ENGINES
    • F02D2200/00—Input parameters for engine control
    • F02D2200/70—Input parameters for engine control said parameters being related to the vehicle exterior
    • F02D2200/703—Atmospheric pressure

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Combined Controls Of Internal Combustion Engines (AREA)

Abstract

The invention discloses a calibration method for volumetric efficiency map of an automobile engine, which comprises the following steps: s1: designing mesh division of the map intake manifold pressure and the engine speed according to the trajectory range of the volumetric efficiency map intake manifold pressure and the engine speed; s2: dividing the actual pressure and the actual rotating speed of the intake manifold of the engine into areas through a grid in the step S1, and acquiring transient working condition data of the engine; s3: and establishing a volumetric efficiency map self-adaptive identification algorithm, and training and estimating volumetric efficiency map parameters through the algorithm. The invention has the beneficial effects that: the calibration method is a new method for calibrating the volumetric efficiency map model of the automobile engine, and can finish the calibration work of the volumetric efficiency map model under the condition of not using steady-state working condition data of the engine.

Description

Method for calibrating volumetric efficiency map of automobile engine
Technical Field
The invention belongs to the technical field of engine production calibration, and particularly relates to a volumetric efficiency MAP rapid calibration method for transient working condition data.
Background
To meet increasingly stringent emission legislation requirements and to achieve good power and fuel economy in automotive engines, air-fuel ratio control is one of the key technologies for current engine emission control. At present, the challenge problems of air-fuel ratio control mainly include high-precision cylinder air inflow estimation, time-varying time lag control and wall wetting effect compensation. For the estimation of cylinder air charge, ideally, the cylinder air charge per cycle is equal to the product of air density and engine displacement. Due to the short cycle time and the flow restriction caused by the air filter, intake manifold and intake valve, the actual cylinder intake air amount is less than the ideal intake air amount. In order to describe the actual cylinder air inflow, the volumetric efficiency is defined as the ratio of the actual cylinder air inflow to the ideal air inflow, and the actual cylinder air inflow can be obtained through a speed density equation of a model of the volumetric efficiency and the ideal air inflow.
At present, the air inflow of an air cylinder cannot be directly measured through a sensor, but the air inflow of the air cylinder is approximately obtained through measurement of an air mass flow sensor when an engine is in a steady state, and a volumetric efficiency measurement value is obtained by combining a speed density equation. In order to identify the model relationship of the volumetric efficiency, a dynamometer is generally adopted to operate the engine at different working points in a steady state, and volumetric efficiency model parameters are identified by measuring a large amount of steady-state input and output data, or higher-precision volumetric efficiency MAP is calibrated. The volumetric efficiency calibration is carried out under the steady-state working condition of the engine, but the volumetric efficiency MAP obtained by calibration based on steady-state data can reflect the air intake amount of the cylinder under the transient working condition. While steady-state conditions can transform complex dynamical model identification problems into parametric measurement problems for algebraic models, steady-state condition data is expensive from a time cost perspective. The engine needs at least 10 minutes [14] to stabilize at a designated grid node, if 15-point two-dimensional maps (input is rotating speed and load) are divided for each input shaft, 225 grid nodes need to be calibrated, and 37 hours are needed for completing complete map calibration. Under the background, a volumetric efficiency MAP rapid calibration method adopting transient working condition data is designed, and the problems of long MAP calibration time and high cost existing at present are solved.
Disclosure of Invention
The calibration method for the volumetric efficiency MAP of the automobile engine is provided for solving the problems that in the prior art, a large amount of time is needed for calibrating the volumetric efficiency MAP to obtain engine steady-state data, the cost is high, and the period is long.
A calibration method for volumetric efficiency map of an automobile engine comprises the following steps:
s1: designing mesh division of the map intake manifold pressure and the engine speed according to the trajectory range of the volumetric efficiency map intake manifold pressure and the engine speed;
s2: dividing the actual pressure and the actual rotating speed of the intake manifold of the engine into areas through a grid in the step S1, and acquiring transient working condition data of the engine;
s3: and establishing a volumetric efficiency map self-adaptive identification algorithm, and training and estimating volumetric efficiency map parameters through the algorithm.
Preferably, in step S2, the adjustment of the intake manifold pressure is controlled by the throttle opening degree.
Preferably, the engine is connected with a dynamometer, and the regulation of the engine speed is realized through the dynamometer.
Preferably, the establishment of the adaptive identification algorithm in step S3 includes the following steps:
s31: establishing an engine air inlet system model:
Figure RE-GDA0002593745570000021
wherein
Figure RE-GDA0002593745570000031
Figure RE-GDA0002593745570000032
Figure RE-GDA0002593745570000033
Figure RE-GDA0002593745570000034
Figure RE-GDA0002593745570000035
In the formula
WthFor throttle mass flow, WeiThe amount of intake air is the amount of the cylinder,
Timfor intake manifold temperature，neIs the rotational speed of the engine and,
pimis intake manifold pressure, uthIs the opening degree of the throttle valve,
d is the diameter of the throttle valve, pi is the pressure ratio,
patm,Tatmis the external atmospheric pressure and temperature, gamma is the adiabatic coefficient,
Rais a gas constant, CdIs the flow coefficient, c0,c1,c2,b0,b1In order to be the parameters of the model,
ηvfor volumetric efficiency, the actual cylinder intake air quantity W is characterizedei,actualAnd the ideal air input Wei,idealI.e. ηv＝Wei,actual/Wei,ideal；
S32 volumetric efficiency η is given across the division of the volumetric efficiency two-dimensional MAP inputvMathematical description of MAP of (1):
Figure RE-GDA0002593745570000036
s33: writing the map mathematical description into the expression form of regression vector, and calculating the approximate error of the volumetric efficiency to obtain the regression model description of the volumetric efficiency map
Figure RE-GDA0002593745570000041
Wherein r isη＝ηv(υ)-ηv,T(θ, upsilon) is the approximation error of the MAP regression model;
s34: using intake manifold pressure pimThe MAP mixed intake system model of the engine obtained by combining the engine intake system model and the MAP regression model of the volumetric efficiency as the system measurement output is as follows:
Figure RE-GDA0002593745570000042
wherein the system state x is the intake manifold pressure pimThe non-linear term g (x, n)e,uth) Mass flow W through the throttlethThe model of (d) is obtained in the form of a regression vector of Φ (v), and (p) is expressed as vim,ne) Is a measurable MAP input, R is the system model error caused by the approximation error R;
s35: aiming at a map mixed air inlet system model of an engine, an adaptive observer is set as follows:
Figure RE-GDA0002593745570000043
wherein the content of the first and second substances,
Figure RE-GDA0002593745570000044
gain of
Figure RE-GDA0002593745570000045
Is a positive-determined diagonal matrix of the matrix,
Figure RE-GDA0002593745570000046
in order to estimate the state of the system,
Figure RE-GDA0002593745570000047
in order to be able to estimate the parameters,
Figure RE-GDA00025937455700000411
is a feedback gain matrix;
according to membership functions
Figure RE-GDA0002593745570000048
By noting that γ has the same sparsity as Φ (v), we get
Figure RE-GDA0002593745570000049
Wherein the vector γlAnd philThe dimension of (v) is the same,
Figure RE-GDA00025937455700000410
local parameter estimation for appropriate dimensionsThe number of the vector of counts is,la positive diagonal matrix of appropriate dimensions.
The invention has the beneficial effects that:
the calibration method is a new method for calibrating the volumetric efficiency map model of the automobile engine, and can finish the calibration work of the volumetric efficiency map model under the condition of not using steady-state working condition data of the engine.
Drawings
FIG. 1 is a block diagram of an engine intake model;
FIG. 2 is a diagram of a volumetric efficiency MAP estimation simulation architecture;
FIG. 3 shows throttle opening uth and engine speed ne in dynamometer mode;
FIG. 4 is intake manifold pressure pim and engine speed ne;
FIG. 5 is a trace plot of data;
FIG. 6 is a graph of volumetric efficiency MAP estimates;
FIG. 7 is a graph of the MAP estimation results after extrapolation completion;
FIG. 8 is a graph comparing estimated MAP to estimated enDYNA;
fig. 9 compares the cylinder intake air amount model with enDYNA.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings in the embodiment of the invention.
Volumetric efficiency of both diesel engines with EGR and VGT and SI gasoline engines is related to intake manifold pressure and engine speed. .
Without loss of generality, the engine air inlet structure is shown in fig. 1, and the air inlet system model is as follows:
Figure RE-GDA0002593745570000051
wherein
Figure RE-GDA0002593745570000061
Figure RE-GDA0002593745570000062
Figure RE-GDA0002593745570000063
Figure RE-GDA0002593745570000064
Figure RE-GDA0002593745570000065
W in the formulae (1-1) and (1-2)thFor throttle mass flow, WeiIs the cylinder intake air quantity, TimIs the intake manifold temperature, neIs the engine speed, pimIs intake manifold pressure, uthIs the opening degree of the throttle valve, D is the diameter of the throttle valve, pi is the pressure ratio, patm,TatmIs the external atmospheric pressure and temperature, gamma is the adiabatic coefficient, RaIs a gas constant, CdIs the flow coefficient, c0,c1,c2,b0,b1η as model parametersvFor volumetric efficiency, the actual cylinder intake air quantity W is characterizedei,actualAnd the ideal air input Wei,idealI.e. ηv＝Wei,actual/Wei,ideal。
To estimate the volumetric efficiency initial MAP, the volumetric efficiency η is given belowvMAP mathematical description of engine volumetric efficiency ηvIs the intake manifold pressure pimAnd engine speed neFunction of (d), noted as ηv(pim,ne). Defining volumetric efficiency two-dimensional MAP input upsilon ═ pim,ne) The method comprises the following steps:
Figure RE-GDA0002593745570000066
wherein the content of the first and second substances,
Figure RE-GDA0002593745570000067
is pimMinimum and maximum of, p1Is in the interval [ a, b]The number of grid points divided.
Figure RE-GDA0002593745570000068
Is neMinimum and maximum of, p2Is in the interval [ c, d]The number of grid points divided.
Let a parameter thetai,jInputting lattice points for MAP
Figure RE-GDA0002593745570000069
Corresponding volumetric efficiency, i.e.
Figure RE-GDA00025937455700000610
Then, for
Figure RE-GDA00025937455700000611
And
Figure RE-GDA00025937455700000612
η can be obtainedvPiecewise bilinear interpolation model ηv,T(θi,j,υ)：
Figure RE-GDA0002593745570000071
By referring to the description method of the MAP regression model, η in the formula (1-5)v,T(θ, upsilon) extends to an undefined interval
Figure RE-GDA0002593745570000072
Above, and written as a regression vector expression form as follows:
ηv,T(θ,υ)＝Φ(υ)·θ (1-6)
wherein
Figure RE-GDA0002593745570000073
Figure RE-GDA0002593745570000074
Figure RE-GDA0002593745570000075
And
Figure RE-GDA0002593745570000076
Figure RE-GDA0002593745570000081
and
Figure RE-GDA0002593745570000082
Figure RE-GDA0002593745570000083
the volumetric efficiency MAP regression model (1-6) is essentially a piecewise bilinear interpolation model, is an approximation to volumetric efficiency, and has an approximation error. The more the MAP input is divided, the smaller the approximation error, but the more the MAP parameters that need to be estimated increase. The less the MAP input is divided, the less the MAP parameters to be estimated, and the more the MAP parameter estimation is affected by the approximation error of the piecewise interpolation model.
To analyze the impact of piecewise interpolation model approximation errors on volumetric efficiency MAP estimation, a mathematical description of the approximation errors is given below. At this time, the regression model of the volumetric efficiency MAP is described as follows:
Figure RE-GDA0002593745570000084
wherein r isη＝ηv(υ)-ηv,TAnd (theta, upsilon) is the approximate error of the MAP regression model.
According to the theory of interpolation, when
Figure RE-GDA0002593745570000085
Time, error rηComprises the following steps:
Figure RE-GDA0002593745570000086
when in use
Figure RE-GDA0002593745570000087
Time, error rηComprises the following steps:
Figure RE-GDA0002593745570000088
when in use
Figure RE-GDA0002593745570000089
Time, error rηComprises the following steps:
Figure RE-GDA0002593745570000091
by combining the formulae (1-15), (1-16) and (1-17), the approximation error r can be obtainedηThe boundaries of (A) are:
Figure RE-GDA0002593745570000092
wherein
Figure RE-GDA0002593745570000093
Figure RE-GDA0002593745570000094
Figure RE-GDA0002593745570000095
Figure RE-GDA0002593745570000096
Using intake manifold pressure pimCombining an engine intake system model (1-1) formula and a MAP regression model (1-14) formula of volumetric efficiency as system measurement output, and finally obtaining a mechanism/MAP mixed intake system model of the engine, wherein the mechanism/MAP mixed intake system model is as follows:
Figure RE-GDA0002593745570000097
wherein
Figure RE-GDA0002593745570000098
Where system state x is intake manifold pressure pimThe non-linear term g (x, n)e,uth) Mass flow W through the throttlethThe model of (d) is obtained in the form of a regression vector of Φ (v), and (p) is expressed as vim,ne) Is the measurable MAP input and R is the system model error caused by the approximation error R.
The system (1-19) shows that the volumetric efficiency is ηv(pim,ne) The calibration problem of (1) to (19) is converted into a joint estimation problem of the state x and the unknown MAP parameter theta of the formula (1).
Adaptive observer design
Unknown MAP parameters theta in the mechanism/MAP mixed model described by the formula (1-19) appear in the state equation, and the error term R exists in the formula (1-19), so that the existing convergence analysis method of the adaptive observer is not applicable any more. For this purpose, the design method of the observer is discussed in terms of the system (1-19). The adaptive observer is provided with the following form:
Figure RE-GDA0002593745570000101
wherein the content of the first and second substances,
Figure RE-GDA0002593745570000102
gain of
Figure RE-GDA0002593745570000103
Is a positive-determined diagonal matrix of the matrix,
Figure RE-GDA0002593745570000104
in order to estimate the state of the system,
Figure RE-GDA0002593745570000105
in order to be able to estimate the parameters,
Figure RE-GDA0002593745570000106
is a feedback gain matrix. The stability of the adaptive observer described by (1-21) is given by the following theorem.
Theorem 1.1 if the system inputs uthAnd a reference vector v such that the pair is arbitrary
Figure RE-GDA0002593745570000107
Is at an initial value
Figure RE-GDA0002593745570000108
The matrix gamma is continuously excited. i.e. in a different order,
Figure RE-GDA0002593745570000109
Figure RE-GDA00025937455700001010
the observer (1-21) equation is then exponentially stable and stable for any initial condition
Figure RE-GDA00025937455700001011
And an arbitrary bounded constant vector
Figure RE-GDA00025937455700001012
When t → ∞ is reached, error
Figure RE-GDA00025937455700001013
And
Figure RE-GDA00025937455700001014
converge respectively to the following tight sets:
Figure RE-GDA00025937455700001015
Figure RE-GDA00025937455700001016
wherein N ismaxIs neL > 0, k, λ > 0.
And (3) proving that: will be wrong
Figure RE-GDA00025937455700001017
Is of substituted type and is provided with
Figure RE-GDA00025937455700001018
After finishing, obtaining:
Figure RE-GDA0002593745570000111
due to the formula (1-25) containing
Figure RE-GDA0002593745570000112
Item, make its stability analysis difficult. To this end, a state estimation error is defined
Figure RE-GDA0002593745570000113
And parameter estimation error
Figure RE-GDA0002593745570000114
Is linearly combined as follows[148]：
Figure RE-GDA0002593745570000115
The two sides of the formula (1-26) are differentiated, and according to the formula (1-25), the derivatives are obtained
Figure RE-GDA0002593745570000116
The solution of equations (1-27) is
Figure RE-GDA0002593745570000117
Can obtain
Figure RE-GDA0002593745570000118
Thus, η is exponential stable and bounded.
As can be seen from the equations (1-27), the autonomous system
Figure RE-GDA0002593745570000119
The index is stable. Thus, it is possible to provide
Figure RE-GDA00025937455700001110
So that
Figure RE-GDA00025937455700001111
Obtained from the formula (1-28)
Figure RE-GDA00025937455700001112
Combining the formulas (1-29) and (1-30), simultaneously noticing the inequality and the inequality that | | phi (upsilon) | | is less than or equal to 1
Figure RE-GDA00025937455700001113
When t → ∞ there are
Figure RE-GDA00025937455700001114
Thus errors
Figure RE-GDA00025937455700001115
The exponent converges to the tight set
Figure RE-GDA00025937455700001116
From (1-2)6) When t → ∞ is expressed, there are
Figure RE-GDA00025937455700001117
Thus errors
Figure RE-GDA00025937455700001118
The exponent converges to the tight set
Figure RE-GDA0002593745570000121
After the syndrome is confirmed.
Note 1.1 tightly assembling omega according to convergence1,Ω2It can be seen that the margin | r of the MAP approximation errorηThe smaller | is, and the larger the feedback gain L is, the estimation error is
Figure RE-GDA0002593745570000122
And
Figure RE-GDA0002593745570000123
the smaller, the parameter estimation
Figure RE-GDA0002593745570000124
The closer to the true value theta. As can be seen from the equations (1-18), by reducing the maximum partition interval h, the margin of model error | r can be reducedηL, improving parameter estimation
Figure RE-GDA0002593745570000125
The accuracy of the estimation of.
Note 1.2 according to membership function
Figure RE-GDA0002593745570000126
As can be seen from the definitions (1-10) and (1-11) of (a), the regression vector ΦηAnd (v) is a sparse vector. Taking gamma (t)0) 0, obtained from the observer (1-21) formula:
Figure RE-GDA0002593745570000127
thus, the vector γ has the same sparsity as Φ (ν). To support by analyzing sparse vector gammaContinuous excitation condition (1-22) expression, judgment parameter estimation
Figure RE-GDA0002593745570000128
First, the sparsity of Φ (ν) is analyzed.
From the MAP input upsilon (p)im,ne) The division (1-3) and regression models (1-14) show that the MAP input upsilon can only fall into one input partition at any time
Figure RE-GDA0002593745570000129
And only the parameters corresponding to the partitions
Figure RE-GDA00025937455700001210
And participating in interpolation calculation. That is, for
Figure RE-GDA00025937455700001211
(1) When (k, l) ∈ {0, p1}×{0,p2At time of the design, corresponding to 1 parameter estimation participating in interpolation calculation
Figure RE-GDA00025937455700001212
Namely, it is
Figure RE-GDA00025937455700001213
(2) When (k, l) ∈ {1,2, … p1-1}×{0,p2At time of the design, there are 2 parameter estimates participating in interpolation calculation
Figure RE-GDA00025937455700001214
(i,j)∈{1,2,…p1-1}×{1,p2I.e. that
Figure RE-GDA00025937455700001215
(3) When (k, l) ∈ {0, p1}×{1,2,…p2-1} for 2 parameter estimates participating in the interpolation
Figure RE-GDA00025937455700001216
(i,j)∈{1,p1}×{1,2,…p2-1}, i.e.
Figure RE-GDA00025937455700001217
(4) When (k, l) ∈ {1,2, … p1-1}×{1,2,…p2-1} for 4 parameter estimates participating in the interpolation
Figure RE-GDA0002593745570000131
(i,j)∈{1,2,…p1-1}×{1,2,…p2-1}, i.e.
Figure RE-GDA0002593745570000132
MAP parameter estimation to facilitate analysis of participating interpolation computations
Figure RE-GDA0002593745570000133
According to the above-mentioned divided regions
Figure RE-GDA0002593745570000134
Defining a local regression vector phiη,lAnd (upsilon) is:
Figure RE-GDA0002593745570000135
wherein
Figure RE-GDA0002593745570000136
i＝1,2,…p1-1；j＝1,2,…p2-1
When in use
Figure RE-GDA0002593745570000137
In the formula of the adaptive observer (1-21), phi (upsilon) can be replaced by phil(υ)＝-(Vd/120Vim)pimne·Φη,l(upsilon) while noting that γ has the same sparsity as Φ (upsilon), we get
Figure RE-GDA0002593745570000138
Wherein the vector γlAnd philThe dimension of (v) is the same,
Figure RE-GDA0002593745570000139
a vector is estimated for the local parameters of the appropriate dimension,la positive diagonal matrix of appropriate dimensions.
As known from theorem 1.1, as long as γlSatisfy the continuous excitation condition (1-22), then corresponding parameter estimation
Figure RE-GDA00025937455700001310
Is convergent. Further, as long as the trace of MAP input upsilon passes through all the partitions
Figure RE-GDA00025937455700001311
While ensuring gammalIf the continuous excitation condition is satisfied, then all MAP parameter estimates are made
Figure RE-GDA00025937455700001312
Convergence is obtained. Therefore, to estimate the volumetric efficiency MAP, the engine operating conditions need to be designed so that the trajectory of upsilon passes through all of the partitions as much as possible
Figure RE-GDA0002593745570000141
Note 1.3 for the region S which is not passed by the trace of the MAP input upsilon, the corresponding parameters of the partition
Figure RE-GDA0002593745570000142
Cannot be obtained by the formula of observer (1-21). In this chapter, the parameters of the S region are obtained by using a second-order bivariate polynomial extrapolation model as follows:
Figure RE-GDA0002593745570000143
wherein, a2,a1,b2,b1,c2,c1Is a polynomial parameter. And fitting the parameters of the polynomial (1-34) by taking the estimated MAP parameters as data, and further calculating the corresponding MAP parameters in the S region according to the polynomial.
Simulation results and analysis
In order to simulate and verify the effectiveness of the method in this chapter, engine simulation software enDYNA is taken as a simulation environment, and a 2.0L four-cylinder SI gasoline engine is taken as a simulation object[171,172]The specific engine parameters include 2L of engine displacement, 4L of intake manifold volume, 1.5L of exhaust manifold volume and 7500rpm of maximum engine speed. The simulation structure is shown in fig. 2.
Dynamometer mode conditions
The dynamometer mode simulates engine bench test results. Intake manifold pressure p in dynamometer modeimCan be opened by a throttle valve opening uthIs regulated, the engine speed neThe adjustment may be performed by a dynamometer. As can be seen from the annotation 1.2, in order to ensure that all MAP parameters are estimated, the MAP input upsilon (p)im,ne) Should cover as much as possible (p) of the data traceim,ne) And (4) a plane. Therefore, in the dynamometer mode, the throttle opening u is designedthDesigning the engine speed n for the periodic signal with 6s as the periodeIs a linear signal, as shown in fig. 3. At this time, the intake manifold pressure p of the enDYNA modelimAnd engine speed neAs shown in fig. 4), MAP input upsilon ═ p (p) accordinglyim,ne) The running locus of (2) is shown in fig. 5). From fig. 5), the intake manifold pressure pimRange of [0,100000]Engine speed neRange of [0,7500]. The average division mode is adopted to divide the materials into [0:10000:100000 ]],[0:250:7500]. Wherein MAP input upsilon ═ (p)im,ne) The S area where the trajectory of (1) does not enter is:
S＝([0,4×104]×[0,500])∪([0,0.8×104]×[500,1500])
according to the formula and theorem 1.1 of the observer (1-21), the feedback gain is L-128 and 10-7I, initial value is
Figure RE-GDA0002593745570000151
Using the 300s data as shown in fig. 3), the volumetric efficiency MAP estimated by the method of this section is shown in fig. 6). As can be seen from the annotation 1.3, since the MAP input upsilon (p)im,ne) The S region is not entered, so that the MAP parameter corresponding to the S region is not estimated and remains as the initial value 0. In order to obtain MAP parameters corresponding to the S region by extrapolation, the MAP parameters estimated in fig. 6) are used as data, and the parameters of the fitting polynomial (1-34) are:
Figure RE-GDA0002593745570000152
calculating the corresponding MAP parameter in the S region according to the polynomial (1-34) is shown in fig. 7), and it can be seen that the corresponding MAP parameter in the S region obtained by extrapolation can approximately reflect the variation trend of the volumetric efficiency MAP.
To verify the effectiveness of the estimated volumetric efficiency MAP, a simulated comparison of the volumetric efficiency MAP in fig. 6) with the volumetric efficiency of the simulation system enDYNA under the same dynamometer mode conditions described above is shown in fig. 8). At this time, the cylinder intake air amount model W based on the volumetric efficiency MAPei,modelIntake air quantity W of cylinder with enDYNAei,actualAs shown in fig. 9). Wherein the average relative error between the volumetric efficiency of the enDYNA and the estimated volumetric efficiency MAP is 0.36%. As can be seen, the estimated MAP can better approximate the true value of the volumetric efficiency and the air intake quantity model W of the cylinderei,modelCan well approximate the true value Wei,actual. At approximately 300s, especially at engine speeds below 800rpm, the accuracy of the MAP estimate is degraded because the mean model cannot describe the reciprocating motion of the engine pistons. However, the cylinder intake air amount model W based on the volumetric efficiency MAPei,modelThe accuracy of (c) is still acceptable.
The invention provides a method for obtaining initial MAP of volumetric efficiency by estimation aiming at the problems of long calibration time and high cost faced by volumetric efficiency calibration. And (3) describing a nonlinear function by using MAP approximation to obtain a regression model of the volumetric efficiency MAP and an error expression method thereof. By establishing an engine air intake system model, the volumetric efficiency MAP estimation problem is converted into a joint estimation problem of system states and unknown MAP parameters. On the basis, according to the characteristic that unknown MAP parameters appear in a state equation, a self-adaptive observer of MAP estimation is designed, and the stable condition of the observer and the convergence compact set of MAP parameter estimation are given. The simulation result of the four-cylinder SI gasoline engine in the high-precision engine model enDYNA as a simulation object shows that the volumetric efficiency MAP estimated by the method in this chapter can better approach the actual value of the volumetric efficiency.
It will be appreciated that although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (4)

1. A calibration method for volumetric efficiency map of an automobile engine is characterized by comprising the following steps:
s1: designing mesh division of the map intake manifold pressure and the engine speed according to the trajectory range of the volumetric efficiency map intake manifold pressure and the engine speed;
s2: dividing the actual pressure and the actual rotating speed of the intake manifold of the engine into areas through a grid in the step S1, and acquiring transient working condition data of the engine;
s3: and establishing a volumetric efficiency map self-adaptive identification algorithm, and training and estimating volumetric efficiency map parameters through the algorithm.
2. The method for calibrating volumetric efficiency map of an automotive engine as recited in claim 1 wherein in said step S2, regulation of intake manifold pressure is controlled by throttle opening.
3. The calibration method for volumetric efficiency map of automotive engine as recited in claim 1, characterized in that the engine is connected with a dynamometer, and the regulation of the engine speed is realized by the dynamometer.
4. The method for calibrating volumetric efficiency map of an automobile engine as recited in claim 1, wherein the step S3 of establishing the adaptive identification algorithm comprises the steps of:
s31: establishing an engine air inlet system model:
Figure FDA0002512418380000011
wherein
Figure FDA0002512418380000021
Figure FDA0002512418380000022
Figure FDA0002512418380000023
Figure FDA0002512418380000024
Figure FDA0002512418380000025
In the formula
WthFor throttle mass flow, WeiThe amount of intake air is the amount of the cylinder,
Timis the intake manifold temperature, neIs the rotational speed of the engine and,
pimis intake manifold pressure, uthIs the opening degree of the throttle valve,
d is the diameter of the throttle valve, pi is the pressure ratio,
patm,Tatmis the external atmospheric pressure and temperature, gamma is the adiabatic coefficient,
Rais a gas constant, CdIs the flow coefficient, c0,c1,c2,b0,b1In order to be the parameters of the model,
ηvfor volumetric efficiency, the actual cylinder intake air quantity W is characterizedei,actualAnd the ideal air input Wei,idealI.e. ηv＝Wei,actual/Wei,ideal；
S32 volumetric efficiency η is given across the division of the volumetric efficiency two-dimensional MAP inputvMathematical description of MAP of (1):
Figure FDA0002512418380000026
s33: writing the map mathematical description into the expression form of regression vector, and calculating the approximate error of the volumetric efficiency to obtain the regression model description of the volumetric efficiency map
Figure FDA0002512418380000031
Wherein r isη＝ηv(υ)-ηv,T(θ, upsilon) is the approximation error of the MAP regression model;
s34: using intake manifold pressure pimThe MAP mixed intake system model of the engine obtained by combining the engine intake system model and the MAP regression model of the volumetric efficiency as the system measurement output is as follows:
Figure FDA0002512418380000032
wherein the system state x is the intake manifold pressure pimThe non-linear term g (x, n)e,uth) Mass flow W through the throttlethThe model of (d) is obtained in the form of a regression vector of Φ (v), and (p) is expressed as vim,ne) Is made byMeasuring MAP input, wherein R is a system model error caused by an approximation error R;
s35: aiming at a map mixed air inlet system model of an engine, an adaptive observer is set as follows:
Figure FDA0002512418380000033
wherein the content of the first and second substances,
Figure FDA0002512418380000034
gain of
Figure FDA0002512418380000035
Is a positive-determined diagonal matrix of the matrix,
Figure FDA0002512418380000036
in order to estimate the state of the system,
Figure FDA0002512418380000037
in order to be able to estimate the parameters,
Figure FDA0002512418380000038
is a feedback gain matrix;
according to membership functions
Figure FDA0002512418380000039
By simultaneously noting that γ and Φ (v) have the same sparsity, we obtained
Figure FDA00025124183800000310
Wherein, vector γlAnd philThe dimension of (v) is the same,
Figure FDA00025124183800000311
a vector is estimated for the local parameters of the appropriate dimension,la positive diagonal matrix of appropriate dimensions.
CN202010465311.0A 2020-05-28 2020-05-28 Method for calibrating volumetric efficiency map of automobile engine Pending CN111648872A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010465311.0A CN111648872A (en) 2020-05-28 2020-05-28 Method for calibrating volumetric efficiency map of automobile engine

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010465311.0A CN111648872A (en) 2020-05-28 2020-05-28 Method for calibrating volumetric efficiency map of automobile engine

Publications (1)

Publication Number Publication Date
CN111648872A true CN111648872A (en) 2020-09-11

Family

ID=72352679

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010465311.0A Pending CN111648872A (en) 2020-05-28 2020-05-28 Method for calibrating volumetric efficiency map of automobile engine

Country Status (1)

Country Link
CN (1) CN111648872A (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102135045A (en) * 2010-01-26 2011-07-27 通用汽车环球科技运作有限责任公司 Adaptive intake oxygen estimation in a diesel engine
US20150128904A1 (en) * 2013-11-11 2015-05-14 Songping Yu Techniques for coordinated variable valve timing and electronic throttle control

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102135045A (en) * 2010-01-26 2011-07-27 通用汽车环球科技运作有限责任公司 Adaptive intake oxygen estimation in a diesel engine
US20150128904A1 (en) * 2013-11-11 2015-05-14 Songping Yu Techniques for coordinated variable valve timing and electronic throttle control

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
梁美等: "基于查表自适应的汽油机气缸进气量估计方法", 《烟台大学学报（自然科学与工程版）》 *

Similar Documents

Publication Publication Date Title
US11506138B2 (en) Engine system with inferential sensor
Desantes et al. Air mass flow estimation in turbocharged diesel engines from in-cylinder pressure measurement
CN111612254B (en) Road motor vehicle exhaust emission prediction method based on improved attention bidirectional long-short term memory network
DE102013202720A1 (en) Estimation device for a cylinder intake air quantity in an internal combustion engine
Stotsky et al. Simple unknown input estimation techniques for automotive applications
Alcan et al. Estimating soot emission in diesel engines using gated recurrent unit networks
Guardiola et al. Cylinder charge composition observation based on in-cylinder pressure measurement
CN106121847B (en) Method for identifying average value model of internal combustion engine
Alcan et al. Diesel engine NOx emission modeling using a new experiment design and reduced set of regressors
CN111648872A (en) Method for calibrating volumetric efficiency map of automobile engine
Orumbayeva et al. On the solvability of the duo-periodic problem for the hyperbolic equation system with a mixed derivative
Ho et al. Hydrogen powered car: Two-stage modelling system
US10067033B2 (en) Systems and methods for in-cylinder pressure estimation using pressure wave modeling
CN114934848B (en) Fuzzy neural network modeling method for optimizing control of combustion performance of diesel engine
Guardiola et al. Representation limits of mean value engine models
EP3368760B1 (en) Systems and methods for detecting anomalies at pressure sensors
Wojnar et al. Nonlinear air-fuel ratio predictive control of spark ignited engines
CN112800588B (en) Simulation calculation method for air intake of cabin heat exchanger under transient driving working condition
Flärdh et al. Nonlinear exhaust pressure control of an SI engine with VGT using partial model inversion
Wahlström et al. Modeling of a diesel engine with VGT and EGR capturing sign reversal and non-minimum phase behaviors
Yang et al. Modelling the exhaust gas recirculation mass flow rate in modern diesel engines
Wahlström et al. Modeling of a diesel engine with VGT and EGR including oxygen mass fraction
Beňo et al. Robust numerical approach to steady-state calibration of mean-value models
Engl The modeling and numerical simulation of gas flow networks
Smallbone et al. Virtual performance and emissions mapping for diesel engine design optimization

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20200911