CN110185791B - Control parameter optimization method for stepped automatic transmission - Google Patents

Control parameter optimization method for stepped automatic transmission Download PDF

Info

Publication number
CN110185791B
CN110185791B CN201910460321.2A CN201910460321A CN110185791B CN 110185791 B CN110185791 B CN 110185791B CN 201910460321 A CN201910460321 A CN 201910460321A CN 110185791 B CN110185791 B CN 110185791B
Authority
CN
China
Prior art keywords
optimized
parameters
parameter
target response
sample points
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.)
Active
Application number
CN201910460321.2A
Other languages
Chinese (zh)
Other versions
CN110185791A (en
Inventor
付尧
雷雨龙
姜赟涛
李兴忠
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.)
Jilin University
Original Assignee
Jilin 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 Jilin University filed Critical Jilin University
Priority to CN201910460321.2A priority Critical patent/CN110185791B/en
Publication of CN110185791A publication Critical patent/CN110185791A/en
Application granted granted Critical
Publication of CN110185791B publication Critical patent/CN110185791B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F16—ENGINEERING ELEMENTS AND UNITS; GENERAL MEASURES FOR PRODUCING AND MAINTAINING EFFECTIVE FUNCTIONING OF MACHINES OR INSTALLATIONS; THERMAL INSULATION IN GENERAL
    • F16H—GEARING
    • F16H61/00—Control functions within control units of change-speed- or reversing-gearings for conveying rotary motion ; Control of exclusively fluid gearing, friction gearing, gearings with endless flexible members or other particular types of gearing
    • F16H61/04—Smoothing ratio shift
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F16—ENGINEERING ELEMENTS AND UNITS; GENERAL MEASURES FOR PRODUCING AND MAINTAINING EFFECTIVE FUNCTIONING OF MACHINES OR INSTALLATIONS; THERMAL INSULATION IN GENERAL
    • F16H—GEARING
    • F16H61/00—Control functions within control units of change-speed- or reversing-gearings for conveying rotary motion ; Control of exclusively fluid gearing, friction gearing, gearings with endless flexible members or other particular types of gearing
    • F16H2061/0075—Control functions within control units of change-speed- or reversing-gearings for conveying rotary motion ; Control of exclusively fluid gearing, friction gearing, gearings with endless flexible members or other particular types of gearing characterised by a particular control method
    • F16H2061/009—Control functions within control units of change-speed- or reversing-gearings for conveying rotary motion ; Control of exclusively fluid gearing, friction gearing, gearings with endless flexible members or other particular types of gearing characterised by a particular control method using formulas or mathematic relations for calculating parameters
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Economics (AREA)
  • Computer Hardware Design (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Mechanical Engineering (AREA)
  • Automation & Control Theory (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Control Of Transmission Device (AREA)

Abstract

The invention discloses a control parameter optimization method for a stepped automatic transmission, which comprises the following steps: step one, selecting parameters to be optimized and target response by taking gear shifting comfort as an optimization target; wherein the parameters to be optimized comprise: the method comprises the following steps of (1) vehicle speed, vehicle acceleration, engine rotating speed, accelerator opening, throttle opening change rate, vehicle quality and main oil pressure of a hydraulic control system; the targets are correspondingly: weighting an acceleration root mean square value, a maximum impact degree and a maximum engine speed variation; determining the value interval of the parameter to be optimized and the value interval of the target response; selecting sample points of the parameters to be optimized in the value interval of the parameters to be optimized to obtain target response values corresponding to the sample points; step four, establishing a parameter optimization model according to the selected parameter sample points to be optimized and the target response values corresponding to the sample points; and fifthly, performing parameter optimization according to the parameter optimization model. The invention can improve the effect and efficiency of parameter optimization.

Description

Control parameter optimization method for stepped automatic transmission
Technical Field
The invention belongs to the technical field of control of a stepped automatic transmission, and particularly relates to a control parameter optimization method of the stepped automatic transmission.
Background
The stepped automatic transmission is developed on the basis of a manual transmission, and an electric control system of the stepped automatic transmission can automatically select a proper gear according to information such as vehicle speed, engine speed, accelerator opening and the like when a vehicle runs so as to ensure the driving power, fuel economy and riding comfort of the vehicle.
When a control strategy of the stepped automatic transmission electric control system is developed, parameter optimization is needed, namely parameters in the control strategy are adjusted according to a required control target so as to achieve the purpose of optimizing control quality. The selection of the control parameters directly determines the characteristics of the automatic transmission and has great influence on the driving performance of the whole vehicle. However, the optimization of the control parameters at present basically depends on the experience adjustment of engineers and continuous experiment, and the method has low efficiency and is difficult to realize global optimization. In addition, in the parameter optimization process of the stepped automatic transmission, the whole vehicle dynamic performance or economic performance index is usually used as a response for optimization, related calibration methods are researched more, but the control parameter optimization method with the shift comfort degree as the target is researched less.
Disclosure of Invention
The invention provides a control parameter optimization method of a stepped automatic transmission, wherein gear shifting comfort level selects parameters to be optimized and target response for optimization targets.
The invention provides a method for optimizing control parameters of a stepped automatic transmission, which is characterized in that after a parameter optimization proxy model is established, the credibility of the parameter optimization proxy model is verified, and when the credibility does not meet the requirement, an optimization parameter sample is reselected.
The technical scheme provided by the invention is as follows:
a method of optimizing control parameters for a stepped automatic transmission, comprising:
step one, selecting parameters to be optimized and target response;
wherein the parameters to be optimized comprise: the method comprises the following steps of (1) vehicle speed, vehicle acceleration, engine rotating speed, accelerator opening, throttle opening change rate, vehicle quality and main oil pressure of a hydraulic control system; and
the target response is: weighting an acceleration root mean square value, a maximum impact degree and a maximum engine speed variation;
secondly, determining the value interval of the parameter to be optimized and the value interval of the target response;
selecting sample points of the parameters to be optimized in the value interval of the parameters to be optimized to obtain target response values corresponding to the sample points;
step four, establishing a parameter optimization model according to the selected parameter sample points to be optimized and the target response values corresponding to the sample points;
and fifthly, performing parameter optimization according to the parameter optimization model.
Preferably, in the second step, after the value interval of the parameter to be optimized is determined, the value interval of the target response is determined according to the following formula:
Figure GDA0002325526290000021
Figure GDA0002325526290000022
Δωemax＝[max(ωe)-min(ωe)]T；
in the formula, awAs weighted acceleration root mean square value, jmaxAt maximum impact, Δ ωemaxIs the maximum engine speed variation; t is the shift process time, a is the vehicle acceleration, ωeIs the engine speed and t is the time.
Preferably, in the third step, the method for selecting the parameter sample points to be optimized comprises:
step 1, optimizing a parameter x to be optimizediThe value interval is divided into m sub-intervals with equal probability, and each sub-interval is marked as
Figure GDA0002325526290000023
k∈[1,m]；
Step 2, random sampling is respectively carried out in each subinterval, and a random number lambda belongs to [0,1], so that the random sampling point is as follows:
Figure GDA0002325526290000031
wherein m is the number of samples extracted, and k is the number of subintervals.
Preferably, in the fourth step, the selected parameter sample point to be optimized is used as an input, the target response corresponding to the parameter sample point to be optimized is used as an output, and an RBF neural network model is established as the parameter optimization model.
Preferably, before the fourth step, the method further includes reducing the number of the parameters to be optimized according to the selected correlation coefficient of the parameters to be optimized;
wherein, the correlation coefficient of the parameter to be optimized is:
Figure GDA0002325526290000032
wherein the content of the first and second substances,
Figure GDA0002325526290000033
x and y respectively represent different parameters to be optimized;
when r isxy>When 0.8, only one parameter to be optimized in x or y is reserved;
when r isxy<And at-0.8, keeping the opposite numbers of x and x as parameters to be optimized, or keeping the opposite numbers of y and y as parameters to be optimized.
Preferably, before the fourth step, the number of the parameters to be optimized is reduced according to the correlation coefficient between the parameters to be optimized and the target response values corresponding to the parameters to be optimized;
wherein, the correlation coefficient of the parameter to be optimized and the target response corresponding to the parameter to be optimized is:
Figure GDA0002325526290000034
wherein the content of the first and second substances,
Figure GDA0002325526290000035
x represents a parameter to be optimized, and z represents a target response corresponding to the parameter to be optimized;
when r isxzAnd when the epsilon is (-0.1,0.1), deleting the corresponding parameter to be optimized.
Preferably, in the fourth step, the correlation degree between the predicted output value and the actual output value of the parameter optimization model is verified, and when the correlation degree is less than 0.9, the third step is repeated;
the calculation method of the correlation degree comprises the following steps:
Figure GDA0002325526290000041
wherein z isiIn order to be the actual output value,
Figure GDA0002325526290000048
in order to actually output the average value,
Figure GDA0002325526290000042
the output values are predicted for the parametric optimization model.
Preferably, in the fifth step, the optimization of the control parameter to be optimized is performed with a goal of optimal comfort, and the method includes the following steps:
step 1, adopting a binary coding mode to carry out comparison on vehicle speed v, vehicle acceleration a and engine rotating speed omegaeAccelerator opening α, and throttle opening change rate
Figure GDA0002325526290000043
The vehicle mass m and the hydraulic control system main oil pressure p are encoded;
step 2, randomly generating a first generation population, wherein each individual in the first generation population comprises v, a and omegae、α、
Figure GDA0002325526290000044
m and p; comfort level
Figure GDA0002325526290000045
Step 3, calculating the corresponding fitness of the individuals in the first generation population; wherein the fitness is as follows:
Figure GDA0002325526290000046
4, sequencing the individuals according to the fitness, selecting the individuals with high fitness, and performing cross and variation to generate a second-generation population;
step 5, calculating the fitness corresponding to the individuals in the second-generation population, and circularly performing the step 4-5 until the set iteration times are reached; selecting an individual corresponding to the maximum fitness as an optimal solution;
and 6, decoding the optimal solution to obtain optimal v, a and omegae、α、
Figure GDA0002325526290000047
And the calibrated values of m and p.
Preferably, in the fifth step, different target response values corresponding to the parameters to be optimized are obtained according to the parameter optimization model, and the parameters to be optimized are optimized and adjusted according to the target response values.
The invention has the beneficial effects that:
according to the control parameter optimization method for the stepped automatic transmission, the gear shifting comfort level is taken as an optimization target, the parameter to be optimized and the target response are selected, the parameter optimization model is established according to the parameter to be optimized and the target response, the parameter optimization is carried out according to the parameter optimization proxy model, and the parameter optimization efficiency can be improved.
According to the control parameter optimization method of the stepped automatic transmission, the reliability of the parameter optimization proxy model is verified after the parameter optimization proxy model is established, and when the reliability does not meet the requirement, the optimization parameter sample is selected again until the reliability of the established parameter optimization proxy model meets the requirement; the invention can improve the effectiveness of parameter sample selection, thereby improving the optimization effect.
Drawings
Fig. 1 is a schematic diagram illustrating an influence of accelerator opening variation on a weighted acceleration root mean square value obtained through a dominant effect analysis in embodiment 1 of the present invention.
Fig. 2 is a schematic diagram illustrating an influence of a vehicle speed change on a weighted acceleration root mean square value obtained through a dominant effect analysis in embodiment 1 of the present invention.
Fig. 3 is a schematic diagram of an influence of the accelerator opening change on the maximum impact degree obtained through the main effect analysis in embodiment 1 of the present invention.
Fig. 4 is a schematic diagram illustrating an influence of a vehicle speed change on a maximum impact degree obtained through a main effect analysis in embodiment 1 of the present invention.
Fig. 5 is a schematic diagram illustrating an influence of a change in accelerator opening on a maximum engine speed change amount obtained through a main effect analysis in embodiment 1 of the present invention.
Fig. 6 is a schematic diagram illustrating an influence of a vehicle speed change on a maximum engine speed change amount obtained through a main effect analysis in embodiment 1 of the present invention.
Detailed Description
The present invention is further described in detail below with reference to the attached drawings so that those skilled in the art can implement the invention by referring to the description text.
As shown in fig. 1, the present invention provides a method for optimizing control parameters of a stepped automatic transmission, comprising the steps of:
step one, determining control parameters to be optimized and target response.
The stepped automatic transmission comprises an electric control mechanical automatic transmission, a hydraulic automatic transmission and a double-clutch automatic transmission, control parameters in control strategies of electric control systems of various automatic transmissions are different, and each transmission has unique control parameters such as control valve oil pressure, clutch master-slave disc rotation speed difference and the like in a gear shifting process. Different automatic transmissions need to select their respective control parameters to be optimized according to their respective constructional features and actuators.
The method takes the gear-shifting comfort degree as an optimization target, and the selected control parameters to be optimized comprise vehicle speed v (km/h), vehicle acceleration a (m/s2), engine speed omega (rpm), accelerator opening α (%), and accelerator opening change rate
Figure GDA0002325526290000061
Vehicle mass m (kg) (when the vehicle load changes between no load and full load, the value directly influences the inertia of the whole vehicle, thereby changing the gear shifting comfort) and the main oil pressure p (kpa) of the hydraulic control system. The parameters can directly influence the gear shifting process in the gear shifting decision and gear shifting execution process of the stepped automatic transmission under various working conditionsThereby affecting comfort.
The invention selects a weighted acceleration root mean square value awMaximum impact jmaxAnd maximum engine speed variation Δ ωemaxAs a targeted response. The shift comfort is targeted, and the above responses are all responses in the vehicle traveling direction, i.e., longitudinal responses. Under the same driving condition, the weighted acceleration root mean square value awMaximum impact jmaxAnd maximum engine speed variation Δ ωemaxThe smaller the shift comfort, the better. Wherein the expression of each target response is:
Figure GDA0002325526290000062
t is the shift schedule time, within which there is:
Figure GDA0002325526290000063
Δωemax＝[max(ωe)-min(ωe)]T(3)。
and step two, preliminarily determining a value interval and a response interval of the control parameter to be optimized.
In the process of developing and optimizing a control strategy of a stepped automatic transmission electric control system, in order to narrow the parameter value range and conveniently and efficiently optimize the parameters, a value interval and a response interval of the control parameters to be optimized are preliminarily determined. The value range of the control parameter to be optimized cannot be obtained through formula calculation, and the value range can be preliminarily determined according to the rated variation range of the corresponding actuating mechanism of each control parameter. And substituting the preliminarily determined value interval of the control parameter to be optimized into the expressions (1), (2) and (3) of the target response, so as to preliminarily determine the target response interval.
And step three, obtaining sample points according to a sampling method.
The parameters to be optimized of various stepped automatic transmissions are more, and each parameter has a value range, so that the experimental parameter dimension is high, the value space is large, and the number of sample points is large. In the embodiment, a sampling design is performed by adopting a latin hypercube, sample points which are uniformly distributed, full of an experiment value space and greatly reduced in number are obtained in a value interval, and the representativeness of the sample points is ensured while the test times are reduced.
For n control parameters to be optimized, m sample points need to be acquired, and for each control parameter x to be optimizediThe selection probability of each subinterval divided into m subintervals with equal probability in the value space is
Figure GDA0002325526290000071
Each subinterval is noted as
Figure GDA0002325526290000072
k∈[1,m](ii) a Independent sampling is respectively carried out in each subinterval, random selection is carried out during independent sampling, and a random number lambda belongs to [0,1]]Then the random sampling points are:
Figure GDA0002325526290000073
after the samples are generated, each parameter sample to be optimized is represented
Figure GDA0002325526290000074
And forming a Latin hypercube sampling design with the sample number of m according to random number arrangement, namely the m multiplied by n LHD.
And step four, obtaining target response according to the selected sample point.
And substituting the selected sample points into expressions (1), (2) and (3) of the substituted target response, and calculating to obtain the target response corresponding to each sample point.
And fifthly, performing correlation analysis and simplifying the quantity of the control parameters to be optimized.
When the parameters of the stepped automatic transmission electric control system are optimized, a plurality of control parameters to be optimized exist, wherein some parameters have small influence on target response or the similarity among some parameters is high, and the control parameters of the type do not need to be optimized and can be directly expressed by constants or similar control parameters.
In the present embodiment, the correlation coefficient r is usedxyAnalyzing and controlling the correlation among the parameters to be optimized:
Figure GDA0002325526290000075
wherein s isxyIs the covariance of sample x and sample y,
Figure GDA0002325526290000076
sxis the standard deviation of the sample x and,
Figure GDA0002325526290000077
coefficient of correlation rxyThe value interval of (a) is between 1 and-1, 1 represents a complete positive correlation, -1 represents a complete negative correlation, and 0 represents a complete non-correlation. Between the parameters to be optimized, the correlation coefficient rxyParameters greater than 0.8 (or less than-0.8) can be represented by their associated parameters (or the inverse of the associated parameters), reducing the number of parameters to be optimized.
In another embodiment, further comprising using the correlation coefficient rxzAnalyzing and controlling the correlation between the parameters to be optimized and the target responses corresponding to the parameters to be optimized:
Figure GDA0002325526290000081
wherein s isxyIs the covariance of the sample x and the target response z;
Figure GDA0002325526290000082
Figure GDA0002325526290000083
x represents a parameter to be optimized, and z represents a target response corresponding to the parameter to be optimized;
when the correlation coefficient r between the parameter to be optimized and the target isxzAnd when the element belongs to (-0.1,0.1), deleting the corresponding parameters to be optimized so as to reduce the number of the parameters to be optimized.
And step six, training to form an optimized agent model.
And constructing an RBF neural network model by taking the obtained sample points as input and the response corresponding to each sample point as output to form a parameter optimization proxy model. The model comprises an input layer, a single-layer hidden layer and an output layer, wherein the input is all parameters to be optimized, the number of neurons in the hidden layer is the same as the number of sample points, the output is 3 target responses, and the RBF neural network model is constructed based on a Gaussian kernel function, namely
Figure GDA0002325526290000084
Preferably, after the optimization agent model is established, verifying the credibility of the agent model is further included. The correlation degree between the predicted output and the actual output of the parameter optimization proxy model can represent the credibility of the calibration proxy model formed by training, and the calculation formula of the correlation degree is
Figure GDA0002325526290000085
Wherein z isiThe actual output value obtained by the experiment is,
Figure GDA0002325526290000086
as an actual output average value obtained by the experiment,
Figure GDA0002325526290000087
the output values are predicted for the optimization proxy model. Degree of correlation R2The closer to 1, the higher the confidence of the prediction output of the proxy model, if the degree of correlation R2If the correlation degree is less than 0.9, acquiring a new sample point and reestablishing the optimization proxy model until the correlation degree R2Until the requirements are met. And establishing an optimization proxy model, namely optimizing the parameters to be optimized by applying the optimization proxy model.
Example 1
In the embodiment, a single-factor main effect analysis method is adopted for analysis, namely, when a certain parameter to be optimized is changed, the influence of all combination conditions generated by the parameter to be optimized and other parameters on target response is analyzed, so that guidance is provided for an engineer to select, adjust and optimize the control parameters.
FIG. 1 and FIG. 2 are respectively the weighted acceleration RMS values a obtained by the analysis of the main effectwThe responses to the accelerator opening α and the vehicle speed v are shown in the figure, and it is understood that the weighted acceleration root mean square value a is calculated as the accelerator opening α is larger in the value range shown in the figurewThe smaller the acceleration, the greater the throttle opening α is taken in the direction of 8 at the calibration time, and the weighted acceleration root mean square value a can be madewReduced, i.e. higher comfort; in the value range of the figure, the vehicle speed v obtains a weighted acceleration root mean square value a at 4.2wThe local minimum value decreases in the direction larger than 8, the vehicle speed v is 4.2 and the direction larger than 8 can be taken at the time of calibration, and the weighted acceleration root mean square value a can be madewThe reduction, i.e. the comfort level becomes high.
FIG. 3 and FIG. 4 are the maximum impact j obtained from the main effect analysismaxIn response to the accelerator opening α and the vehicle speed v, the direction in which the accelerator opening α is greater than 20 and the direction in which the vehicle speed v is less than 10 or greater than 20 are taken at the time of calibration for the purpose of improving comfort.
FIG. 5 and FIG. 6 show the maximum engine speed variation Δ ω, respectively, obtained from the main effect analysisemaxIn response to the accelerator opening α and the vehicle speed v, the direction in which the accelerator opening α is greater than 20 and the direction in which the vehicle speed v is less than 10 are calibrated for comfort.
Example 2
And analyzing the parameters to be optimized by using the constructed parameter optimization agent model and using a genetic algorithm, so as to obtain a global minimum value, namely the maximum comfort level in the gear shifting process. The method comprises the following steps:
(1) firstly, according to experience, the size of a population is set to be 50, the maximum algebra, namely the optimization iteration number, is set to be 1000, the cross rate is set to be 1 so as to ensure that the full evolution of the population is approximate to the optimal solution, and the probability of variation caused by the optimization parameters is set to be 0.1, so that the probability of variation is low;
(2) n control parameters v, a to be optimized of each group of solutions、ωe、α、
Figure GDA0002325526290000091
After m and p are normalized, coding is carried out according to binary system to obtain codes of each group of solutions in a feasible domain, namely the chromosomes;
(3) randomly creating an initial population;
(4) setting an objective function, i.e. a function of comfort level with respect to each parameter to be optimized, with comfort level C, then
Figure GDA0002325526290000092
The comfort level is higher when the function value is larger, and a target function is max (C), namely the comfort level index is maximum;
(5) judging the fitness of the individuals, wherein the fitness refers to the deviation degree of the target response obtained by each individual and the optimal response finally obtained, and the larger the deviation degree is, the worse the response is; the smaller the deviation degree, the better the response, and the optimal solution is obtained without deviation.
Formula for fitness
Figure GDA0002325526290000101
And (4) characterization, namely, the larger the F is, the closer the F is to the optimal solution, and the maximum individual corresponding to the F is the optimal solution.
(6) Sorting according to the fitness of individuals, selecting the individuals with high fitness as parents, eliminating the individuals with low fitness, crossing the chromosomes of the parents according to a certain method to generate offspring and carrying out mutation on the offspring chromosomes.
(7) And generating a new generation of population by crossing and variation, and calculating the individual fitness of the new generation of population until the iteration times are reached, wherein the comfort level corresponding to the individual with the maximum fitness is max (C).
(8) And taking the individual with the maximum fitness as an optimal solution, and decoding and reverse normalizing the optimal solution to obtain an optimized parameter.
The algorithm is used for carrying out combined optimization on the parameters to be optimized, and the comprehensive optimal gear shifting control parameters under the optimal comfort level can be obtained.
While embodiments of the invention have been described above, it is not limited to the applications set forth in the description and the embodiments, which are fully applicable in various fields of endeavor to which the invention pertains, and further modifications may readily be made by those skilled in the art, it being understood that the invention is not limited to the details shown and described herein without departing from the general concept defined by the appended claims and their equivalents.

Claims (8)

1. A method of optimizing control parameters for a stepped automatic transmission, comprising the steps of:
step one, selecting parameters to be optimized and target response;
wherein the parameters to be optimized comprise: the method comprises the following steps of (1) vehicle speed, vehicle acceleration, engine rotating speed, accelerator opening, throttle opening change rate, vehicle quality and main oil pressure of a hydraulic control system; and
the target response is: weighting an acceleration root mean square value, a maximum impact degree and a maximum engine speed variation;
secondly, determining the value interval of the parameter to be optimized and the value interval of the target response;
selecting sample points of the parameters to be optimized in the value interval of the parameters to be optimized to obtain target response values corresponding to the sample points;
step four, establishing a parameter optimization model according to the selected parameter sample points to be optimized and the target response values corresponding to the sample points;
fifthly, parameter optimization is carried out according to the parameter optimization model;
before the fourth step, the number of the parameters to be optimized is reduced according to the selected correlation coefficient of the parameters to be optimized;
wherein, the correlation coefficient of the parameter to be optimized is:
Figure FDA0002359166010000011
wherein the content of the first and second substances,
Figure FDA0002359166010000012
x and y respectively represent different parameters to be optimized; r isxyFor the correlation coefficient, s, of the parameters x and y to be optimizedxyFor the covariance, s, of the parameters x and y to be optimizedxThe standard deviation of the parameter x to be optimized;
when r isxy>When 0.8, only one parameter to be optimized in x or y is reserved;
when r isxy<And at-0.8, keeping the opposite numbers of x and x as parameters to be optimized, or keeping the opposite numbers of y and y as parameters to be optimized.
2. The method according to claim 1, wherein in the second step, after the interval of the parameter to be optimized is determined, the interval of the target response is determined according to the following formula:
Figure FDA0002359166010000021
Figure FDA0002359166010000022
Δωemax＝[max(ωe)-min(ωe)]T；
in the formula, awAs weighted acceleration root mean square value, jmaxAt maximum impact, Δ ωemaxIs the maximum engine speed variation; t is the shift process time, a is the vehicle acceleration, ωeIs the engine speed and t is the time.
3. The method for optimizing control parameters of a stepped automatic transmission according to claim 2, wherein in the third step, the method for selecting the sample points of the parameters to be optimized is as follows:
step 1, optimizing a parameter x to be optimizediThe value interval is divided into m sub-intervals with equal probability, and each sub-interval is marked as
Figure FDA0002359166010000023
Step 2, random sampling is respectively carried out in each subinterval, and a random number lambda belongs to [0,1], so that the random sampling point is as follows:
Figure FDA0002359166010000024
wherein m is the number of samples extracted, and k is the number of subintervals.
4. The method as claimed in claim 3, wherein in the fourth step, the selected parameter sample points to be optimized are used as input, the target responses corresponding to the parameter sample points to be optimized are used as output, and an RBF neural network model is established as the parameter optimization model.
5. The method of optimizing control parameters for a stepped automatic transmission according to claim 4, wherein prior to said step four, further comprising reducing the number of parameters to be optimized in accordance with a correlation coefficient of the parameters to be optimized with respect to a target response value corresponding to the parameters to be optimized;
wherein, the correlation coefficient of the parameter to be optimized and the target response corresponding to the parameter to be optimized is:
Figure FDA0002359166010000025
wherein the content of the first and second substances,
Figure FDA0002359166010000026
x represents a parameter to be optimized, and z represents a target response corresponding to the parameter to be optimized; r isxzThe correlation coefficient of the parameter x to be optimized and the target response z corresponding to the parameter x to be optimized; sxzThe covariance of the parameter x to be optimized and the target response z corresponding to the parameter x to be optimized;
when r isxz∈(-0.1,0.1), deleting the corresponding parameters to be optimized.
6. The stepped automatic transmission control parameter optimization method according to claim 5, further comprising verifying a degree of correlation between a predicted output value and an actual output value of the parameter optimization model in the fourth step, and repeating the third step when the degree of correlation is less than 0.9;
the calculation method of the correlation degree comprises the following steps:
Figure FDA0002359166010000031
wherein z isiIn order to be the actual output value,
Figure FDA0002359166010000032
in order to actually output the average value,
Figure FDA0002359166010000033
predicting an output value, R, for a parametric optimization model2Is the degree of correlation.
7. The method of optimizing control parameters for a stepped automatic transmission according to claim 6, wherein in said step five, the parameters to be optimized are optimized with the objective of comfort optimization, comprising the steps of:
step 1, adopting a binary coding mode to carry out comparison on vehicle speed v, vehicle acceleration a and engine rotating speed omegaeAccelerator opening α, and throttle opening change rate
Figure FDA0002359166010000034
The vehicle mass m and the hydraulic control system main oil pressure p are encoded;
step 2, randomly generating a first generation population, wherein each individual in the first generation population comprises v, a and omegae、α、
Figure FDA0002359166010000035
m and p; comfort level
Figure FDA0002359166010000036
Step 3, calculating the corresponding fitness of the individuals in the first generation population; wherein the fitness is as follows:
Figure FDA0002359166010000037
4, sequencing the individuals according to the fitness, selecting the individuals with high fitness, and performing cross and variation to generate a second-generation population;
step 5, calculating the fitness corresponding to the individuals in the second-generation population, and circularly performing the step 4-5 until the set iteration times are reached; selecting an individual corresponding to the maximum fitness as an optimal solution;
and 6, decoding the optimal solution to obtain optimal v, a and omegae、α、
Figure FDA0002359166010000038
And the calibrated values of m and p.
8. The method according to claim 7, wherein in the fifth step, target response values corresponding to different parameters to be optimized are obtained according to the parameter optimization model, and the parameters to be optimized are optimized and adjusted according to the target response values.
CN201910460321.2A 2019-05-30 2019-05-30 Control parameter optimization method for stepped automatic transmission Active CN110185791B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910460321.2A CN110185791B (en) 2019-05-30 2019-05-30 Control parameter optimization method for stepped automatic transmission

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910460321.2A CN110185791B (en) 2019-05-30 2019-05-30 Control parameter optimization method for stepped automatic transmission

Publications (2)

Publication Number Publication Date
CN110185791A CN110185791A (en) 2019-08-30
CN110185791B true CN110185791B (en) 2020-03-20

Family

ID=67718746

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910460321.2A Active CN110185791B (en) 2019-05-30 2019-05-30 Control parameter optimization method for stepped automatic transmission

Country Status (1)

Country Link
CN (1) CN110185791B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114321360B (en) * 2022-01-04 2023-07-14 一汽解放汽车有限公司 Method and device for identifying gear of manual gear gearbox and computer equipment

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101118620A (en) * 2007-09-18 2008-02-06 吉林大学 Vehicle gear shifting quality evaluation method based on nerval net
CN106763724A (en) * 2017-01-19 2017-05-31 西华大学 The optimal Shift Strategy system of many performance synthesises of automobile geared automatic transmission
JP6686929B2 (en) * 2017-02-21 2020-04-22 トヨタ自動車株式会社 Vehicle control device

Also Published As

Publication number Publication date
CN110185791A (en) 2019-08-30

Similar Documents

Publication Publication Date Title
Eckert et al. Gear shifting multi-objective optimization to improve vehicle performance, fuel consumption, and engine emissions
CN106641227B (en) A kind of geared automatic transmission shift schedule optimization method
JP7136073B2 (en) VEHICLE CONTROL DATA GENERATION METHOD, VEHICLE CONTROL DEVICE, VEHICLE CONTROL SYSTEM, AND VEHICLE LEARNING DEVICE
Zhang et al. ELM-based driver torque demand prediction and real-time optimal energy management strategy for HEVs
EP0969385A2 (en) Evaluation method for a hereditary algorithm
CN110185791B (en) Control parameter optimization method for stepped automatic transmission
Fofana et al. Multi-objective genetic algorithm for an automatic transmission gear shift map
CN112429005A (en) Pure electric vehicle personalized gear shifting rule optimization method considering transmission efficiency and application
CN113103971A (en) Method for generating vehicle control data, vehicle control device, vehicle control system, and vehicle learning device
CN113006951A (en) Method for generating vehicle control data, vehicle control device, vehicle control system, and vehicle learning device
CN106828500A (en) Electric automobile geared automatic transmission schedule optimization method
Wu et al. IOT and cloud computing based parallel implementation of optimized RBF neural network for loader automatic shift control
CN115828438B (en) Method, medium and equipment for predicting ultimate performance of automobile
Debata et al. Efficient energy management strategies for hybrid electric vehicles using shuffled frog-leaping algorithm
Yin et al. Multi-performance optimization of the shift schedule for stepped automatic transmissions
CN116227311A (en) Improved Henry gas solubility optimization method based on differential evolution
Yin et al. Multi-performance optimal gearshift schedule of stepped automatic transmissions adaptive to road slope
CN112100909B (en) Parallel configurable intelligent optimization method based on collaborative optimization strategy
Xu et al. Automating shift-scheduling calibration by using bionic optimization and personalized driver models
Liu et al. Shift quality optimization based on assessment system for dual clutch transmission
Wang et al. Smart shift decision method based on stacked autoencoders
CN113537620A (en) Vehicle speed prediction method based on Markov model optimization and working condition recognition
Boehme et al. Multi-objective optimal design of parallel plug-in hybrid powertrain configurations with respect to fuel consumption and driving performance
Divakarla et al. Artificial neural network based adaptive control for plug-in hybrid electric vehicles
Zhang et al. Shift quality evaluation system based on neural network for DCT vehicles

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
GR01 Patent grant
GR01 Patent grant