CN107220405A - A kind of automobile chassis system integration Multipurpose Optimal Method based on improvement cell membrane optimized algorithm - Google Patents

A kind of automobile chassis system integration Multipurpose Optimal Method based on improvement cell membrane optimized algorithm Download PDF

Info

Publication number
CN107220405A
CN107220405A CN201710267334.9A CN201710267334A CN107220405A CN 107220405 A CN107220405 A CN 107220405A CN 201710267334 A CN201710267334 A CN 201710267334A CN 107220405 A CN107220405 A CN 107220405A
Authority
CN
China
Prior art keywords
automobile
steering
suspension
mul
optimization
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.)
Granted
Application number
CN201710267334.9A
Other languages
Chinese (zh)
Other versions
CN107220405B (en
Inventor
赵万忠
崔滔文
王春燕
徐志江
孔祥创
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.)
Nanjing University of Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
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 Nanjing University of Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN201710267334.9A priority Critical patent/CN107220405B/en
Publication of CN107220405A publication Critical patent/CN107220405A/en
Application granted granted Critical
Publication of CN107220405B publication Critical patent/CN107220405B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • 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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/002—Biomolecular computers, i.e. using biomolecules, proteins, cells
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Geometry (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biophysics (AREA)
  • General Engineering & Computer Science (AREA)
  • Molecular Biology (AREA)
  • Organic Chemistry (AREA)
  • Data Mining & Analysis (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Automation & Control Theory (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Chemical & Material Sciences (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Vehicle Body Suspensions (AREA)

Abstract

The present invention relates to a kind of based on the automobile chassis integrated system Multipurpose Optimal Method for improving cell membrane optimized algorithm, the optimization with suspension system is turned to, gearratio, torque sensor rigidity and the motor rotation inertia that optimization steering turns to servomotor including choosing are used as design variable；Suspension system optimization includes choosing suspension rate, suspension QS angular rigidity, suspension damping and tire cornering stiffness are used as design variable, using the ride comfort of the steering response, steering sensitivity and active suspension system of electric servo steering system as integrated optimization target, with steering stability and steering sensitivity, and suspension dynamic deflection is limited to integrated constraints, the cell membrane optimized algorithm of application enhancements optimizes design to integrated system；Finally according to the optimal compromise solution under Pareto solution selection extent functions as optimum results, so as to realize the complex optimum of vehicle ride performance, operational stability and security.

Description

It is a kind of excellent based on the automobile chassis system integration multiple target for improving cell membrane optimized algorithm Change method
Technical field
The present invention relates to automobile chassis system regions, particularly the automobile chassis system integration and based on many of the integrated system Purpose optimal method.
Background technology
Steering and the two big critical systems that suspension system is influence body gesture in automobile chassis system, are to ensure car The important composition part of ride performance, operational stability and security.Generally in the performance of analysis suspension system and steering system When, people's custom comes their influences to vehicle ride performance and control stability are relatively independent, i.e., to suspension system with Steering system sets up independent non-interfering kinetic model to be analyzed.Assuming that horizontal stroke of the input of vertical direction to automobile The motion of pendular motion and transverse movement without influence, i.e. suspension does not influence steering system；The side force that same tire is produced is only limited to behaviour The scope that handing stability considers, motion of the motion without influence, i.e. steering system on automotive vertical direction does not influence Suspension movement. So it is assumed that can with Simplified analysis scope, but in the actual driving process of automobile, road surface provided to vehicle turn to it is lateral While power, also to also have input a vertical interference to suspension system, therefore, automobile is in vertical and horizontal motion Influence each other, intercouple, it is necessary to this influence is taken in, the modeling to vehicle actual motion can be just fully achieved Analysis.The effect of suspension is in addition to support vehicle, isolation road agitation, also by body gesture when controlling to turn to, and transmission comes from The power of tire.I.e. same body movement can be caused by traveling input, the body roll as caused by Uneven road；Also can be by grasping Vertical aspect causes, body roll caused by during as turned to.Both independent studies, this is constantly walked with modern vehicle control technology Disagreed to the developing direction of integrated control.On the other hand, it was found from automobile theory, ride comfort and control stability can be considered Conflict, both are generally in opposite variation tendency, and an improvement inevitably results in another deterioration, and this is after all Because both kinetic models are strictly distinguished, it is impossible to while the result analyzed and optimized.Thus chassis system is carried out Analysis and research, set up the Integrated Optimization Model of steering system and suspension system and propose that corresponding integrated optimization algorithm and strategy just become Obtain extremely urgent.
In addition, the research of multi-objective Algorithm is significant to the combination property for lifting automobile chassis system.Cell membrane Optimized algorithm is a kind of new bionic intelligence algorithm, and it enters the process of cell membrane by studying material, by cell membrane pair The selection mechanism of material is applied in the selection to optimum results, and by experimental check, with good effect of optimization.But Traditional cell membrane optimized algorithm, on the one hand have the shortcomings that optimization convergence it is slower, before optimization the phase be easily trapped into part Optimal solution, after optimization the phase, with approaching for optimal solution of considering, optimization efficiency is reduced；On the other hand, traditional cell membrane optimization Algorithm is in terms of the diversity of disaggregation and NSGA-II, and archipelago genetic algorithm etc. has certain gap, and its diversity is needed into one Step improves.
The content of the invention
The technical problems to be solved by the invention are to be directed to defect involved in background technology to be based on changing there is provided one kind Enter the automobile chassis system integration Multipurpose Optimal Method of cell membrane optimized algorithm.
The present invention uses following technical scheme to solve above-mentioned technical problem：
The Multipurpose Optimal Method, is comprised the following steps：
1) electric servo steering system model, Full Vehicle Dynamics model are set up, active suspension system model, road surface is defeated Enter model, tire model, wherein electric servo steering system model include steering wheel input model, turn to servomotor Model, assist motor model, output shaft model, rack-and-pinion model；
2) by the ride comfort of the steering response, steering sensitivity and active suspension system of electric servo steering system As the Performance Evaluating Indexes of electric power-assisted steering, and set up quantitative formula；
Wherein, the quantitative formula of steering response is：
In formula：X1=Je+n1Jp1+Jp2+G2Jm；
Y1=Be+n1Bp+Bp+G2Bm；
Z1=n1Ks+Gn1KKaKs。
Wherein, s is Laplace operator, and E (s) is steering response, ThFor steering wheel input torque, corresponding Th(s) it is warp Steering wheel input torque after Laplace transform, TwTo act on the anti-torque on output shaft, Tw(s) it is through La Pula The anti-torque on output shaft, n are acted on after this conversion1To turn to the stator corner of servomotor and the ratio of angle of rotor Value. KsFor steering column equivalent stiffness, JeFor the rotary inertia of output shaft, Jp1Rotation for steering servo motor stator part is used to Amount, Jp2To turn to the rotary inertia of servo motor rotor part；BeFor the damped coefficient of output shaft；G slows down for worm-gear-worm The speed reducing ratio of mechanism；JmFor assist motor and the rotary inertia of clutch, BmFor assist motor viscous damping coefficient, BpTo turn to Viscous damping when servomotor is turned to, K is assist motor power-assisted gain, KaFor motor torque coefficient.
The quantitative formula of steering sensitivity is：
Wherein,
X2=n2G2Jm+n1n2Jp1+n2Jp2+n2Je；
Y2=n2G2Bm+n1n2Bp+n2Bp+n2Be；
A2=-IxzLβYδ+IxzLδYβ-IxNβYδ+IxNδYβ+muLpNδ+mshLβNδ-mshLδNβ
A1=LpNβYδ-LpNδYβ-muLδNφ+muLφNδ+mshuNδYφ-mshuNφYδ
A0=-LβNφYδ+LβNδYφ-LδNβYφ+LδNφYβ+LφNβYδ-LφNδYβ
B1=IzLβYφ-IzLφYβ+IxzNβYφ-IxzNφYβ-LpNβYr+LpNrYβ
+muLpNβ-muLφNr+muLrNφ+mshuNφYr-mshuNrYφ
B0=LβNφYr-LβNrYφ-LφNβYr+LφNrYβ+LrNβYφ-LrNφYβ
-muLβNφ+muLφNβ+mshuNβYφ-mshuNφYβ
F3=I2 xzYδ-IxIzYδ-mshIzLδ-mshIxzNδ
F1=-IzLδYφ+IzLφYδ-IxzNδYφ+IxzNφYδ+LpNδYr-LpNrYδ-muLpNδ
F0=-LδNφYr+LδNrYφ+LφNδYr-LφNrYδ-LrNδYφ+LrNφYδ
+muLδNφ-muLφNδ+mshuNφYδ-mshuNδYφ
H2=-muIzLδ-muIxzNδ-mshuIzYδ
H1=-IzLβYδ+IzLδYβ-IxzNβYδ+IxzNδYβ
+muLδNr-muLrNδ-mshuNδYr+mshuNrYδ
H0=-LβNδYr+LβNrYδ+LδNβYr-LδNrYβ-LrNβYδ+LrNδYβ
+muLβNδ-muLδNβ+mshuNδYβ-mshuNβYδ
Nβ=-a (k1+k2)+b(k3+k4)
Nφ=-aE1(k1+k2)+bE2(k3+k4)
Nδ=a (k1+k2)；
Yβ=-(k1+k2+k3+k4)
Yφ=-(k1+k2)E1-(k3+k4)E2
Yδ=k1+k2；
Lβ=-(k1+k2+k3+k4)h
Lθ=-[(C21-C22)a+(C23-C24)b]d
Lδ=(k1+k2)h
Lp=-(D21+D22+D23+D24)d2
Le=-[(D21-D22)a+(D23-D24)b]d
In formula (2), δ (s) is the front wheel angle after Laplace transform, θs(s) it is the side after Laplace transform To disk corner, β (s) is the yaw acceleration after Laplace transform, and φ (s) is the barycenter side after Laplace transform Drift angle, wr(s) it is the yaw velocity after Laplace transform, n2For the gearratio from steering screw to front-wheel, u is automobile Speed, b is barycenter to vehicle rear axle distance, and d is the wheelspan of vehicle 1/2, E1For roll steer coefficient, k1、k2Respectively automobile is left The cornering stiffness of front-wheel and off-front wheel；H is the rolling moment arm of automobile；M is the complete vehicle quality of automobile；msFor the spring charge material of automobile Amount；IxFor automobile sprung mass to the rotary inertia of x-axis；IyFor automobile sprung mass to the rotary inertia of y-axis；IzFor vapour Rotary inertia of the sprung mass of car to z-axis；IxzFor automobile sprung mass to x, the product of inertia of z-axis；E1For the front side of automobile Incline steer coefficient；E2For the rear roll steer coefficient of automobile；Ca1For the front suspension QS angular rigidity of automobile；Ca2For vapour The rear suspension QS angular rigidity of car；C21, C22 are respectively the left front suspension rate and right front suspension rigidity of automobile； C23, C24 are respectively the left rear suspension rigidity and right rear suspension rigidity of automobile；D21、 D22The respectively left front suspension damping of automobile Coefficient and right front suspension damped coefficient；D23、D24The respectively left rear suspension damped coefficient and right rear suspension damped coefficient of automobile.
Steering stability condition：
Select the denominator of steering sensitivity transmission function：
Q6s6+Q5s5+Q4s4+Q3s3+Q2s2+Q1s1+Q0 (3)
Wherein：
Q6=X2B4
Q5=X2B3+Y2B4
Q4=X2B2+Y2B3+Z2B4
List its Routh table as follows：
According to Routh criterions, it is desirable to which respectively value is just to first row in Routh tables；
Automobile ride index includes body vibrations acceleration, the dynamic deflection and dynamic wheel load of suspension, wherein, suspension System model is：
In formula (5), m11,m12Unsprung mass, m respectively before the left front unsprung mass of automobile and the right side13,m14Respectively Unsprung mass, Z after the left back unsprung mass of automobile and the right side11,Z12It is respectively vertical at the left front unsprung mass of automobile Vertical displacement at displacement and right preceding unsprung mass, Z13,Z14Vertical displacement respectively at the left back unsprung mass of automobile With the vertical displacement behind the right side at unsprung mass.z1,z2Road at road surface input displacement and off-front wheel respectively at the near front wheel of automobile Face input displacement, z3,z4Road surface input displacement at road surface input displacement and off hind wheel respectively at the left rear wheel of automobile,For car Body vibration acceleration；
Body vibrations acceleration transmission function：
Suspension dynamic deflection transmission function：
Dynamic wheel load transmission function：
Inputted for road roughness, Z21,Z22The respectively vertical displacement at the left front sprung mass of automobile and right front overhang Hang the vertical displacement at quality, Z23,Z24The respectively vertical displacement at the left back sprung mass of automobile and right rear-mounted quality The vertical displacement at place.
Body vibrations accelerationThe dynamic deflection δ of suspensiondWith dynamic wheel load FdThree vibratory response amounts, their power The power spectral density of spectrum density and road surface input quantity is represented by：
In formula：For frequency,As amplitude versus frequency characteGx(f) it is response quautity power Spectrum density；For the power spectral density of road surface input quantity.
According to random vibration theory, response mean-square value is：
In formula：σxFor the standard deviation of vibratory response amount.
3) according to DOE experimental results, choosing influences larger parameter as optimized variable above-mentioned optimization aim, chooses and turns To the gearratio n of servomotor1, torque sensor stiffness KsAnd motor rotation inertia J (N.m/rad)m(kg.m2), suspension Rigidity C21, C22, C23, C24(N/m), suspension QS angular rigidity Ca, Ca2(N.m/rad), suspension damping D21, D22, D23, D24And tire cornering stiffness k (N.s/m)1, k2, k3, k4(N/rad) as design variable.Assuming that vehicle fore suspension and rear suspension both sides have There are identical rigidity and damping, if front and rear tire both sides cornering stiffness is equal, in summary, it is excellent that integrated optimization system is chosen Changing design variable is：
X=[n1、Ks、Jm、k1、k3、C21、C23、D21、D23、Ca1、Ca2]； (11)
4) steering response, steering sensitivity and ride comfort using integrated system are as optimization aim, with steering stability And steering sensitivity, and suspension dynamic deflection is limited to constraints, sets up integrated system Model for Multi-Objective Optimization, its target letter Number is expressed as
In formula, f1(X) effective frequency range (0, ω of the road feel in information of road surface is represented0) frequency domain energy average value；ω0 Represent to take ω in the maximum frequency values of useful signal in information of road surface, optimization design0=40Hz, f2(X) represent sensitivity on road The effective frequency range (0, ω of face information0) frequency domain energy average value, f3(X) ride comfort, w are represented0、wi、wj+4For weight, s0、si、sj+4For scale factor.
5) first, steering needs to meet Routh criterions in integrated system, i.e.,：
Q6＞ 0；Q5＞ 0；a1＞ 0；b1＞ 0；c1＞ 0；d1＞ 0；Q0＞ 0
In addition, it is desirable to which steering sensitivity energy is in certain scope, to ensure that driver obtains good steering sensitivity, I.e.
0.0008≤f(x2)≤0.0099；
Finally, the dynamic deflection δ of suspensiondIt is required that within the specific limits, this optimal design-aside is:
0.06≤δd≤0.1；
6) above-mentioned parameter is optimized using improved cell membrane fusion algorithm, drawn according to optimum results optimal Pareto disaggregation, and choose optimal solution of compromising；
7) each parameter setting is taken off as optimum results using optimal compromise, completes optimization.
Further, improved cell membrane optimized algorithm of the present invention is comprised the following steps that：
Step A), randomly generate the primary population P that population scale is N0。
Step B), population is divided
Population is ranked up using TOPSIS methods, the material of Ps ratios is liposoluble substance before taking, and is come below All as non-fat-soluble material；Further according to material concentration height, non-fat-soluble material is divided into two classes：The non-liposoluble of high concentration Property material and low concentration non-fat-soluble material；
For something Y, the material number that the concentration residing for it is defined as being included in its contiguous range accounts for total material number Percentage：
Wherein, Con is concentration around the material, and m is the quantity of total material, and n represents XiIn to Y distance be less than r × (u-l) number, wherein, r is the radius factor that substance for calculation concentration is used, and u is the upper dividing value of solution space, and l is solution space Floor value；
The average value MeanCon of all concentration is：
If not the concentration residing for liposoluble substance is more than MeanCon, then the material is divided into high concentration non-fat-soluble Material, is otherwise divided into low concentration non-fat-soluble material；
Step C), liposoluble substance diffusion
For each liposoluble substanceCentered on the material, raFor in the region of search of radius, random generation is new K1Individual materialAnd it is rightPart beyond hunting zone is modified；
Then search radius vector carries out contraction rnew a=ra×u； (15)
Search radius vector r during beginningaComputational methods are：
Wherein,R ∈ [0,1], b=3, T set maximum algebraically to evolve, and t is to optimize current algebraically, t Initial value be 0；When t is smaller, S (t) ≈ 1, search radius is relatively large, when t is larger, S (t) ≈ 0, search radius phase To smaller, accelerate search speed；
The modification method of hunting zone is：For any K, ifThen makeIfThen makeWherein, ukFor the upper dividing value of kth dimension of solution space, lkFloor value is tieed up for the kth of solution space；
Step D), the high concentration non-fat-soluble motion of matter
The probability for making each high concentration non-fat-soluble material there is carrier is P1If, rand [0,1]≤P1, then the material Diffusion can be assisted, from high concentration lateral movement to low concentration side, and makes new position be Local Search centerOtherwise make former Position is Local Search center；
Then, the material can carry out the Local Search motion of K times, before this, it is necessary to initialize search radius vector
After search, K is recorded2Individual novel substance
Step E), the low concentration non-fat-soluble motion of matter
The probability for making each low concentration non-fat-soluble material there is carrier is P2；
The energy of low concentration non-fat-soluble material is made to be in [0,1], and by linear distribution；
The functional value of each low concentration non-fat-soluble material is calculated firstAgain by these functional values carry out from it is small to Big sequence；
Its ENERGY E of the material of functional value minimumiFor Emin, its ENERGY E of the material of functional value maximumiFor Emax, other materials ENERGY EiBetween EminWith EmaxBetween, and calculated according to the order linear of sequence, wherein, EminWith EmaxTo be normal in [0,1] Count, herein EminIt is taken as 0, EmaxIt is taken as 1；
If somethingThere is carrier and enough energy, then it can carry out active transport, from low concentration side to High concentration side, and make new position be Local Search centerOtherwise it is Local Search center to make in situ；
Low concentration non-fat-soluble materialNew position after active transport
Then, search radius is initialized
Then, withCentered on and with rcFor in the region of search of radius, random motion K3It is secondary, and to their model Enclose and be modified；
Record K3Individual novel substance
Step F), generation reference point, reference line
According to the quantity M of the optimization aim and search hop count g artificially divided, determine equally distributed in optimization space The quantity H of reference point
The reference point of formation is connected with optimizing the initial point (may be configured as origin) in space, connected line segment is referred to as reference Line.
Step G), generate new population
Calculate a new generation produced by liposoluble substance, high concentration non-fat-soluble material, low concentration non-fat-soluble material Body, if number of individuals is less than or equal to N, all individuals are designated as P as next filial generationt+1；Otherwise need to sieve filial generation Choosing, is operated according to the microhabitat of genetic algorithm, is chosen individual and is entered population of future generation, its screening mode is：According to non-branch With sequence, the high non-dominant layer individual of grade is preferentially entered, until population quantity for the first time up to or over N (assuming that now The number of plies is dominated l), just to reach N, then all individuals of above non-dominant layer enter population of new generation, otherwise first l-1 layers New population is fully entered, l layers need to screen again, be ranked up according to the vertical range of l layers of individual and reference line, while than Whether it is further associated compared with the l layers of associated reference line of individual with other preceding l-1 layers of individuals, if so, be compared again, When l layers of individual and reference lines at a distance of closer to, then remain into population of future generation, otherwise will take in order next l layer it is individual, directly To making individual amount of future generation reach N, if compromised in l layers of search less than enough individuals, by away from reference line closer to Individual income population of future generation, makes population quantity reach N.
Step H), judge whether optimization terminates
If t<T, then Jia 1 to t progress and operate, and correct search radius ra、rb、rc, return to step B) and enter iteration, if T=T, then be ranked up to disaggregation, forms Pareto disaggregation, provides optimization solution, and optimization terminates.
The present invention uses above technical scheme compared with prior art, with following technique effect：
1. automobile chassis system Multipurpose Optimal Method proposed by the present invention, to lift two aspects of steering and suspension Combination property be target, in terms of steering, by improving steering response to lift driver's driving experience, improve and turn To system sensitivity to improve the maneuvering performance of automobile；In terms of suspension system, using the ride comfort of steering as optimization mesh Mark.Multiple optimization aims are optimized, can be effectively improved automobile chassis system by the comprehensive performance requirement turned to suspension Combination property, meanwhile, be that the optimization of automobile other systems lays the foundation that there is provided method, theoretical direction.
2. improved cell membrane optimized algorithm proposed by the present invention, by non-dominated ranking, the choosing of the microhabitat based on reference point The system of selecting a good opportunity is blended with traditional cell membrane optimized algorithm, offspring individual is produced by conventional cell film optimized algorithm, by non- Dominated Sorting is screened with microhabitat operation to offspring individual, can effectively improve the various of multiple-objection optimization result solution Property, lift the efficiency of multi-objective optimization algorithm.
Brief description of the drawings
Fig. 1 is chassis integrated system optimization method flow chart；
Fig. 2 is improved cell membrane optimized algorithm flow chart.
Embodiment
Technical scheme is described in further detail below in conjunction with the accompanying drawings：
As shown in figure 1, the Multipurpose Optimal Method, it is characterised in that comprise the following steps：
1) electric servo steering system model, Full Vehicle Dynamics model are set up, active suspension system model, road surface is defeated Enter model, tire model, wherein electric servo steering system model include steering wheel input model, turn to servomotor Model, assist motor model, output shaft model, rack-and-pinion model；
Method for establishing model referring to《The research of automobile power steering pump and control valve》(Shandong university of science etc., colleges and universities' reason Section is studied),《The design studies of Electro-Hydraulic Power Steering System》(monarch Mr. Zhang, Jiangsu University),《Electric hydraulic power-assisted steering System control strategy and its energy consumption analysis method》The method disclosed in document such as (Su Jiankuan etc., machine design and manufacture)；
2) by the ride comfort of the steering response, steering sensitivity and active suspension system of electric servo steering system As the Performance Evaluating Indexes of electric power-assisted steering, and set up quantitative formula；
Wherein, the quantitative formula of steering response is：
In formula：X1=Je+n1Jp1+Jp2+G2Jm；
Y1=Be+n1Bp+Bp+G2Bm；
Z1=n1Ks+Gn1KKaKs。
S is Laplace operator, and E (s) is steering response, ThFor steering wheel input torque, corresponding Th (s) is general through drawing Steering wheel input torque after the conversion of Lars, TwTo act on the anti-torque on output shaft, Tw(s) it is to become through Laplce The anti-torque acted on after changing on output shaft, n1To turn to the stator corner of servomotor and the ratio of angle of rotor；Ks For steering column equivalent stiffness, s, JeFor the rotary inertia of output shaft, Jp1To turn to the rotary inertia of servo motor stator part, Jp2To turn to the rotary inertia of servo motor rotor part；BeFor the damped coefficient of output shaft；G is worm-gear-worm reducing gear Speed reducing ratio；JmFor assist motor and the rotary inertia of clutch, BmFor assist motor viscous damping coefficient, BpTo turn to servo Viscous damping during motor steering, K is assist motor power-assisted gain, KaFor motor torque coefficient.
The quantitative formula of steering sensitivity is：
Wherein,
X2=n2G2Jm+n1n2Jp1+n2Jp2+n2Je；
Y2=n2G2Bm+n1n2Bp+n2Bp+n2Be；
A2=-IxzLβYδ+IxzLδYβ-IxNβYδ+IxNδYβ+muLpNδ+mshLβNδ-mshLδNβ
A1=LpNβYδ-LpNδYβ-muLδNφ+muLφNδ+mshuNδYφ-mshuNφYδ
A0=-LβNφYδ+LβNδYφ-LδNβYφ+LδNφYβ+LφNβYδ-LφNδYβ
B1=IzLβYφ-IzLφYβ+IxzNβYφ-IxzNφYβ-LpNβYr+LpNrYβ
+muLpNβ-muLφNr+muLrNφ+mshuNφYr-mshuNrYφ
B0=LβNφYr-LβNrYφ-LφNβYr+LφNrYβ+LrNβYφ-LrNφYβ
-muLβNφ+muLφNβ+mshuNβYφ-mshuNφYβ
F3=I2 xzYδ-IxIzYδ-mshIzLδ-mshIxzNδ
F1=-IzLδYφ+IzLφYδ-IxzNδYφ+IxzNφYδ+LpNδYr-LpNrYδ-muLpNδ
F0=-LδNφYr+LδNrYφ+LφNδYr-LφNrYδ-LrNδYφ+LrNφYδ
+muLδNφ-muLφNδ+mshuNφYδ-mshuNδYφ
H2=-muIzLδ-muIxzNδ-mshuIzYδ
H1=-IzLβYδ+IzLδYβ-IxzNβYδ+IxzNδYβ
+muLδNr-muLrNδ-mshuNδYr+mshuNrYδ
H0=-LβNδYr+LβNrYδ+LδNβYr-LδNrYβ-LrNβYδ+LrNδYβ
+muLβNδ-muLδNβ+mshuNδYβ-mshuNβYδ
Nβ=-a (k1+k2)+b(k3+k4)
Nφ=-aE1(k1+k2)+bE2(k3+k4)
Nδ=a (k1+k2)；
Yβ=-(k1+k2+k3+k4)
Yφ=-(k1+k2)E1-(k3+k4)E2
Yδ=k1+k2；
Lβ=-(k1+k2+k3+k4)h
Lθ=-[(C21-C22)a+(C23-C24)b]d
Lδ=(k1+k2)h
Lp=-(D21+D22+D23+D24)d2
Le=-[(D21-D22)a+(D23-D24)b]d
δ (s) is the front wheel angle after Laplace transform, θs(s) it is the steering wheel angle after Laplace transform, β (s) is the yaw acceleration after Laplace transform, and φ (s) is the side slip angle after Laplace transform, wr(s) For the yaw velocity after Laplace transform, n2For the gearratio from steering screw to front-wheel, u is automobile speed, and d is The wheelspan of vehicle 1/2, E1For roll steer coefficient, k1、k2The respectively cornering stiffness of automobile the near front wheel and off-front wheel；H is automobile Rolling moment arm；M is the complete vehicle quality of automobile；msFor the spring carried mass of automobile；IxFor rotation of the sprung mass to x-axis of automobile Inertia；IyFor automobile sprung mass to the rotary inertia of y-axis；IzFor automobile sprung mass to the rotary inertia of z-axis； Ixz For automobile sprung mass to x, the product of inertia of z-axis；E1For the preceding roll steer coefficient of automobile；E2For the rear roll steer of automobile Coefficient；Ca1For the front suspension QS angular rigidity of automobile；Ca2For the rear suspension QS angular rigidity of automobile； C21、 C22 is respectively the left front suspension rate and right front suspension rigidity of automobile；C23, C24 be respectively automobile left rear suspension rigidity and Right rear suspension rigidity；D21、D22The respectively left front suspension damping coefficient and right front suspension damped coefficient of automobile；D23、 D24Respectively For the left rear suspension damped coefficient and right rear suspension damped coefficient of automobile.
Steering stability condition：
Select the denominator of steering sensitivity transmission function：
Q6s6+Q5s5+Q4s4+Q3s3+Q2s2+Q1s1+Q0 (3)
Wherein：
Q6=X2B4
Q5=X2B3+Y2B4
Q4=X2B2+Y2B3+Z2B4
List its Routh table as follows：
According to Routh criterions, it is desirable to which respectively value is just to first row in Routh tables.
Automobile ride index includes body vibrations acceleration, the dynamic deflection and dynamic wheel load of suspension, wherein, suspension System model is：
In formula, m11,m12Unsprung mass, m respectively before the left front unsprung mass of automobile and the right side13,m14Respectively automobile Left back unsprung mass and it is right after unsprung mass, Z11,Z12Vertical position respectively at the left front unsprung mass of automobile Vertical displacement before moving and being right at unsprung mass, Z13,Z14Vertical displacement respectively at the left back unsprung mass of automobile With the vertical displacement behind the right side at unsprung mass.z1,z2Respectively at the near front wheel of automobile at road surface input displacement and off-front wheel Road surface input displacement, z3,z4Road surface input displacement at road surface input displacement and off hind wheel respectively at the left rear wheel of automobile,For Body vibrations acceleration
Body vibrations acceleration transmission function：
Suspension dynamic deflection transmission function：
Dynamic wheel load transmission function：
Inputted for road roughness, Z21,Z22The respectively vertical displacement at the left front sprung mass of automobile and right front overhang Hang the vertical displacement at quality, Z23,Z24The respectively vertical displacement at the left back sprung mass of automobile and right rear-mounted quality The vertical displacement at place.
Body vibrations accelerationThe dynamic deflection δ of suspensiondWith dynamic wheel load FdThree vibratory response amounts, their power The power spectral density of spectrum density and road surface input quantity is represented by：
In formula：For frequency,As amplitude versus frequency characteGx(f) it is response quautity power Spectrum density；For the power spectral density of road surface input quantity.
According to random vibration theory, response mean-square value is：
In formula：σxFor the standard deviation of vibratory response amount.
3) according to DOE experimental results, choosing influences larger parameter as optimized variable above-mentioned optimization aim, chooses and turns To the gearratio n of servomotor1, torque sensor stiffness KsAnd motor rotation inertia J (N.m/rad)m(kg.m2), suspension Rigidity C21, C22, C23, C24(N/m), suspension QS angular rigidity Ca, Ca2(N.m/rad), suspension damping D21, D22, D23, D24And tire cornering stiffness k (N.s/m)1, k2, k3, k4(N/rad) as design variable.Assuming that vehicle fore suspension and rear suspension both sides have There are identical rigidity and damping, if front and rear tire both sides cornering stiffness is equal, in summary, it is excellent that integrated optimization system is chosen Changing design variable is：
X=[n1、Ks、Jm、k1、k3、C21、C23、D21、D23、Ca1、Ca2]。 (11)
4) steering response, steering sensitivity and ride comfort using integrated system are as optimization aim, with steering stability And steering sensitivity, and suspension dynamic deflection is limited to constraints, sets up integrated system Model for Multi-Objective Optimization, its target letter Number is expressed as
In formula, f1(X) effective frequency range (0, ω of the road feel in information of road surface is represented0) frequency domain energy average value；ω0 Represent to take ω in the maximum frequency values of useful signal in information of road surface, optimization design0=40Hz, f2(X) represent sensitivity on road The effective frequency range (0, ω of face information0) frequency domain energy average value, f3(X) ride comfort, w are represented0、wi、wj+4For weight, s0、si、sj+4For scale factor.
5) first, steering needs to meet Routh criterions in integrated system, i.e.,：
Q6＞ 0；Q5＞ 0；a1＞ 0；b1＞ 0；c1＞ 0；d1＞ 0；Q0＞ 0
In addition, it is desirable to which steering sensitivity energy is in certain scope, to ensure that driver obtains good steering sensitivity, I.e.
0.0008≤f(x2)≤0.0099
Finally, the dynamic deflection δ of suspensiondIt is required that within the specific limits, this optimal design-aside is:
0.06≤δd≤0.1
6) above-mentioned parameter is optimized using improved cell membrane fusion algorithm, drawn according to optimum results optimal Pareto disaggregation, and choose optimal solution of compromising；
7) each parameter setting is taken off as optimum results using optimal compromise, completes optimization.
In the present embodiment, as shown in Fig. 2 the improved cell membrane optimized algorithm is it is characterised in that it includes following steps：
Step A), randomly generate the primary population P that population scale is N0。
Step B), population is divided
Population is ranked up using TOPSIS methods, the material of Ps ratios is liposoluble substance before taking, and is come below All as non-fat-soluble material；Further according to material concentration height, non-fat-soluble material is divided into two classes：The non-liposoluble of high concentration Property material and low concentration non-fat-soluble material；
For something Y, the material number that the concentration residing for it is defined as being included in its contiguous range accounts for total material number Percentage：
Wherein, Con is concentration around the material, and m is the quantity of total material, and n represents XiIn to Y distance be less than r × (u-l) number, wherein, r is the radius factor that substance for calculation concentration is used, and u is the upper dividing value of solution space, and l is solution space Floor value；
The average value MeanCon of all concentration is：
If not the concentration residing for liposoluble substance is more than MeanCon, then the material is divided into high concentration non-fat-soluble Material, is otherwise divided into low concentration non-fat-soluble material；
Step C), liposoluble substance diffusion
For each liposoluble substanceCentered on the material, raFor in the region of search of radius, random generation is new K1Individual materialAnd it is rightPart beyond hunting zone is modified；
Then search radius vector carries out contraction rnew a=ra×u； (15)
Search radius vector r during beginningaComputational methods are：
Wherein,R ∈ [0,1], b=3, T set maximum algebraically to evolve, and t is to optimize current algebraically, T initial value is 0；When t is smaller, S (t) ≈ 1, search radius is relatively large, when t is larger, S (t) ≈ 0, search radius phase To smaller, accelerate search speed；
The modification method of hunting zone is：For any K, ifThen makeIfThen makeWherein, ukFor the upper dividing value of kth dimension of solution space, lkFloor value is tieed up for the kth of solution space；
Step D), the high concentration non-fat-soluble motion of matter
The probability for making each high concentration non-fat-soluble material there is carrier is P1If, rand [0,1]≤P1, then the material Diffusion can be assisted, from high concentration lateral movement to low concentration side, and makes new position be Local Search centerOtherwise make in situ For Local Search center；
Then, the material can carry out the Local Search motion of K times, before this, it is necessary to initialize search radius vector
After search, K is recorded2Individual novel substance
Step E), the low concentration non-fat-soluble motion of matter
The probability for making each low concentration non-fat-soluble material there is carrier is P2；
The energy of low concentration non-fat-soluble material is made to be in [0,1], and by linear distribution；
The functional value of each low concentration non-fat-soluble material is calculated firstAgain by these functional values carry out from it is small to Big sequence；
Its ENERGY E of the material of functional value minimumiFor Emin, its ENERGY E of the material of functional value maximumiFor Emax, other materials ENERGY EiBetween EminWith EmaxBetween, and calculated according to the order linear of sequence, wherein, EminWith EmaxTo be normal in [0,1] Count, herein EminIt is taken as 0, EmaxIt is taken as 1；
If somethingThere is carrier and enough energy, then it can carry out active transport, from low concentration side to High concentration side, and make new position be Local Search centerOtherwise it is Local Search center to make in situ；
Low concentration non-fat-soluble materialNew position after active transport
Then, search radius is initialized
Then, withCentered on and with rcFor in the region of search of radius, random motion K3It is secondary, and to their model Enclose and be modified；
Record K3Individual novel substance
Step F), generation reference point, reference line
According to the quantity M of the optimization aim and search hop count g artificially divided, determine equally distributed in optimization space The quantity H of reference point:
The reference point of formation is connected with optimizing the initial point (may be configured as origin) in space, connected line segment is referred to as reference Line.
Step G), generate new population
Calculate a new generation produced by liposoluble substance, high concentration non-fat-soluble material, low concentration non-fat-soluble material Body, if number of individuals is less than or equal to N, all individuals are designated as P as next filial generationt+1；Otherwise need to sieve filial generation Choosing, is operated according to the microhabitat of genetic algorithm, is chosen individual and is entered population of future generation, its screening mode is：According to non-branch With sequence, the high non-dominant layer individual of grade is preferentially entered, until population quantity for the first time up to or over N (assuming that now The number of plies is dominated l), just to reach N, then all individuals of above non-dominant layer enter population of new generation, otherwise first l-1 layers New population is fully entered, l layers need to screen again, be ranked up according to the vertical range of l layers of individual and reference line, while than Whether it is further associated compared with the l layers of associated reference line of individual with other preceding l-1 layers of individuals, if so, be compared again, When l layers of individual and reference lines at a distance of closer to, then remain into population of future generation, otherwise will take in order next l layer it is individual, directly To making individual amount of future generation reach N, if compromised in l layers of search less than enough individuals, by away from reference line closer to Individual income population of future generation, makes population quantity reach N.
Step H), judge whether optimization terminates
If t<T, then Jia 1 to t progress and operate, and correct search radius ra、rb、rc, return to step B) and enter iteration, if T=T, then be ranked up to disaggregation, forms Pareto disaggregation, provides optimization solution, and optimization terminates.
Those skilled in the art of the present technique are it is understood that unless otherwise defined, all terms used herein (including skill Art term and scientific terminology) with the general understanding identical meaning with the those of ordinary skill in art of the present invention.Also It should be understood that those terms defined in such as general dictionary should be understood that with the context with prior art In the consistent meaning of meaning, and unless defined as here, will not be solved with idealization or excessively formal implication Release.
Above-described embodiment, has been carried out further to the purpose of the present invention, technical scheme and beneficial effect Describe in detail, should be understood that the embodiment that the foregoing is only the present invention, be not limited to this hair Bright, within the spirit and principles of the invention, any modification, equivalent substitution and improvements done etc. should be included in this hair Within bright protection domain.

Claims (2)

1. it is a kind of based on the automobile chassis integrated system Multipurpose Optimal Method for improving cell membrane optimized algorithm, it is characterised in that Step is as follows：
1) electric servo steering system model, Full Vehicle Dynamics model, active suspension system model, road surface input mould are set up Type and tire model；
The electric servo steering system model includes steering wheel input model, turns to servomotor model, assist motor Model, output shaft model and rack-and-pinion model；
2) using the ride comfort of the steering response, steering sensitivity and active suspension system of electric servo steering system as The Performance Evaluating Indexes of electric power-assisted steering, and set up quantitative formula；
Wherein, the quantitative formula of steering response is：
In formula：X1=Je+n1Jp1+Jp2+G2Jm；
Y1=Be+n1Bp+Bp+G2Bm；
Z1=n1Ks+Gn1KKaKs；
S is Laplace operator, and E (s) is steering response, ThFor steering wheel input torque, corresponding Th(s) it is through Laplce Steering wheel input torque after conversion, TwTo act on the anti-torque on output shaft, Tw(s) it is after Laplace transform The anti-torque acted on output shaft, n1For the gearratio of steering steering wheel angle to front wheel angle, KsFor steering column Equivalent stiffness, JeFor the rotary inertia of output shaft, Jp1To turn to the rotary inertia of servo motor stator part, Jp2To turn to servo The rotary inertia of rotor part；BeFor the damped coefficient of output shaft；G is the speed reducing ratio of worm-gear-worm reducing gear；JmFor The rotary inertia of assist motor and clutch, BmFor assist motor viscous damping coefficient, BpIt is viscous when servomotor is turned to turn to Property damping, K be assist motor power-assisted gain, KaFor motor torque coefficient；
The quantitative formula of steering sensitivity is：
Wherein,
X2=n2G2Jm+n1n2Jp1+n2Jp2+n2Je；
Y2=n2G2Bm+n1n2Bp+n2Bp+n2Be；
A2=-IxzLβYδ+IxzLδYβ-IxNβYδ+IxNδYβ+muLpNδ+mshLβNδ-mshLδNβ
A1=LpNβYδ-LpNδYβ-muLδNφ+muLφNδ+mshuNδYφ-mshuNφYδ
A0=-LβNφYδ+LβNδYφ-LδNβYφ+LδNφYβ+LφNβYδ-LφNδYβ
B1=IzLβYφ-IzLφYβ+IxzNβYφ-IxzNφYβ-LpNβYr+LpNrYβ
+muLpNβ-muLφNr+muLrNφ+mshuNφYr-mshuNrYφ
B0=LβNφYr-LβNrYφ-LφNβYr+LφNrYβ+LrNβYφ-LrNφYβ
-muLβNφ+muLφNβ+mshuNβYφ-mshuNφYβ
F3=I2 xzYδ-IxIzYδ-mshIzLδ-mshIxzNδ
F1=-IzLδYφ+IzLφYδ-IxzNδYφ+IxzNφYδ+LpNδYr-LpNrYδ-muLpNδ
F0=-LδNφYr+LδNrYφ+LφNδYr-LφNrYδ-LrNδYφ+LrNφYδ
+muLδNφ-muLφNδ+mshuNφYδ-mshuNδYφ
H2=-muIzLδ-muIxzNδ-mshuIzYδ
H1=-IzLβYδ+IzLδYβ-IxzNβYδ+IxzNδYβ
+muLδNr-muLrNδ-mshuNδYr+mshuNrYδ
H0=-LβNδYr+LβNrYδ+LδNβYr-LδNrYβ-LrNβYδ+LrNδYβ
+muLβNδ-muLδNβ+mshuNδYβ-mshuNβYδ
Nβ=-a (k1+k2)+b(k3+k4)
Nφ=-aE1(k1+k2)+bE2(k3+k4)
Nδ=a (k1+k2)；
Yβ=-(k1+k2+k3+k4)
Yφ=-(k1+k2)E1-(k3+k4)E2
Yδ=k1+k2；
Lβ=-(k1+k2+k3+k4)h
Lθ=-[(C21-C22)a+(C23-C24)b]d
Lδ=(k1+k2)h
Lp=-(D21+D22+D23+D24)d2
Le=-[(D21-D22)a+(D23-D24)b]d
δ (s) is the front wheel angle after Laplace transform, θs(s) it is the steering wheel angle after Laplace transform, β (s) For the yaw acceleration after Laplace transform, φ (s) is the side slip angle after Laplace transform, wr(s) it is warp Yaw velocity after Laplace transform, n2For the gearratio from steering screw to front-wheel, u is automobile speed, and b arrives for barycenter Vehicle rear axle distance, d is the wheelspan of vehicle 1/2, E1For roll steer coefficient, k1、k2The respectively side of automobile the near front wheel and off-front wheel Inclined rigidity；H is the rolling moment arm of automobile；M is the complete vehicle quality of automobile；msFor the spring carried mass of automobile；IxFor the suspension matter of automobile Measure the rotary inertia to x-axis；IyFor automobile sprung mass to the rotary inertia of y-axis；IzFor automobile sprung mass to z-axis Rotary inertia；IxzFor automobile sprung mass to x, the product of inertia of z-axis；E1For the preceding roll steer coefficient of automobile；E2For automobile Rear roll steer coefficient；Ca1For the front suspension QS angular rigidity of automobile；Ca2For the rear suspension QS of automobile Angular rigidity；C21、C22The respectively left front suspension rate of automobile and right front suspension rigidity；C23、C24The respectively left rear suspension of automobile Rigidity and right rear suspension rigidity；D21、D22The respectively left front suspension damping coefficient and right front suspension damped coefficient of automobile；D23、D24 The respectively left rear suspension damped coefficient and right rear suspension damped coefficient of automobile；
Steering stability condition：
Select the denominator of steering sensitivity transmission function：
Q6s6+Q5s5+Q4s4+Q3s3+Q2s2+Q1s1+Q0 (3)
Wherein：
Q6=X2B4
Q5=X2B3+Y2B4
Q4=X2B2+Y2B3+Z2B4
List its Routh table as follows：
According to Routh criterions, it is desirable to which respectively value is just to first row in Routh tables；
Automobile ride index includes body vibrations acceleration, the dynamic deflection and dynamic wheel load of suspension, wherein, suspension system Model is：
In formula, m11,m12Unsprung mass, m respectively before the left front unsprung mass of automobile and the right side13,m14A respectively left side for automobile Afterwards unsprung mass and it is right after unsprung mass, Z11,Z12The respectively vertical displacement at the left front unsprung mass of automobile and the right side Vertical displacement at preceding unsprung mass, Z13,Z14It is non-after vertical displacement and the right side respectively at the left back unsprung mass of automobile Vertical displacement at sprung mass.z1,z2Road surface input bit at road surface input displacement and off-front wheel respectively at the near front wheel of automobile Move, z3,z4Road surface input displacement at road surface input displacement and off hind wheel respectively at the left rear wheel of automobile,For body vibrations plus Speed；
Body vibrations acceleration transmission function：
Suspension dynamic deflection transmission function：
Dynamic wheel load transmission function：
Inputted for road roughness, Z21,Z22The respectively vertical displacement at the left front sprung mass of automobile and right forward mounting matter Vertical displacement at amount, Z23,Z24Hanging down at the vertical displacement and right rear-mounted quality respectively at the left back sprung mass of automobile Straight displacement；
Body vibrations accelerationThe dynamic deflection δ of suspensiondWith dynamic wheel load FdThree vibratory response amounts, their power spectrum The power spectral density spent with road surface input quantity is expressed as：
In formula：For frequency,As amplitude versus frequency characteGx(f) it is response quautity power spectrum Degree；For the power spectral density of road surface input quantity；
According to random vibration theory, response mean-square value is：
In formula：σxFor the standard deviation of vibratory response amount；
3) according to DOE experimental results, choose influences larger parameter as optimized variable to above-mentioned optimization aim, chooses steering and watches Take the gearratio n of motor1, torque sensor stiffness KsWith motor rotation inertia Jm, suspension rate C21, C22, C23, C24, suspension QS angular rigidity Ca, Ca2, suspension damping D21, D22, D23, D24With tire cornering stiffness k1, k2, k3, k4Become as design Amount.Assuming that vehicle fore suspension and rear suspension both sides have identical rigidity and damping, if front and rear tire both sides cornering stiffness is equal, to sum up institute State, the optimization design variable that integrated optimization system is chosen is：
X=[n1、Ks、Jm、k1、k3、C21、C23、D21、D23、Ca1、Ca2]； (11)
4) using the steering response, steering sensitivity and ride comfort of integrated system as optimization aim, with steering stability and turn Constraints is limited to sensitivity, and suspension dynamic deflection, integrated system Model for Multi-Objective Optimization is set up, its object function is represented For
In formula, f1(X) effective frequency range (0, ω of the road feel in information of road surface is represented0) frequency domain energy average value；ω0Represent In information of road surface ω is taken in the maximum frequency values of useful signal, optimization design0=40Hz, f2(X) represent sensitivity in information of road surface Effective frequency range (0, ω0) frequency domain energy average value, f3(X) ride comfort, w are represented0、wi、wj+4For weight, s0、si、sj+4 For scale factor；
5) first, steering needs to meet Routh criterions in integrated system, i.e.,：
Q6＞ 0；Q5＞ 0；a1＞ 0；b1＞ 0；c1＞ 0；d1＞ 0；Q0＞ 0
Secondly, it is desirable to which steering sensitivity energy is in certain scope, to ensure that driver obtains good steering sensitivity, i.e.,
0.0008≤f(x2)≤0.0099；
Finally, the dynamic deflection δ of suspensiondIt is required that within the specific limits, this optimal design-aside is:
0.06≤δd≤0.1；
6) above-mentioned parameter is optimized using improved cell membrane fusion algorithm, show that optimal Pareto is solved according to optimum results Collection, and choose optimal solution of compromising；
7) each parameter setting is taken off as optimum results using optimal compromise, completes optimization.
2. according to claim 1 in the automobile chassis system integration Multipurpose Optimal Method for improving cell membrane optimized algorithm, Characterized in that, step 6) the improved cell membrane optimized algorithm comprises the following steps that：
Step A), randomly generate the primary population P that population scale is N0；
Step B), population is divided：
Population is ranked up using TOPSIS methods, the material of Ps ratios is liposoluble substance before taking, and comes all works below For non-fat-soluble material；Further according to material concentration height, non-fat-soluble material is divided into two classes：High concentration non-fat-soluble material With low concentration non-fat-soluble material；
For something Y, the concentration residing for it is defined as the percentage that the interior material number included of its contiguous range accounts for total material number Than：
Wherein, Con is concentration around the material, and m is the quantity of total material, and n represents XiIn be less than r × (u-l) to Y distance Number, wherein, r is the radius factor that substance for calculation concentration is used, and u is the upper dividing value of solution space, and l is the floor value of solution space；
The average value MeanCon of all concentration is：
If not the concentration residing for liposoluble substance is more than MeanCon, then the material is divided into high concentration non-fat-soluble material, Otherwise it is divided into low concentration non-fat-soluble material；
Step C), liposoluble substance diffusion：
For each liposoluble substanceCentered on the material, raFor in the region of search of radius, new K is generated at random1 Individual materialAnd it is rightPart beyond hunting zone is modified；
Then search radius vector carries out contraction rnew a=ra×u； (15)
Search radius vector r during beginningaComputational methods are：
Wherein,R ∈ [0,1], b=3, T set maximum algebraically for evolution, and t is optimizes current algebraically, and t's is first Initial value is 0；When t is smaller, S (t) ≈ 1, search radius is relatively large, when t is larger, S (t) ≈ 0, and search radius is relatively It is small, accelerate search speed；
The modification method of hunting zone is：For any K, ifThen makeIfThen makeWherein, ukFor the upper dividing value of kth dimension of solution space, lkFloor value is tieed up for the kth of solution space；
Step D), the high concentration non-fat-soluble motion of matter：
The probability for making each high concentration non-fat-soluble material there is carrier is P1If, rand [0,1]≤P1, then the materialCan be with Diffusion is assisted, from high concentration lateral movement to low concentration side, and makes new position be Local Search centerOtherwise it is office to make in situ Portion's search center；
Then, the material can carry out the Local Search motion of K times, before this, it is necessary to initialize search radius vector
After search, K is recorded2Individual novel substance
Step E), the low concentration non-fat-soluble motion of matter：
The probability for making each low concentration non-fat-soluble material there is carrier is P2；
The energy of low concentration non-fat-soluble material is made to be in [0,1], and by linear distribution；
The functional value of each low concentration non-fat-soluble material is calculated firstThese functional values are arranged from small to large again Sequence；
Its ENERGY E of the material of functional value minimumiFor Emin, its ENERGY E of the material of functional value maximumiFor Emax, the energy of other materials EiBetween EminWith EmaxBetween, and calculated according to the order linear of sequence, wherein, EminWith EmaxFor the constant in [0,1], at this In EminIt is taken as 0, EmaxIt is taken as 1；
If somethingThere is carrier and enough energy, then it can carry out active transport, from low concentration side to highly concentrated Side is spent, and makes new position be Local Search centerOtherwise it is Local Search center to make in situ；
Low concentration non-fat-soluble materialNew position after active transport
Then, search radius is initialized
Then, withCentered on and with rcFor in the region of search of radius, random motion K3It is secondary, and their scope is carried out Amendment；
Record K3Individual novel substance
Step F), generation reference point, reference line：
According to the quantity M of the optimization aim and search hop count g artificially divided, the equally distributed reference in optimization space is determined The quantity H of point
The reference point of formation is connected with optimizing the initial point (may be configured as origin) in space, connected line segment is referred to as reference line；
Step G), generate new population：
A new generation's individual produced by liposoluble substance, high concentration non-fat-soluble material, low concentration non-fat-soluble material is calculated, if Number of individuals is less than or equal to N, then all individuals are designated as P as next filial generationt+1；Otherwise need to screen filial generation, root Operated according to the microhabitat of genetic algorithm, choose individual and enter population of future generation, its screening mode is：According to non-dominated ranking, The high non-dominant layer individual of grade is preferentially entered, until population quantity for the first time up to or over N (assuming that now dominating the number of plies L), just to reach N, then all individuals of above non-dominant layer enter population of new generation, and otherwise first l-1 layers fully enters newly Population, l layers need to screen again, are ranked up according to the vertical range of l layers of individual and reference line, while comparing and l layers of individual Whether associated reference line is further associated with other preceding l-1 layers of individuals, if so, being compared again, when l layers of individual and ginseng Examine line at a distance of closer to, then remain into population of future generation, next l layer individual otherwise will be taken in order, it is of future generation individual until making Quantity reaches N, if compromised in l layers of search less than enough individuals, by away from reference line closer to individual income of future generation plant Group, makes population quantity reach N；
Step H), judge whether optimization terminates：
If t<T, then Jia 1 to t progress and operate, and correct search radius ra、rb、rc, return to step B) and enter iteration, if t=T, Then disaggregation is ranked up, Pareto disaggregation is formed, optimization solution is provided, optimization terminates.
CN201710267334.9A 2017-04-21 2017-04-21 Automobile chassis system integration multi-objective optimization method based on improved cell membrane optimization algorithm Active CN107220405B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710267334.9A CN107220405B (en) 2017-04-21 2017-04-21 Automobile chassis system integration multi-objective optimization method based on improved cell membrane optimization algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710267334.9A CN107220405B (en) 2017-04-21 2017-04-21 Automobile chassis system integration multi-objective optimization method based on improved cell membrane optimization algorithm

Publications (2)

Publication Number Publication Date
CN107220405A true CN107220405A (en) 2017-09-29
CN107220405B CN107220405B (en) 2020-04-24

Family

ID=59944632

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710267334.9A Active CN107220405B (en) 2017-04-21 2017-04-21 Automobile chassis system integration multi-objective optimization method based on improved cell membrane optimization algorithm

Country Status (1)

Country Link
CN (1) CN107220405B (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109933886A (en) * 2019-03-11 2019-06-25 桂林电子科技大学 A kind of commercial-vehicle cab suspension arrangement optimization method
CN111553021A (en) * 2020-04-26 2020-08-18 贵州理工学院 Design method of active suspension system based on cascade disturbance observer
CN111709160A (en) * 2020-03-31 2020-09-25 桂林电子科技大学 Method and system for analyzing and optimizing driving dynamic performance based on truck chassis
CN111985044A (en) * 2019-05-23 2020-11-24 上海汽车集团股份有限公司 Method and device for analyzing rigidity of transverse stabilizer bar
CN112832848A (en) * 2021-03-05 2021-05-25 湖南科技大学 Construction method for preventing drilling and spraying holes in drilling construction process of extremely-soft coal seam
CN116859725A (en) * 2023-06-25 2023-10-10 盐城工学院 Genetic algorithm-based optimization method for automobile chassis control system

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101308386A (en) * 2008-07-10 2008-11-19 同济大学 Automobile chassis integrated controller hardware in-the-loop simulation test stand
CN104385873A (en) * 2014-09-24 2015-03-04 湖南大学 Multi-objective optimization method of car suspension system
US20160046166A1 (en) * 2012-06-25 2016-02-18 Ford Global Technologies, Llc Ride performance optimization in an active suspension system
CN105718607A (en) * 2014-12-02 2016-06-29 广州汽车集团股份有限公司 Suspension hard point optimization method based on K and C characteristics

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101308386A (en) * 2008-07-10 2008-11-19 同济大学 Automobile chassis integrated controller hardware in-the-loop simulation test stand
US20160046166A1 (en) * 2012-06-25 2016-02-18 Ford Global Technologies, Llc Ride performance optimization in an active suspension system
CN104385873A (en) * 2014-09-24 2015-03-04 湖南大学 Multi-objective optimization method of car suspension system
CN105718607A (en) * 2014-12-02 2016-06-29 广州汽车集团股份有限公司 Suspension hard point optimization method based on K and C characteristics

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109933886A (en) * 2019-03-11 2019-06-25 桂林电子科技大学 A kind of commercial-vehicle cab suspension arrangement optimization method
CN111985044A (en) * 2019-05-23 2020-11-24 上海汽车集团股份有限公司 Method and device for analyzing rigidity of transverse stabilizer bar
CN111985044B (en) * 2019-05-23 2023-08-04 上海汽车集团股份有限公司 Analysis method and device for rigidity of transverse stabilizer bar
CN111709160A (en) * 2020-03-31 2020-09-25 桂林电子科技大学 Method and system for analyzing and optimizing driving dynamic performance based on truck chassis
CN111709160B (en) * 2020-03-31 2023-08-11 桂林电子科技大学 Driving dynamic performance analysis optimization method and system based on truck chassis
CN111553021A (en) * 2020-04-26 2020-08-18 贵州理工学院 Design method of active suspension system based on cascade disturbance observer
CN112832848A (en) * 2021-03-05 2021-05-25 湖南科技大学 Construction method for preventing drilling and spraying holes in drilling construction process of extremely-soft coal seam
CN112832848B (en) * 2021-03-05 2022-05-20 湖南科技大学 Construction method for preventing drilling and spraying holes in drilling construction process of extremely-soft coal seam
CN116859725A (en) * 2023-06-25 2023-10-10 盐城工学院 Genetic algorithm-based optimization method for automobile chassis control system

Also Published As

Publication number Publication date
CN107220405B (en) 2020-04-24

Similar Documents

Publication Publication Date Title
CN107220405A (en) A kind of automobile chassis system integration Multipurpose Optimal Method based on improvement cell membrane optimized algorithm
CN104385873B (en) A kind of Multipurpose Optimal Method of automobile suspension system
CN101005981B (en) Control device for vehicle
CN106585709B (en) A kind of automobile chassis integrated system and its optimization method
DE102012216985A1 (en) Vehicle movement control device and suspension control device
CN106970524B (en) Design method of vehicle roll motion safety linear quadratic form optimal LQG controller for active suspension
CN102303602A (en) Coordination method and control device for smooth running and control stability of passenger car
Ariff et al. Optimal control strategy for low speed and high speed four-wheel-active steering vehicle
CN112373459B (en) Method for controlling upper-layer motion state of four-hub motor-driven vehicle
CN111391595A (en) Vehicle rollover prevention active tilt model prediction control method
CN112793560A (en) Unmanned vehicle safety and operation stability control method based on torque vector control
CN110164124B (en) Longitudinal following control method for vehicles in running of highway heavy truck fleet
CN206900467U (en) A kind of automobile chassis integrated system
Van Tan Preventing rollover phenomenon with an active anti-roll bar system using electro-hydraulic actuators: a full car model
Wang et al. Unsprung mass effects on electric vehicle dynamics based on coordinated control scheme
Gáspár et al. Robust Reconfgurable Control for In-wheel Electric Vehicles
Zulkarnain et al. Controller design for an active anti-roll bar system
jie Yue et al. Research on vehicle suspension systems based on fuzzy logic control
Omar et al. Evaluation of effect of in-wheel electric motors mass on the active suspension system performance using linear quadratic regulator control method
CN114056027A (en) Vehicle height and damping cooperative control method for air suspension
CN111814258B (en) Design method for transmission ratio of four-wheel independent electric drive vehicle steer-by-wire system
Yan et al. The design of anti-roll moment distribution for dual-channel active stabilizer bar system
Luo et al. Vehicle stability and attitude improvement through the coordinated control of longitudinal, lateral and vertical tyre forces for electric vehicles
Raksincharoensak et al. Robust vehicle handling dynamics of light-weight vehicles against variation in loading conditions
Jianhua et al. Coordinated control of AFS and ESP based on fuzzy logic method

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