CN104408265B - Vehicle running state and the method for tire magic formula parametric joint estimation - Google Patents
Vehicle running state and the method for tire magic formula parametric joint estimation Download PDFInfo
- Publication number
- CN104408265B CN104408265B CN201410764137.4A CN201410764137A CN104408265B CN 104408265 B CN104408265 B CN 104408265B CN 201410764137 A CN201410764137 A CN 201410764137A CN 104408265 B CN104408265 B CN 104408265B
- Authority
- CN
- China
- Prior art keywords
- particle
- mrow
- magic formula
- tire
- msub
- 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.)
- Expired - Fee Related
Links
Abstract
Patent of the present invention is related to the discrimination method of a kind of tire key parameter and vehicle running state, belongs to information science field.This method includes：Set up multiple degrees of freedom dynamics of vehicle equation, the side force determined in tire magic formula is substituted into wherein, it is using imperial lattice-storehouse tower method that dynamics of vehicle is equations turned for Markov Chain form, the Combined estimator of vehicle running state and magic formula parameter is realized by particle filter, introduce auxiliary variable, make it that the change of particle weights is more stable by secondary weighted operation, particle resampling is carried out by multinomial sampling method, Random Perturbation method is introduced during resampling, increase the diversity of particle while improving sample degeneracy, estimate while realizing tire magic formula parameter and vehicle running state, preferable estimated result can be obtained, further coefficient of road adhesion can be calculated by the parameter estimated, determine road surface characteristic, manipulated for following intelligent vehicle traveling and foundation is provided.
Description
Technical field
The invention belongs to the system mode in information technology and key parameter estimation field, more particularly to a kind of vehicle traveling
State and the method for tire magic formula parametric joint estimation.
Background technology
State estimation in vehicle traveling process, is the major issue in automobile dynamics, its object is to determine automobile
The important state variable such as longitudinal speed, yaw velocity, side slip angle, yaw angular displacement, is to realize vapour under transport condition
One of key technology of chassis active control.The longitudinal dynamics control of automobile, dependent on the accurate estimation to longitudinal speed；
The lateral dynamics control of automobile is dependent on the accurate estimation to yaw velocity or side slip angle.These states to be estimated
Variable, although equal available sensors direct measurement, but expensive special test equipment is had to rely on, and need specific peace
Fixed form is filled, is unsuitable for configuration on volume production car, is suitable only for the development experiments stage.Therefore, should from cost-effective and reality
From the perspective of, it is necessary to according to the sensor configuration on volume production car, by vehicle state estimation technology, accurate calculate obtains it
Needed for him but immesurable status information.
On the other hand, new structure, the development of new control system in automobile product, for describing automobile operation
There is uncertain factor in the mathematical modeling of dynamic behavior.Main uncertain factor comes from by tire and ground in model
Side force that is before nonlinear, being determined jointly by parameters such as side slip angle, vertical load.It is intended to determine side force of tire etc.
The tire model that the handling dynamics analysis of dynamics of vehicle behavior is applied primarily to is the magic formula that professor Pacejka proposes
Model：
FTA=DAsin(CAarctan(BAαA-EA(BAαA-arctan(BAαA))))
FTP=DPsin(CParctan(BPαP-EP(BPαP-arctan(BPαP))))
F in above formulaTA、FTPRepresent the side force of front and rear wheel, αA、αPRepresent the angle of heel of front and rear wheel, DA、CA、BA、EAPoint
Not Biao Shi front-wheel peak factor, form factor, the gradient factor, Curvature factor；DP、CP、BP、EPThe peak value of trailing wheel is represented respectively
The factor, form factor, the gradient factor, Curvature factor；Above-mentioned eight parameters are the key parameters for calculating side force of tire.
This is the semiempirical tire model based on test data, and the foundation of model depends on extensive data rather than tire
Structure is in itself, it is therefore desirable to which substantial amounts of test data constantly is corrected the parameters in magic formula, and finally determines this
A little parameters, can determine the dynamic behavior of vehicle, and this just greatly adds the cost for determining these parameters.
The algorithm for estimating more commonly used at present has Kalman filtering algorithm, Luenberger observer, Robust Observers, sliding formwork to see
Survey device, Fuzzy Observer and the nonlinear observer based on Leah Pu Nuofu theory deductions.
Common Kalman filter is merely able to carry out state estimation to linear Vehicle dynamics.EKF filter
(EKF) state estimation and its innovatory algorithm, can be carried out to the automobile operation kinetic model comprising non-linear factor, but will
Premised on sacrificing precision；Luenberger observer is to reach one kind side of state estimation purpose by the POLE PLACEMENT USING of system
Method, higher requirement is proposed to feedback matrix；The estimation effect of Robust Observers and synovial membrane observer is not ideal enough, therefore seeks
A kind of breakthrough system Gauss, linear constraint are looked for, is highly desirable to while ensure that the algorithm of degree of precision just seems.
Particle filter algorithm is by finding one group of random sample propagated in state space come approximate expression probability
Density function, integral operation is replaced with sample average, so that the process of the minimum variance estimate of system mode is obtained, to non-thread
Property, the state estimation of non-Gaussian filtering have preferable effect.
The domestic research in terms of vehicle state estimation problem is in the starting stage.That studies focuses on by Kalman
Yaw velocity, the speed of wave filter and its innovatory algorithm to estimate vehicle etc., and the simulation study stage is rested on mostly, due to
The limitation of Kalman filter itself, the precision and stability of estimation is also poor；And on vehicle key parameter especially tire
Key parameter identification problem research, current research is also less, and the connection of vehicle tyre parameter and vehicle running state
The research for closing estimation is even more important again for following intelligent driving technology.
The content of the invention
The present invention be directed to current functional need, reduce experiment quantity needed for the parameters for determining magic formula and
Data, a kind of method for proposing vehicle running state and the estimation of tire magic formula parametric joint, this method comprises the steps：
Step 1, multiple degrees of freedom dynamics of vehicle equation is set up, by Runge-Kutta methods, equation depression of order, general side is realized
Journey changes into markovian form；
Step 2, physical quantity, the prior information estimated as needed are initialized to particle, complete granular Weights Computing；
Step 3, by Auxiliary Particle Filter technology, by secondary weighted, make particle distribution more reasonable；
Step 4, stochastic pertUrbationtechnique is employed during resampling, to the particle superposition Random Perturbation of maximum weight,
While sample degeneracy is solved, increase particle diversity；
Step 5, realize peak factor D in vehicle running state and magic formula, Curvature factor E, form factor C, tiltedly
The estimation of factor B, and the magic formula parameter by estimating to obtain are spent, the estimation of vehicle running state is furthermore achieved that.
Advantages of the present invention effect is as follows：
The test data easily measured is obtained using the sensor of lower cost, using these data, can be realized simultaneously
The estimation of multiple key parameters of the magic formula of tire dynamics behavior described in driving vehicle is directed to, and then realizes vehicle
The estimation of transport condition and the indirect estimation of side force of tire, and by the change of side force of tire, surface conditions are grasped in real time,
Reference is provided for the further operation of following intelligent driving system.
Brief description of the drawings
Fig. 1 is the FB(flow block) of the inventive method.
Fig. 2 is the particle filter calculation process block diagram of the inventive method.
Embodiment
The inventive method, is described with reference to the accompanying drawings as follows：
1) multiple degrees of freedom dynamics of vehicle equation is set up, by Runge-Kutta methods, equation depression of order is realized, equation is turned
It is melted into markovian form；
2) physical quantity, the prior information estimated as needed are initialized to particle, complete granular Weights Computing；
3) by Auxiliary Particle Filter technology, by secondary weighted, make particle distribution more reasonable；
4) stochastic pertUrbationtechnique is employed during resampling, to the particle superposition Random Perturbation of maximum weight, in solution
Certainly while sample degeneracy, increase particle diversity.
5) realize peak factor D in vehicle running state and magic formula, Curvature factor E, form factor C, gradient because
Sub- B estimation, and the magic formula parameter by estimating to obtain, furthermore achieved that the estimation of vehicle running state.
6) dynamics of vehicle equation is set up, it is determined that needing the transport condition variable and key parameter estimated.
7) design factor k1=f (tn,yn)
8) design factor
9) design factor
10) design factor k4=f (tn+h,yn+hk3)
11) and using Runge Kutta method by equations turned into following iteration form：
12) h represents the time interval of algorithm in step 11, using fixed step algorithm, and multiple repetition test is needed herein, finds
Suitable step-length.
13) determination process noise and initial error covariance；
14) number of particles is determined, particle is initialized；
15) particle initial distribution is determined；
16) constraints of front and rear wheel peak factor D in magic formula is determined；
17) constraints of front and rear wheel Curvature factor E in magic formula is determined；
18) constraints of front and rear wheel form factor C in magic formula is determined；
19) constraints of front and rear wheel gradient factor B in magic formula is determined；
20) weight is calculated using measurement data；
21) by μk(i)~p (xk/xk-1(i) μ) is calculatedk(i), wherein i=1,2,3 ... num, num represent number of particles, k
At the moment where representing, pass through μk(i) introducing, by secondary weighted operation, makes the distribution of weight more reasonable.
22) priori conditions probability sampling x is passed throughk(i)~p (xk)；
23) formula is utilizedWeights are calculated, are normalized simultaneously.
24) normalized weight value；
25) resampling is realized using multinomial method for resampling, it is to avoid sample degeneracy；
26) N is judgedeff＜ N/3；
27) by particleAccording to weightRealize that descending is arranged；
28) keep effective particle overall number constant；
29) Random Perturbation re-sampling is carried out to failure particle, i.e.,
30) weights are assigned again to the particle that the particle after re-sampling is concentrated
31) tire parameter estimated is brought into state equation again, vehicle row is realized using Runge-Kutta methods
Sail the estimation of middle side slip angle, yaw velocity.
32) it is real using data such as estimated obtained tire magic formula parameter and side slip angle, yaw velocities
Existing front-wheel side force, the estimation of trailing wheel side force.
Claims (5)
1. a kind of vehicle running state and the method for tire magic formula parametric joint estimation, it is characterised in that including following steps
Suddenly：
1) set up multiple degrees of freedom dynamics of vehicle equation, by Runge-Kutta methods, realize equation depression of order, by it is equations turned into
Markovian form；
2) physical quantity, the prior information estimated as needed are initialized to particle, complete granular Weights Computing；
3) by Auxiliary Particle Filter technology, by secondary weighted, make particle distribution more reasonable；
4) stochastic pertUrbationtechnique is employed during resampling, to the particle superposition Random Perturbation of maximum weight, grain is being solved
While son is degenerated, increase particle diversity；
5) peak factor D in vehicle running state and magic formula, Curvature factor E, form factor C, gradient factor B are realized
Estimation, and the magic formula parameter by estimating to obtain, furthermore achieved that the estimation of vehicle running state.
2. vehicle running state according to claim 1 and the method for tire magic formula parametric joint estimation, its feature
It is in described step 1, sets up kinetics equation, and utilize Runge Kutta method by equations turned into markovian shape
Formula, specific method is as follows：
1) dynamics of vehicle equation is set up, it is determined that needing the transport condition variable and tire parameter estimated；
2) design factor k1=f (tn,yn)；
5) design factor k4=f (tn+h,yn+hk3)；
6) and using Runge Kutta method by equations turned into markovian form：
<mrow>
<msub>
<mi>y</mi>
<mrow>
<mi>n</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>=</mo>
<msub>
<mi>y</mi>
<mi>n</mi>
</msub>
<mo>+</mo>
<mfrac>
<mi>h</mi>
<mn>6</mn>
</mfrac>
<mrow>
<mo>(</mo>
<msub>
<mi>k</mi>
<mn>1</mn>
</msub>
<mo>+</mo>
<mn>2</mn>
<msub>
<mi>k</mi>
<mn>2</mn>
</msub>
<mo>+</mo>
<mn>2</mn>
<msub>
<mi>k</mi>
<mn>3</mn>
</msub>
<mo>+</mo>
<msub>
<mi>k</mi>
<mn>4</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
Wherein h represents the time interval of algorithm, and using fixed step algorithm, i.e. h is determined, state equation is converted into iteration form.
3. vehicle running state according to claim 1 and the method for tire magic formula parametric joint estimation, its feature
It is in described step 2, it is determined that the physical quantity of estimation is realized, to parameters according to its actual physical significance to particle
After initialization, specific method is as follows：
1) determination process noise and initial error covariance；
2) number of particles is determined, particle is initialized；
3) particle initial distribution is determined；
4) constraints of front and rear wheel peak factor D in magic formula is determined；
5) constraints of front and rear wheel Curvature factor E in magic formula is determined；
6) constraints of front and rear wheel form factor C in magic formula is determined；
7) constraints of front and rear wheel gradient factor B in magic formula is determined；
8) weight is calculated using measurement data；
9) normalized weight value；
10) resampling is realized using multinomial method for resampling, it is to avoid sample degeneracy；
11) tire parameter estimated is brought into state equation again, realized using Runge-Kutta methods in vehicle traveling
The estimation of side slip angle, yaw velocity；
12) before being realized using data such as estimated obtained tire magic formula parameter and side slip angle, yaw velocities
Take turns side force, the estimation of trailing wheel side force.
4. the method for the vehicle running state and the estimation of tire magic formula parametric joint according to claim 1, it is special
Levy and be in described step 3, by secondary weighted operation, make the distribution of weight more reasonable, specific method is as follows：
1) by μk(i)~p (xk/xk-1(i) μ) is calculatedk(i), wherein i=1,2,3 ... num, num represent number of particles, and k represents institute
At the moment, pass through μk(i) introducing, by secondary weighted operation, makes the distribution of weight more reasonable；
2) priori conditions probability sampling x is passed throughk(i)~p (xk)；
3) formula is utilizedWeights are calculated, are normalized simultaneously.
5. vehicle running state according to claim 1 and the method for tire magic formula parametric joint estimation, its feature
It is in described step 4, specific method is as follows：
1) N is judgedeff＜ N/3, NeffFor number of effective particles, N is total number of particles；
2) by particle xk(i) according to weight wk(i) realize that descending is arranged；
3) keep effective particle overall number constant；
4) Random Perturbation re-sampling is carried out to failure particle；
<mrow>
<msubsup>
<mover>
<mi>x</mi>
<mo>^</mo>
</mover>
<mi>k</mi>
<mrow>
<mo>&lsqb;</mo>
<mrow>
<mo>(</mo>
<mi>N</mi>
<mi>e</mi>
<mi>f</mi>
<mi>f</mi>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>:</mo>
<mi>N</mi>
<mo>&rsqb;</mo>
</mrow>
</msubsup>
<mo>=</mo>
<msubsup>
<mover>
<mi>x</mi>
<mo>^</mo>
</mover>
<mi>k</mi>
<mi>i</mi>
</msubsup>
<mo>+</mo>
<mi>r</mi>
<mi>a</mi>
<mi>n</mi>
<mi>d</mi>
<mi>n</mi>
</mrow>
5) weights are assigned again to the particle that the particle after re-sampling is concentrated
<mrow>
<msubsup>
<mi>w</mi>
<mi>k</mi>
<mrow>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>:</mo>
<mi>N</mi>
<mo>&rsqb;</mo>
</mrow>
</msubsup>
<mo>=</mo>
<mn>1</mn>
<mo>/</mo>
<mi>N</mi>
<mo>.</mo>
</mrow>
2
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410764137.4A CN104408265B (en) | 2014-12-10 | 2014-12-10 | Vehicle running state and the method for tire magic formula parametric joint estimation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410764137.4A CN104408265B (en) | 2014-12-10 | 2014-12-10 | Vehicle running state and the method for tire magic formula parametric joint estimation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104408265A CN104408265A (en) | 2015-03-11 |
| CN104408265B true CN104408265B (en) | 2017-10-31 |
Family
ID=52645896
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410764137.4A Expired - Fee Related CN104408265B (en) | 2014-12-10 | 2014-12-10 | Vehicle running state and the method for tire magic formula parametric joint estimation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104408265B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104809292B (en) * | 2015-04-28 | 2017-11-28 | 西安理工大学 | A kind of on-line identification method of bullet train non-linear dynamic model parameter |
| CN106203684A (en) * | 2016-06-29 | 2016-12-07 | 长安大学 | A kind of parameter identification for tire magic formula and optimization method |
| CN110884499B (en) * | 2019-12-19 | 2021-03-19 | 北京理工大学 | Method and system for determining vehicle mass center slip angle |
| CN111241692B (en) * | 2020-01-16 | 2022-11-08 | 南京航空航天大学 | Parameter identification method for tire magic formula |
| CN112784355A (en) * | 2020-12-21 | 2021-05-11 | 吉林大学 | Fourteen-degree-of-freedom vehicle dynamics model modeling method based on multi-body dynamics |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103310044A (en) * | 2013-05-27 | 2013-09-18 | 上海工程技术大学 | Railway vehicle suspension system parameter estimation method based on improved particle filtering algorithm |
| CN104182991A (en) * | 2014-08-15 | 2014-12-03 | 辽宁工业大学 | Vehicle running state estimation method and vehicle running state estimation device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102202949B (en) * | 2008-10-29 | 2014-11-26 | 日产自动车株式会社 | Device and method for estimating frictional condition of ground surface with which vehicle is in contact |
-
2014
- 2014-12-10 CN CN201410764137.4A patent/CN104408265B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103310044A (en) * | 2013-05-27 | 2013-09-18 | 上海工程技术大学 | Railway vehicle suspension system parameter estimation method based on improved particle filtering algorithm |
| CN104182991A (en) * | 2014-08-15 | 2014-12-03 | 辽宁工业大学 | Vehicle running state estimation method and vehicle running state estimation device |
Non-Patent Citations (2)
| Title |
|---|
| Optimal SIR algorithm vs. fully adapted auxiliary particle filter: a non asymptotic analysis;Yohan Petetin· François Desbouvries.;《Stat Comput(2013)》;20120811;第759-775页 * |
| 基于粒子滤波算法的汽车状态估计技术;林棻等;《农业机械学报》;20110228;第42卷(第2期);第22-27页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104408265A (en) | 2015-03-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104408265B (en) | Vehicle running state and the method for tire magic formula parametric joint estimation | |
| Singh et al. | Literature review and fundamental approaches for vehicle and tire state estimation | |
| CN108482379B (en) | Wheel-hub motor driven vehicle coefficient of road adhesion and the synchronous real-time estimation system and method for road gradient | |
| Zhao et al. | Design of a nonlinear observer for vehicle velocity estimation and experiments | |
| US6508102B1 (en) | Near real-time friction estimation for pre-emptive vehicle control | |
| CN104354700B (en) | A kind of vehicle parameter On-line Estimation method based on Unscented kalman filtering | |
| CN103909933B (en) | A kind of front wheel side of distributed electro-motive vehicle is to force evaluating method | |
| CN104590276B (en) | Recognition method for rotational inertia around z axis and tire cornering stiffness of automobile | |
| US20090177346A1 (en) | Dynamic estimation of vehicle inertial parameters and tire forces from tire sensors | |
| CN110884499B (en) | Method and system for determining vehicle mass center slip angle | |
| US20190263421A1 (en) | Determining driving state variables | |
| CN104077459A (en) | Automobile chassis operation stability analysis model based on suspension K and C characteristics | |
| CN103754218B (en) | Coefficient of road adhesion method of estimation under a kind of motor tire lateral deviation operating mode | |
| CN109606378A (en) | Vehicle running state estimation method towards non-Gaussian noise environment | |
| CN103279675B (en) | Tire-road attachment coefficient and the method for estimation of slip angle of tire | |
| CN111547059A (en) | Distributed driving electric automobile inertia parameter estimation method | |
| Cheng et al. | A new nonlinear observer using unscented Kalman filter to estimate sideslip angle, lateral tire road forces and tire road friction coefficient | |
| Chen et al. | Longitudinal-lateral-cooperative estimation algorithm for vehicle dynamics states based on adaptive-square-root-cubature-Kalman-filter and similarity-principle | |
| CN111959516B (en) | Method for jointly estimating vehicle state and road adhesion coefficient | |
| CN102980580A (en) | No-gyro satellite gesture determination method based on tensor product multi-cell robust heavy hydrogen (H2) filtering | |
| Ichalal et al. | Observer design for motorcycle lean and steering dynamics estimation: A Takagi-Sugeno approach | |
| CN117272525A (en) | Intelligent electric automobile road adhesion coefficient estimation method | |
| CN106671985A (en) | Electric vehicle dynamics system modeling method | |
| Li et al. | Vehicle velocity estimation for real-time dynamic stability control | |
| Slimi et al. | Motorcycle lateral dynamic estimation and lateral tire-road forces reconstruction using sliding mode observer |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171031 Termination date: 20211210 |