CN108284841A - A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation - Google Patents

A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation Download PDF

Info

Publication number
CN108284841A
CN108284841A CN201711304213.3A CN201711304213A CN108284841A CN 108284841 A CN108284841 A CN 108284841A CN 201711304213 A CN201711304213 A CN 201711304213A CN 108284841 A CN108284841 A CN 108284841A
Authority
CN
China
Prior art keywords
observer
estimation
vehicle
kalman filter
formula
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201711304213.3A
Other languages
Chinese (zh)
Inventor
陈特
陈龙
徐兴
江浩斌
蔡英凤
江昕炜
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.)
Jiangsu University
Original Assignee
Jiangsu University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiangsu University filed Critical Jiangsu University
Priority to CN201711304213.3A priority Critical patent/CN108284841A/en
Publication of CN108284841A publication Critical patent/CN108284841A/en
Pending legal-status Critical Current

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/10—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to vehicle motion
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/10—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to vehicle motion
    • B60W40/105—Speed
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/10—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to vehicle motion
    • B60W40/112—Roll movement
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W2050/0001—Details of the control system
    • B60W2050/0043—Signal treatments, identification of variables or parameters, parameter estimation or state estimation
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W2050/0001—Details of the control system
    • B60W2050/0043—Signal treatments, identification of variables or parameters, parameter estimation or state estimation
    • B60W2050/0052—Filtering, filters
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2520/00—Input parameters relating to overall vehicle dynamics
    • B60W2520/14—Yaw
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W2720/00—Output or target parameters relating to overall vehicle dynamics
    • B60W2720/10—Longitudinal speed

Abstract

A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation of the present invention, includes the following steps：Step S1：Distributed-driving electric automobile vehicle models；Step S2：Longitudinal force Design of Observer based on nonlinear observer and Extended Kalman filter；Step S3：Vehicle running state adaptive iteration method of estimation designs, step S4：Longitudinal force observer and transport condition adaptive iteration method of estimation experimental verification.Unknown worm is reconstructed the present invention is based on the mode that nonlinear observer and Kalman filter combine and designs longitudinal force observer, and longitudinal force estimation is realized by input quantity of inexpensive sensor information.In addition, based on longitudinal force estimated information, vehicle running state adaptive iteration method of estimation in the present invention, estimator is tested before design vehicle transport condition, in conjunction with preceding testing estimated result, further more accurate Posterior estimator is obtained based on Extended Kalman filter, helps to improve estimated accuracy, anti-interference ability and the adaptivity of estimator.

Description

A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation
Technical field
The invention belongs to electric vehicle research fields, and in particular to a kind of distributed-driving electric automobile transport condition is adaptive Answer iterative estimate method.
Background technology
Distributed-driving electric automobile relies on its degree of freedom in Full Vehicle Dynamics control and good energy-saving potential, It is one of the Main way of automobile future development generally acknowledged at present, to obtain the concern of many researchers in the industry, wherein Distributed-driving electric automobile transport condition estimation problem is an important research topic.The common algorithm of vehicle state estimation Including Kalman filter, sliding mode observer, nonlinear observer and Robust Observers etc., wherein Kalman filter and its improvement is calculated Method is applied the most extensive.With going deep into for research, researcher starts Kalman filtering and other estimation theories being combined, By the mutual iteration between model or observer, estimated accuracy is improved using the redundancy of Given information.If can combine Distributed-driving electric automobile drives feature to carry out the research of transport condition estimation with intrinsic advantage, while considering the practical row of vehicle The interference that some complex environment factors are brought when sailing, can further increase the precision of estimation.In addition, in recent years, intelligent vehicle And unmanned research obtained the concern of many domestic scholars, wherein traffic environment perception and important vehicle state estimation It is an important ring for Vehicular intelligent control.Based on considerations above, it is necessary to carry out the research of vehicle running state estimation.
Invention content
The purpose of the present invention is provide a kind of distributed-driving electric automobile transport condition regarding to the issue above adaptively to change For method of estimation.Include the following steps：Step S1：Distributed-driving electric automobile vehicle models；Step S2：Based on nonlinear riew Survey the longitudinal force Design of Observer of device and Extended Kalman filter；Step S3：Vehicle running state adaptive iteration method of estimation Design；Step S4：Longitudinal force observer and transport condition adaptive iteration method of estimation experimental verification.The present invention is based on non-linear The mode that observer and Kalman filter combine reconstructs Unknown worm and designs longitudinal force observer, is believed with inexpensive sensor Breath is that input quantity realizes longitudinal force estimation.In addition, being based on longitudinal force estimated information, the present invention devises a kind of distributed driving electricity Electrical automobile transport condition adaptive iteration method of estimation, help to improve the estimated accuracy of estimator, anti-interference ability and from Adaptability.
The technical scheme is that：A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation, Include the following steps：
Step S1：Distributed-driving electric automobile vehicle models, including vehicle Nonlinear dynamic models, electric driving wheel are built Mould and tire modeling；
Step S2：Longitudinal force Design of Observer based on nonlinear observer and Extended Kalman filter, first with electricity The state estimation equation of driving wheel Construction of A Model longitudinal force, to Nonlinear Observer Design, by choosing suitable gain square Battle array so that observer meets stability condition, in addition, nonlinear observer and Extended Kalman filter are combined, reduces Influence of the noise to estimation；
Step S3：Vehicle running state adaptive iteration method of estimation designs, and is directed to yaw rate respectively first Design Justin Lemberg observer with lateral speed, estimation tested before obtaining yaw velocity and lateral speed, test before this estimation be In the case of not needing longitudinal acceleration of the vehicle and transverse acceleration, estimation is tested before described using indulging estimated by the step S2 Be calculated to force information and by the lateral force that tire model obtains, thus comparatively estimated result be easy to be affected and There are certain errors；Further according to vehicle non-linear dynamic model, vehicle running state is carried out based on Extended Kalman filter Estimation, obtains the Posterior estimator of vehicle running state, i.e. Kalman Filter Estimation, wherein before test estimated information and be used as expanding Open up Kalman filtering measurement updaue, obtained Posterior estimator simultaneously be used as before test estimate designed by Justin Lemberg observer puppet Sensor input improves estimation to realize the iterative estimate of car status information by the iterative compensation of redundancy The anti-interference ability of precision and estimated information；In addition, fuzzy rule has also been devised in the step, according to when driving speed and front-wheel Corner tests estimation weight shared in Extended Kalman filter measurement updaue before dynamically adjusting, so that vehicle-state is estimated The iterative process of meter has the adaptivity based on vehicle driving-cycle；
Step 4：Longitudinal force observer and transport condition adaptive iteration method of estimation experimental verification.
In said program, the kinetics equation of vehicle non-linear dynamic model is in the step S1：
Wherein, vyFor lateral speed, γ is yaw velocity, and β is vehicle centroid side drift angle, vxFor longitudinal speed,For vehicle centroid speed and in view of lateral speed is smaller, v is equal to v in one c of formulax, m is car mass, and δ is Front wheel angle, IzFor around the rotary inertia of z-axis, lfFor distance of the barycenter away from front axle, lrFor distance of the barycenter away from rear axle, bfIt is preceding The 1/2, b of wheelspanrIt is the 1/2, F of rear treadxjAnd Fyj(j=1,2,3,4) be respectively number it is longitudinal force corresponding to the tire of j And lateral force.
In said program, the spin dynamics equation of single wheel is in the step S1 electric driving wheels modeling：
In formula two, ωjFor longitudinal force FxjThe rotating speed of corresponding wheel；J1For vehicle wheel rotation inertia；R is wheel effectively half Diameter；TLjTo be installed on the loading moment of wheel inner wheel hub motor；
Torque balance equation on wheel hub motor output shaft is：
The dynamic electric voltage equilibrium equation of wheel hub motor equivalent circuit is：
In formula three, four, J2For the rotary inertia of rotor；B is damped coefficient；KtFor motor torque constant；ijFor line electricity Stream；ujFor line voltage；R is the equivalent line resistance of winding；L is winding equivalent inductance；KaFor back EMF coefficient.
In said program, the step S1 tires modeling includes the following steps：
Tire model using semiempirical magic formula estimates side force of tire formula is：
Fy=Dsin { Carctan [B α-E (B α-arctan (B α))] } formula five
In formula five, B is stiffness factor, and C is the curve shape factor, and D is peak factor, and E is the curvature of curve factor, and α is vehicle Take turns side drift angle；
Tire model parameter B, C, D, E are related to the vertical load of tire, and the vertical load of each tire is：
In formula six, Fz1、Fz2、Fz3、Fz4, for the vertical load of corresponding tire, h is height of center of mass, and g is acceleration of gravity.
Each slip angle of tire is：
In formula seven, α1、α2、α3、α4For the side drift angle of corresponding tire.
In said program, the longitudinal force state estimation equation that electric driving wheel Construction of A Model described in the step S2 goes out can table It is shown as：
14 b of y=Cx+Fv formula
Wherein, x, u, d, y, respectively system state amount, it is known that input, Unknown worm and measured value, w and v are mutual not phase The zero-mean white noise sequence of pass, and have
Since there are Unknown worms for formula 14, by constructing Unknown worm state equation, and Nonlinear Observer Design Carry out longitudinal force estimation：
Wherein,For system state estimation,For Unknown inputs, KpFor the gain matrix of longitudinal force observer, KiFor The integration matrix of Unknown inputs；
Formula 15 is further converted to：
Definition status evaluated error：
To derive：
Wherein,Choose rational gain matrix KpAnd Ki, make matrix AePOLE PLACEMENT USING is to coordinate It is Left half-plane.
In said program, the step S3 specifically includes following steps：
It is directed to yaw rate and lateral speed design Justin Lemberg observer 1 and Justin Lemberg observer 2, γ respectivelyLO And vyLORespectively use the elder generation of the yaw velocity and lateral speed of Justin Lemberg observer 1 and the estimation gained of Justin Lemberg observer 2 Test estimation, KγAnd KvyThe suitable observer respectively chosen in 2 design process of Justin Lemberg observer 1 and Justin Lemberg observer increases Benefit, vyEKF, γEKFAnd βEKFDivide vyLOIt is divided into and uses the lateral speed obtained estimated by extended Kalman filter, yaw velocity And side slip angle, wherein vyEKFAnd γEKFIt is considered as posterior estimate, λ is the fuzzy weighted values coefficient of fuzzy controller output, vyMAnd γMInput is measured to be input to the pseudo- of extended Kalman filter after FUZZY WEIGHTED,
According to formula one, vehicle non-linear dynamic model is expressed as：
Kalman filter is designed using vehicle non-linear dynamic model, input vector is：
Wherein, uvIt is the input vector collection of whole vehicle model, uv1-uv10The respectively number of each subvector of input vector collection；
The quantity of state of whole vehicle model is：
xv(t)=[xv1 xv2 xv3]T=[vyEKF γEKF βEKF]TFormula 21
Wherein, xvIt is the state vector collection of whole vehicle model, xv1-xv3The respectively number of each subvector of state vector collection；
The measurement vector of whole vehicle model is：
yv=[vyM γM]T=[yv1 yv2]TFormula 22
Wherein, yvIt is the measurement vector set of whole vehicle model, yv1And yv2The number of each subvector of vector set is respectively measured, vγMAnd γMIt is considered as the pseudo- measuring value of lateral speed and yaw velocity, the puppet measuring value is estimated by Extended Kalman filter Meter and Justin Lemberg observer estimate Weighted Fusion and obtain that fuzzy weighted values coefficient therein is travelled according to vehicle driving-cycle and vehicle Degree of stability dynamic regulation puppet measuring value, be expressed as to the state equation based on whole vehicle model：
And it measures equation and is：
Wherein, priori estimates vyMAnd γMIt is considered as the pseudo- measuring value of Extended Kalman filter here as known quantity.
In said program, the design of the Justin Lemberg observer 1 and Justin Lemberg observer 2 includes the following steps：
According to one b of formula, prior estimate of the following Justin Lemberg observer 1 for yaw velocity is designed：
Wherein, γLOFor the yaw velocity that Justin Lemberg observer 1 is estimated, γEKFEstimate for extended Kalman filter The yaw velocity arrived, KγFor the observer feedback oscillator of Justin Lemberg observer 1；
According to the longitudinal force Design of Observer method in step S2, longitudinal force observer is separately designed for electric driving wheel, Gained longitudinal force estimated value is used for the input quantity of Justin Lemberg observer, meanwhile, the side force of tire and biography obtained by tire model The collected front wheel angle of the sensor also input quantity as Justin Lemberg observer 1, and by the yaw angle of 1 gained of Justin Lemberg observer Speed prior estimate is γLO；
According to one a of formula, design Justin Lemberg observer 2 is used for the prior estimate of lateral speed：
Wherein, vyLOFor the lateral speed that Justin Lemberg observer 2 is estimated, vuEKFIt is estimated for extended Kalman filter Lateral speed, KvyFor the observer feedback oscillator of Justin Lemberg observer 2；Longitudinal force estimated value, lateral tire force and preceding rotation Input quantity of the angle as Justin Lemberg observer 2, meanwhile, longitudinal speed and the yaw angle for estimating to obtain by Justin Lemberg observer 1 The speed prior estimate also input quantity as Justin Lemberg observer 2, by the lateral speed prior estimate of 2 gained of Justin Lemberg observer It is denoted as vyLO, for Justin Lemberg observer 1 and Justin Lemberg observer 2, based on the yaw angle speed obtained by Extended Kalman filter The known pseudo- measurement input value of degree and the Posterior estimator of lateral speed as Justin Lemberg observer 1 and Justin Lemberg observer 2.
In said program, in the design of the Justin Lemberg observer 1 and Justin Lemberg observer 12, the step S2 is utilized Middle institute's Extended Kalman filter method designs the Kalman filter based on vehicle non-linear dynamic model, realizes vehicle traveling The further unbiased esti-mator of state；In extended Kalman filter, prior estimate γLOAnd vyLOAs Extended Kalman filter Pseudo- sensor collection value.
In said program, fuzzy weighted values are devised in fuzzy rule described in the step S3 for automatic adjusument priori Estimate γLOAnd vyLOConfidence level in Extended Kalman filter measurement updaue,
It is based on the extension measurement updaue amalgamation mode obtained by fuzzy weighted values coefficient：
Wherein, λ is the fuzzy weighted values coefficient that designed fuzzy controller inputs, and the fuzzy controller input quantity is vertical To speed and front wheel angle, export as real-time fuzzy weighted values coefficient.
Compared with prior art, the beneficial effects of the invention are as follows：
(1) present invention proposes the longitudinal force method of estimation suitable for In-wheel motor driving electric vehicle, based on non-linear The mode that observer and Kalman filter combine reconstructs Unknown worm and designs longitudinal force observer, is believed with inexpensive sensor Breath is that input quantity realizes that longitudinal force estimation, the method realize longitudinal force estimation using the inexpensive sensor such as electric current and rotating speed, The advantage for taking full advantage of distributed-driving electric automobile, reduces estimated cost.
(2) the present invention is based on longitudinal force estimated informations, it is proposed that a kind of vehicle running state adaptive iteration method of estimation, Originally estimator is tested before design vehicle transport condition first, yaw velocity and lateral vehicle has been obtained using less sensor information Speed estimation, in conjunction with estimated result is preceding tested, obtains further more accurate Posterior estimator, and devise based on Extended Kalman filter Fuzzy rule comes before adjusting according to vehicle driving-cycle dynamic to test the power of estimation and Posterior estimator in vehicle-state iterative estimate Weight, helps to improve estimated accuracy, anti-interference ability and the adaptivity of estimator.This method is filled using less sensor It sets and realizes vehicle running state estimation, and improve the precision of estimation by way of information fusion, devise fuzzy weighted values Coefficient improves the adaptive ability of estimator.
Description of the drawings
Fig. 1 is whole vehicle running state adaptive iteration method of estimation schematic diagram；
Fig. 2 is vertical relations device estimation effect proof diagram；
Fig. 3,4,5 are using lateral speed, yaw velocity and the barycenter side obtained by adaptive iteration method of estimation respectively The estimation effect proof diagram of drift angle.
Specific implementation mode
Invention is further described in detail with reference to the accompanying drawings and detailed description, but protection scope of the present invention It is not limited to this.
A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation of the present invention, specific method Flow chart is as shown in Figure 1, include the following steps：
Step S1：Distributed-driving electric automobile vehicle models.
(a) vehicle Nonlinear dynamic models
The moving coordinate system xoy origins being fixed on automobile are overlapped with automobile barycenter, x-axis be automobile longitudinal symmetry axis, it is specified that It is just forward；Y-axis is by automobile barycenter, it is specified that being just to the left；Angle and torque in all coordinate planes is with side counterclockwise Xiang Weizheng, the component of all vectors is to be just in the same direction with reference axis.Ignore suspension and automobile catenary motion, ignores automobile around y The pitching movement of axis and roll motion around x-axis, it is believed that the mechanical property of each tire of automobile is identical.Wheel 1,2,3,4 is distinguished It is corresponding it is left front, right before, left back, right rear wheel.The kinetics equation of vehicle non-linear dynamic model is
Wherein, vyFor lateral speed, γ is yaw velocity, and β is vehicle centroid side drift angle, vxFor longitudinal speed,For vehicle centroid speed and in view of lateral speed is smaller, v is approximately equal to v in one c of formulax, m is automobile matter Amount, δ is front wheel angle, IzFor around the rotary inertia of z-axis, lfFor distance of the barycenter away from front axle, lrFor distance of the barycenter away from rear axle, bfIt is the 1/2, b of front treadrIt is the 1/2, F of rear treadxjAnd Fyj(j=1,2,3,4) respectively number is corresponding to the tire of j Longitudinal force and lateral force.
(b) electric driving wheel models
The each wheel of four motorized wheels electric vehicle is respectively operated alone by a wheel hub motor, by wheel hub motor and wheel The driving wheel of composition is an independent driving unit, and driving wheel model is as shown in Figure 2.The spin dynamics equation of single wheel For
In formula, ωjFor longitudinal force FxjThe rotating speed of corresponding wheel；J1For vehicle wheel rotation inertia；R is wheel effective radius； TLjTo be installed on the loading moment of wheel inner wheel hub motor.Torque on wheel hub motor output shaft
Equilibrium equation is
The dynamic electric voltage equilibrium equation of wheel hub motor equivalent circuit is
In formula three, four, J2For the rotary inertia of rotor；B is damped coefficient；KtFor motor torque constant；ijFor line electricity Stream；ujFor line voltage；R is the equivalent line resistance of winding；L is winding equivalent inductance；KaFor back EMF coefficient.
(c) tire models
Tire model using semiempirical magic formula estimates side force of tire formula is
Fy=Dsin { Carctan [B α-E (B α-arctan (B α))] } formula five
In formula, B is stiffness factor, and C is the curve shape factor, and D is peak factor, and E is the curvature of curve factor, and α is wheel Side drift angle.
Tire model parameter B, C, D, E are related to the vertical load of tire, and the vertical load of each tire is
In formula, Fz1、Fz2、Fz3、Fz4, for the vertical load of corresponding tire, h is height of center of mass, and g is acceleration of gravity.
Each slip angle of tire is
In formula, α1、α2、α3、α4For the side drift angle of corresponding tire.
Step 2：Longitudinal force Design of Observer based on nonlinear observer and Extended Kalman filter；
Expanded Kalman filtration algorithm can be divided into prediction process and correction course two parts.Prediction process according to it is current when The system mode at quarter obtains the predictive estimation to subsequent time；Correction course is by observed result and predictive estimation combination acquisition system Optimal estimation.Expanded Kalman filtration algorithm step is
(a) process is predicted
CALCULATING PREDICTION value
CALCULATING PREDICTION varivance matrix
Wherein,
(b) correction course
Calculate gain matrix
Calculate filter value
Calculate filtering error variance matrix
Pk+1/k+1=(I-Kk+1Hk+1)Pk+1/kFormula 12
Wherein,
It can be obtained by two, three, four simultaneous of formula
Wherein, J=J1+J2.Then the state estimation equation of electric driving wheel Construction of A Model longitudinal force is expressed as
14 b of y=Cx+Fv formula
Wherein, x, u, d, y, respectively system state amount, it is known that input, Unknown worm and measured value, w and v are mutual not phase The zero-mean white noise sequence of pass.And have
Since formula 14 is there are Unknown worm, common Justin Lemberg observer is unable to get accurate state estimation.Cause This is by constructing Unknown worm state equation, and the longitudinal force observer of Nonlinear Observer Design design
Wherein,For system state estimation,For Unknown inputs, KpFor the gain matrix of longitudinal force observer, KiFor The integration matrix of Unknown inputs.Formula (15) can be further converted to
Definition status evaluated errorIt can obtain
So as to derive
Wherein,Choose rational gain matrix KpAnd Ki, matrix A can be madeePOLE PLACEMENT USING to sit Mark system Left half-plane.By formula 16 it is found that making longitudinal force become quantity of state to be estimated by designing longitudinal force observer, but Noise present in system can influence estimated accuracy.In order to inhibit the evaluated error of grass, in the base of longitudinal force observer Expanded Kalman filtration algorithm is introduced on plinth, extended Kalman filter is designed according to formula 16, to realize longitudinal force Unbiased esti-mator.
Step 3：Vehicle running state adaptive iteration method of estimation designs；
In the existing research about vehicle running state estimation, it is based on vehicle dynamic model, is filtered using spreading kalman It is very common that the other improvements form of wave and Kalman filtering, which carries out vehicle running state estimation, but it is most of before vehicle In state Kalman Filter Estimation, it is required for acquiring longitudinal acceleration and side acceleration as karr using related sensor The measurement updaue of graceful filtering.The present invention proposes a kind of vehicle suitable under longitudinal acceleration and side acceleration unknown situation Method for estimating state.Overall estimation strategy is as shown in Figure 1.In Fig. 1, γLOAnd vyLORespectively use Justin Lemberg observer 1 and human relations The prior estimate of the yaw velocity and lateral speed of the estimation gained of Burger observer 2.KγAnd KvyRespectively Justin Lemberg observer 1 With the suitable observer gain chosen in 2 design process of Justin Lemberg observer.vyEKF, γEKFAnd βEKFDivide vyLOIt is divided into using expansion Lateral speed, yaw velocity and the side slip angle obtained estimated by exhibition Kalman filter, wherein vyEKFAnd γEKFDepending on For posterior estimate.λ is the fuzzy weighted values coefficient of fuzzy controller output, vyMAnd γMTo be input to expansion after FUZZY WEIGHTED It opens up the pseudo- of Kalman filter and measures input.
According to formula one, vehicle non-linear dynamic model is represented by
Kalman filter is designed using vehicle non-linear dynamic model, input vector is：
Wherein, uvIt is the input vector collection of whole vehicle model, uv1-uv10The respectively number of each subvector of input vector collection；
The quantity of state of whole vehicle model is：
xv(t)=[xv1 xv2 xv3]T=[vyEKF γEKF βEKF]TFormula 21
Wherein, xvIt is the state vector collection of whole vehicle model, xv1-xv3The respectively number of each subvector of state vector collection；
The measurement vector of whole vehicle model is：
yv=[vyM γM]T=[yv1 yv2]TFormula 22
Wherein, yvIt is the measurement vector set of whole vehicle model, yv1And yv2The number of each subvector of vector set is respectively measured, vγMAnd γMIt is considered as the pseudo- measuring value of lateral speed and yaw velocity, the puppet measuring value is estimated by Extended Kalman filter Meter and Justin Lemberg observer estimate Weighted Fusion and obtain that fuzzy weighted values coefficient therein is travelled according to vehicle driving-cycle and vehicle Degree of stability dynamic regulation puppet measuring value, be expressed as to the state equation based on whole vehicle model：
And it measures equation and is
Wherein priori estimates vyMAnd γMIt is considered as the pseudo- measuring value of Extended Kalman filter here as known quantity.
As Fig. 1 designs prior estimate of the following Justin Lemberg observer 1 for yaw velocity according to one b of formula
Wherein, γLOFor the yaw velocity that Justin Lemberg observer 1 is estimated, γEKFEstimate for extended Kalman filter The yaw velocity arrived, KγFor the observer feedback oscillator of Justin Lemberg observer 1；
According to the longitudinal force Design of Observer method in step S2, longitudinal force observation is separately designed for four electric driving wheels Device, gained longitudinal force estimated value are used for the input quantity of Justin Lemberg observer.Meanwhile the side force of tire that is obtained by tire model and The collected front wheel angle of the sensor also input quantity as Justin Lemberg observer 1, and by the sideway of 1 gained of Justin Lemberg observer Angular speed prior estimate is denoted as γLO.Similarly, according to one a of formula, priori of the design Justin Lemberg observer 2 for lateral speed is estimated Meter：
Wherein, vyLOFor the lateral speed that Justin Lemberg observer 2 is estimated, vuEKFIt is estimated for extended Kalman filter Lateral speed, KvyFor the observer feedback oscillator of Justin Lemberg observer 2；
Longitudinal force estimated value, lateral tire force and front wheel angle similarly as Justin Lemberg observer 2 input quantity, together When, longitudinal speed and the yaw velocity prior estimate obtained by the estimation of Justin Lemberg observer 1 are also used as Justin Lemberg observer 2 Input quantity, v is denoted as by the lateral speed prior estimate of the gained of Justin Lemberg observer 2yLO.In addition, for Justin Lemberg observer 1 For 2, the Posterior estimator based on yaw velocity and lateral speed obtained by Extended Kalman filter is observed as Justin Lemberg The known pseudo- measurement input value of device 1 and 2.
In the design of Justin Lemberg observer 1 and 2, sideway is only considered respectively and lateral dynamics of vehicle is free Degree, therefore under the influence of noise and unknown disturbance, it is easy to the integral for error occur is accumulated so as to cause estimated accuracy reduction. Therefore it can be designed based on vehicle non-linear dynamic model using the Extended Kalman filter method described in step 2 Kalman filter realizes the further unbiased esti-mator of vehicle running state.In extended Kalman filter, prior estimate γLOAnd vyLOAs the pseudo- sensor collection value of Extended Kalman filter, and by the Posterior estimator obtained by Extended Kalman filter γEKFAnd vyEKFIt is more more accurate than prior estimate.
With the change of vehicle driving-cycle, due to the uncertainty of auto model, ginseng is not modeled such as non-linear factor and Several variations, vehicle state estimation deviation can aggravate.For the evaluated error for inhibiting above-mentioned factor to bring, the present invention devises mould It pastes weight and is used for automatic adjusument prior estimate γLOAnd vyLOConfidence level in Extended Kalman filter measurement updaue.Based on mould Pasting the extension measurement updaue amalgamation mode obtained by weight coefficient is
Wherein, λ is the fuzzy weighted values coefficient that designed fuzzy controller inputs.Fuzzy controller of the present invention is defeated It is longitudinal speed and front wheel angle to enter amount, is exported as real-time fuzzy weighted values coefficient.
Step 4：Longitudinal force observer and transport condition adaptive iteration method of estimation experimental verification.
The verification of longitudinal force observer estimation effect is carried out first on chassis dynamometer experimental stand.Experiment vehicle is one The distributed-driving electric automobile converted realizes the control of vehicle based on rapid prototyping platform.With corresponding sensor Electric current, rotating speed and the voltage of wheel hub motor are measured, and is acquired and records by CAN bus, utilizes chassis dynamometer data Acquisition system obtains longitudinal force.Using collected electric current, rotating speed and voltage as the input quantity of longitudinal force observer, by observer The longitudinal force estimated and the longitudinal force actually measured estimate that the results are shown in Figure 2.Designed longitudinal direction as can be seen from Figure 2 Force observer has preferable estimation effect.After demonstrating longitudinal force observer, carries out vehicle transport condition adaptive iteration and estimate The verification of meter method.Real vehicle roadway experiment has been carried out, shape is travelled using GPS navigation navigation device and inertial device collection vehicle State corresponds to the estimation effect of lateral speed, yaw velocity and side slip angle respectively as a result such as Fig. 3,4,5.From Fig. 3,4,5 As can be seen that the lateral speed, yaw velocity and the matter that are obtained estimated by adaptive iteration method of estimation designed by the present invention Heart side drift angle is all more nearly actual measured value, to prove the estimation effect of the adaptive iteration method of estimation designed by the present invention Fruit is more preferable.
The series of detailed descriptions listed above is illustrated only for possible embodiments of the invention, They are all without departing from equivalent embodiment made by technical spirit of the present invention or change not to limit the scope of the invention It should all be included in the protection scope of the present invention.

Claims (9)

1. a kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation, which is characterized in that including walking as follows Suddenly：
Step S1：Distributed-driving electric automobile vehicle model, including vehicle Nonlinear dynamic models, electric driving wheel modeling with And tire modeling；
Step S2：Longitudinal force Design of Observer based on nonlinear observer and Extended Kalman filter, first with electric drive The state estimation equation for taking turns Construction of A Model longitudinal force, to Nonlinear Observer Design, by nonlinear observer and extension karr Graceful filtering is combined the influence for reducing noise to estimation；
Step S3：Vehicle running state adaptive iteration method of estimation designs, and is directed to yaw rate and side respectively first Justin Lemberg observer is designed to speed, estimation is tested before obtaining yaw velocity and lateral speed, estimation is tested before described and utilizes institute It states longitudinal force information estimated by step S2 and is calculated by the lateral force that tire model obtains；It is non-linear further according to vehicle Kinetic model carries out vehicle running state estimation based on Extended Kalman filter, obtains the Posterior estimator of vehicle running state, That is Kalman Filter Estimation, wherein before test the measurement updaue that estimated information is used as Extended Kalman filter, after obtaining Test estimation simultaneously be used as before test estimation designed by Justin Lemberg observer pseudo- sensor input, to realize vehicle-state letter The iterative estimate of breath；Fuzzy rule has also been devised, according to when driving speed and front wheel angle dynamically adjust before test estimation exist Shared weight in Extended Kalman filter measurement updaue, so that the iterative process of vehicle state estimation, which has, is based on vehicle The adaptivity of driving cycle；
Step 4：Longitudinal force observer and transport condition adaptive iteration method of estimation experimental verification.
2. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 1, feature It is, the kinetics equation of vehicle non-linear dynamic model is in the step S1：
Wherein, vyFor lateral speed, γ is yaw velocity, and β is vehicle centroid side drift angle, vxFor longitudinal speed, For vehicle centroid speed and in view of lateral speed is smaller, v is equal to v in one c of formulax, m is car mass, and δ is preceding rotation Angle, IzFor around the rotary inertia of z-axis, lfFor distance of the barycenter away from front axle, lrFor distance of the barycenter away from rear axle, bfFor front tread 1/2, brIt is the 1/2, F of rear treadxjAnd Fyj(j=1,2,3,4) be respectively number it is longitudinal force corresponding to the tire of j and lateral Power.
3. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 2, feature It is, the spin dynamics equation of single wheel is in the step S1 electric driving wheels modeling：
In formula two, ωjFor longitudinal force FxjThe rotating speed of corresponding wheel；J1For vehicle wheel rotation inertia；R is wheel effective radius；TLj To be installed on the loading moment of wheel inner wheel hub motor；
Torque balance equation on wheel hub motor output shaft is：
The dynamic electric voltage equilibrium equation of wheel hub motor equivalent circuit is：
In formula three, four, J2For the rotary inertia of rotor；B is damped coefficient；KtFor motor torque constant；ijFor line current；uj For line voltage；R is the equivalent line resistance of winding；L is winding equivalent inductance；KaFor back EMF coefficient.
4. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 3, feature It is, the step S1 tires modeling includes the following steps：
Tire model using semiempirical magic formula estimates side force of tire formula is：
Fy=Dsin { Carctan [B α-E (B α-arctan (B α))] } formula five
In formula five, B is stiffness factor, and C is the curve shape factor, and D is peak factor, and E is the curvature of curve factor, and α is wheel side Drift angle；
Tire model parameter B, C, D, E are related to the vertical load of tire, and the vertical load of each tire is：
In formula six, Fz1、Fz2、Fz3、Fz4, for the vertical load of corresponding tire, h is height of center of mass, and g is acceleration of gravity；
Each slip angle of tire is：
In formula seven, α1、α2、α3、α4For the side drift angle of corresponding tire.
5. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 4, feature It is, the state estimation equation for the longitudinal force that electric driving wheel Construction of A Model goes out described in the step S2 is expressed as：
14 b of y=Cx+Fv formula
Wherein, x, u, d, y, respectively system state amount, it is known that input, Unknown worm and measured value, w and v are orthogonal Zero-mean white noise sequence, and have
Since there are Unknown worms for formula 14, by constructing Unknown worm state equation, and Nonlinear Observer Design carries out Longitudinal force is estimated：
Wherein,For system state estimation,For Unknown inputs, KpFor the gain matrix of longitudinal force observer, KiIt is unknown defeated Enter the integration matrix of estimation；
Formula 15 is further converted to：
Definition status evaluated error：
To derive：
Wherein,Choose rational gain matrix KpAnd Ki, make matrix AePOLE PLACEMENT USING is left to coordinate system Half-plane.
6. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 5, feature It is, the step S3 specifically includes following steps：
It is directed to yaw rate and lateral speed design Justin Lemberg observer 1 and Justin Lemberg observer 2, γ respectivelyLOAnd vyLO The priori of the yaw velocity and lateral speed that respectively use Justin Lemberg observer 1 and the estimation gained of Justin Lemberg observer 2 is estimated Meter, KγAnd KvyThe suitable observer gain respectively chosen in 2 design process of Justin Lemberg observer 1 and Justin Lemberg observer, vyEKF, γEKFAnd βEKFDivide vyLOBe divided into use the lateral speed obtained estimated by extended Kalman filter, yaw velocity with And side slip angle, wherein vyEKFAnd γEKFIt is considered as posterior estimate, λ is the fuzzy weighted values coefficient of fuzzy controller output, vyM And γMInput is measured to be input to the pseudo- of extended Kalman filter after FUZZY WEIGHTED,
According to formula one, vehicle non-linear dynamic model is expressed as：
Kalman filter is designed using vehicle non-linear dynamic model, input vector is：
Wherein, uvIt is the input vector collection of whole vehicle model, uv1-uv10The respectively number of each subvector of input vector collection；
The quantity of state of whole vehicle model is：
xv(t)=[xv1 xv2 xv3]T=[vyEKF γEKF βEKF]TFormula 21
Wherein, xvIt is the state vector collection of whole vehicle model, xv1-xv3The respectively number of each subvector of state vector collection；
The measurement vector of whole vehicle model is：
yv=[vyM γM]T=[yv1 yv2]TFormula 22
Wherein, yvIt is the measurement vector set of whole vehicle model, yv1And yv2Respectively measure the number of each subvector of vector set, vγM And γMIt is considered as the pseudo- measuring value of lateral speed and yaw velocity, the puppet measuring value is estimated by Extended Kalman filter Estimate Weighted Fusion with Justin Lemberg observer and obtain, fuzzy weighted values coefficient therein is travelled according to vehicle driving-cycle and vehicle Degree of stability dynamic regulation puppet measuring value, to which the state equation based on whole vehicle model is expressed as：
And it measures equation and is：
yv1=vyMFormula 24
yv2=γM
Wherein, priori estimates vyMAnd γMIt is considered as the pseudo- measuring value of Extended Kalman filter here as known quantity.
7. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 6, feature It is, the design of the Justin Lemberg observer 1 and Justin Lemberg observer 2 includes the following steps：
According to one b of formula, prior estimate of the following Justin Lemberg observer 1 for yaw velocity is designed：
Wherein, γLOFor the yaw velocity that Justin Lemberg observer 1 is estimated, γEKFIt is estimated for extended Kalman filter Yaw velocity, KγFor the observer feedback oscillator of Justin Lemberg observer 1；
According to the longitudinal force Design of Observer method in step S2, longitudinal force observer, gained are separately designed for electric driving wheel Longitudinal force estimated value is used for the input quantity of Justin Lemberg observer, meanwhile, the side force of tire and sensor obtained by tire model The collected front wheel angle also input quantity as Justin Lemberg observer 1, and by the yaw velocity of 1 gained of Justin Lemberg observer Prior estimate is γLO；
According to one a of formula, design Justin Lemberg observer 2 is used for the prior estimate of lateral speed：
Wherein, vyLOFor the lateral speed that Justin Lemberg observer 2 is estimated, vuEKFIt is estimated for extended Kalman filter lateral Speed, KvyFor the observer feedback oscillator of Justin Lemberg observer 2；
The input quantity of longitudinal force estimated value, lateral tire force and front wheel angle as Justin Lemberg observer 2, meanwhile, longitudinal speed And the obtained yaw velocity prior estimate also input quantity as Justin Lemberg observer 2 is estimated by Justin Lemberg observer 1, by The lateral speed prior estimate of 2 gained of Justin Lemberg observer is denoted as vyLO, for Justin Lemberg observer 1 and Justin Lemberg observer 2 It says, the Posterior estimator based on yaw velocity and lateral speed obtained by Extended Kalman filter is as 1 He of Justin Lemberg observer The known pseudo- measurement input value of Justin Lemberg observer 2.
8. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 7, feature It is, in the design of the Justin Lemberg observer 1 and Justin Lemberg observer 12, utilizes institute's spreading kalman in the step S2 Filtering method design the Kalman filter based on vehicle non-linear dynamic model, realize vehicle running state further without Estimation partially；In extended Kalman filter, prior estimate γLOAnd vyLOPseudo- sensor as Extended Kalman filter acquires Value.
9. distributed-driving electric automobile transport condition adaptive iteration method of estimation according to claim 8, feature It is, fuzzy weighted values is devised in fuzzy rule described in the step S3 for automatic adjusument prior estimate γLOAnd vyLO Confidence level in Extended Kalman filter measurement updaue,
It is based on the extension measurement updaue amalgamation mode obtained by fuzzy weighted values coefficient：
vyM=(1- λ) vyLo+λvyEKFFormula 27
γM=(1- λ) γLo+λγEKF
Wherein, λ is the fuzzy weighted values coefficient that designed fuzzy controller inputs, and the fuzzy controller input quantity is longitudinal vehicle Speed and front wheel angle, export as real-time fuzzy weighted values coefficient.
CN201711304213.3A 2017-12-11 2017-12-11 A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation Pending CN108284841A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711304213.3A CN108284841A (en) 2017-12-11 2017-12-11 A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711304213.3A CN108284841A (en) 2017-12-11 2017-12-11 A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation

Publications (1)

Publication Number Publication Date
CN108284841A true CN108284841A (en) 2018-07-17

Family

ID=62831865

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711304213.3A Pending CN108284841A (en) 2017-12-11 2017-12-11 A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation

Country Status (1)

Country Link
CN (1) CN108284841A (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109606378A (en) * 2018-11-19 2019-04-12 江苏大学 Vehicle running state estimation method towards non-Gaussian noise environment
CN109910617A (en) * 2019-03-27 2019-06-21 武汉理工大学 A kind of diagnostic method of distribution wheel-hub motor driven vehicle failure of removal
CN109962647A (en) * 2019-03-21 2019-07-02 上海交通大学 Electric machine control system and method for estimating state with Two-mode Coupling structure observer
CN110422175A (en) * 2019-07-31 2019-11-08 上海智驾汽车科技有限公司 Vehicle state estimation method and device, electronic equipment, storage medium, vehicle
CN111547059A (en) * 2020-04-23 2020-08-18 上海大学 Distributed driving electric automobile inertia parameter estimation method
CN111703429A (en) * 2020-05-29 2020-09-25 北京理工大学重庆创新中心 Method for estimating longitudinal speed of wheel hub motor driven vehicle
CN111845775A (en) * 2020-07-20 2020-10-30 上海大学 Joint estimation method for driving state and inertia parameters of distributed driving electric automobile
CN112590803A (en) * 2020-12-16 2021-04-02 北理慧动（常熟）车辆科技有限公司 Online estimation method for finished vehicle mass of single-shaft parallel hybrid power commercial vehicle
CN113203422A (en) * 2021-04-14 2021-08-03 武汉理工大学 Freight car state inertia parameter joint estimation method based on size measurement device
CN113771857A (en) * 2021-09-24 2021-12-10 北京易航远智科技有限公司 Longitudinal speed estimation method and system for vehicle control
CN113830094A (en) * 2021-09-16 2021-12-24 江苏大学 Vehicle mass center slip angle self-adaptive fusion and compensation method considering multi-source input information
CN114348002A (en) * 2021-12-29 2022-04-15 江苏大学 System and method for estimating speed of hub motor driven electric automobile
CN116232282A (en) * 2023-01-12 2023-06-06 湖南大学无锡智能控制研究院 Time-varying time delay estimation method, device and system based on adaptive all-pass filter
CN116923428A (en) * 2023-09-07 2023-10-24 华东交通大学 Combined estimation method for electric automobile centroid side deflection angle and tire side force
CN113830094B (en) * 2021-09-16 2024-04-30 常州工学院 Vehicle centroid slip angle self-adaptive fusion and compensation method considering multi-source input information

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102556075A (en) * 2011-12-15 2012-07-11 东南大学 Vehicle operating state estimation method based on improved extended Kalman filter
CN102582626A (en) * 2012-02-16 2012-07-18 吉林大学 Method for estimating heavy semitrailer status
US20140277926A1 (en) * 2013-03-12 2014-09-18 The Goodyear Tire & Rubber Company Dynamic tire slip angle estimation system and method
US20170101108A1 (en) * 2015-10-09 2017-04-13 The Goodyear Tire & Rubber Company Method for estimating tire forces from can-bus accessible sensor inputs

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102556075A (en) * 2011-12-15 2012-07-11 东南大学 Vehicle operating state estimation method based on improved extended Kalman filter
CN102582626A (en) * 2012-02-16 2012-07-18 吉林大学 Method for estimating heavy semitrailer status
US20140277926A1 (en) * 2013-03-12 2014-09-18 The Goodyear Tire & Rubber Company Dynamic tire slip angle estimation system and method
US20170101108A1 (en) * 2015-10-09 2017-04-13 The Goodyear Tire & Rubber Company Method for estimating tire forces from can-bus accessible sensor inputs

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陈特: "分布式驱动电动汽车行驶状态估计与转矩节能优化研究", 《CNKI优秀硕士学位论文库》 *

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109606378A (en) * 2018-11-19 2019-04-12 江苏大学 Vehicle running state estimation method towards non-Gaussian noise environment
CN109962647A (en) * 2019-03-21 2019-07-02 上海交通大学 Electric machine control system and method for estimating state with Two-mode Coupling structure observer
CN109910617A (en) * 2019-03-27 2019-06-21 武汉理工大学 A kind of diagnostic method of distribution wheel-hub motor driven vehicle failure of removal
CN109910617B (en) * 2019-03-27 2022-05-06 武汉理工大学 Method for diagnosing failure fault of distributed hub motor driven vehicle
CN110422175B (en) * 2019-07-31 2021-04-02 上海智驾汽车科技有限公司 Vehicle state estimation method and device, electronic device, storage medium, and vehicle
CN110422175A (en) * 2019-07-31 2019-11-08 上海智驾汽车科技有限公司 Vehicle state estimation method and device, electronic equipment, storage medium, vehicle
CN111547059A (en) * 2020-04-23 2020-08-18 上海大学 Distributed driving electric automobile inertia parameter estimation method
CN111703429A (en) * 2020-05-29 2020-09-25 北京理工大学重庆创新中心 Method for estimating longitudinal speed of wheel hub motor driven vehicle
CN111703429B (en) * 2020-05-29 2022-05-10 北京理工大学重庆创新中心 Method for estimating longitudinal speed of wheel hub motor driven vehicle
CN111845775B (en) * 2020-07-20 2022-01-07 上海大学 Joint estimation method for driving state and inertia parameters of distributed driving electric automobile
CN111845775A (en) * 2020-07-20 2020-10-30 上海大学 Joint estimation method for driving state and inertia parameters of distributed driving electric automobile
CN112590803B (en) * 2020-12-16 2022-02-22 北理慧动（常熟）车辆科技有限公司 Online estimation method for finished vehicle mass of single-shaft parallel hybrid power commercial vehicle
CN112590803A (en) * 2020-12-16 2021-04-02 北理慧动（常熟）车辆科技有限公司 Online estimation method for finished vehicle mass of single-shaft parallel hybrid power commercial vehicle
CN113203422A (en) * 2021-04-14 2021-08-03 武汉理工大学 Freight car state inertia parameter joint estimation method based on size measurement device
CN113830094B (en) * 2021-09-16 2024-04-30 常州工学院 Vehicle centroid slip angle self-adaptive fusion and compensation method considering multi-source input information
CN113830094A (en) * 2021-09-16 2021-12-24 江苏大学 Vehicle mass center slip angle self-adaptive fusion and compensation method considering multi-source input information
CN113771857A (en) * 2021-09-24 2021-12-10 北京易航远智科技有限公司 Longitudinal speed estimation method and system for vehicle control
CN114348002A (en) * 2021-12-29 2022-04-15 江苏大学 System and method for estimating speed of hub motor driven electric automobile
CN116232282B (en) * 2023-01-12 2023-12-19 湖南大学无锡智能控制研究院 Time-varying time delay estimation method, device and system based on adaptive all-pass filter
CN116232282A (en) * 2023-01-12 2023-06-06 湖南大学无锡智能控制研究院 Time-varying time delay estimation method, device and system based on adaptive all-pass filter
CN116923428A (en) * 2023-09-07 2023-10-24 华东交通大学 Combined estimation method for electric automobile centroid side deflection angle and tire side force
CN116923428B (en) * 2023-09-07 2023-11-28 华东交通大学 Combined estimation method for electric automobile centroid side deflection angle and tire side force

Similar Documents

Publication Publication Date Title
CN108284841A (en) A kind of distributed-driving electric automobile transport condition adaptive iteration method of estimation
CN106394561B (en) A kind of method of estimation and device of longitudinal speed of vehicle
Nam et al. Estimation of sideslip and roll angles of electric vehicles using lateral tire force sensors through RLS and Kalman filter approaches
CN105151047B (en) A kind of automobile side slip angle measurement method
CN102673569B (en) Vehicle-state is calculated device, method and is used the vehicle of this device
CN111845775B (en) Joint estimation method for driving state and inertia parameters of distributed driving electric automobile
CN104354700A (en) Vehicle parameter on-line estimation method based on unscented Kalman filtering
CN103921786B (en) A kind of nonlinear model predictive control method of electric vehicle process of regenerative braking
CN109606378B (en) Vehicle running state estimation method for non-Gaussian noise environment
CN108597058A (en) Distributed-driving electric automobile state based on pseudo- measurement information cascades method of estimation
CN103909933B (en) A kind of front wheel side of distributed electro-motive vehicle is to force evaluating method
CN108594652B (en) Observer information iteration-based vehicle state fusion estimation method
US6754615B1 (en) Method of simulating the performance of a vehicle on a road surface
CN102556075B (en) Vehicle operating state estimation method based on improved extended Kalman filter
CN102529976B (en) Vehicle running state nonlinear robust estimation method based on sliding mode observer
CN104180818B (en) A kind of monocular vision mileage calculation device
CN110884499B (en) Method and system for determining vehicle mass center slip angle
CN107696915A (en) A kind of wheeled driving control system of electric automobile based on hierarchical control and its control method
Zhang et al. Real-time estimation of vehicle mass and road grade based on multi-sensor data fusion
Lin et al. A novel H∞ and EKF joint estimation method for determining the center of gravity position of electric vehicles
Hsu et al. Vehicle full-state estimation and prediction system using state observers
CN108819950A (en) The vehicle speed estimation method and system of automobile stability control system
CN108287934B (en) Vehicle mass center slip angle robust estimation method based on longitudinal force observer
CN113650619B (en) Four-wheel drive electric vehicle tire force soft measurement method
Chen et al. Design of vehicle running states-fused estimation strategy using Kalman filters and tire force compensation 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
RJ01 Rejection of invention patent application after publication

Application publication date: 20180717

RJ01 Rejection of invention patent application after publication