CN112417365A - Automatic driving truck quality estimation method based on extended Kalman filtering - Google Patents

Automatic driving truck quality estimation method based on extended Kalman filtering Download PDF

Info

Publication number
CN112417365A
CN112417365A CN202011206433.4A CN202011206433A CN112417365A CN 112417365 A CN112417365 A CN 112417365A CN 202011206433 A CN202011206433 A CN 202011206433A CN 112417365 A CN112417365 A CN 112417365A
Authority
CN
China
Prior art keywords
calculating
extended kalman
equation
estimation method
quality estimation
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
CN202011206433.4A
Other languages
Chinese (zh)
Inventor
胡子牛
江浩斌
张利
刘擎超
边有钢
肖雅金
俞越
江绍康
吴明康
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 CN202011206433.4A priority Critical patent/CN112417365A/en
Publication of CN112417365A publication Critical patent/CN112417365A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Operations Research (AREA)
  • Computing Systems (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Abstract

The invention discloses an automatic driving truck quality estimation method based on extended Kalman filtering, which comprises the following steps: step 1, establishing a vehicle longitudinal dynamic model; step 2, acquiring a state transition equation; step 3, obtaining an observation equation; step 4, calculating a Jacobian matrix Hk+1(ii) a And 5, calculating the estimated mass of the truck. According to the automatic driving truck quality estimation method based on the extended Kalman filtering, the dynamic model can be effectively established firstly through the steps 1 to 5, and then the truck quality is estimated.

Description

Automatic driving truck quality estimation method based on extended Kalman filtering
Technical Field
The invention relates to a quality estimation technology, in particular to an automatic driving truck quality estimation method and system based on extended Kalman filtering.
Background
The truck has the advantages of large freight volume, low cost, high economic benefit and the like, and occupies an important position in road transportation. The automatic driving truck can greatly improve the safety of the vehicle road and the economic benefit. In automatic driving of a truck, control parameters need to be designed by using vehicle quality, and control performance is greatly influenced, so that accurate vehicle quality estimation is of great significance to design of an automatic driving system.
At present, there are some patents focusing on vehicle mass estimation methods. For example, CN108225502B has designed a method and a system for estimating the quality of ore loaded on a truck, which automatically estimate the quality of ore loaded on the truck by combining a computer vision technique with a deep convolutional neural network learning method, thereby avoiding the influence of human factors on ore measurement. CN106740870B designs a vehicle mass estimation method considering shift factors, which takes into account the abrupt change of the transmission ratio brought during shifting and the change of the conversion coefficient of the rotating mass of the vehicle after shifting, and adopts a weighted least square recursive estimation method with multiple forgetting factors to realize the real-time estimation of the vehicle mass. CN108394415B designs a method and a system for estimating vehicle mass, which consider air resistance and rolling resistance as known quantities, reduce data operation quantity, estimate the mass of the vehicle by using a simple least square method, and take accuracy and real-time into account.
The technology has positive significance for improving the vehicle mass estimation, but the method only considers the hopper ore mass estimation or considers the influence of gear shifting factors, air resistance and rolling resistance on the vehicle mass estimation in a unilateral way, and does not consider the condition that the mass of the whole truck is greatly changed. Accordingly, a technical solution is desired to overcome or at least alleviate at least one of the above-mentioned drawbacks of the prior art.
Disclosure of Invention
In view of the deficiencies of the prior art, it is an object of the present invention to provide an extended kalman filter based method for automated mass estimation of a driven truck that overcomes or at least mitigates at least one of the above-identified deficiencies in the prior art.
In order to achieve the purpose, the invention provides the following technical scheme: an automatic driving truck quality estimation method based on extended Kalman filtering is characterized in that: the method comprises the following steps:
step 1, establishing a vehicle longitudinal dynamic model;
step 2, acquiring a state transition equation;
step 3, obtaining an observation equation;
step 4, calculating a Jacobian matrix Hk+1；
And 5, calculating the estimated mass of the truck.
As a further improvement of the present invention, the vehicle longitudinal dynamics model in step 1 is as follows:
Figure BDA0002757231460000021
wherein, FTAs a driving force, rwheelIs the wheel radius, TwheelIs the torque applied to the wheels.
As a further improvement of the present invention, the specific steps of acquiring the state transition equation in step 2 are as follows:
step 21, describing a vehicle system by using an extended Kalman filter algorithm:
x(k+1)＝f(x(k))
step 22, calculating the vehicle acceleration from the vehicle dynamics equation obtained in step 1:
Figure BDA0002757231460000022
step 23, defining a first-order state process of the system as:
Figure BDA0002757231460000023
and 24, substituting each parameter to obtain a final state equation:
Figure BDA0002757231460000031
in the above-mentioned formula,
Figure BDA0002757231460000032
representing vehicle speed in m/s;
Figure BDA0002757231460000033
m represents mass in Kg;
Figure BDA0002757231460000034
Tsrepresenting the time step in units s.
As a further improvement of the present invention, the step 22 further includes a simplified step, specifically:
let the rolling resistance coefficient mu be tan betaμThen, the formula:
Figure BDA0002757231460000035
can be simplified as follows:
Figure BDA0002757231460000036
note the book
Figure BDA0002757231460000037
The formula can be further simplified as:
Figure BDA0002757231460000038
as a further improvement of the present invention, the step of obtaining the observation equation in step 3 is as follows:
step 31, defining the process noise of the vehicle speed as WVMass process noise WmCourse noise W of a rampβ. The system state equation may instead be:
Figure BDA0002757231460000039
step 32, the changed state equation is differentiated for the process noise to obtain a Jacobian matrix W of the process noise signalk+1Comprises the following steps:
Figure BDA0002757231460000041
the covariance matrix of the process noise is:
Figure BDA0002757231460000042
wherein,
Figure BDA0002757231460000043
representing the process noise variance of the corresponding state, which can be calibrated;
step 33, the relationship between the measured values and the one-step prediction state, and the observation equation h is obtained as follows:
Figure BDA0002757231460000044
as a specific implementation mode of improvement, the Jacobian matrix H is calculated in the step 4k+1The method comprises the following specific steps:
the observation equation is derived to obtain a measured Jacobian matrix Hk+1：
Figure BDA0002757231460000045
As a further improvement of the present invention, the specific steps of calculating the estimated mass of the truck in step 5 are as follows:
step 51, calculating a one-step prediction variance according to the following formula:
Figure BDA0002757231460000046
step 52, calculating the kalman gain according to the following formula:
Figure BDA0002757231460000047
step 53, the covariance error is calculated according to the following formula:
Figure BDA0002757231460000048
step 54, calculating the kalman state estimate according to the following formula:
Figure BDA0002757231460000049
the method has the advantages of avoiding the problem of larger estimation error caused by larger difference of the mass of the whole truck when the truck is in no load and full load, and effectively improving the accuracy of the mass estimation of the automatic driving truck.
Drawings
FIG. 1 is a vehicle force diagram;
fig. 2 is a system architecture diagram.
Detailed Description
The invention will be further described in detail with reference to the following examples, which are given in the accompanying drawings.
Referring to fig. 1, the method for estimating the mass of an autonomous truck based on extended kalman filtering according to the present embodiment includes the following steps:
step 1, establishing a vehicle longitudinal dynamics model.
And 2, acquiring a state transition equation.
And step 3, obtaining an observation equation.
Step 4, calculating a Jacobian matrix Hk+1。
And 5, calculating the estimated mass of the truck.
In step 1, the stress condition of the vehicle running on the slope is analyzed as shown in fig. 1. The force applied is mainly driving force FTAir resistance FairRamp resistance FgAnd rolling resistance Fμ。
The driving force is the positive force for driving the vehicle forward, and the magnitude of the positive force is defined by the radius r of the wheelwheel(unit m) and torque T applied to the wheelwheel(unit Nm) determination.
Figure BDA0002757231460000051
In practice, wheel torque is generally calculated from engine torque:
Twheel＝Teng*ig*η*i0*ηwheel
wherein, TengAs engine torque, igIs the transmission ratio of the gearbox, eta is the transmission efficiency, i0Is a main reduction ratio, ηwheelThe tire transmission efficiency is improved.
Air resistance is represented by air density ρ (unit kgm)-2) Coefficient of aerodynamics CdLongitudinal vehicle speed V, and vehicle frontal area Af(unit m)2) Determining:
Figure BDA0002757231460000052
the ramp resistance is a component of gravity along the ramp direction, and when climbing a slope, the ramp resistance is a negative force, and when descending a slope, the ramp resistance can be regarded as a positive force in the same direction as the driving force.
Fg＝mgsinβ
Wherein, beta represents a slope, the uphill is positive, the downhill is negative, and g is the gravity acceleration.
The rolling resistance is produced by the friction and deformation of the tires, mainly by the rolling resistance coefficient μ, the road slope β, the vehicle mass m (in Kg), and the gravitational acceleration g (in ms)2) Determining:
Fμ＝μmgcosβ
in step 2, the Extended Kalman Filter (EKF) describes the vehicle system with a nonlinear difference equation:
x(k+1)＝f(x(k))
calculating vehicle acceleration from the vehicle dynamics equation:
Figure BDA0002757231460000061
let the rolling resistance coefficient mu be tan betaμThen the above equation can be simplified as:
Figure BDA0002757231460000062
note the book
Figure BDA0002757231460000063
The formula can be further simplified as:
Figure BDA0002757231460000064
the first order state process of the system is defined as:
Figure BDA0002757231460000065
substituting each parameter to obtain a final state equation:
Figure BDA0002757231460000071
wherein,
Figure BDA0002757231460000072
representing vehicle speed in m/s;
Figure BDA0002757231460000073
m represents mass in Kg;
Figure BDA0002757231460000074
Tsrepresenting the time step in units s.
In particular, selection
Figure BDA0002757231460000075
Instead of m, as one state of the system equation, is to simplify the linearity of the kalman filter.
The state equation is derived to obtain a state transition matrix, namely a Jacobian matrix Ak+1：
Figure BDA0002757231460000076
In step 3, defining the process noise of the vehicle speed as WVMass process noise WmCourse noise W of a rampβ. The system state equation may instead be:
Figure BDA0002757231460000077
the above equation of state is used to derive the process noise to obtain the Jacobian matrix W of the process noise signalk+1Comprises the following steps:
Figure BDA0002757231460000078
the covariance matrix of the process noise is:
Figure BDA0002757231460000081
wherein,
Figure BDA0002757231460000082
the process noise variance, which represents the corresponding state, can be calibrated.
The relationship between the measured value and the one-step predicted state, namely the observation equation h, is:
Figure BDA0002757231460000083
in step 4, derivation is carried out on the observation equation to obtain a measured Jacobian matrix Hk+1：
Figure BDA0002757231460000084
The measurement vector is:
Figure BDA0002757231460000085
wherein v iskRepresenting the actual vehicle speed tested in kmh-1；akRepresenting the measured longitudinal acceleration measured by the accelerometer, in ms-2。
The measurement noise covariance matrix is:
Figure BDA0002757231460000086
wherein R is1The process noise deviation representing the vehicle speed can be calibrated; r2The process noise bias, which represents the longitudinal acceleration, can be calibrated.
In step 5, the one-step prediction variance calculation formula is as follows:
Figure BDA0002757231460000087
the Kalman gain calculation formula is as follows:
Figure BDA0002757231460000088
the covariance error calculation formula is:
Figure BDA0002757231460000089
the Kalman state estimator has the formula:
Figure BDA0002757231460000091
the present embodiment further provides an automatic driving truck quality estimation system based on extended kalman filter, the system architecture is shown in fig. 2, and the system architecture includes an input signal processing module, an EFK filtering module, and an output control module, where the input signal mainly includes: longitudinal acceleration, engine torque signals, brake signals, gear signals, vehicle speed signals, and the like. The automatic driving truck quality estimation method based on the extended Kalman filtering is adopted.
Due to the adoption of the technical scheme, the method has the following advantages: the problem of large estimation error caused by large difference of the whole truck mass when the truck is in no-load and full-load is avoided, and the mass estimation accuracy of the automatic driving truck is effectively improved
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that modifications and embellishments within the scope of the invention may occur to those skilled in the art without departing from the principle of the invention, and are considered to be within the scope of the invention.

Claims (7)

1. An automatic driving truck quality estimation method based on extended Kalman filtering is characterized in that: the method comprises the following steps:
step 1, establishing a vehicle longitudinal dynamic model;
step 2, acquiring a state transition equation;
step 3, obtaining an observation equation;
step 4, calculating a Jacobian matrix Hk+1；
And 5, calculating the estimated mass of the truck.
2. The extended kalman filter-based automatic driving truck quality estimation method according to claim 1, wherein: the vehicle longitudinal dynamic model in the step 1 is as follows:
Figure FDA0002757231450000011
wherein, FTAs a driving force, rwheelIs the wheel radius, TwheelIs the torque applied to the wheels.
3. The extended kalman filter-based automatic driving truck quality estimation method according to claim 2, wherein: the specific steps for acquiring the state transition equation in the step 2 are as follows:
step 21, describing a vehicle system by using an extended Kalman filter algorithm:
x(k+1)＝f(x(k))
step 22, calculating the vehicle acceleration from the vehicle dynamics equation obtained in step 1:
Figure FDA0002757231450000012
step 23, defining a first-order state process of the system as:
Figure FDA0002757231450000013
and 24, substituting each parameter to obtain a final state equation:
Figure FDA0002757231450000021
in the above-mentioned formula,
Figure FDA0002757231450000022
representing vehicle speed in m/s;
Figure FDA0002757231450000023
m represents mass in Kg;
Figure FDA0002757231450000024
Tsrepresenting the time step in units s.
4. The extended kalman filter-based automatic driving truck quality estimation method according to claim 3, wherein: the step 22 further comprises a simplification step, specifically:
let the rolling resistance coefficient mu be tan betaμThen, the formula:
Figure FDA0002757231450000025
can be simplified as follows:
Figure FDA0002757231450000026
note the book
Figure FDA0002757231450000027
The formula can be further simplified as:
Figure FDA0002757231450000028
5. the extended kalman filter-based automatic driving truck quality estimation method according to claim 4, wherein: the step of obtaining the observation equation in the step 3 is as follows:
step 31, defining the process noise of the vehicle speed as WVMass process noise WmCourse noise W of a rampβ. The system state equation may instead be:
Figure FDA0002757231450000029
step 32, the changed state equation is differentiated for the process noise to obtain a Jacobian matrix W of the process noise signalk+1Comprises the following steps:
Figure FDA0002757231450000031
the covariance matrix of the process noise is:
Figure FDA0002757231450000032
wherein,
Figure FDA0002757231450000033
representing the process noise variance of the corresponding state, which can be calibrated;
step 33, the relationship between the measured values and the one-step prediction state, and the observation equation h is obtained as follows:
Figure FDA0002757231450000034
6. the extended kalman filter-based automated driving truck quality estimation method according to claim 5, wherein: calculating the Jacobian matrix H in the step 4k+1The method comprises the following specific steps:
the observation equation is derived to obtain a measured Jacobian matrix Hk+1：
Figure FDA0002757231450000035
7. The extended kalman filter-based automated driving truck quality estimation method according to claim 6, wherein: the specific steps of calculating the estimated mass of the truck in step 5 are as follows:
step 51, calculating a one-step prediction variance according to the following formula:
Figure FDA0002757231450000036
step 52, calculating the kalman gain according to the following formula:
Figure FDA0002757231450000037
step 53, the covariance error is calculated according to the following formula:
Figure FDA0002757231450000038
step 54, calculating the kalman state estimate according to the following formula:
Figure FDA0002757231450000039
CN202011206433.4A 2020-11-03 2020-11-03 Automatic driving truck quality estimation method based on extended Kalman filtering Pending CN112417365A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011206433.4A CN112417365A (en) 2020-11-03 2020-11-03 Automatic driving truck quality estimation method based on extended Kalman filtering

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011206433.4A CN112417365A (en) 2020-11-03 2020-11-03 Automatic driving truck quality estimation method based on extended Kalman filtering

Publications (1)

Publication Number Publication Date
CN112417365A true CN112417365A (en) 2021-02-26

Family

ID=74828127

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011206433.4A Pending CN112417365A (en) 2020-11-03 2020-11-03 Automatic driving truck quality estimation method based on extended Kalman filtering

Country Status (1)

Country Link
CN (1) CN112417365A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113392518A (en) * 2021-06-08 2021-09-14 阿波罗智联(北京)科技有限公司 Method and device for estimating vehicle weight

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003023334A1 (en) * 2001-09-06 2003-03-20 Siemens Aktiengesellschaft Method and device for determining the mass of a vehicle and road ascent by a recursive state estimation algorithm
CN108128308A (en) * 2017-12-27 2018-06-08 长沙理工大学 A kind of vehicle state estimation system and method for distributed-driving electric automobile
CN109684677A (en) * 2018-12-04 2019-04-26 西安法士特汽车传动有限公司 A kind of gradient evaluation method based on Kalman filtering algorithm
CN111507019A (en) * 2020-05-06 2020-08-07 北京理工大学 Vehicle mass and road gradient iterative type joint estimation method based on MMR L S and SH-STF

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2003023334A1 (en) * 2001-09-06 2003-03-20 Siemens Aktiengesellschaft Method and device for determining the mass of a vehicle and road ascent by a recursive state estimation algorithm
CN108128308A (en) * 2017-12-27 2018-06-08 长沙理工大学 A kind of vehicle state estimation system and method for distributed-driving electric automobile
CN109684677A (en) * 2018-12-04 2019-04-26 西安法士特汽车传动有限公司 A kind of gradient evaluation method based on Kalman filtering algorithm
CN111507019A (en) * 2020-05-06 2020-08-07 北京理工大学 Vehicle mass and road gradient iterative type joint estimation method based on MMR L S and SH-STF

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
任志英;沈亮量;黄伟;刘星星;: "基于AEKF的车辆质量与道路坡度实时估计", 振动.测试与诊断, no. 04, 15 August 2020 (2020-08-15) *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113392518A (en) * 2021-06-08 2021-09-14 阿波罗智联(北京)科技有限公司 Method and device for estimating vehicle weight
CN113392518B (en) * 2021-06-08 2024-05-17 阿波罗智联(北京)科技有限公司 Method and apparatus for estimating vehicle weight

Similar Documents

Publication Publication Date Title
JP5214738B2 (en) Vehicle load estimation method
CN107100993B (en) A kind of fluid drive shift modification method of vehicle mass and hill gradient identification
CN111806449A (en) Method for estimating total vehicle mass and road surface gradient of pure electric vehicle
US20160332633A1 (en) Automotive control unit programmed to estimate road slope and vehicle mass, vehicle with such a control unit and corresponding program product
CN107247824A (en) Consider the car mass road grade combined estimation method of brake and influence of turning
CN106740870A (en) A kind of vehicle mass method of estimation for considering gearshift factor
CN107117178A (en) Consider the vehicle mass method of estimation of gearshift and road grade factor
CN103946679A (en) Vehicle mass estimation method and system
CN101678768A (en) The vehicle speed control system of vehicle and method for controlling driving speed
CN102486400A (en) Vehicle mass identification method and device
CN107458380A (en) A kind of road grade real-time estimation method being applied under comprehensive driving cycles
CN111605559B (en) Vehicle mass estimation method, torque control method and device
Torabi et al. Road grade and vehicle mass estimation for heavy-duty vehicles using feedforward neural networks
CN105849514A (en) Method and vehicle with arrangement for estimating mass of the vehicle
CN113859253B (en) Real-time estimation method for mass in vehicle driving process
CN115503737A (en) Vehicle mass estimation method, device, medium, equipment and vehicle
CN112417365A (en) Automatic driving truck quality estimation method based on extended Kalman filtering
CN113954846B (en) Estimation and system for ramp information in vehicle running
CN116674571A (en) Real-time estimation method for automobile quality and gradient based on data confidence factor
CN110027537B (en) Hydraulic retarder control method based on quality identification
CN111497866A (en) Steering wheel corner sensor fault diagnosis method based on improved Mahalanobis distance
CN116821775A (en) Load estimation method based on machine learning
CN115503736A (en) Automobile mass estimation method based on engine torque estimation
CN115402331A (en) Steep descent method based on dynamic ramp estimation
CN115329579A (en) Commercial vehicle mass and gradient joint estimation method considering whole vehicle mass change

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