CN110920626B - Data-driven electric drive vehicle attachment stability identification method and device - Google Patents

Data-driven electric drive vehicle attachment stability identification method and device Download PDF

Info

Publication number
CN110920626B
CN110920626B CN201911260714.5A CN201911260714A CN110920626B CN 110920626 B CN110920626 B CN 110920626B CN 201911260714 A CN201911260714 A CN 201911260714A CN 110920626 B CN110920626 B CN 110920626B
Authority
CN
China
Prior art keywords
adhesion
road surface
vehicle
slip ratio
parameter
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201911260714.5A
Other languages
Chinese (zh)
Other versions
CN110920626A (en
Inventor
徐坤
李慧云
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.)
Shenzhen Institute of Advanced Technology of CAS
Original Assignee
Shenzhen Institute of Advanced Technology of CAS
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 Shenzhen Institute of Advanced Technology of CAS filed Critical Shenzhen Institute of Advanced Technology of CAS
Priority to CN201911260714.5A priority Critical patent/CN110920626B/en
Publication of CN110920626A publication Critical patent/CN110920626A/en
Priority to PCT/CN2020/129524 priority patent/WO2021115063A1/en
Application granted granted Critical
Publication of CN110920626B publication Critical patent/CN110920626B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • B60W40/02—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 ambient conditions
    • B60W40/06—Road conditions
    • B60W40/064—Degree of grip
    • 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
    • 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
    • B60W2050/0019—Control system elements or transfer functions
    • B60W2050/0028—Mathematical models, e.g. for simulation
    • B60W2050/0031—Mathematical model of the vehicle
    • B60W2050/0034—Multiple-track, 2D vehicle model, e.g. four-wheel model
    • 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
    • B60W2510/00—Input parameters relating to a particular sub-units
    • B60W2510/08—Electric propulsion units
    • 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
    • B60W2520/10—Longitudinal speed
    • B60W2520/105—Longitudinal acceleration

Landscapes

  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)

Abstract

The invention relates to a method and a device for identifying the adhesion stability of an electric drive vehicle based on directly measured parameters such as driving force, acceleration, road surface identification types and the like. The whole technical scheme does not need to estimate or calculate signals which are difficult to accurately estimate or measure, and has the advantages of simplicity, reliability, quick response and good practicability.

Description

Data-driven electric drive vehicle attachment stability identification method and device
Technical Field
The invention belongs to the technical field of vehicle control, particularly relates to monitoring of vehicle adhesion stability, and particularly relates to a vehicle adhesion stability identification method and a device for implementing the method.
Background
Vehicle stability recognition in low adhesion road conditions is critical to ensure safe vehicle travel. When a vehicle runs on a low-adhesion road surface, such as a rain, snow, ice, or muddy road surface, the vehicle easily enters an unstable adhesion state, and further, the driving wheels slip or are locked. Description figure 1 shows the adhesion steady state condition of a vehicle under different road surfaces, wherein the abscissa of the graph is the slip ratio, and the ordinate of the graph is the adhesion coefficient, wherein the slip ratio is the proportion of a sliding component in the movement of wheels. It can be seen that there is an optimum working point or optimum slip ratio, corresponding to the maximum adhesion coefficient, for either the dry asphalt pavement at the top of the figure or the rain pavement at the bottom. When the slip ratio is smaller than the optimal slip ratio, the adhesion state is stable, and the adhesion coefficient and the adhesion force are increased along with the increase of the slip ratio, so that the driving stability of the vehicle is facilitated; when the slip ratio is greater than the optimal slip ratio, the adhesion state is unstable, the adhesion coefficient and the adhesion force can be reduced along with the increase of the slip ratio, the wheel speed and the slip ratio can be further increased along with the reduction of the adhesion force of the driving wheel, and the slip degree of the driving wheel is aggravated. Therefore, the position of the optimal operating point or the optimal slip ratio can be defined as the adhesion boundary, that is, the stable adhesion region of the vehicle corresponding to the point toward the direction of smaller slip ratio and the unstable adhesion region of the vehicle corresponding to the point toward the direction of larger slip ratio, or called unstable slip region. If the vehicle is in an unstable adhesion region, the adhesion force is reduced, the vehicle handling performance is deteriorated, and the stability is deteriorated. As can be seen, the boundary between the stable and unstable attachment areas of the vehicle varies with the type of road surface due to the complexity of the contact between the tire and the road surface. Therefore, when different road conditions are faced, the accurate identification of the unstable vehicle attachment area has important significance for the safe and stable control of the vehicle. After the adhesion boundary, i.e. the optimal operating point in fig. 1, is identified, a reference may be provided to the control system, for example, after a stable adhesion boundary in an ice condition is identified, the operating point corresponding to the stable boundary, e.g. the optimal slip ratio, the maximum adhesion coefficient, etc., may be used as control reference information, so that the maximum adhesion capability of the low adhesion road surface may be fully utilized while the vehicle is ensured to be stable. However, accurate identification of stable adhesion boundaries under such uncertain road conditions has remained a challenging problem to date.
Currently, an active safety control system in a conventional fuel commercial vehicle, such as a Traction Control System (TCS), an anti-lock brake control system (ABS), etc., is generally controlled according to a worst-adhesion road surface, that is, a slip ratio of a driving wheel is consistently controlled around a certain value, for example, a target slip ratio is controlled to be 0.2, by applying an external friction braking force to each wheel regardless of a current road surface. The method has the defects that the stable adhesion boundary under different road surface conditions cannot be accurately identified, and further the optimal adhesion control cannot be realized.
In order to recognize the adhesion stability state of different road surfaces, the determination can be made using an adhesion coefficient-slip ratio curve of the contact between the tire and the road surface as shown in fig. 1. However, as described above, the optimal slip ratios corresponding to the adhesion coefficient curves of different road surfaces are different, and identifying the optimal slip ratio is a difficult problem in practical application.
In order to avoid the problem that the adhesion stability is difficult to judge due to the difficulty in accurately identifying the optimal slip ratio, the adhesion stability can be judged by using the ratio of the change rate of the adhesion coefficient to the change rate of the slip ratio, namely, the following formula is calculated:
Figure BDA0002311524100000021
when the ratio k of the change rate of the adhesion coefficient to the change rate of the slip rate is greater than 0, the adhesion state is stable; when k is less than 0, the attachment is unstable; when k is equal to 0, it is a boundary state of the stable adhesion and the unstable slip state. The method is consistent for all tires and road surface relation curves, so that the problem of consistency of adhesion stability judgment under different road surface conditions is solved. However, since the measurement accuracy and bandwidth of the adhesion coefficient and slip ratio are not high, and the measurement is easily interfered by noise, high frequency noise is introduced into the differential signal for calculating them, and thus the practicability of the above-mentioned discrimination method is not good.
In order to overcome the problem, patent ZL201410783813.2 uses driving force, wheel speed and estimated adhesion to determine whether the vehicle is stable in a traction state, and aims to solve the problem that the adhesion coefficient, slip ratio and differential signal thereof are easy to interfere and cause difficulty in practical application faced by the above method. However, the method still needs to estimate the adhesion, which increases the computational complexity on one hand, and on the other hand, the adhesion estimation accuracy is also easily affected by the wheel speed measurement signal, so that the accuracy is difficult to guarantee.
Therefore, in the prior art of vehicle adhesion stability identification, measurement of parameters such as slip rate, optimal slip rate, adhesion coefficient, wheel speed and the like is generally required, or signals are estimated based on the parameters, and the measurement bandwidth of the signals is low, the measurement or estimation accuracy is not high, and the overall scheme also has the defects of high calculation complexity and the like, so that the prior adhesion stability identification technical scheme is easy to be effective in practical application.
Disclosure of Invention
In view of the above-mentioned drawbacks of the prior art, the present invention provides a method for identifying adhesion stability of a data-driven electrically-driven vehicle, comprising the following steps: measuring the driving torque of the motor; measuring the acceleration of the vehicle along the longitudinal axis of the vehicle body; identifying a road surface type parameter; inputting the trained recognizer based on the three types of signals to obtain vehicle attachment stability parameters; and judging the vehicle stable state according to the obtained adhesion stable state.
Wherein the recognizer is a neural network model capable of implementing nonlinear mapping. Preferably, the recognizer is trained as follows: collecting parameters of vehicle operation in various different scenes, wherein the parameters comprise vehicle motor driving force, road surface type and vehicle acceleration signals; meanwhile, collecting a slip rate signal and an optimal slip rate signal, and calculating an adhesion stability parameter according to the relationship between the slip rate and the optimal slip rate so as to describe quantitative adhesion stability states under different motor driving forces, road surface types and accelerations; and training the network model parameters by using the data obtained in the steps as a training set.
Preferably, the calculation of the adhesion stability parameter is performed according to the following method:
when the slip ratio is greater than the optimum slip ratio, the adhesion stabilization parameter is normalized to the interval (0, -1 ]:
Figure BDA0002311524100000041
when the slip ratio is equal to the optimum slip ratio, the adhesion stability parameter σ is 0;
when the slip ratio is less than the optimum slip ratio, the adhesion stabilization parameter is normalized to the interval (0,1 ]:
Figure BDA0002311524100000042
based on the training process, the vehicle stable state is judged according to the obtained attachment stable state according to the following concrete basis: stable adhesion is achieved when the adhesion stability parameter σ >0, optimal adhesion is achieved when the adhesion stability parameter σ is 0, and unstable slippage is achieved when the adhesion stability parameter σ < 0.
Further, in the above scheme, the motor driving torque measurement signal is a signal sequence at the current time and in a period of time before; the acceleration measurement signal is a signal sequence preceding the current time. Preferably, the motor drive torque measurement signal and the acceleration measurement signal input to the identifier are normalized signals.
Preferably, the road surface type parameter is a value between [0, 1 ]. Further, the road surface type parameters can be fuzzified according to the road surface condition. The road surface type parameter may be obtained by classifying the road surface type based on an image.
The invention also provides an adhesion stability recognition device for the electrically driven vehicle, which is used for implementing the adhesion stability recognition method for the electrically driven vehicle, and comprises the following steps:
a motor driving torque acquisition unit for acquiring a motor driving torque signal;
an acceleration acquisition unit for acquiring an acceleration of the vehicle along a longitudinal axis of the vehicle body;
the road surface type parameter acquisition unit is used for acquiring road surface type parameters;
a memory in which computer-executable instructions are stored, and a processor that calls the instructions to acquire an adhesion stability parameter from data of the motor driving torque acquisition unit, the acceleration acquisition unit, and the road surface type parameter acquisition unit through a trained recognizer, thereby determining an adhesion stability state of a vehicle.
The invention provides an identification method and device for estimating adhesion stability of an electrically driven vehicle based on directly measured parameters such as driving force, acceleration and road surface identification types. Its main advantage lies in:
(1) signals which are difficult to accurately estimate or measure such as an adhesion coefficient, a slip rate and an optimal slip rate do not need to be estimated or calculated, driving force and acceleration signals which can be directly measured are adopted, and the method is simple, reliable and good in practicability.
(2) The invention has the advantages of higher identification speed, capability of being matched with the fast response control characteristic (millisecond level) of the electrically driven vehicle, contribution to fast adjustment when an unstable attachment state is found, and contribution to the safety and stability of the vehicle.
(3) In the existing method, a mechanism model needs to be considered, particularly, an estimator needs to be designed according to a mechanism mathematical model when an estimation signal is related, and the precision, unmodeled dynamics and the like of the model can influence the estimation precision and robustness. The invention adopts a data driving method, has good generalization capability, does not need a mechanism model, and can cope with the influence of system nonlinearity uncertainty, so that the recognizer is more stable and has better adaptability.
Drawings
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings;
FIG. 1 is a schematic view of the adhesion stability under different road surfaces;
FIG. 2 is a schematic diagram of a vehicle model and vehicle parameters provided in an embodiment of the present disclosure;
FIG. 3 is a flow chart of recognizer training provided by embodiments of the present specification;
FIG. 4 is an attachment state recognition model structure provided by embodiments of the present description;
FIG. 5 is a graph of the adhesion stability parameter σ versus slip ratio provided in the examples herein;
fig. 6 is a diagram of a vehicle attachment stability recognition apparatus provided in an embodiment of the present specification.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of protection of the present invention.
In electrically driven vehicles, the driving torque T of the drive motor is applied to the drive wheels, the tires coming into direct contact with the road surface, this contact producing a friction force, i.e. a grip force F, of the road surface against the tiresdThe massage isThe friction force, otherwise known as the adhesion force, is the driving force that drives the vehicle in motion. This relationship can be described by two equations, a wheel dynamics model and a vehicle dynamics model:
T-rFd＝Jωω (2)
Fd-Fdr＝MV (3)
where ω is the driving wheel rotational angular velocity, FdrIs the resistance of the whole vehicle, M is the mass of the vehicle, and the longitudinal acceleration of the vehicle is V, JωIs the equivalent rotational inertia of the wheel, r is the rolling radius of the wheel, and the specific parameters can be referred to as the description and shown in the attached figure 2.
Further, the slip ratio is represented by λ, which represents a degree of difference between the translational velocity equivalent to the wheel speed and the vehicle speed,
Figure BDA0002311524100000061
where ε is a small positive constant to avoid a denominator of zero. And adhesion force FdIs the friction force generated by the contact action of the tyre and the road surface, can be described by the longitudinal adhesion coefficient mu and the vertical load N of the wheel as follows,
Fd＝μ(λ)N (5)
therefore, the motor driving force T influences the wheel speed and further influences the change of the slip ratio lambda; aiming at different road surfaces, the boundary of the stable attachment area and the unstable slipping area is different, namely the optimal slip rate is different in size. Under different road surfaces, the friction force generated by the contact action between the tire and the road surface is different, and the curve relation between the slip ratio and the adhesion coefficient is also different, so that the generated adhesion force F is generateddThe acceleration of the vehicle differs from one another.
According to the above formula and the known knowledge, the motor driving torque and the road surface type factor are main external factors causing the change of the adhesion state, the acceleration is a physical quantity generated under the combined action of the external factors, and the change of the three physical quantities causes the change of the vehicle adhesion working point, such as the slip ratio and the adhesion coefficient, thereby influencing whether the vehicle is in a stable adhesion state or an unstable slip state. Therefore, based on the above analysis, by the three signals of the motor driving force, the road surface type, and the acceleration, a mapping relation with the vehicle adhesion state can be established. Due to the characteristics of nonlinearity, complexity and coupling of the model, it is difficult to express the mapping relationship by using a mathematical formula, and the attachment state is identified by complex calculation. Thus, the present invention employs a data-driven approach. The data-driven method of the invention means that the recognizer adopts a data-driven model, not a mechanism model. A preferred data-driven recognizer is to use a neural network model that can implement non-linear mapping and to determine parameters within the network by performing training.
As shown in the attached figures 3 and 4 of the specification, the data-driven method adopted by the invention mainly comprises the following steps of training the recognizer:
and S01, collecting the vehicle running parameters in different scenes, wherein the parameters comprise vehicle motor driving force, road surface type and vehicle acceleration signals.
The motor driving force CAN be measured by a special torque sensor of a motor driving system, or CAN be directly given by a motor driver through a CAN bus message; the acceleration signal measurement may be measured directly by the acceleration sensor. In the data acquisition process, the driving torque of the motor can adopt various input signal types such as slope input, constant value input, sine input and the like, and the pavement types can be set to various pavement types such as ice, rain, snow, dry asphalt pavement and the like, so that the coverage and richness of test data are enhanced.
Specifically, the motor driving torque measurement signal is a signal sequence at the current moment and in a period before the current moment. If the current time is k, the motor driving torque measurement signal refers to a signal time sequence of m +1 dimensions: { Tk-m,...,Tk-1,TkAnd represents the measured values of the motor at the time k, the time k-1, … and the time k-m. The acceleration measurement signal is also a signal sequence at the current moment and a period of time before. If the current time is k, the motor driving torque measurement signal refers to a signal time sequence of m +1 dimensions: { ak-m,...,ak-1,ak}. In order to eliminate the influence of numerical values on the network and facilitate network training, the normalization of the current road surface type parameter is taken as [0, 1]]The numerical value in between. For example, where 0 represents the most slippery road surface, 1 represents the strongest road surface adhesion, and the intermediate value is set according to the specific road surface condition. One solution is to classify the road surface image obtained by the vision sensor, such as ice road surface, snow road surface, rain road surface, semi-dry road surface, etc., and the corresponding road surface type parameter can be set to 0.2, 0.4, 0.6, 0.8, 1.0. Further, the road surface type parameters can be fuzzified, for example, the ice and snow mixed road surface is 0.3.
Because the signal ranges of the motor driving force and the acceleration measurement data are different, network training is not facilitated, and therefore consistency processing is required. The motor driving force and the acceleration are normalized and mapped to an interval of [ -1, 1] according to the following formula:
Figure BDA0002311524100000081
Figure BDA0002311524100000082
and S02, collecting the slip ratio signal and the optimal slip ratio signal at the same time, and calculating the adhesion stability parameter according to the relationship between the slip ratio and the optimal slip ratio, so that the quantitative adhesion stability under different motor driving forces, road surface types and accelerations can be described.
The slip rate signal is obtained by measuring the wheel speed and the vehicle speed and calculating according to the slip rate definition; the optimal slip rate is related to the type of the road surface and can be obtained by off-line test measurement; the slip rate and the optimal slip rate are only used for determining the attachment stability parameters in the training data set, and when the model is actually applied after being trained well, the two parameters do not need to be measured.
According to the slip rate data and the optimal slip rate data, calculating an adhesion stability parameter sigma reflecting the current adhesion stable state according to the following mode:
when the slip ratio is larger than the optimum slip ratio, the adhesion stability parameter is normalized to the interval (0, -1 ]:
Figure BDA0002311524100000083
when the slip ratio is equal to the optimal slip ratio and is in the optimal adhesion state, the adhesion stability parameter sigma is 0;
when the slip ratio is less than the optimum slip ratio, the adhesion stability parameter is normalized to the interval (0,1 ]:
Figure BDA0002311524100000091
in the above formula, λ is the slip ratio, λoptThe calculation result of the adhesion stability parameter σ is shown in fig. 5 in the specification.
S03, establishing a nonlinear neural network model which can realize a complex nonlinear mapping relation, and training network model parameters by using the data obtained in the steps S01 and S02 as a training set.
And training the network by taking the signals (T ', a', theta) as input signals and the sigma as output signals until the network is converged, and determining the weight of the network parameters. Wherein the input signals T ' and a ' are normalized (m +1) -dimensional time-series signals, i.e., T ' ═ Tk′-m,...,Tk′-1,Tk′]，a′＝[a′k-m,...,a′k-1,a′k]。
After the trained network model is used, the attachment stability parameters can be predicted on line.
The invention provides a data-driven electric drive vehicle attachment stability identification method, which comprises the following steps:
s11, measuring the driving torque T of the motor;
s12, measuring the acceleration a of the vehicle along the longitudinal axis of the vehicle body;
s13, identifying a road surface type parameter sigma;
s14, inputting the signals into a trained recognizer based on the three types of signals to obtain a vehicle attachment stability parameter sigma;
s15, determining the adhesion stable state according to the following conditions: σ >0 is stable adhesion, σ ═ 0 is optimal adhesion, and σ <0 is unstable slippage.
In this embodiment, the motor driving torque measurement signal is a signal sequence at the current time and a period before the current time. If the current time is k, the motor driving torque measurement signal refers to a signal time sequence of m +1 dimensions: { Tk-m,...,Tk-1,TkAnd represents the measured values of the motor at the time k, the time k-1, … and the time k-m. The size of m can be adjusted according to specific sampling period and identification effect, for example, the sampling period is 10ms, and m is 9, which represents that the motor driving torque signal sequence is 10 times of measured data in the past 100 milliseconds. The larger the m is selected, the more obvious the trend of the signal is, but the sensitivity of detection response can be correspondingly influenced, and the better balance value can be found by actually adjusting according to the result. Preferably, the measurement data can be obtained 10-20 times, and if the sampling period is 10ms, the data of the motor driving torque 100 and 200ms are preferably obtained.
In this embodiment, the acceleration measurement signal is also a signal sequence before the current time. If the current time is k, the motor driving torque measurement signal refers to a signal time sequence of m +1 dimensions: { ak-m,...,ak-1,ak}. The size of m can be adjusted according to specific sampling period and identification effect, for example, the sampling period is 10ms, and m is 9, which represents that the motor driving torque signal sequence is 10 times of measurement data in the past 100 milliseconds. Similarly, the driving torque signal sequence is the same as that of a motor, the larger the m is selected, the more obvious the trend of the signal is, but the sensitivity of detection response is influenced correspondingly, and the driving torque signal sequence can be actually adjusted according to the result to find a better balance value. Preferably, the measurement data can be obtained 10-20 times, and if the sampling period is 10ms, the data of 100-200ms of the acceleration signal is preferably obtained。
Preferably, since the signal ranges of the motor driving force and the acceleration measurement data are different, the consistency processing is required. The motor driving force and acceleration are normalized and mapped to the [ -1, 1] interval in the calculation manner of the above equations (6) and (7).
In this embodiment, for the requirement of network training, eliminating the influence of the value, and at the same time facilitating the training of the network, the normalization of the current road surface type parameter θ is taken as a value between [0, 1 ]. For example, where 0 represents the most slippery road surface, 1 represents the strongest road surface adhesion, and the intermediate value is set according to the specific road surface condition. One solution is to classify the road surface image obtained by the vision sensor, such as ice road surface, snow road surface, rain road surface, semi-dry road surface, etc., and the corresponding road surface type parameter can be set to 0.2, 0.4, 0.6, 0.8, 1.0. Further, the road surface type parameters can be fuzzified, for example, the ice and snow mixed road surface is 0.3.
In this embodiment, the motor driving force may be measured by a dedicated torque sensor of the motor driving system, or may be directly given by the motor driver through a CAN bus message. The acceleration signal measurement may be measured directly by the acceleration sensor. The road surface type parameters are mainly used for fuzzy representation of road surface adhesion characteristics, a specific obtaining mode of the road surface type parameters is not limited, and the method for obtaining the road surface type parameters by classifying the road surface type based on the image is only one method. In addition, other special sensors, such as optical, acoustic, etc. special sensors, may be used to obtain the road surface type.
Description of the drawings fig. 6 shows a schematic diagram of an apparatus for implementing the method for identifying adhesion stability of an electrically driven vehicle provided by the foregoing embodiment, which includes: a motor driving torque acquisition unit for acquiring a motor driving torque signal; an acceleration acquisition unit for acquiring an acceleration of the vehicle along a longitudinal axis of the vehicle body; and the road surface type parameter acquisition unit is used for acquiring the road surface type parameters. A memory and a processor, wherein the memory stores computer executable instructions, the instructions are called by the processor to identify an attachment stability parameter σ through a trained network model, so as to determine an attachment stability state: σ >0 is stable adhesion, σ ═ 0 is optimal adhesion, and σ <0 is unstable slippage. The above result can provide control basis for other vehicle control units.
The invention provides an identification method and device for estimating adhesion stability of an electrically driven vehicle based on directly measured parameters such as driving force, acceleration and road surface identification types. The driving force and the acceleration signal which can be directly measured are adopted, so that the attachment coefficient can be obtained through the recognizer, signals which are difficult to accurately estimate or measure such as the attachment coefficient, the slip rate and the optimal slip rate do not need to be estimated or calculated, speed measurement signals such as wheel speed do not need to be obtained, the measurement step of obtaining parameters with low bandwidth or inaccurate measurement is avoided, the whole method is simple and reliable, the response is fast, and the practicability is good. By adopting a data driving method, the generalization capability is good, a mechanism model is not needed, and the influence of system nonlinearity uncertainty can be responded, so that the whole system is more stable.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (10)

1. An electric drive vehicle attachment stability identification method is characterized by comprising the following steps:
s11, measuring the driving torque of the motor;
s12, measuring the acceleration of the vehicle along the longitudinal axis of the vehicle body;
s13, identifying road surface type parameters;
s14, inputting the trained recognizer based on the three types of signals to obtain vehicle attachment stability parameters;
the recognizer is trained according to the following steps:
s01, collecting vehicle operation parameters in different scenes, wherein the parameters comprise vehicle motor driving force, road surface types and vehicle acceleration signals;
s02, collecting slip rate signals and optimal slip rate signals at the same time, and calculating adhesion stability parameters according to the relationship between the slip rate and the optimal slip rate, so as to describe quantitative adhesion stability states under different motor driving forces, road surface types and accelerations;
s03, training the network model parameters by using the data obtained in the steps S01 and S02 as a training set;
and S15, judging the vehicle stable state according to the obtained adhesion stable state.
2. The method of claim 1, wherein the identifier is a neural network model that enables non-linear mapping.
3. The method according to claim 1, wherein said calculating the adhesion stability parameter in step S02 is performed according to the following method:
when the slip ratio is greater than the optimum slip ratio, the adhesion stabilization parameter is normalized to the interval (0, -1 ]:
Figure FDA0002907593800000021
when the slip ratio is equal to the optimum slip ratio, the adhesion stability parameter σ is 0;
when the slip ratio is less than the optimum slip ratio, the adhesion stabilization parameter is normalized to the interval (0,1 ]:
Figure FDA0002907593800000022
in the above formula, λ is the slip ratio, λopt Is the optimum slip ratio.
4. The method according to claim 3, wherein said judging operation of said step S15 is: stable adhesion is achieved when the adhesion stability parameter σ >0, optimal adhesion is achieved when the adhesion stability parameter σ is 0, and unstable slippage is achieved when the adhesion stability parameter σ < 0.
5. Method according to claim 1, characterized in that the motor drive torque measurement signal is a signal sequence at the present moment and a time preceding it; the acceleration measurement signal is a signal sequence at the current moment and a period of time before.
6. The method according to claim 1, wherein the motor driving torque measurement signal and the acceleration measurement signal input to the trained recognizer at step S14 are normalized signals.
7. The method according to claim 1, characterized in that said road surface type parameter is a value between [0, 1 ].
8. The method of claim 7, wherein the road surface type parameters are fuzzified according to road surface conditions.
9. The method of claim 1, wherein the road surface type parameter is obtained by classifying the road surface type based on an image.
10. An electric-drive-vehicle attachment-stability identifying device for implementing the electric-drive-vehicle attachment-stability identifying method of any one of claims 1 to 9, comprising:
a motor driving torque acquisition unit for acquiring a motor driving torque signal;
an acceleration acquisition unit for acquiring an acceleration of the vehicle along a longitudinal axis of the vehicle body;
the road surface type parameter acquisition unit is used for acquiring road surface type parameters;
a memory in which computer-executable instructions are stored, and a processor that calls the instructions to acquire an adhesion stability parameter from data of the motor driving torque acquisition unit, the acceleration acquisition unit, and the road surface type parameter acquisition unit through a trained recognizer, thereby determining an adhesion stability state of a vehicle.
CN201911260714.5A 2019-12-10 2019-12-10 Data-driven electric drive vehicle attachment stability identification method and device Active CN110920626B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201911260714.5A CN110920626B (en) 2019-12-10 2019-12-10 Data-driven electric drive vehicle attachment stability identification method and device
PCT/CN2020/129524 WO2021115063A1 (en) 2019-12-10 2020-11-17 Data-driven method and apparatus for identifying adhesion stability of electric-driven vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911260714.5A CN110920626B (en) 2019-12-10 2019-12-10 Data-driven electric drive vehicle attachment stability identification method and device

Publications (2)

Publication Number Publication Date
CN110920626A CN110920626A (en) 2020-03-27
CN110920626B true CN110920626B (en) 2021-06-04

Family

ID=69858721

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911260714.5A Active CN110920626B (en) 2019-12-10 2019-12-10 Data-driven electric drive vehicle attachment stability identification method and device

Country Status (2)

Country Link
CN (1) CN110920626B (en)
WO (1) WO2021115063A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110920626B (en) * 2019-12-10 2021-06-04 中国科学院深圳先进技术研究院 Data-driven electric drive vehicle attachment stability identification method and device
CN112699924A (en) * 2020-12-22 2021-04-23 安徽卡思普智能科技有限公司 Method for identifying lateral stability of vehicle
CN112660135B (en) * 2020-12-25 2022-11-15 浙江吉利控股集团有限公司 Road surface adhesion coefficient estimation method and device
CN114987443B (en) * 2022-06-29 2023-07-18 小米汽车科技有限公司 Wheel slip rate control method, device, vehicle and storage medium
CN116443022B (en) * 2023-06-19 2023-08-15 成都赛力斯科技有限公司 Method and device for adjusting vehicle based on road surface tire adhesion coefficient

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105654067A (en) * 2016-02-02 2016-06-08 北京格灵深瞳信息技术有限公司 Vehicle detection method and device
CN109910897A (en) * 2019-01-30 2019-06-21 江苏大学 A kind of calculation of safety distance method based on road surface ahead peak adhesion coefficient

Family Cites Families (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4413931B2 (en) * 2007-01-18 2010-02-10 株式会社日立製作所 Vehicle and vehicle control device
JP2009113717A (en) * 2007-11-08 2009-05-28 Fuji Heavy Ind Ltd State estimation system
US9043113B2 (en) * 2012-05-31 2015-05-26 Caterpillar Inc. Drive system having ongoing pull-slip learning
GB2508462B (en) * 2012-08-16 2015-01-07 Jaguar Land Rover Ltd System and method for controlling vehicle speed
CN103612634B (en) * 2013-12-03 2016-01-06 北京交通大学 The evaluation method of distributing In-wheel motor driving electronlmobil coefficient of road adhesion
CN104210383B (en) * 2014-09-18 2016-05-25 上海工程技术大学 A kind of four-wheel drive electric automobile torque distribution control method and system
KR101594730B1 (en) * 2015-07-09 2016-02-23 고성석 Device for preventing sudden unexpected acceleration of automobiles
US9914475B2 (en) * 2015-10-05 2018-03-13 Delphi Technologies, Inc. Humanized steering model for automated vehicles
KR101827152B1 (en) * 2016-10-04 2018-03-22 현대자동차주식회사 Method for deciding a road surface using vehicle data
CN106599495B (en) * 2016-12-20 2017-09-05 湖北文理学院 Optimal slip ratio recognition methods based on glowworm swarm algorithm Optimized BP Neural Network
EP3562702A1 (en) * 2016-12-30 2019-11-06 Elaphe Propulsion Technologies Ltd. Arrangement for determining maximum allowable torque
DE102017203171A1 (en) * 2017-02-27 2018-08-30 Zf Friedrichshafen Ag Determination of a maximum traction coefficient
GB2562308B (en) * 2017-05-12 2022-10-26 Arrival Ltd Regenerative braking control system
CN107253482B (en) * 2017-05-22 2019-07-23 上海大学 A kind of Driving safety assistant system based on pavement image analysis
DE102017124990B4 (en) * 2017-10-25 2021-11-04 Dr. Ing. H.C. F. Porsche Aktiengesellschaft Method and system for operating a motor vehicle
DE102017221887A1 (en) * 2017-12-05 2019-06-06 Zf Friedrichshafen Ag Method for operating a drive train of a motor vehicle
US10800423B2 (en) * 2018-02-20 2020-10-13 Deere & Company Monitoring steering conditions of an off-road vehicle
DE102018202918A1 (en) * 2018-02-27 2019-08-29 Zf Friedrichshafen Ag Method for operating a drive train of a motor vehicle
CN108715166B (en) * 2018-04-28 2023-05-12 南京航空航天大学 Vehicle stability index estimation method based on deep learning
CN108622101A (en) * 2018-05-09 2018-10-09 南京航空航天大学 A kind of coefficient of road adhesion method of estimation under motor turning operating mode
CN109466558B (en) * 2018-10-26 2020-08-04 重庆邮电大学 Road adhesion coefficient estimation method based on EKF (extended Kalman Filter) and BP (Back propagation) neural network
CN109774717B (en) * 2018-12-24 2020-09-11 科大讯飞股份有限公司 Tire friction force regulation and control method and system
CN109849899B (en) * 2019-02-01 2021-01-15 江苏大学 Electro-hydraulic composite vehicle body stability control system and method for electric wheel vehicle
CN110203205B (en) * 2019-05-31 2021-01-12 上海大学 Method for detecting adhesion stability and quasi-optimal slip rate of electric automobile
CN110497916B (en) * 2019-08-15 2021-03-05 太原科技大学 Automobile driving state estimation method based on BP neural network
CN110509916B (en) * 2019-08-30 2021-06-29 的卢技术有限公司 Vehicle body posture stabilizing method and system based on deep neural network
CN110920626B (en) * 2019-12-10 2021-06-04 中国科学院深圳先进技术研究院 Data-driven electric drive vehicle attachment stability identification method and device

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105654067A (en) * 2016-02-02 2016-06-08 北京格灵深瞳信息技术有限公司 Vehicle detection method and device
CN109910897A (en) * 2019-01-30 2019-06-21 江苏大学 A kind of calculation of safety distance method based on road surface ahead peak adhesion coefficient

Also Published As

Publication number Publication date
CN110920626A (en) 2020-03-27
WO2021115063A1 (en) 2021-06-17

Similar Documents

Publication Publication Date Title
CN110920626B (en) Data-driven electric drive vehicle attachment stability identification method and device
JP6921096B2 (en) Devices and methods to improve the performance of vehicle anti-lock braking and anti-slip regulation
US6508102B1 (en) Near real-time friction estimation for pre-emptive vehicle control
US6324461B1 (en) Road surface condition estimating apparatus and variation reduction processing apparatus
EP2995520B1 (en) Tire grounded state estimation method
US9988057B2 (en) Method and device for determining the mass of a motor vehicle, and a motor vehicle with a device of this type
US6839615B2 (en) Method and apparatus for estimating mass of vehicle, and method and apparatus for estimating gradient employing the method
US11650134B2 (en) Determining tread depth using data from a tire-mounted sensor
JP5339121B2 (en) Slip rate estimation device and method, and slip rate control device and method
CN107933562B (en) Method and system for calculating road friction estimates
CN110203205B (en) Method for detecting adhesion stability and quasi-optimal slip rate of electric automobile
CN111114551B (en) Vehicle ramp gradient identification method and device
CN109159787B (en) Real-time detection system and method for tire adhesion steady state of electric automobile
CN109080636A (en) Commerial vehicle uphill starting auxiliary control method based on real-time quality identification
US20230202241A1 (en) Enhanced tracking of tire tread wear
EP4061679B1 (en) System for determining a wheel-rail adhesion value for a railway vehicle
JPH11334555A (en) Road surface friction coefficient estimating device
US11453400B2 (en) Method for estimating an index representative of the frictional behavior of a vehicle on a road
JP5183275B2 (en) Road surface state determination device and method, and road surface state determination program
CN110509780B (en) System and method for controlling road surface sudden change stability and optimizing maximum driving/braking torque of electric automobile
JP2016217065A (en) Measurement device for road surface roughness
WO2022066867A1 (en) Adjustment of indirectly determined values of a tire monitoring system
CN209022906U (en) A kind of electric car tire attachment stable state real-time detecting system
KR101912515B1 (en) System and method for detecting condition of driver by applying torque
CN211995173U (en) System for controlling stability and optimizing torque of electric automobile during road surface sudden 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
GR01 Patent grant
GR01 Patent grant