CN110531740B - Intelligent degree quantitative evaluation method for intelligent vehicle - Google Patents

Intelligent degree quantitative evaluation method for intelligent vehicle Download PDF

Info

Publication number
CN110531740B
CN110531740B CN201910843325.9A CN201910843325A CN110531740B CN 110531740 B CN110531740 B CN 110531740B CN 201910843325 A CN201910843325 A CN 201910843325A CN 110531740 B CN110531740 B CN 110531740B
Authority
CN
China
Prior art keywords
intelligent vehicle
intelligent
tested
vehicle
degree
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
CN201910843325.9A
Other languages
Chinese (zh)
Other versions
CN110531740A (en
Inventor
王建强
黄荷叶
郑讯佳
涂茂然
许庆
李研强
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.)
Tsinghua University
Original Assignee
Tsinghua 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 Tsinghua University filed Critical Tsinghua University
Priority to CN201910843325.9A priority Critical patent/CN110531740B/en
Publication of CN110531740A publication Critical patent/CN110531740A/en
Application granted granted Critical
Publication of CN110531740B publication Critical patent/CN110531740B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B23/00—Testing or monitoring of control systems or parts thereof
    • G05B23/02—Electric testing or monitoring
    • G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
    • G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
    • G05B23/0224—Process history based detection method, e.g. whereby history implies the availability of large amounts of data
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/20—Pc systems
    • G05B2219/24—Pc safety
    • G05B2219/24065—Real time diagnostics

Abstract

The invention discloses an intelligent degree quantitative evaluation method for an intelligent vehicle, which comprises the following steps: s1, carrying out hierarchical division on the intelligent degree of the intelligent vehicle according to the autonomous driving function; s2, constructing an intelligent degree evaluation environment of the intelligent vehicle; s3, setting an intelligent vehicle evaluation scheme, wherein the scheme comprises a test environment, a test task, an evaluation object and an evaluation index; s4, selecting an intelligent evaluation index, and constructing a proper evaluation index system by analyzing key parameters representing and evaluating the driving process of the intelligent vehicle; s5, acquiring multiple groups of quantitative evaluation basis data of the tested intelligent vehicle under different test task scenes; and S6, obtaining the corresponding intelligentization degree when the highest-level unmanned driving is realized, and evaluating the intelligentization degree of the tested automatic driving system of the tested intelligent vehicle by taking the intelligentization degree as a reference. The invention can completely and truly evaluate the driving auxiliary system and the intelligent vehicle to different intelligent degrees.

Description

Intelligent degree quantitative evaluation method for intelligent vehicle
Technical Field
The invention relates to the technical field of driving assistance systems and intelligent vehicle evaluation, in particular to an intelligent degree quantitative evaluation method for an intelligent vehicle.
Background
Smart vehicles offer the potential to increase vehicle productivity, the safety and efficiency of transportation systems, and as the demand for people increases, autonomous driving systems become more and more complex, and must therefore be tested effectively to verify whether smart vehicles can better serve people. The evaluation of the degree of intellectualization is considered to be an important component of the research of intelligent vehicles; the intelligent vehicle can evaluate the intelligent degree from each component system of the intelligent vehicle to the whole, is beneficial to finding research problems, and improves the safety of autonomous driving. The existing intelligent vehicle test and evaluation project is gradually carried out, a plurality of specific test centers, such as the McBashi of the Michigan university transport school, have already carried out the work of testing autonomous driving, and various software simulation tests and hardware-in-the-loop test technologies are also gradually applied.
However, these test approaches are closed, simulated, limited to specific traffic scenarios, and difficult to reproduce or simulate real complex traffic with human-vehicle-environment interactions. The project for carrying out the intelligent vehicle field test is mainly unmanned vehicle challenge games of various countries, such as European land robot test robot competition, unmanned vehicle competition organized by the United states DARPA, future intelligent vehicle challenge games of China and the like, and the challenge games are dedicated to developing an active vehicle safety test and evaluation method and testing the vehicle cooperative driving level in a simulated urban environment. Through years of experience accumulation, intelligent vehicle challenge races gradually form a relatively perfect intelligent behavior evaluation system suitable for unmanned driving, however, related tests of the intelligent vehicles do not cover all real traffic scenes, different intelligent levels corresponding to the safety levels of the intelligent vehicles are not obtained, and whether the intelligent degree of the test vehicles can ensure that the safe driving of the intelligent vehicles is still to be testified. Meanwhile, in the whole development process, including function development and testing, system integration and verification, test driving and verification, and the like, a complete autonomous vehicle testing method is still very needed.
Therefore, some problems of the existing intelligent vehicle intelligent degree quantitative evaluation method are summarized:
1. the definition of "intelligence" of the intelligent degree of an intelligent vehicle is not clear. The existing evaluation method does not provide a specific measure for intellectualization of evaluation, and indexes covered by intelligence are not clearly defined and the capability of taking adaptive decision and action in the environment is not quantified.
2. The intelligent degree evaluation technology of the intelligent vehicle lacks systematicness. Due to the particularity of the intelligent degree of the intelligent vehicle (the unknown of the test environment and the object and the comprehensiveness of the evaluation index), the evaluation method of the system needs to comprehensively realize the functions of test environment construction, system identification, collision process analysis and the like, and ensures that the requirements of expandability, high precision, quick operation, repeatable results and the like are met, but the existing evaluation method has independent functions and is not systematic.
3. The intelligent degree evaluation index of the intelligent vehicle is single. Because the intelligent vehicle belongs to the integrated intelligent body, the evaluation method of the existing autonomous unmanned system mainly comprises the following steps: grade evaluation method, biaxial method, triaxial method, table look-up method, formula method, and spider web evaluation model method. The autonomous evaluation method of the unmanned system must consider diversity, multi-dimensionality, hierarchy, and primary and secondary application objectives and the nature of the system itself. These evaluation methods not only do not evaluate a single index sufficiently, but also evaluation indexes driven by similar tasks may not meet the driving requirements of the driver.
Therefore, the performance evaluation method of the existing intelligent vehicle and the related technology are difficult to meet the development requirements of the intelligent vehicle technology. Therefore, in order to solve the above problems, it is necessary to develop a method and an apparatus for quantitatively evaluating the degree of intelligence of an intelligent vehicle.
Disclosure of Invention
The invention aims to provide an intelligent degree quantitative evaluation method for an intelligent vehicle, which can completely and truly evaluate different intelligent degrees of a driving auxiliary system and the intelligent vehicle.
In order to achieve the purpose, the invention provides an intelligent degree quantitative evaluation method for an intelligent vehicle, which comprises the following steps: s1, according to the autonomous driving function, the intelligent degree of the intelligent vehicle is divided into layers, and the layers are mainly divided into an environment perception layer, a behavior decision layer and a motion planning layer; s2, testing the intelligent degree of the equipment on the same layer fused by the single autonomous driving function driving auxiliary system or the multiple autonomous driving function systems, and building an intelligent degree evaluation environment of the intelligent vehicle corresponding to the layers divided in S1; s3, setting a corresponding intelligent vehicle evaluation scheme according to the intelligent degree evaluation environment corresponding to each level provided by S2, wherein the scheme comprises a test environment, a test task, an evaluation object and an evaluation index; s4, aiming at the evaluation indexes provided by S3, selecting intelligent evaluation indexes, and constructing a proper evaluation index system by analyzing and characterizing key parameters of the driving process of the intelligent vehicle; s5, acquiring multiple groups of quantitative evaluation basis data of the tested intelligent vehicle in different test task scenes in the test task scene set in the S3; and S6, performing statistical analysis on each group of quantitative evaluation basis data obtained in S5, obtaining the corresponding intelligent degree when the highest-level unmanned driving is realized according to the intelligent evaluation index in S4 to which each statistical analysis result belongs, and evaluating the intelligent degree of the tested automatic driving system of the tested intelligent vehicle by taking the intelligent degree as a reference.
Furthermore, the intelligent evaluation indexes in the S4 comprise safety, high efficiency, rationality and comfort; in a traffic system with n road users, the general indicator S provided by the formula (12) is usedintThe method represents the multi-target pursuit of the tested intelligent vehicle i on the intelligent evaluation indexes in the driving process:
Figure BDA0002194405290000031
Figure BDA0002194405290000032
in the formulae (12) and (13), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifM is the termination time of the driving process of the tested intelligent vehicle iiFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredjFor the speed of the other road users j,
Figure BDA0002194405290000038
for the lateral impact of the tested intelligent vehicle i,
Figure BDA0002194405290000039
for the longitudinal impact, R, of the intelligent vehicle i to be testediLongitudinal constraint resistance to the driver for traffic regulations, GiFor virtual gravitation, vi，xFor the longitudinal speed of the intelligent vehicle i to be measured, FIi，1And FIi，2The transverse restraining force v of the driver is respectively the two side lane linesi，yFor the transverse speed of the intelligent vehicle i to be measured, FjiThe external force on the tested intelligent vehicle i caused by other road users j is provided.
Further, the safety is mainly determined by the driving risk level U of the tested intelligent vehicle i in a specific traffic scene represented by the formula (1)riskAnd calculating to obtain:
Figure BDA0002194405290000033
Figure BDA0002194405290000034
Figure BDA0002194405290000035
Figure BDA0002194405290000036
Figure BDA0002194405290000037
in formulae (1) to (5), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifM is the termination time of the driving process of the tested intelligent vehicle iiFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredlimitFor lane speed limit,. tau.for calibration parameter,. vi，xFor the longitudinal speed, v, of the intelligent vehicle i to be measuredi，yFor the lateral speed, v, of the intelligent vehicle i to be measuredjSpeed of other road users,/tOf the lane type, rliFor the distance between the measured intelligent vehicle i and the road boundary, lwIs the width of the lane, dijIs the straight-line distance between the measured intelligent vehicle i and the central point of other road users j, vijIs the relative speed, theta, between the measured intelligent vehicle i and other road users jijIs from dijTo vijThe angle of,
Figure BDA0002194405290000041
is from dijTo viAnd the counterclockwise direction is positive, mjBeing the actual physical mass of the other road users j,
Figure BDA0002194405290000042
to relate to the j speed v of other road usersjFunction of, TjThe type of other road users j.
Further, the efficiency is mainly represented by the formula (7) of the efficiency U of the task completed by the intelligent vehicle to be testedeffCarrying out weighing:
Figure BDA0002194405290000043
Gi＝mig sinθi(8)
Figure BDA0002194405290000044
in formulae (7) to (9), miFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredi，xThe longitudinal speed of the intelligent vehicle i to be measured is shown, and g is the gravity acceleration; k is a constant, vderIs the desired speed, v, of the driverlimitThe speed limit of the lane is realized.
Further, the rationality is mainly represented by the driver psychological expectation U assumed/actually present by the expression (10)comAnd calculating to obtain:
Figure BDA0002194405290000045
Figure BDA0002194405290000046
Figure BDA0002194405290000047
in the formulae (4), (5) and (10), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifFor the intelligent vehicle i under testEnd time of driving process, miFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredjSpeed, v, of other road usersijFor the relative speed between the measured intelligent vehicle i and other road users j, dijIs the straight-line distance theta between the measured intelligent vehicle i and the central point of other road users jijIs from dijTo vijThe angle of,
Figure BDA0002194405290000051
is from dijTo viAnd the counterclockwise direction is positive, mjBeing the actual physical mass of the other road users j,
Figure BDA0002194405290000052
to relate to the j speed v of other road usersjFunction of, TjThe type of other road users j.
Further, the comfort is reflected by a vehicle performance index, which is defined as a comfort soft constraint function U represented by equation (11)smooth：
Figure BDA0002194405290000053
In the formula (11), the reaction mixture is,
Figure BDA0002194405290000054
for the lateral impact of the tested intelligent vehicle i,
Figure BDA0002194405290000055
the longitudinal impact degree of the intelligent vehicle i to be measured.
Further, "representing the degree of safety, efficiency, comfort and reasonable intelligence of the driving process" in S6 is defined as DintThe higher the value is, the higher the intelligentization degree is, the data is uniformly mapped to 0, 1]In the interval:
Figure BDA0002194405290000056
in the formula (15), SintThe comprehensive index of the intelligent degree of the multi-target pursuit of the safety, the high efficiency, the rationality and the comfort of the tested intelligent vehicle i in the driving process is provided,
Figure BDA0002194405290000057
the minimum action amount of the whole test theory is completed when the intelligent vehicle reaches the highest intelligent level.
Further, the air conditioner is provided with a fan,
Figure BDA0002194405290000058
obtained by using a data mining average method, or obtained by using the following formula (14):
Figure BDA0002194405290000059
furthermore, on an environment perception layer, the motion states and physical attributes of different objects are predicted and filtered according to the reliability of data fused by the equipment with the environment perception function, and meanwhile, the relative distance, the motion direction, the speed, the posture and the positioning estimation of a road user are identified, so that the environment perception capability of the equipment with the environment perception function carried on the tested intelligent vehicle is independently evaluated;
in the behavior decision layer, when the tested intelligent vehicle runs in a structured or unstructured road environment, whether the intellectualization degree of the running process is improved or not is judged through a strategy after decision so as to independently evaluate the behavior decision capability of the tested intelligent vehicle;
on the motion planning layer, a safe and feasible driving track is generated through speed planning and local path planning of the driving process of the tested intelligent vehicle, and whether the track planning meets the set judgment indexes of safety, efficiency, rationality and comfort or not is independently evaluated.
Due to the adoption of the technical scheme, the invention has the following advantages: 1. the novel intelligent vehicle intelligent degree quantitative evaluation method provided by the invention can be used for openly, quantitatively and fairly measuring the intelligent degree of an intelligent vehicle from a single driving auxiliary system to an integral intelligent vehicle, and specifically can be used for evaluating the performance of the vehicle and detecting the task execution condition of the vehicle from three levels of environment perception, behavior decision and motion planning. 2. The intelligent vehicle intelligent degree quantitative evaluation method is suitable for a human-vehicle-road complex environment interaction system, considers the driving track (speed, acceleration and the like), the driving time, specific braking operation and the like of the intelligent vehicle in the whole test process, quantifies indexes of the operations, and enables the quantitative evaluation result of the intelligent level to be more reasonable and reliable based on real operation behaviors. 3. The intelligent degree quantitative evaluation method for the intelligent vehicle comprehensively considers the indexes of driving safety, high efficiency, reasonability, comfort and the like, comprehensively models the intelligent vehicle evaluation according to the traffic environment, defines the intelligent level of the intelligent vehicle to be evaluated based on the theoretical minimum action quantity, establishes a complete evaluation index system of the intelligent level of the unmanned vehicle, and can directly support the intelligent vehicle evaluation as a result.
Drawings
FIG. 1 is a frame diagram of intelligent vehicle intelligent degree multi-level quantitative evaluation provided by the present invention;
FIG. 2 is a flow chart for evaluating the intelligent degree of the intelligent vehicle provided by the invention;
FIG. 3 is a schematic diagram of a test driving task design of the intelligent vehicle provided by the invention;
fig. 4 to 11 are schematic diagrams of different test scenarios for testing the intelligent degree of the intelligent vehicle provided by the invention.
Detailed Description
In the drawings, the same or similar reference numerals are used for the same or similar elements or elements having the same or similar functions. Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the intelligent degree quantitative evaluation method for the intelligent vehicle provided by this embodiment includes the following steps:
and S1, according to the autonomous driving function, carrying out hierarchical division on the intelligent degree of the intelligent vehicle, wherein the hierarchical division is mainly divided into an environment perception layer, a behavior decision layer and a motion planning layer.
And S2, testing the intelligent degree of the equipment on the same layer fused by the single autonomous driving function driving assistance system or the multiple autonomous driving function systems, and building an intelligent degree evaluation environment of the intelligent vehicle corresponding to the layers divided in the S1.
1. On the environment perception layer, the environment perception capability of the tested intelligent vehicle can be independently evaluated, namely equipment with the environment perception function, such as multiple sensors including a laser radar, a millimeter wave radar and a camera, carried on the tested intelligent vehicle is quantitatively evaluated, and the fusion detection precision of the multiple sensors is evaluated. The main detection objects include structured roads, traffic signs, dynamic and static road users. In the test process, the tested intelligent vehicle predicts and filters the motion states and physical attributes of different objects according to the reliability of data after the multi-sensor fusion, and simultaneously identifies the relative distance, motion direction, speed, posture and positioning estimation of other road users, including the number statistics and attribute categories (such as pedestrians, traffic signs, traffic lights, lane lines, other vehicles and other related elements) of the other road users.
2. And in the behavior decision layer, the behavior decision capability of the tested intelligent vehicle can be independently evaluated. The evaluation behavior decision system mainly comprises a driving behavior decision module which can make reasonable driving behaviors under complex and changeable traffic scenes when the tested intelligent vehicle runs in a structured or unstructured road environment. The main assessment indexes comprise whether the intellectualization degree of the driving process is improved by the strategy after decision making, and preliminary assessment can be carried out by recording the reaction time, the operation correctness and the like. The main decision-making behaviors include lane keeping, lane changing, overtaking, passing through an intersection, and the like. On the basis of meeting the output accuracy of the environment perception layer, the evaluation of the behavior decision layer can be realized. On the contrary, if the perception performance of the measured intelligent vehicle is insufficient, the interference of a behavior decision layer can be caused.
3. And on the motion planning layer, the motion planning capability of the tested intelligent vehicle can be independently evaluated. The motion planning layer is mainly used for carrying out speed planning and local path planning on the running process of the tested intelligent vehicle to generate a safe and feasible running track. And receiving a command into the motion planning layer through the behavior decision layer, and reasonably planning the trajectory of the tested intelligent vehicle from the starting point to the target point. The main assessment indexes comprise the judgment indexes of whether the driving track meets the intelligent requirement or not, namely whether the track planning meets the set safety, high efficiency, rationality, comfort and the like.
The evaluation of the functions can be output through a sensor and an actuator of the tested intelligent vehicle, and the driving auxiliary system with single function can be evaluated in a test driving or simulation mode before the whole-course test.
And S3, setting a corresponding intelligent vehicle evaluation scheme according to the intelligent degree evaluation environment corresponding to each level provided by S2, wherein the scheme comprises a test environment, a test task, an evaluation object and an evaluation index.
Further, the setting of the evaluation scheme of the intelligent vehicle to be tested in S3 specifically includes:
the construction of the tested intelligent vehicle evaluation experiment platform needs to meet the requirements of a closed test field, repeatable experiments, remote operation and the like, so that a real driving scene is restored by arranging the closed test field to provide a test environment for the tested intelligent vehicle test, and meanwhile, the test scene can cover dangerous scenes, accident scenes, natural driving scenes and the like; the specific evaluation object covers intelligent driving auxiliary systems with different functions, intelligent vehicles to be tested with different grades, vehicles of different types (trucks, SUVs and cars) and the like. The evaluation indexes include a safety degree, an efficient degree, a rationality degree, a comfort degree and the like. In the test process, a referee vehicle is arranged behind each tested intelligent vehicle and is responsible for emergency stop of the tested intelligent vehicle in front, and measurement of observed traffic regulations and evaluation indexes. On the key evaluation road segment, there is a live judge whether it is responsible for measurement (time, speed and distance), recording and measurement.
In the testing process, the tested intelligent vehicle needs to pass through each testing scene one by one, corresponding tasks are completed in different testing scenes in sequence, and the whole driving track (speed, acceleration and the like), the driving time, specific braking operation and the like are recorded. These can be measured by an instrument device mounted on the following judgment vehicle. The vehicle odometer and the wheel encoder are used for remotely measuring the running distance of the measured intelligent vehicle. The calibrated camera is used for short range distance estimation. The speedometer is used to estimate the vehicle speed and the timer is used to record the time to complete the mission. For the index that is difficult to measure, the present embodiment employs observation and recording by a live reviewer, or obtaining by analyzing recorded video after a match.
And S4, selecting an intelligent evaluation index according to the evaluation index provided by S3, and constructing a suitable evaluation index system by analyzing and characterizing key parameters of the driving process of the intelligent vehicle. Furthermore, by selecting a proper evaluation method, the driving track and the operation behavior in the test process of the tested intelligent vehicle are analyzed according to the excellent manned behavior standard, the behavior of the tested intelligent vehicle is effectively evaluated through the established evaluation index system, and a basis is provided for the intelligent behavior evaluation system of the tested intelligent vehicle.
Further, the selecting of the intelligent evaluation indexes in S4 specifically includes:
and the intelligent degree of the tested intelligent vehicle is evaluated according to the output in the test task through the established intelligent degree standard. The intelligent indexes specifically included in the intelligent degree standard can cover safety, high efficiency, reasonability and comfort in the driving process of the vehicle. In the comprehensive road environment, due to the fact that driving scenes are complex and changeable, behaviors of traffic participants are difficult to predict, and requirements of people on driving safety, efficiency, rationality, comfort and the like are improved, for the whole dynamic complex process, an intelligent decision system of an intelligent vehicle is equivalent to the brain of a human driver, the human driver is in an objective state, and the driver divides a driving path into operations of a steering wheel, an accelerator and a brake pedal, so that a planning target is executed. In the whole driving process, the tested intelligent vehicle embodies high real-time and intelligent characteristics.
The safety in the intelligent evaluation index is mainly determined by the driving risk level U of the tested intelligent vehicle i in the specific traffic scene represented by the formula (1)riskAnd is calculated. When the tested intelligent vehicle i runs in the road environmentAnd as the road environment also comprises other road users j, the dynamic interaction among different road users brings hidden dangers to the safe driving of the tested intelligent vehicle. The degree of safety of the tested intelligent vehicle depends on the operation behavior of the tested intelligent vehicle. In this embodiment, the action process of the external force of the surrounding driving environment, which is the risk caused by the manipulation of the tested smart car, is defined, and in a traffic system with n road users, the driving risk U is setriskSpecifically characterized by formula (1):
Figure BDA0002194405290000091
in the formula (1), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifFor the end time of the driving process of the tested intelligent vehicle i, RiLongitudinal constraint resistance for the driver for traffic regulations; fli，1And Fli，2Representing the lateral restraining force of the lane lines on the two sides on the driver; v. ofiFor the speed, v, of the intelligent vehicle i to be measuredi，xFor the longitudinal speed (along the lane line direction), v, of the intelligent vehicle i to be measuredi，yIs the transverse speed (vertical lane line direction) v of the tested intelligent vehicle ijThe speed of other road users. FjiAnd (4) potential driving risks brought by external force of other road users j on the i, namely characteristic of interaction between vehicles.
Defining the constraint resistance R of traffic rules (mainly speed limit rules) to driversiSatisfies the following formula:
Figure BDA0002194405290000092
in the formula (2), τ is a calibration parameter, and in this example, τ is 1.
Meanwhile, the lane boundary (such as a lane line) has a constraint effect on the transverse motion of the intelligent vehicle i to be detected. Typically researchers use a spring model to describe this constraint:
Figure BDA0002194405290000093
in the formula (3), ltOf the lane line type (e.g. dashed line l)tWith 2, the solid line is given byt＝3)；rliThe distance between the intelligent vehicle i to be detected and the road boundary is obtained; lwIs the lane width.
Defining the external force F caused by other road users j to the tested intelligent vehicle ijiSatisfies formula (4):
Figure BDA0002194405290000094
in the formula (4), EjiThe sum of the relative kinetic energy between the tested intelligent vehicle i and other road users j; dijThe distance is the straight line distance between the measured intelligent vehicle i and the central points of other road users j, wherein when the other road users j indicate vehicles, the central points of the other road users j correspond to the axes of the vehicles; when the other road users j indicate other objects than the vehicle, the center points of the other road users j correspond to the geometric center of the object. v. ofi、vijAre all vectors, vijThe relative speed between the measured intelligent vehicle i and the relative speed between other road users j; thetaijIs from dijTo vijThe angle of,
Figure BDA0002194405290000095
is from dijTo viThe included angle and the anticlockwise direction are positive; mjIs the virtual mass of a road user j, which is equivalent to a field source in an electric field, and is expressed as follows:
Figure BDA0002194405290000101
in the formula (5), mjIs the actual physical mass, v, of other road users jjIs the speed of the other road users j.
Figure BDA0002194405290000102
Is about other road usersj speed vjThe function of (a) describing the impact of speed on driving risk can be calibrated by analyzing the relationship between the loss of accident and the speed that α, β and gamma are all undetermined constantsjIs the type of other road users j and is determined in the following manner TiThe determination method is the same:
firstly, selecting a class of objects as a reference, and recording a T value corresponding to the reference object as 1; then, the T values for other types of objects are calculated as follows:
Figure BDA0002194405290000103
in formula (6), ξ*Average number of deaths from accident for reference objects (for simplicity, the average number of deaths from accident is used herein to measure the accident loss), ξiIs TiThe average number of deaths from accidents caused by type objects. Wherein, the 'average number of deaths in accident' can be obtained by inquiring the national traffic accident data statistical table.
The efficiency of the intelligent vehicle to be tested in the intelligent evaluation index is mainly balanced by the efficiency of the intelligent vehicle to be tested for completing tasks, which is represented by the formula (7).
Figure BDA0002194405290000104
In the formula (7), the virtual gravitational force GiThe virtual gravitation caused by the traffic environment representing the requirement of the running intelligent vehicle i on mobility can be represented by the following formula (8):
Gi＝mig sinθi(8)
in the formula (8), miThe quality of the intelligent vehicle i to be tested; g is the acceleration of gravity; thetaiIn the invention, theta is defined as the inclination angle of a U-shaped groove in a physical model, which is related to the pursuit of a driving speed by a driveriSatisfies formula (9):
Figure BDA0002194405290000105
in the formula (9), k is a constant, and k is 0.2 in the invention; v. ofderIs the driver's desired speed; v. oflimitThe speed limit of the lane is realized.
The reasonableness in the intelligent evaluation index mainly reflects the assumed/actually existing psychological expectation degree of a driver, namely the difference between the actual operation behavior and the expected behavior of the tested intelligent vehicle, and particularly when other road users exist, the tested intelligent vehicle strives to make the operation behavior of the tested intelligent vehicle and surrounding vehicles consistent as much as possible for the safety and the high efficiency of driving, so that the traffic flow is stable. Specifically defined by formula (10):
Figure BDA0002194405290000111
in the formula (10), miAnd viRespectively the mass and speed, v, of the intelligent vehicle i to be testedjThe speed of other road users. FjiExternal force, t, on i caused by j for other road users0And tfRespectively the starting time and the ending time of the driving process.
The comfort in the intelligent evaluation index is reflected by a vehicle performance index, the mechanical output of the automobile is measured, and the driving experience of drivers and passengers is mainly involved. Mainly aiming at the change of the psychophysical comfort of passengers caused by horizontal vibration of the mechanical structure and assembly manufacturing of the vehicle caused by the maneuverability of decision making of an automatic driving vehicle, wherein the passenger is obviously impacted by the maneuvering behaviors such as rapid acceleration, rapid deceleration and the like, and a comfortable soft constraint function U is definedsmoothIs represented by formula (11):
Figure BDA0002194405290000112
in the formula (11), the reaction mixture is,
Figure BDA0002194405290000113
the transverse (vertical lane line direction) impact degree of the tested intelligent vehicle i,
Figure BDA0002194405290000114
the longitudinal (along the lane line direction) impact degree of the tested intelligent vehicle i is shown.
Furthermore, in a traffic system with n road users, based on the principle of minimum action quantity, the multi-target pursuit of the tested intelligent vehicle i on safety, high efficiency, rationality and comfort in the driving process can be used as a comprehensive index S for evaluating the intelligent degree of the tested intelligent vehicleintSpecifically, the formula (12):
Figure BDA0002194405290000115
in the formula (12), LintThe Lagrange quantity in the test process of the tested intelligent vehicle i is expressed by an expression (13):
Figure BDA0002194405290000116
and S5, acquiring multiple groups of quantitative evaluation basis data of the tested intelligent vehicle in different test task scenes in the test task scene set in the S3.
And S6, performing statistical analysis on each group of quantitative evaluation basis data obtained in S5, obtaining the corresponding intelligent degree when the highest-level unmanned driving is realized according to the intelligent evaluation index in S4 to which each statistical analysis result belongs, and evaluating the intelligent degree of the tested automatic driving system of the tested intelligent vehicle by taking the intelligent degree as a reference.
Further, the "evaluating the intelligence level of the measured automatic driving system" in S6 specifically includes:
s61, calculating the corresponding intelligent degree when the highest-level unmanned driving is realized, namely calculating the minimum action amount of the tested intelligent vehicle which finishes the whole test theory when reaching the highest-level intelligent level
Figure BDA0002194405290000125
Such as: theoretical calculation method of formula (14):
Figure BDA0002194405290000121
meanwhile, when the data mining average method is adopted, calculation can be carried out
Figure BDA0002194405290000122
The method needs to calculate the average value of the action quantity of the driving vehicle to complete the whole test by screening a large number of excellent drivers, and the average value is usually calculated by a data mining average method
Figure BDA0002194405290000123
It is small and both methods are feasible.
S62, the evaluation of the intelligent level of the tested automatic driving system is to quantify the difference between the actual action quantity and the theoretical action quantity of the tested intelligent vehicle in the test process, and in order to evaluate the safe, efficient, comfortable and reasonable intelligent degree of the tested intelligent vehicle in the driving process, in the embodiment, D is definedintIn order to represent the safe, efficient, comfortable and reasonable intelligent degree of the driving process, the higher the value of the intelligent degree is, the higher the intelligent degree is, the raw data is subjected to normalization processing, namely, the data is uniformly mapped to [0, 1 ]]On the interval. The expression is formula (15):
Figure BDA0002194405290000124
as shown in fig. 2, the following process of evaluating the intelligent degree of the intelligent vehicle to be tested mainly includes:
firstly, a closed test field is built, a real driving scene is restored to provide a test environment for the test of the tested intelligent vehicle, meanwhile, the test scene can cover dangerous scenes, accident scenes, natural driving scenes and the like, elements including traffic light equipment, traffic signs, street lamps, other obstacle vehicles, riders, pedestrians and the like need to be covered in the environment, and the real traffic scene is simulated and restored.
The method comprises the following steps of selecting evaluation indexes aiming at different evaluation objects (such as intelligent driving auxiliary systems with different functions, intelligent vehicles to be tested with different grades, vehicles of different types (trucks, SUVs, cars and the like)) and the like, wherein the evaluation indexes comprise safety degree, high-efficiency degree, rationality degree, comfort degree and the like. The intelligent degree of the intelligent vehicle to be tested is output by recording the driving track and the operation behavior in the driving process, calculating the actual action quantity and the theoretical action quantity in the driving process. In the testing process, the tested intelligent vehicle needs to pass through each testing scene one by one, each corresponding task is completed in different testing scenes in sequence, the whole driving track (speed, acceleration and the like), the driving time, the specific braking operation and the like are recorded, meanwhile, a judge vehicle is arranged behind each tested intelligent vehicle and is responsible for emergency stop of the tested intelligent vehicle in front, measurement of observed traffic regulations and evaluation indexes, and the key effect is played on the task completion degree, emergency rescue and auxiliary judgment of the whole testing process.
In this embodiment, the evaluation results of the whole evaluation process are shown in table 1, specifically for different scenes and different obstacles, and multiple sets of quantitative evaluation basis data of the intelligent vehicle under test in each scene are obtained through a set intelligent driving test scene. The normalized result is divided into various intervals, each interval corresponds to an intelligent level, and the intelligent level is sequentially 'very high', 'medium', 'low' and 'very low'.
TABLE 1 Intelligent level of the intelligent vehicle to be tested
Figure BDA0002194405290000131
As shown in fig. 3, the tested intelligent vehicle test driving task needs to cover a complex and variable driving scene, covers road users with different attributes, and enables the tested intelligent vehicle to use different levels of functions of environment perception, behavior decision and motion planning in the whole test task completion process. And the tested intelligent vehicle is required to travel from the starting point to the end point of the designed driving task test route. In the embodiment, the intelligent vehicle to be tested is divided according to different driving tasks, firstly the intelligent vehicle to be tested performs obstacle detection on a first section of road which passes through, including attribute type, speed, relative distance, motion direction, posture and positioning estimation of the obstacles, gives sensing information and inputs the sensing information into a behavior decision layer, and then performs operation behaviors such as straight line driving, intersection left turn, overtaking decision, parking waiting, pedestrian avoidance, turning around and the like on the rest sections of road, and meanwhile, in the process of completing a series of test operations, the vehicle-mounted unit carried by the intelligent vehicle to be tested acquires information of all objects in the surrounding environment in real time.
In this embodiment, the testing driving task of the tested intelligent vehicle needs to cover "necessary testing conditions", that is, dynamic changes of the testing conditions caused by uncertainty of information of other road users, and in the intelligent testing field, the road traffic environment information can be changed, so as to establish different testing conditions.
The traffic elements in the "test scenario" include: as shown in fig. 3, the tested smart car, other vehicles, pedestrians, bicycles, traffic signs, different structure test roads, test facilities and test sites. In this embodiment, the test scenario is set as: the intelligent vehicle to be tested runs from the starting point of the test field to the end point of the test field, a series of operation processes such as obstacle detection and lane changing, traffic light detection and left turning, pedestrian avoidance and parking, acceleration overtaking, crossing head dropping and the like are needed in the middle, and a solid single arrow in the figure represents the direction in which the vehicle or the pedestrian is going to run or walk.
As shown in fig. 4 to 11, different test scenes for testing the intelligent degree of the tested intelligent vehicle are designed, so that the intelligent degree of the tested intelligent vehicle is tested in different scenes. The test scene shown in fig. 4 is that a straight-going lane main vehicle runs straight with a vehicle, the test scene shown in fig. 5 is that pedestrians cross and cut in adjacent lanes on the left side of the straight-going lane main vehicle, the test scene shown in fig. 6 is that obstacles on two sides of the straight-going lane main vehicle run straight, the test scene shown in fig. 7 is that motor vehicles cut in and turn off behind adjacent lanes on the right side of the straight-going lane main vehicle run straight, the test scene shown in fig. 8 is that the main vehicle of the straight-going lane is cut in by the straight-going lane, the test scene shown in fig. 9 is that vehicles change and cut in front of adjacent lanes on the right side of the straight-going lane main vehicle run straight, the test scene shown in fig. 10 is that the main vehicle of a crossroad stops waiting for the red light to pass through and then runs straight, and the test scene shown in fig. In this embodiment, the following types are set for specific scenes:
1. avoidance of static obstacles
1.1. Straight road and a plurality of subsequent obstacles
As shown, a vehicle detection scene traveling along a road with 3 consecutive obstacles is displayed. When testing, the tested intelligent vehicle can reach the highest level of intelligence, namely, according to the comfort level and the consistency cost in the proposed method, the track is smooth, continuous and safe. This allows the vehicle to easily travel on the road without sharp steering.
'S' shaped road
As shown, the moment when the vehicle avoids an obstacle on the top of the "S" shaped road when there are two obstacles. When the intelligent vehicle to be tested can reach the highest-level intelligent degree during testing, the intelligent vehicle to be tested can select the safest, efficient, comfortable and reasonable path from the planned candidate tracks according to the method so as to avoid the obstacles. At the same time, the speed and acceleration vary slightly, and when driving on the slope of the "S" shaped road, the vehicle does not select the centerline as the best path, but rather the path to the left of the "S" shaped road because of comfort requirements.
1.3. Multi-lane road
As shown, a multi-lane road having two traffic lanes and two opposite lanes is designed. When the detected intelligent vehicle detects a large obstacle, the candidate route of the current lane is dangerous. In terms of collision rates for different lane lines, the vehicle chooses to change the driving lane to avoid the obstacle instead of crossing the opposite lane because the cost of crossing the opposite lane is higher than the cost of changing the driving lane.
2. Avoiding moving obstacles
2.1. Overtaking scene
As shown, a road scenario with a two lane road is designed to test cut-in performance.
2.2. Scene following vehicle
As shown in the figure, a scene with other vehicles in front and at the back of the same road scene is designed to test the following performance.
3. Avoiding moving and stationary obstacles simultaneously
Challenging driving scenarios can arise when static and moving obstacles are present simultaneously. As shown, the tested intelligent vehicle runs on the central lane of the road. The static barrier is located on the current lane and the moving barrier moves on the left lane.
The intelligent vehicle platform is provided with sensors for testing, and the sensors can specifically comprise an inertial navigator, an embedded system, an industrial personal computer, a micro control unit, a differential positioning system, a speedometer, a deflection angle sensor, an ultrasonic radar, a millimeter wave radar, a camera, a GPS antenna, a backward radar, a look-around laser radar and the like. The hardware requirements of the automatic driving system on independent perception in the test process can be met by arranging the devices, and the software layer can be connected and disconnected with the automatic driving system to be tested in the vehicle. The method comprises the steps that test data of a road environment can be obtained through a test vehicle, wherein the test data comprises road roughness, road gradient, direction and speed of the road environment, the position of a lane line or a road boundary, the position and the size of a static obstacle, and the like, and meanwhile, the test data also comprises obstacle information, road environment information, lane line information and the like in the surrounding environment where the tested intelligent vehicle is located, and sensors such as LiDAR, radar, cameras and the like provide sensing information of the surrounding environment in real time; the vehicle position and navigation information comprises the position and speed information of the tested intelligent vehicle and other vehicles, and the positioning information of the vehicle can be acquired by combining GPS with inertial navigation.
Meanwhile, the intelligent degree quantitative evaluation device of the tested intelligent vehicle, which is provided by the invention, is also carried in the platform for testing the tested intelligent vehicle, and the device is a vehicle-mounted device and mainly records relevant parameters in the running process of the tested intelligent vehicle so as to calculate the action quantity of the whole testing process and output an intelligent degree index after the whole process is finished.
Finally, it should be pointed out that: the above examples are only for illustrating the technical solutions of the present invention, and are not limited thereto. Those of ordinary skill in the art will understand that: modifications can be made to the technical solutions described in the foregoing embodiments, or some technical features may be equivalently replaced; such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (8)

1. The intelligent degree quantitative evaluation method for the intelligent vehicle is characterized by comprising the following steps of:
s1, according to the autonomous driving function, the intelligent degree of the intelligent vehicle is divided into layers, and the layers are mainly divided into an environment perception layer, a behavior decision layer and a motion planning layer;
s2, testing the intelligent degree of the equipment on the same layer fused by the single autonomous driving function driving auxiliary system or the multiple autonomous driving function systems, and building an intelligent degree evaluation environment of the intelligent vehicle corresponding to the layers divided in S1;
s3, setting a corresponding intelligent vehicle evaluation scheme according to the intelligent degree evaluation environment corresponding to each level provided by S2, wherein the scheme comprises a test environment, a test task, an evaluation object and an evaluation index;
s4, aiming at the evaluation indexes provided by S3, selecting intelligent evaluation indexes, and constructing a proper evaluation index system by analyzing and characterizing key parameters of the driving process of the intelligent vehicle; the intelligent evaluation indexes in the S4 comprise safety, high efficiency, rationality and comfort; in a traffic system with n road users, the general indicator S provided by the formula (12) is usedintThe method represents the multi-target pursuit of the tested intelligent vehicle i on the intelligent evaluation indexes in the driving process:
Figure FDA0002385382750000011
Figure FDA0002385382750000012
in the formulae (12) and (13), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifM is the termination time of the driving process of the tested intelligent vehicle iiFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredjFor the speed of the other road users j,
Figure FDA0002385382750000013
for the lateral impact of the tested intelligent vehicle i,
Figure FDA0002385382750000014
for the longitudinal impact, R, of the intelligent vehicle i to be testediLongitudinal constraint resistance to the driver for traffic regulations, GiFor virtual gravitation, viX is the longitudinal speed of the intelligent vehicle i to be measured, Fli，1And Fli，2The transverse restraining force v of the driver is respectively the two side lane linesi，yFor the transverse speed of the intelligent vehicle i to be measured, FjiExternal force on the tested intelligent vehicle i caused by other road users j;
s5, acquiring multiple groups of quantitative evaluation basis data of the tested intelligent vehicle in different test task scenes in the test task scene set in the S3;
and S6, performing statistical analysis on each group of quantitative evaluation basis data obtained in S5, obtaining the corresponding intelligent degree when the highest-level unmanned driving is realized according to the intelligent evaluation index in S4 to which each statistical analysis result belongs, and evaluating the intelligent degree of the tested automatic driving system of the tested intelligent vehicle by taking the intelligent degree as a reference.
2. The intelligent vehicle intelligent degree quantitative evaluation method according to claim 1, wherein the safety is mainly determined by a driving risk level U of the tested intelligent vehicle i in a specific traffic scene, which is characterized by formula (1)riskAnd calculating to obtain:
Figure FDA0002385382750000021
Figure FDA0002385382750000022
Figure FDA0002385382750000023
Figure FDA0002385382750000024
Figure FDA0002385382750000025
in formulae (1) to (5), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifM is the termination time of the driving process of the tested intelligent vehicle iiFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredlimitFor lane speed limit,. tau.for calibration parameter,. vi，xFor the longitudinal speed, v, of the intelligent vehicle i to be measuredi，yFor the lateral speed, v, of the intelligent vehicle i to be measuredjSpeed of other road users j,/tOf the lane type, rliFor the distance between the measured intelligent vehicle i and the road boundary, lwIs the width of the lane, dijIs the straight-line distance between the measured intelligent vehicle i and the central point of other road users j, vijIs the relative speed, theta, between the measured intelligent vehicle i and other road users jijIs from dijTo vijThe angle of,
Figure FDA0002385382750000026
is from dijTo viAnd the counterclockwise direction is positive, mjBeing the actual physical mass of the other road users j,
Figure FDA0002385382750000027
for j speed of other road usersDegree vjFunction of, TjThe type of other road users j.
3. The intelligent vehicle intelligentization degree quantitative evaluation method according to claim 1, wherein the efficiency is mainly the efficiency U of the task completed by the intelligent vehicle to be tested, which is expressed by formula (7)effCarrying out weighing:
Figure FDA0002385382750000031
Gi＝migsinθi(8)
Figure FDA0002385382750000032
in formulae (7) to (9), miFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredi，xThe longitudinal speed of the intelligent vehicle i to be measured is shown, and g is the gravity acceleration; k is a constant, vderIs the desired speed, v, of the driverlimitThe speed limit of the lane is realized.
4. The intelligent vehicle intellectualization degree quantitative evaluation method according to claim 1, wherein the reasonableness is mainly determined by embodying the assumed/actually existing driver psychological expectation U represented by the formula (10)comAnd calculating to obtain:
Figure FDA0002385382750000033
Figure FDA0002385382750000034
Figure FDA0002385382750000035
in the formulae (4), (5) and (10), t0To be measuredStarting time t of driving process of intelligent vehicle ifM is the termination time of the driving process of the tested intelligent vehicle iiFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredjSpeed, v, of other road users jijFor the relative speed between the measured intelligent vehicle i and other road users j, dijIs the straight-line distance theta between the measured intelligent vehicle i and the central point of other road users jijIs from dijTo vijThe angle of,
Figure FDA0002385382750000036
is from dijTo viAnd the counterclockwise direction is positive, mjBeing the actual physical mass of the other road users j,
Figure FDA0002385382750000037
to relate to the j speed v of other road usersjFunction of, TjThe type of other road users j.
5. The intelligent vehicle intellectualization degree quantitative evaluation method according to claim 1, wherein the comfort is reflected by a vehicle performance index defined as a comfort soft constraint function U expressed by an equation (11)smooth：
Figure FDA0002385382750000041
In the formula (11), the reaction mixture is,
Figure FDA0002385382750000042
for the lateral impact of the tested intelligent vehicle i,
Figure FDA0002385382750000043
the longitudinal impact degree of the intelligent vehicle i to be measured.
6. The smart vehicle of any of claims 1-5The quantitative evaluation method of the intelligent degree is characterized in that the intelligent degree representing the safety, high efficiency, comfort and reasonable driving process is defined as DintThe higher the value is, the higher the intelligentization degree is, the data is uniformly mapped to 0, 1]In the interval:
Figure FDA0002385382750000044
in the formula (15), SintThe comprehensive index of the intelligent degree of the multi-target pursuit of the safety, the high efficiency, the rationality and the comfort of the tested intelligent vehicle i in the driving process is provided,
Figure FDA0002385382750000045
the minimum action amount of the whole test theory is completed when the intelligent vehicle reaches the highest intelligent level.
7. The intelligent vehicle intellectualization degree quantitative evaluation method according to claim 6,
Figure FDA0002385382750000046
obtained by using a data mining average method, or obtained by using the following formula (14):
Figure FDA0002385382750000047
8. the intelligent vehicle intelligent degree quantitative evaluation method according to claim 1, wherein in an environment perception layer, motion states and physical attributes of different objects are predicted and filtered according to the reliability of fused data of the equipment with the environment perception function, and meanwhile, relative distances, motion directions, speeds, postures and positioning estimations of road users are identified so as to independently evaluate the environment perception capability of the equipment with the environment perception function carried on the intelligent vehicle to be evaluated;
in the behavior decision layer, when the tested intelligent vehicle runs in a structured or unstructured road environment, whether the intellectualization degree of the running process is improved or not is judged through a strategy after decision so as to independently evaluate the behavior decision capability of the tested intelligent vehicle;
on a motion planning layer, generating a safe and feasible driving track by planning the speed and the local path of the driving process of the tested intelligent vehicle, and independently evaluating whether the track planning meets the set judgment indexes of safety, efficiency, rationality and comfort; the intelligent evaluation indexes comprise safety, high efficiency, rationality and comfort; in a traffic system with n road users, the general indicator S provided by the formula (12) is usedintThe method represents the multi-target pursuit of the tested intelligent vehicle i on the intelligent evaluation indexes in the driving process:
Figure FDA0002385382750000051
Figure FDA0002385382750000052
in the formulae (12) and (13), t0Is the starting moment t of the driving process of the tested intelligent vehicle ifM is the termination time of the driving process of the tested intelligent vehicle iiFor the quality of the intelligent vehicle i to be tested, viFor the speed, v, of the intelligent vehicle i to be measuredjFor the speed of the other road users j,
Figure FDA0002385382750000053
for the lateral impact of the tested intelligent vehicle i,
Figure FDA0002385382750000054
for the longitudinal impact, R, of the intelligent vehicle i to be testediLongitudinal constraint resistance to the driver for traffic regulations, GiFor virtual gravitation, vi，xFor the longitudinal speed of the intelligent vehicle i to be measured, Fli，1And Fli，2The transverse restraining force v of the driver is respectively the two side lane linesi，yFor the intelligent vehicle i under testTransverse velocity of (F)jiThe external force on the tested intelligent vehicle i caused by other road users j is provided.
CN201910843325.9A 2019-09-06 2019-09-06 Intelligent degree quantitative evaluation method for intelligent vehicle Active CN110531740B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910843325.9A CN110531740B (en) 2019-09-06 2019-09-06 Intelligent degree quantitative evaluation method for intelligent vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910843325.9A CN110531740B (en) 2019-09-06 2019-09-06 Intelligent degree quantitative evaluation method for intelligent vehicle

Publications (2)

Publication Number Publication Date
CN110531740A CN110531740A (en) 2019-12-03
CN110531740B true CN110531740B (en) 2020-05-05

Family

ID=68667484

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910843325.9A Active CN110531740B (en) 2019-09-06 2019-09-06 Intelligent degree quantitative evaluation method for intelligent vehicle

Country Status (1)

Country Link
CN (1) CN110531740B (en)

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111178735B (en) * 2019-12-24 2024-02-02 国汽(北京)智能网联汽车研究院有限公司 Test evaluation method, device and system for automatic driving function
CN111581102A (en) * 2020-05-11 2020-08-25 中国人民解放军陆军研究院装甲兵研究所 Test question bank system based on environmental data
CN112330122A (en) * 2020-10-28 2021-02-05 中国计量大学 Floor sweeping robot intelligent degree quantitative evaluation method and system
CN112382086A (en) * 2020-10-30 2021-02-19 上海炬宏信息技术有限公司 Application method of open road test evaluation system of automatic driving automobile
CN112362356B (en) * 2020-11-02 2021-08-10 吉林大学 Intelligent vehicle braking and parking capacity testing method considering passenger comfort
CN112744223B (en) * 2021-01-18 2022-04-15 北京智能车联产业创新中心有限公司 Method and system for evaluating intersection performance of automatic driving vehicle
CN113190921B (en) * 2021-05-14 2021-11-19 上海交通大学 Automatic evaluation method and system for intelligent automobile driving performance test
CN113379333A (en) * 2021-07-22 2021-09-10 交通运输部公路科学研究所 Intelligent grade evaluation method for auxiliary driving operation vehicle
CN113610166B (en) * 2021-08-10 2023-12-26 吉林大学 Method for establishing test scene library for intelligent vehicle
CN114434466B (en) * 2022-03-14 2022-09-20 交通运输部公路科学研究所 Automobile intelligent cockpit performance evaluation simulation robot
CN114638420B (en) * 2022-03-22 2022-10-25 交通运输部公路科学研究所 Road intelligence evaluation method and hazardous chemical substance vehicle road-level navigation method
CN115729810B (en) * 2022-11-02 2024-01-02 北京华龙宏达科技有限公司 Vehicle-mounted test system for intelligent automobile road test

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103234763A (en) * 2013-04-09 2013-08-07 北京理工大学 System and method for quantitatively evaluating unmanned vehicles
EP3379502A2 (en) * 2017-03-22 2018-09-26 Kabushiki Kaisha Toshiba Paper sheet processing system, paper sheet processing apparatus, and paper sheet processing method
CN108647437A (en) * 2018-05-09 2018-10-12 公安部交通管理科学研究所 A kind of autonomous driving vehicle evaluation method and evaluation system
CN108829087A (en) * 2018-07-19 2018-11-16 山东省科学院自动化研究所 A kind of intelligent test system and test method of autonomous driving vehicle
CN109141920A (en) * 2018-08-27 2019-01-04 山东省科学院自动化研究所 Automatic driving vehicle rainy day environment sensing recognition capability test evaluation system and method
CN110196994A (en) * 2019-04-23 2019-09-03 同济大学 A kind of autonomous driving vehicle traffic coordinating evaluation and test model and method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103234763A (en) * 2013-04-09 2013-08-07 北京理工大学 System and method for quantitatively evaluating unmanned vehicles
EP3379502A2 (en) * 2017-03-22 2018-09-26 Kabushiki Kaisha Toshiba Paper sheet processing system, paper sheet processing apparatus, and paper sheet processing method
CN108647437A (en) * 2018-05-09 2018-10-12 公安部交通管理科学研究所 A kind of autonomous driving vehicle evaluation method and evaluation system
CN108829087A (en) * 2018-07-19 2018-11-16 山东省科学院自动化研究所 A kind of intelligent test system and test method of autonomous driving vehicle
CN109141920A (en) * 2018-08-27 2019-01-04 山东省科学院自动化研究所 Automatic driving vehicle rainy day environment sensing recognition capability test evaluation system and method
CN110196994A (en) * 2019-04-23 2019-09-03 同济大学 A kind of autonomous driving vehicle traffic coordinating evaluation and test model and method

Also Published As

Publication number Publication date
CN110531740A (en) 2019-12-03

Similar Documents

Publication Publication Date Title
CN110531740B (en) Intelligent degree quantitative evaluation method for intelligent vehicle
US11554785B2 (en) Driving scenario machine learning network and driving environment simulation
Best et al. Autonovi-sim: Autonomous vehicle simulation platform with weather, sensing, and traffic control
RU2733015C1 (en) Real-time vehicle control
Wang et al. Cooperative ramp merging system: Agent-based modeling and simulation using game engine
US20210339741A1 (en) Constraining vehicle operation based on uncertainty in perception and/or prediction
CN110562258B (en) Method for vehicle automatic lane change decision, vehicle-mounted equipment and storage medium
Lattarulo et al. Urban motion planning framework based on n-bézier curves considering comfort and safety
CN111746555A (en) Identification and avoidance of collision behavior
US11628850B2 (en) System for generating generalized simulation scenarios
CN114077541A (en) Method and system for validating automatic control software for an autonomous vehicle
US20220198107A1 (en) Simulations for evaluating driving behaviors of autonomous vehicles
US20230150549A1 (en) Hybrid log simulated driving
González et al. Control architecture for cybernetic transportation systems in urban environments
US11774259B2 (en) Mapping off-road entries for autonomous vehicles
GB2600188A (en) Metric back-propagation for subsystem performance evaluation
EP3869341A1 (en) Play-forward planning and control system for an autonomous vehicle
CN114430722A (en) Security analysis framework
Huang et al. An integrated architecture for intelligence evaluation of automated vehicles
CN114415672A (en) Dynamic model evaluation for autonomously driven vehicles
US20230031130A1 (en) Open door reconstruction for sensor simulation
CN110426215B (en) Model establishing method for vehicle ride comfort test and intelligent driving system
CN114813157A (en) Test scene construction method and device
Sun et al. The Development of a Self-Driving Bus
Peng Connected and automated vehicles

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