CN111930026A - Test method and device - Google Patents
Test method and device Download PDFInfo
- Publication number
- CN111930026A CN111930026A CN202010842387.0A CN202010842387A CN111930026A CN 111930026 A CN111930026 A CN 111930026A CN 202010842387 A CN202010842387 A CN 202010842387A CN 111930026 A CN111930026 A CN 111930026A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- vehicle
- intelligent driving
- vehicle model
- real
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000010998 test method Methods 0.000 title claims abstract description 9
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 160
- 238000012360 testing method Methods 0.000 claims abstract description 110
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000004088 simulation Methods 0.000 claims abstract description 26
- 238000006243 chemical reaction Methods 0.000 claims description 16
- 230000008569 process Effects 0.000 claims description 13
- 206010000117 Abnormal behaviour Diseases 0.000 claims description 4
- 230000001960 triggered effect Effects 0.000 claims description 4
- 230000006870 function Effects 0.000 abstract description 90
- 238000012795 verification Methods 0.000 abstract description 13
- 238000004519 manufacturing process Methods 0.000 abstract description 9
- 238000010586 diagram Methods 0.000 description 11
- 238000011156 evaluation Methods 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000006399 behavior Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000002360 preparation method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
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
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a test method and a test device, wherein the method comprises the following steps: triggering an intelligent driving twinning algorithm under a preset condition, wherein the intelligent driving twinning algorithm is consistent with an algorithm model of the intelligent driving algorithm but different in triggering condition; acquiring real vehicle sensor data and inputting the data into a vehicle model and a real vehicle; driving a vehicle model by adopting an intelligent driving twinning algorithm based on real vehicle sensor data; and generating a test result based on the running condition and the real vehicle condition of the vehicle model. According to the testing method and the testing device, a real vehicle scene and a sensor are used as sources of functional verification data, so that the accuracy of the functional verification data is superior to that of simulation test verification, the real vehicle scene does not need to be constructed, and the investment of a large amount of manpower and material resources is avoided; the realization can be deployed in mass-production vehicle controller products, then the execution conditions of the actual functions of mass-production vehicles are collected, the test result is obtained, and the method has the advantages of low cost and high accuracy.
Description
Technical Field
The invention relates to the technical field of simulation testing, in particular to a testing method and a testing device.
Background
Simulation test is an important link in the development process of intelligent driving or unmanned driving algorithms, and the test of the intelligent driving algorithms or controllers is realized by constructing a scene library, so that the simulation test becomes a main verification means and method in the development process of intelligent driving.
At present, simulation tests of intelligent driving algorithms are realized based on some constructed models, and the test method has certain defects mainly due to the following points: 1) the simulation degree of a sensor model used for intelligent driving simulation cannot completely reach an ideal state, namely, the simulation degree is different from a real sensor model acquisition signal of a real vehicle; 2) the real scene is different from the scene simulated by the simulation software to a certain extent, the scene simulated by the simulation software mostly comes from laws, regulations, functional specifications and experiences, and the scenes cannot be generalized with the scenes encountered by the real road condition; 3) the vehicle model and the simulation line control have certain difference with the real vehicle. Therefore, in practical application, on the basis of carrying out simulation test on the intelligent driving algorithm, a large number of real vehicle tests are required to be carried out for cross validation.
Currently, real vehicle testing of an intelligent driving algorithm is limited by development cost and construction difficulty of real vehicle testing scenes, and the problems that testing cannot be completed in some extreme scenes and the accuracy of testing results is poor due to the fact that the number of the testing scenes is single are solved. Therefore, how to provide a testing method can conveniently and accurately test the intelligent driving algorithm, which is a problem to be solved urgently by technical personnel in the field.
Disclosure of Invention
In view of this, the present invention provides the following technical solutions:
a method of testing, comprising:
triggering an intelligent driving twinning algorithm under a preset condition, wherein the intelligent driving twinning algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the triggering condition of the intelligent driving twinning algorithm is different from that of the intelligent driving algorithm;
acquiring real vehicle sensor data and inputting the data into a vehicle model and a real vehicle;
driving the vehicle model by adopting the intelligent driving twinning algorithm based on the real vehicle sensor data;
and generating a test result based on the running condition and the real vehicle condition of the vehicle model.
Optionally, the driving the vehicle model by using the intelligent driving twin algorithm based on the real vehicle sensor data includes:
acquiring first positioning information of a real vehicle and second positioning information of the vehicle model, and determining a positioning difference of the first positioning information and the second positioning information;
determining a coordinate conversion parameter between the first positioning information and the second positioning information based on the positioning difference;
based on the coordinate conversion parameters, converting the positions of the target traffic vehicles, the obstacles, the road signs and the lane lines detected in the real vehicle sensor coordinate system to be in the vehicle model sensor coordinate system;
and driving the vehicle model by adopting the intelligent driving twinning algorithm based on the positions of the target traffic vehicle, the obstacle, the road sign and the lane line in the coordinate system of the vehicle model sensor.
Optionally, the generating a test result based on the operating condition and the actual condition of the vehicle model includes:
and judging whether the vehicle model driven by the intelligent driving algorithm has abnormal behaviors or not based on the running condition and the real vehicle condition of the vehicle model, and determining comfort data of a driver, wherein the running data of the vehicle model comprises a running track and dynamic information.
Optionally, after the generating the test result based on the operating condition and the real vehicle condition of the vehicle model, the method further includes:
and uploading the test result and the data in the test process to a server.
Optionally, the triggering of the intelligent driving twinning algorithm under the preset condition includes:
triggering an intelligent driving twin algorithm when a driver triggers an intelligent driving function and the function to be tested belongs to a function item which is subjected to a simulation test and is not subjected to a real vehicle test and/or a function item which is not started by the driver;
or the like, or, alternatively,
when a driver triggers the intelligent driving function, the real vehicle sensor detects that the surrounding environment meets the triggering condition for testing the corresponding intelligent driving function, and the function to be tested belongs to the preset function set, the intelligent driving twin algorithm is triggered.
Optionally, the method further includes:
and automatically quitting the intelligent driving twinning algorithm when the quitting condition is met.
Optionally, when the function to be tested belongs to a function item that has undergone a simulation test but has not undergone an actual vehicle test, the exit condition includes:
the Euclidean distance between the vehicle model and the centroid coordinate of the real vehicle is larger than the radius of a functional domain sensed by a real vehicle sensor, wherein the functional domain is the region where the real vehicle executes the function to be tested by the intelligent driving twin algorithm;
when the function to be tested belongs to a function item which is not started by the driver, the exit condition comprises the following steps:
the Euclidean distance between the vehicle model and the centroid coordinate of the real vehicle is larger than the radius of a functional domain sensed by a real vehicle sensor;
or the like, or, alternatively,
the driver actively turns on the function to be tested that was not previously activated.
Optionally, the vehicle model is a three-freedom six-state dual-input nonlinear system.
A test apparatus, comprising:
the intelligent driving twinborn algorithm comprises an algorithm triggering module, a driving algorithm judging module and a driving algorithm judging module, wherein the algorithm triggering module is used for triggering an intelligent driving twinborn algorithm under a preset condition, the intelligent driving twinborn algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the intelligent driving twinborn algorithm is different from the triggering condition of the intelligent driving algorithm;
the data input module is used for acquiring data of the real vehicle sensor and inputting the data into the vehicle model and the real vehicle;
the driving control module is used for driving the vehicle model by adopting the intelligent driving twinning algorithm based on the real vehicle sensor data;
and the result determining module is used for generating a test result based on the running condition and the real vehicle condition of the vehicle model.
Optionally, the driving control module includes:
the information acquisition module is used for acquiring first positioning information of a real vehicle and second positioning information of the vehicle model and determining a positioning difference of the first positioning information and the second positioning information;
a parameter determination module configured to determine a coordinate conversion parameter between the first positioning information and the second positioning information based on the positioning difference;
the data conversion module is used for converting the positions of the target traffic vehicle, the obstacle, the road sign and the lane line detected in the real vehicle sensor coordinate system to the vehicle model sensor coordinate system based on the coordinate conversion parameters;
and the driving control submodule is used for driving the vehicle model by adopting the intelligent driving twinning algorithm based on the positions of the target traffic vehicle, the obstacle, the road sign and the lane line under the coordinate system of the vehicle model sensor.
As can be seen from the above technical solutions, compared with the prior art, the embodiment of the present invention discloses a testing method and apparatus, where the method includes: triggering an intelligent driving twinning algorithm under a preset condition, wherein the intelligent driving twinning algorithm is consistent with an algorithm model of the intelligent driving algorithm but different in triggering condition; acquiring real vehicle sensor data and inputting the data into a vehicle model and a real vehicle; driving a vehicle model by adopting an intelligent driving twinning algorithm based on real vehicle sensor data; and generating a test result based on the running condition and the real vehicle condition of the vehicle model. According to the testing method and the testing device, a real vehicle scene and a sensor are used as sources of functional verification data, so that the accuracy of the functional verification data is superior to that of simulation test verification, the real vehicle scene does not need to be constructed, and the investment of a large amount of manpower and material resources is avoided; the realization can be deployed in mass-production vehicle controller products, then the execution conditions of the actual functions of mass-production vehicles are collected, the test result is obtained, and the method has the advantages of low cost and high accuracy.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flow chart of a testing method disclosed in the embodiments of the present invention;
FIG. 2 is a schematic diagram of a conventional intelligent driving vehicle;
FIG. 3 is a schematic diagram of a realization link of an intelligent driving vehicle adopting an intelligent driving twinning algorithm, which is disclosed by the implementation of the invention;
FIG. 4 is a schematic flow chart illustrating operation control of a vehicle model according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a control link of a conventional intelligent driving algorithm;
FIG. 6 is a control link diagram of an intelligent driving twin algorithm disclosed in an embodiment of the present invention;
FIG. 7 is a flow chart of another testing method disclosed in the embodiments of the present invention;
FIG. 8 is a schematic structural diagram of a testing apparatus according to an embodiment of the present disclosure;
fig. 9 is a schematic structural diagram of an operation control module according to an embodiment of the present invention.
Detailed Description
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 derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a flowchart of a testing method according to an embodiment of the present invention, and referring to fig. 1, the testing method may include:
step 101: triggering an intelligent driving twinning algorithm under a preset condition, wherein the intelligent driving twinning algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the triggering condition of the intelligent driving twinning algorithm is different from that of the intelligent driving algorithm.
The intelligent driving twin algorithm is completely consistent with an algorithm model of the intelligent driving algorithm, and the intelligent driving algorithm is used for controlling the real vehicle to run after the intelligent driving function is started; the intelligent driving twins algorithm is used for controlling the vehicle model under the condition that preset conditions are met, and the execution effect of each function of the intelligent driving algorithm is determined through the execution effect of each function of the intelligent driving twins algorithm.
For all function items controlled by the real vehicle, because some function items are already subjected to mature simulation tests or the execution results of the function items can be obtained when the intelligent driving algorithm operates, in the implementation, all function items do not need to be tested, so that the intelligent driving twin algorithm can be started to operate at the opportunity instead of operating all the time under the condition of meeting the preset conditions. The preset conditions have different implementations, and specific contents will be described in detail in the following contents, which are not described too much herein.
Fig. 2 is a schematic diagram of an implementation link of a conventional intelligent driving vehicle, fig. 3 is a schematic diagram of an implementation link of an intelligent driving vehicle adopting an intelligent driving twin algorithm, which is disclosed in the implementation of the present invention, and in combination with fig. 2 and fig. 3, a control link adopting the intelligent driving twin algorithm is added with the intelligent driving twin algorithm and a vehicle model compared with the conventional control link, and the two modules can be installed and deployed in an intelligent driving controller as a test verification mode in parallel with a simulation test or an actual vehicle test. Specifically, the intelligent driving twin algorithm is an application layer control algorithm which is completely the same as the intelligent driving algorithm built in the intelligent driving controller, and is equivalent to copying the intelligent driving algorithm by the same part; the vehicle model refers to a vehicle dynamics model, the vehicle dynamics model is controlled through an intelligent driving twinning algorithm, and the kinematic behavior of the vehicle model can be expressed in a digital form, which is similar to on-line real-time simulation; the motion result of the vehicle is expressed not by the motion behavior of the real vehicle but by the motion behavior of the vehicle model. Specific applications of the embodiments of the present application can be understood with reference to fig. 2 and 3 and the above description.
After step 101, the process proceeds to step 102.
Step 102: real vehicle sensor data is acquired and input to the vehicle model and the real vehicle.
In the embodiment of the application, after the driver triggers the intelligent driving function, the intelligent driving twin algorithm can control the operation of the vehicle model synchronously at a certain function execution node of the real vehicle controlled by the intelligent driving algorithm. Different from the simulation test process of the existing intelligent driving algorithm, the sensing data acquired by the vehicle model, namely the real vehicle sensor data, is not generated by the simulation of a virtual sensor model but is the real sensor data acquired by the real vehicle sensor, so that the reality and accuracy of the sensor data are ensured; the test based on the real sensor data realizes the semi-real vehicle test of the intelligent driving algorithm, and the test result has good practical significance compared with a simulation test and a real vehicle test with larger limitation.
Step 103: and driving the vehicle model by adopting the intelligent driving twinning algorithm based on the real vehicle sensor data.
Step 103 is that the vehicle model is controlled through the intelligent driving twin algorithm on the basis of the real vehicle sensor data, and through the process, whether the intelligent driving twin algorithm can control the vehicle model to accurately avoid the obstacle to continue to move forward under certain specific scenes, such as scenes with the obstacle in front of the vehicle in operation can be determined.
Step 104: and generating a test result based on the running condition and the real vehicle condition of the vehicle model.
The real vehicle condition may be some basic conditions of the real vehicle, such as whether the vehicle is in a starting state, whether the vehicle is in a moving state, and the like.
Continuing with the example of the scene in which the obstacle exists in front of the running vehicle, if the intelligent driving twinning algorithm can accurately avoid the obstacle and continue to move ahead, determining that the obstacle avoidance function of the intelligent driving twinning algorithm passes the test; if the intelligent driving twinning algorithm does not avoid the obstacle, the vehicle continues to move forwards after colliding with the obstacle and stopping or the side of the vehicle model frictionally collides with the obstacle, and the obstacle avoidance function of the intelligent driving twinning algorithm is determined to fail the test.
According to the test method provided by the embodiment, the real vehicle scene and the sensor are used as the source of the functional verification data, so that the accuracy of the functional verification data is superior to that of the simulation test verification, the real vehicle scene does not need to be constructed, and the investment of a large amount of manpower and material resources is avoided; the realization can be deployed in mass-production vehicle controller products, then the execution conditions of the actual functions of mass-production vehicles are collected, the test result is obtained, and the method has the advantages of low cost and high accuracy.
Fig. 4 is a schematic flow chart of operation control on a vehicle model according to an embodiment of the present invention, and referring to fig. 4, in the embodiment, the driving a vehicle model by using the intelligent driving twin algorithm based on the real vehicle sensor data may include:
step 401: acquiring first positioning information of a real vehicle and second positioning information of the vehicle model, and determining a positioning difference of the first positioning information and the second positioning information.
The first positioning information of the real vehicle is the position information of the real vehicle; the second positioning information of the vehicle model is virtual position information.
Step 402: determining a coordinate conversion parameter between the first positioning information and the second positioning information based on the positioning difference.
Because the real vehicle coordinate system is different from the coordinate system adopted by the vehicle model, the positioning difference between the first positioning information and the second positioning information needs to be determined, the coordinate conversion parameter between the first positioning information and the second positioning information is further determined, and the corresponding relation between data in the two coordinate systems is determined.
Step 403: and based on the coordinate conversion parameters, converting the positions of the target traffic vehicle, the obstacles, the road signs and the lane lines detected in the real vehicle sensor coordinate system into the vehicle model sensor coordinate system.
Step 404: and driving the vehicle model by adopting the intelligent driving twinning algorithm based on the positions of the target traffic vehicle, the obstacle, the road sign and the lane line in the coordinate system of the vehicle model sensor.
Fig. 5 is a schematic control link diagram of a conventional intelligent driving algorithm, fig. 6 is a schematic control link diagram of an intelligent driving twin algorithm disclosed in an embodiment of the present invention, and with reference to fig. 5 and 6, a coordinate transformation module is disposed behind a sensing module of the intelligent driving twin algorithm; because the intelligent driving twinborn algorithm adopts the original sensing signal to carry out sensing identification, the obtained sensing result is the sensing result of the surrounding environment of the real vehicle, but not the sensing result of the surrounding environment of the vehicle model, so the sensing fusion result of the running semi-virtual environment of the vehicle model is obtained by carrying out positioning difference through the positioning of the real vehicle and the positioning of the vehicle model and then carrying out coordinate transformation. The semi-virtual environment refers to that the surrounding scene is a real scene, but the motion trail of the vehicle is calculated by a vehicle model, so the motion trail is virtual. The coordinate transformation requires transforming the positions of the target traffic vehicles, obstacles, road signs and lane lines detected by the real vehicle sensor coordinate system to be under the vehicle model sensor coordinate system.
On the basis of the above embodiment disclosed by the present invention, fig. 7 is a flowchart of another testing method disclosed by the embodiment of the present invention, and the method shown in fig. 7 is also applied to an intelligent driving controller. As shown in fig. 7, the testing method may include:
step 701: triggering an intelligent driving twinning algorithm under a preset condition, wherein the intelligent driving twinning algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the triggering condition of the intelligent driving twinning algorithm is different from that of the intelligent driving algorithm.
Step 702: and acquiring real vehicle sensor data and inputting the data into the vehicle model and the real vehicle.
Step 703: and driving the vehicle model by adopting the intelligent driving twinning algorithm based on the real vehicle sensor data.
Step 704: and generating a test result based on the running condition and the real vehicle condition of the vehicle model.
Step 705: and uploading the test result and the data in the test process to a server.
Wherein the data during the test process may include, but is not limited to, the running trajectory and the functional execution data of the vehicle model during the entire test process.
In the implementation, when a driver controls the real vehicle, the intelligent driving twin algorithm servo work controls the vehicle model to work, the test result of the intelligent driving twin algorithm is obtained according to the operation control result, and then the track, the function execution condition and the test result of the whole control result can be uploaded to a rear-end server of a research and development enterprise through 4G or 5G wireless mobile communication means and the like.
In the above embodiment, the triggering of the intelligent driving twin algorithm under the preset condition may include: triggering an intelligent driving twin algorithm when a driver triggers an intelligent driving function and the function to be tested belongs to a function item which is subjected to a simulation test and is not subjected to a real vehicle test and/or a function item which is not started by the driver; or when the driver triggers the intelligent driving function, the real vehicle sensor detects that the surrounding environment meets the triggering condition for testing the corresponding intelligent driving function, and the function to be tested belongs to the preset function set, triggering the intelligent driving twin algorithm.
Specifically, the intelligent driving algorithm generally has the following starting mode, 1) when the driver selects the intelligent driving function, or when the driver switches the intelligent driving function, the intelligent driving algorithm enters the Standby mode (the function is switched to the preparation mode, but the function is not completely activated, which is equivalent to a preparation activation state, but when the sensor detects a trigger signal, the function is completely activated), and when the sensor detects that the surrounding environment meets the trigger condition of a certain function, the intelligent driving algorithm is started; 2) the driver selects the intelligent driving function, and the intelligent driving function is directly and automatically executed. The intelligent driving twin algorithm is started and the above conditions are also satisfied.
Besides the two starting modes, the starting mode of the intelligent driving twin algorithm needs to simultaneously meet the selective execution of the function items, namely the function to be executed by the intelligent driving algorithm belongs to a preset function set. Let the full set A of intelligent driving function items which can be executed include function items which can be executed independently by { A1, A2, A3, …, Am }, let the driver start the activated function item set B include functions which can be executed independently by { B1, B2, B3, …, Bn }, wherein B is contained in A, and in addition, the test function items C to be tested by OEM (Original Engineered manufacturing) or supplier include the functions that { C1, C2, C3, …, Ck } item can independently execute, if the function items are not completely in the executable intelligent driving function item set, the function (corresponding to the preset function set) that can be started is C e (C-a ≧ C) U (a-B), wherein, C-A and C represent the function which is only subjected to the simulation test but not subjected to the real vehicle test due to the function maturity problem, and A-B represents the function which is not started by the driver.
The function items included in the preset function set are the function items which are not tested by the real vehicle and the function items which are not started by the driver.
On the basis of the disclosure of the above embodiment, the testing method may further include: automatically quitting the intelligent driving twin algorithm when the quitting condition is met; when the function to be tested belongs to a function item which is subjected to the simulation test and is not subjected to the real vehicle test, the exit condition comprises the following steps: and the Euclidean distance between the vehicle model and the mass center coordinate of the real vehicle is larger than the radius of a functional domain sensed by a real vehicle sensor, wherein the functional domain is the region where the real vehicle is located when the intelligent driving twin algorithm executes the function to be tested. When the function to be tested belongs to a function item which is not started by the driver, the exit condition comprises the following steps: the Euclidean distance between the vehicle model and the centroid coordinate of the real vehicle is larger than the radius of a functional domain sensed by a real vehicle sensor; or the driver actively starts the function to be tested which is not started before.
Specifically, the traditional exit condition of the intelligent driving algorithm is that the driver actively exits, and the exit condition of the intelligent driving twin algorithm is different from the exit condition of the intelligent driving algorithm for controlling the real vehicle. In a preset function set C e (C-A n C) U (A-B), a e (A-B) is set for an automobile executing a function in the A-B function set, the condition P is generally met or the function a is actively started by a driver, and the function a in the intelligent driving twin algorithm exits; for a car executing a function in the C-A ≦ C function set, the condition P is typically satisfied and all functions in the function set exit.
The functions of the intelligent driving twin algorithm cannot work or do not need to work all the way, and the reasons include: 1) the scene signal obtained by the real sensor of the real vehicle is spatially limited, and generally works in the radius range of 150-300m around the vehicle, and if the sensing area (sensing area for short) of the real sensor and the sensing area required by the vehicle model cannot be overlapped due to the fact that the vehicle model executes the function different from the operation of the driver, even the function execution area (functional area for short) cannot be overlapped, the intelligent driving twin algorithm cannot provide the sensing input required by function execution for the controlled vehicle model. 2) The vehicle model and the road model can not be completely consistent with the real vehicle and road precision, and after long-time driving, even if the operation of a driver is completely consistent with the control of an intelligent driving algorithm, the positions of the two sides can also cause great deviation. Therefore, the intelligent driving twin algorithm generally only needs to execute some functions with short running time, verify whether the functions can be normally triggered, verify whether the function execution is normal in a short time, and simultaneously need to have a function early exit condition; the centroid coordinate of a vehicle model controlled by an intelligent driving twin algorithm is positioned as a point A, the real centroid coordinate of a real vehicle is positioned as a point B, the Euclidean distance between the point A and the point B is set as L, the radius of a functional domain sensed by a sensor is set as L, and when L is greater than L, a condition P is triggered.
Furthermore, since the intelligent driving twin algorithm controls the vehicle model, the start-stop conditions of the vehicle model are derived from or follow the start-stop of the intelligent driving twin algorithm.
Limited by the limited vehicle driving state information which can be provided by real vehicle test data, a vehicle model can adopt a three-freedom six-state dual-input nonlinear system shown in an equation (1), wherein model inputs, states, parameters and external interference terms involved in the equation are shown in a table 1.
TABLE 1 detailed information of inputs, states and parameters involved in a vehicle model
By combining the simplified three-degree-of-freedom dual-input six-state multi-output nonlinear vehicle system model with initial conditions and by means of linearization, discretization and other technical means, prediction of dynamic response of different input combinations can be realized. The initial conditions required to simplify the vehicle model described above are as equation (2). In the formula, the initial conditions of each model and the acquisition mode of the real-time road surface longitudinal gradient interference information are shown in table two.
x(0)＝{X(0),vx(0),Y(0),vy(0),ψ(0),γ(0)}T (2)
TABLE 1 detailed information of inputs, states and parameters involved in a vehicle model
In the above embodiment, the generating a test result based on the operating condition and the actual condition of the vehicle model may include: and judging whether the vehicle model driven by the intelligent driving algorithm has abnormal behaviors or not based on the running condition and the real vehicle condition of the vehicle model, and determining comfort data of a driver, wherein the running data of the vehicle model comprises a running track and dynamic information.
Wherein determining whether the vehicle model has abnormal behavior may include: determining whether the vehicle model collides with a surrounding object based on the running track of the vehicle model, and determining whether the vehicle model violates a traffic regulation based on map information and dynamics information of the vehicle model. Determining driver comfort data may include: determining comfort data for a driver based on the dynamics information of the vehicle model.
For example, in one specific implementation, after the test result is obtained, the test result can be evaluated, and the result evaluation criteria can be used for screening abnormal working conditions of function execution by an intelligent driving developer, so that the test result can be wirelessly transmitted back to a DT department of the developer to check the operation and test results. The specific evaluation criteria can refer to the following items:
whether or not to violate traffic regulations
1. Is a collision with a traffic vehicle or a traffic obstacle?
a. Whether a vehicle model controlled by an intelligent driving twin algorithm collides with a traffic vehicle and a traffic obstacle or not needs to be detected; the moving track of the vehicle model is detailed in a calculation mode (formula 1) of vehicle kinematic information, the position of a surrounding traffic obstacle is obtained by data and coordinate transformation after sensing by a real vehicle sensor, and whether collision occurs is obtained by an existing algorithm (such as a rectangular bounding box intersection algorithm) for judging whether collision occurs.
b. Is there a violation of traffic light running, traffic sign violation, line pressing running, etc.?
And the rule judgment can be carried out by reading the high-precision map and the vehicle kinematics information.
2. Evaluation criterion for driver comfort
The driver comfort is judged through the dynamic information (calculated by the formula 1) of the vehicle model, and the specific judgment method can be executed according to the existing comfort evaluation mode of the automobile passengers.
While, for purposes of simplicity of explanation, the foregoing method embodiments have been described as a series of acts or combination of acts, it will be appreciated by those skilled in the art that the present invention is not limited by the illustrated ordering of acts, as some steps may occur in other orders or concurrently with other steps in accordance with the invention. Further, those skilled in the art should also appreciate that the embodiments described in the specification are preferred embodiments and that the acts and modules referred to are not necessarily required by the invention.
The method is described in detail in the embodiments disclosed above, and the method of the present invention can be implemented by various types of apparatuses, so that the present invention also discloses an apparatus, and the following detailed description will be given of specific embodiments.
Fig. 8 is a schematic structural diagram of a testing apparatus according to an embodiment of the present invention, and as shown in fig. 8, the testing apparatus 80 may include:
the intelligent driving twinborn algorithm triggering module 801 is used for triggering an intelligent driving twinborn algorithm under a preset condition, the intelligent driving twinborn algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the intelligent driving twinborn algorithm is different from the triggering condition of the intelligent driving algorithm.
And a data input module 802 for acquiring real vehicle sensor data and inputting the data into the vehicle model and the real vehicle.
And the driving control module 803 is configured to drive the vehicle model by using the intelligent driving twinning algorithm based on the real vehicle sensor data.
A result determination module 804 configured to generate a test result based on the operation condition of the vehicle model and the real vehicle condition.
The testing device of the embodiment adopts a real vehicle scene and a sensor as the source of functional verification data, so that the testing device has accuracy superior to that of simulation test verification, and the real vehicle scene does not need to be constructed, thereby avoiding the investment of a large amount of manpower and material resources; the realization can be deployed in mass-production vehicle controller products, then the execution conditions of the actual functions of mass-production vehicles are collected, the test result is obtained, and the method has the advantages of low cost and high accuracy.
In the above embodiment, referring to fig. 9, a specific implementation of the driving control module 803 may be shown, as shown in fig. 9, the driving control module 803 may include:
an information obtaining module 901, configured to obtain first positioning information of an actual vehicle and second positioning information of the vehicle model, and determine a positioning difference between the first positioning information and the second positioning information.
A parameter determining module 902, configured to determine a coordinate conversion parameter between the first positioning information and the second positioning information based on the positioning difference.
And the data conversion module 903 is used for converting the positions of the target traffic vehicle, the obstacle, the road sign and the lane line detected in the real vehicle sensor coordinate system into the vehicle model sensor coordinate system based on the coordinate conversion parameters.
And the driving control sub-module 904 is configured to drive the vehicle model by using the intelligent driving twinning algorithm based on the positions of the target traffic vehicle, the obstacle, the road sign, and the lane line in the vehicle model sensor coordinate system.
The specific implementation and achieved effect of the testing device and the above modules can be described with reference to the content of relevant parts in the method embodiment, and will not be described again here.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
It is further noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A method of testing, comprising:
triggering an intelligent driving twinning algorithm under a preset condition, wherein the intelligent driving twinning algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the triggering condition of the intelligent driving twinning algorithm is different from that of the intelligent driving algorithm;
acquiring real vehicle sensor data and inputting the data into a vehicle model and a real vehicle;
driving the vehicle model by adopting the intelligent driving twinning algorithm based on the real vehicle sensor data;
and generating a test result based on the running condition and the real vehicle condition of the vehicle model.
2. The testing method of claim 1, wherein said driving the vehicle model with the intelligent driving twin algorithm based on the real vehicle sensor data comprises:
acquiring first positioning information of a real vehicle and second positioning information of the vehicle model, and determining a positioning difference of the first positioning information and the second positioning information;
determining a coordinate conversion parameter between the first positioning information and the second positioning information based on the positioning difference;
based on the coordinate conversion parameters, converting the positions of the target traffic vehicles, the obstacles, the road signs and the lane lines detected in the real vehicle sensor coordinate system to be in the vehicle model sensor coordinate system;
and driving the vehicle model by adopting the intelligent driving twinning algorithm based on the positions of the target traffic vehicle, the obstacle, the road sign and the lane line in the coordinate system of the vehicle model sensor.
3. The testing method of claim 1, wherein generating test results based on the operating conditions and real-world conditions of the vehicle model comprises:
and judging whether the vehicle model driven by the intelligent driving algorithm has abnormal behaviors or not based on the running condition and the real vehicle condition of the vehicle model, and determining comfort data of a driver, wherein the running data of the vehicle model comprises a running track and dynamic information.
4. The testing method of claim 1, after generating the test results based on the operating conditions and the real-world conditions of the vehicle model, further comprising:
and uploading the test result and the data in the test process to a server.
5. The testing method of claim 1, wherein the triggering of the intelligent driving twin algorithm under the preset conditions comprises:
triggering an intelligent driving twin algorithm when a driver triggers an intelligent driving function and the function to be tested belongs to a function item which is subjected to a simulation test and is not subjected to a real vehicle test and/or a function item which is not started by the driver;
or the like, or, alternatively,
when a driver triggers the intelligent driving function, the real vehicle sensor detects that the surrounding environment meets the triggering condition for testing the corresponding intelligent driving function, and the function to be tested belongs to the preset function set, the intelligent driving twin algorithm is triggered.
6. The test method of claim 5, further comprising:
and automatically quitting the intelligent driving twinning algorithm when the quitting condition is met.
7. The test method according to claim 6,
when the function to be tested belongs to a function item which is subjected to the simulation test and is not subjected to the real vehicle test, the exit condition comprises the following steps:
the Euclidean distance between the vehicle model and the centroid coordinate of the real vehicle is larger than the radius of a functional domain sensed by a real vehicle sensor, wherein the functional domain is the region where the real vehicle executes the function to be tested by the intelligent driving twin algorithm;
when the function to be tested belongs to a function item which is not started by the driver, the exit condition comprises the following steps:
the Euclidean distance between the vehicle model and the centroid coordinate of the real vehicle is larger than the radius of a functional domain sensed by a real vehicle sensor;
or the like, or, alternatively,
the driver actively turns on the function to be tested that was not previously activated.
8. The testing method of any one of claims 1-7, wherein the vehicle model is a three-free six-state dual-input nonlinear system.
9. A test apparatus, comprising:
the intelligent driving twinborn algorithm comprises an algorithm triggering module, a driving algorithm judging module and a driving algorithm judging module, wherein the algorithm triggering module is used for triggering an intelligent driving twinborn algorithm under a preset condition, the intelligent driving twinborn algorithm is consistent with an algorithm model of the intelligent driving algorithm, and the intelligent driving twinborn algorithm is different from the triggering condition of the intelligent driving algorithm;
the data input module is used for acquiring data of the real vehicle sensor and inputting the data into the vehicle model and the real vehicle;
the driving control module is used for driving the vehicle model by adopting the intelligent driving twinning algorithm based on the real vehicle sensor data;
and the result determining module is used for generating a test result based on the running condition and the real vehicle condition of the vehicle model.
10. The test device of claim 9, wherein the drive control module comprises:
the information acquisition module is used for acquiring first positioning information of a real vehicle and second positioning information of the vehicle model and determining a positioning difference of the first positioning information and the second positioning information;
a parameter determination module configured to determine a coordinate conversion parameter between the first positioning information and the second positioning information based on the positioning difference;
the data conversion module is used for converting the positions of the target traffic vehicle, the obstacle, the road sign and the lane line detected in the real vehicle sensor coordinate system to the vehicle model sensor coordinate system based on the coordinate conversion parameters;
and the driving control submodule is used for driving the vehicle model by adopting the intelligent driving twinning algorithm based on the positions of the target traffic vehicle, the obstacle, the road sign and the lane line under the coordinate system of the vehicle model sensor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010842387.0A CN111930026A (en) | 2020-08-20 | 2020-08-20 | Test method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010842387.0A CN111930026A (en) | 2020-08-20 | 2020-08-20 | Test method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111930026A true CN111930026A (en) | 2020-11-13 |
Family
ID=73306192
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010842387.0A Pending CN111930026A (en) | 2020-08-20 | 2020-08-20 | Test method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111930026A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113050455A (en) * | 2021-03-27 | 2021-06-29 | 上海智能新能源汽车科创功能平台有限公司 | Digital twin test system for intelligent networked automobile and control method |
| CN113885496A (en) * | 2021-09-29 | 2022-01-04 | 三一专用汽车有限责任公司 | Intelligent driving simulation sensor model and intelligent driving simulation method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110221546A (en) * | 2019-05-21 | 2019-09-10 | 武汉理工大学 | The ship intelligence control system test platform of virtual reality fusion |
| CN110348103A (en) * | 2019-07-04 | 2019-10-18 | 北京航空航天大学 | A kind of vehicle twin based on number is anti-to injure appraisal procedure |
| CN110488629A (en) * | 2019-07-02 | 2019-11-22 | 北京航空航天大学 | A kind of management-control method of the hybrid vehicle based on the twin technology of number |
| US20190382003A1 (en) * | 2018-06-13 | 2019-12-19 | Toyota Jidosha Kabushiki Kaisha | Collision avoidance for a connected vehicle based on a digital behavioral twin |
| CN110716558A (en) * | 2019-11-21 | 2020-01-21 | 上海车右智能科技有限公司 | Automatic driving system for non-public road based on digital twin technology |
| US20200126415A1 (en) * | 2018-10-19 | 2020-04-23 | Toyota Jidosha Kabushiki Kaisha | Digital behavioral twin system for intersection management in connected environments |
-
2020
- 2020-08-20 CN CN202010842387.0A patent/CN111930026A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190382003A1 (en) * | 2018-06-13 | 2019-12-19 | Toyota Jidosha Kabushiki Kaisha | Collision avoidance for a connected vehicle based on a digital behavioral twin |
| US20200126415A1 (en) * | 2018-10-19 | 2020-04-23 | Toyota Jidosha Kabushiki Kaisha | Digital behavioral twin system for intersection management in connected environments |
| CN110221546A (en) * | 2019-05-21 | 2019-09-10 | 武汉理工大学 | The ship intelligence control system test platform of virtual reality fusion |
| CN110488629A (en) * | 2019-07-02 | 2019-11-22 | 北京航空航天大学 | A kind of management-control method of the hybrid vehicle based on the twin technology of number |
| CN110348103A (en) * | 2019-07-04 | 2019-10-18 | 北京航空航天大学 | A kind of vehicle twin based on number is anti-to injure appraisal procedure |
| CN110716558A (en) * | 2019-11-21 | 2020-01-21 | 上海车右智能科技有限公司 | Automatic driving system for non-public road based on digital twin technology |
Non-Patent Citations (1)
| Title |
|---|
| 李亚楠: "智能网联汽车数字孪生测试理论和技术研究", CNKI硕士学位论文全文数据库工程科技II辑, pages 4 - 12 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113050455A (en) * | 2021-03-27 | 2021-06-29 | 上海智能新能源汽车科创功能平台有限公司 | Digital twin test system for intelligent networked automobile and control method |
| CN113885496A (en) * | 2021-09-29 | 2022-01-04 | 三一专用汽车有限责任公司 | Intelligent driving simulation sensor model and intelligent driving simulation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11513523B1 (en) | Automated vehicle artificial intelligence training based on simulations | |
| Huang et al. | Autonomous vehicles testing methods review | |
| JP7064303B2 (en) | Autonomous system verification method | |
| Gietelink et al. | Development of advanced driver assistance systems with vehicle hardware-in-the-loop simulations | |
| US20190235521A1 (en) | System and method for end-to-end autonomous vehicle validation | |
| CN112613169B (en) | Expected function safety analysis method for misoperation of automatic driving vehicle | |
| Belbachir et al. | Simulation-driven validation of advanced driving-assistance systems | |
| Alghodhaifi et al. | Autonomous vehicle evaluation: A comprehensive survey on modeling and simulation approaches | |
| EP3958129A1 (en) | Method and system for validating autonomous control software for a self-driving vehicle | |
| CN111795832A (en) | Intelligent driving vehicle testing method, device and equipment | |
| Beglerovic et al. | Model-based safety validation of the automated driving functio highway pilot | |
| CN111930026A (en) | Test method and device | |
| US20230365145A1 (en) | Method, system and computer program product for calibrating and validating a driver assistance system (adas) and/or an automated driving system (ads) | |
| US11702044B2 (en) | Vehicle sensor cleaning and cooling | |
| US20220289248A1 (en) | Vehicle autonomous mode operating parameters | |
| US20230343153A1 (en) | Method and system for testing a driver assistance system | |
| CN114763150A (en) | Ranking fault conditions | |
| Son et al. | A simulation-based testing and validation framework for ADAS development | |
| US11657635B2 (en) | Measuring confidence in deep neural networks | |
| EP4198804A1 (en) | Simulation obstacle vehicles with driving styles | |
| US20220223047A1 (en) | Computer-implemented method for determining similarity values of traffic scenarios | |
| US11738771B2 (en) | Dynamic model with actuation latency | |
| CN117413257A (en) | Method and system for testing driver assistance system for vehicle | |
| Dixit et al. | Modelling and testing of emergency braking in autonomous vehicles | |
| Kong et al. | Simulation based methodology for assessing forced merging strategies for autonomous 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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 4 / F, building 1, No.14 Jiuxianqiao Road, Chaoyang District, Beijing 100020 Applicant after: Beijing Jingwei Hengrun Technology Co., Ltd Address before: 8 / F, block B, No. 11, Anxiang Beili, Chaoyang District, Beijing 100101 Applicant before: Beijing Jingwei HiRain Technologies Co.,Ltd. |