CN113918615A - Simulation-based driving experience data mining model construction method and system - Google Patents
Simulation-based driving experience data mining model construction method and system Download PDFInfo
- Publication number
- CN113918615A CN113918615A CN202110993321.6A CN202110993321A CN113918615A CN 113918615 A CN113918615 A CN 113918615A CN 202110993321 A CN202110993321 A CN 202110993321A CN 113918615 A CN113918615 A CN 113918615A
- Authority
- CN
- China
- Prior art keywords
- data
- scene
- driving experience
- driving
- vehicle
- 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.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Abstract
The invention relates to a driving experience data mining model construction method and system based on simulation, which comprises the following steps: extracting and defining a traffic scene needing to be supported by driving experience data, and recording the location and the occurrence condition described by the traffic scene to form a driving experience scene library; selecting a specific traffic scene from a driving experience scene library, and constructing a simulator of the specific traffic scene by utilizing a simulation technology; simulating walking in a simulator according to set conditions and plans in a multi-person simulated driving or automatic driving mode, and outputting a walking data file by the simulator; the walking data file records the information of the self vehicle and each time of the front vehicle and the rear vehicle; constructing a driving experience data mining model of the specific traffic scene based on the walking data; the vehicle driving data is acquired based on the simulation environment, the scene is convenient to customize, a large amount of running data can be generated in a short time, the data acquisition convenience and the use effectiveness are considered, and the problems that the vehicle data acquisition is difficult and the cost is high are effectively solved.
Description
Technical Field
The invention relates to the technical field of automatic driving, in particular to a driving experience data mining model construction method and system based on simulation.
Background
Automatic driving refers to the driving behavior of human beings to machines, and is also the main traffic pattern in the future. Although artificial intelligence can be given to the machine, there may be a way for the machine to completely replace human driving operations. For example, scenes such as confluence, where driving is difficult, or other complex scenes, may currently require a human driver to take over. That is, to achieve the intelligence equivalent to manual driving for automatic driving, it may be necessary to provide human driving experience to the vehicle in a specific scene.
The driving experience data can be generally understood as: the best practice selected from a large number of driving practices is oriented to a particular scene. The measuring standard is that the safety, comfort and economy of driving are met on the premise that the destination is successfully reached. The driving experience data describes the driving pattern of the excellent practice, including the selection of the driving route, the control of the speed, the control of the steering wheel, the control of acceleration and deceleration, and the like.
However, currently, there are few relevant methods for obtaining driving experience data. The following methods are commonly used:
the method is used for modeling the traffic and vehicles in a specific scene, deducing the vehicle space time by using a mathematical method, and calculating the suggested speed, acceleration, steering angle and the like of the vehicle motion in the specific scene. And (3) large data mining, wherein the method obtains driving experience through analysis and mining of traffic large data, and has high usability due to reality. But practical use is limited due to the way the data is accumulated and acquired. The traffic big data is the running data of the vehicle, and the traditional method for acquiring the running data is that the real vehicle runs on an actual road and acquires the running data in a CAN (controller area network) device or vehicle-mounted PC (personal computer) mode.
However, this method of acquiring data has the following problems:
1. the data analysis authority problem is that for safety reasons, the data of the current vehicle body uses a private protocol, and the protocols of vehicles of different brands and even vehicles of the same brand and different models are different, and only a vehicle manufacturer finally has the analysis authority. 2. The data analysis workload is different from vehicle factories and vehicles of different brands in different data formats, and even if the analysis authority is obtained, the analysis workload is also large. 3. The effective data is difficult to obtain, the running data of the vehicle serves the vehicle running, and the traffic big data analysis is not oriented. The data itself is huge in information amount, and it is difficult to find specific scene (for example, confluence) data from the huge information block, match the location, and synchronize the control information, attitude information and current running environment information of the vehicle. 4. The period of effective data accumulation is long, the cost is high, large data analysis and mining are implemented, a large amount of effective data need to be accumulated, and in a traditional mode, an acquisition module needs to be installed and a data processing platform needs to be constructed in a real vehicle acquisition mode, so that the period can be very long. 5. Under extreme conditions, data cannot be acquired, and under normal conditions, traveling data of a vehicle is generally easy to acquire, but under extreme conditions, such as a scene related to driving under dangerous conditions, traveling using a real vehicle is unrealistic, and the traveling data may not be acquired. That is, the prior art can not solve the problem of acquiring driving experience data well.
Disclosure of Invention
The invention provides a driving experience data mining model construction method and system based on simulation aiming at the technical problems in the prior art, vehicle driving data are obtained based on a simulation environment, a scene is convenient to customize, a large amount of traveling data can be generated in a short time, the convenience of data obtaining and the effectiveness of data use are considered, and the problems that the vehicle data are difficult to obtain and the cost is high in a general method are effectively solved.
According to a first aspect of the invention, a driving experience data mining model construction method based on simulation is provided, and comprises the following steps: step 1, extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene, and forming a driving experience scene library;
step 2, selecting a specific traffic scene from the driving experience scene library, and constructing a simulator of the specific traffic scene by using a simulation technology;
step 3, simulating walking is implemented in the simulator according to set conditions and plans in a mode of multi-person simulated driving or automatic driving, and the simulator outputs walking data files; the walking data file records information of each time of the self vehicle and the front and rear vehicles;
and 4, constructing a driving experience data mining model of the specific traffic scene based on the walking data.
On the basis of the technical scheme, the invention can be improved as follows.
Optionally, the traffic scenario requiring the support of the driving experience data in step 1 includes:
the traffic scene that the autopilot function is low, the traffic scene that autopilot is difficult, the traffic scene that accident is frequent and the traffic scene of the complicated place of traffic.
Optionally, step 2 includes:
step 201, constructing a virtual scene based on real road data of a place where the specific traffic scene is located, wherein the virtual scene comprises a road network, road marked lines, road facilities and other traffic participants;
step 202, constructing virtual vehicles and traffic flows based on occurrence conditions recorded in the specific traffic scene;
step 203, setting the running conditions according with the actual traffic rules under the specific traffic scene, and making a simulated running plan.
Optionally, the information of the own vehicle and the preceding and following vehicles at each time includes:
the vehicle control information, the vehicle position information, the vehicle dynamics information, the vehicle posture information, the correlation information of the front and rear vehicles and the running index information, wherein the running index information comprises: whether the driving purpose, the safety index, the comfort index and the economic index are achieved.
Optionally, the process of constructing the driving experience data mining model of the specific traffic scene in step 4 includes:
step 401, sequentially extracting parameter values of each influence factor from the walking data, respectively establishing a scatter diagram between the parameter values of each influence factor and the parameter values of the driving index, and analyzing to obtain main influence factors influencing the driving index in each influence factor;
step 402, carrying out segmented cumulative statistics on all the walking results according to the parameter values of the main influence factors, and taking the segmented numerical value with the largest ratio as the driving experience data of the main influence factors;
and 403, constructing a driving experience data mining model of the specific traffic scene based on the traveling data, the analysis method and the corresponding result.
Optionally, after the step 4, the method further includes:
and 5, collecting vehicle traveling data in the specific traffic scene, and excavating driving experience data according to the driving experience data mining model.
Optionally, step 5 further includes:
judging whether the walking data output by the simulator is representative or not according to whether the specific traffic scene is a simple scene or not;
and when the traveling data are not representative, erecting observation equipment on an automatic driving vehicle or a road after the construction of the driving experience data mining model is completed, collecting parameter data in the driving experience data mining model, and mining the driving experience data in real time on line according to a method described in the driving experience data mining model.
According to a second aspect of the present invention, there is provided a simulation-based driving experience data mining model construction system, comprising: the system comprises a scene library generating module, a simulator generating module, a walking data simulation generating module and a model building module;
the scene library generation module is used for extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene and forming a driving experience scene library;
the simulator generation module is used for selecting a specific traffic scene from the driving experience scene library and constructing a simulator of the specific traffic scene by utilizing a simulation technology;
the walking data simulation generation module is used for implementing simulated walking in the simulator according to set conditions and plans in a multi-person simulated driving or automatic driving mode, and the simulator outputs walking data files; the walking data file records information of each time of the self vehicle and the front and rear vehicles;
and the model building module is used for building a driving experience data mining model of the specific traffic scene based on the walking data.
According to a third aspect of the present invention, there is provided an electronic device comprising a memory, a processor for implementing the steps of the simulation-based driving experience data mining model construction method when executing a computer management class program stored in the memory.
According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having stored thereon a computer management-like program which, when executed by a processor, implements the steps of the simulation-based driving experience data mining model construction method.
According to the driving experience data mining model construction method based on simulation, the system, the electronic equipment and the storage medium, the limitation that the traveling data is protected by a private protocol in the traditional method is broken through, and the vehicle traveling data can be obtained quickly and conveniently; the walking scene, walking conditions and simulation parameters can be customized at will; the actual vehicle is not needed to travel on the spot, so that the cost for acquiring the driving experience data is greatly reduced; the contents of the walking data can be abundant enough, and the dimensionality related to the driving experience data can be expanded; data under an extreme dangerous scene that a real vehicle cannot run on the spot can be acquired; if the simple scene is oriented, the data of the simulated walking is evaluated to be representative, and the driving experience data can be directly generated through the data of the simulated walking. In other cases, after the construction of the driving experience data mining model is completed, observation equipment is erected on the automatic driving vehicle or the road, the parameter data in the model are collected, the driving experience data is mined on line in real time according to the method described in the model, and the development period of the driving experience data is effectively shortened.
Drawings
FIG. 1 is a flow chart of an embodiment of a method for constructing a driving experience data mining model based on simulation according to the present invention;
fig. 2 is a distribution diagram of successful cumulative times of merging flows under different speed differences according to an embodiment of the present invention;
FIG. 3 is a graph illustrating the successful cumulative times of confluence at different accelerations for the same speed difference according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a hardware structure of a possible electronic device provided in the present invention;
fig. 5 is a schematic diagram of a hardware structure of a possible computer-readable storage medium according to the present invention.
Detailed Description
The principles and features of this invention are described below in conjunction with the following drawings, which are set forth by way of illustration only and are not intended to limit the scope of the invention.
At present, automatic driving is taken over manually at places with low automatic driving function or difficult driving. If the corresponding scene is recognized and the driving experience of the human driver is recorded, the knowledge is provided to the machine, so that the problem can be solved and the automatic driving range can be expanded.
In order to solve the problem of acquiring the driving experience data in a specific scene, the invention provides a driving experience data mining model construction method based on simulation in view of the defects and shortcomings of the prior general technology, which comprises the following steps: traffic data for mining driving experience data is acquired. Modeling is carried out by taking a specific scene as an example, and traffic data is recorded for analysis and mining. And mining driving experience data by using the established model. Fig. 1 is a flowchart of a simulation-based driving experience data mining model construction method provided by the present invention, and as shown in fig. 1, the construction method includes:
step 1, extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene, and forming a driving experience scene library.
The automatic driving is difficult to realize in a scene difficult to drive or other complex scenes, so that the automatic driving achieves the intelligence equivalent to manual driving, and the driving experience of human beings in a specific scene needs to be provided for a vehicle, so that the traffic scene needing the driving experience data for support needs to be manually defined firstly.
And 2, selecting a specific traffic scene from the driving experience scene library, and constructing a simulator of the specific traffic scene by utilizing a simulation technology.
Step 3, simulating walking is implemented in a simulator according to set conditions and plans in a mode of multi-person simulated driving or automatic driving, and the simulator outputs a large number of walking data files; the travel data file records information of the own vehicle and the preceding and following vehicles at each time.
And 4, constructing a driving experience data mining model of the specific traffic scene based on the walking data.
According to the method for constructing the driving experience data mining model based on simulation, the vehicle driving data is obtained based on the simulation environment, the scene is convenient to customize, a large amount of traveling data can be generated in a short time, the convenience and the effectiveness of data obtaining are considered, and the problems that the vehicle data is difficult to obtain and the cost is high in a general method are effectively solved.
Example 1
Embodiment 1 provided by the present invention is an embodiment of a method for constructing a driving experience data mining model based on simulation provided by the present invention, and as shown in fig. 1, is a flowchart of an embodiment of a method for constructing a driving experience data mining model based on simulation provided by the present invention, and as can be seen with reference to fig. 1, the embodiment includes:
step 1, extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene, and forming a driving experience scene library.
It is understood that the sources of the scene include, but are not limited to: brainstorming, traffic accidents, and testing of scenarios defined by regulations.
The types of traffic scenes in the scene library that need to be supported by the driving experience data include, but are not limited to:
traffic scenarios with difficulty in autopilot. For example: and (5) merging the scenes.
A traffic scenario with reduced autopilot functionality, comprising: unsafe, uncomfortable, energy-consuming, inefficient, etc. scenarios, such as: sudden acceleration, deceleration, discomfort caused by sudden bending and the like.
The accident frequently occurs, the traffic is complicated scene.
Scene library the content of each traffic scene record includes but is not limited to:
and the scene name and the scene number are convenient to search.
The specific conditions that occur in a traffic scenario, for example: time, location, weather conditions, road surface characteristics, environmental factors, etc.
Driving characteristics of the vehicle, for example: travel speed, acceleration, steering wheel angle, etc.
Characteristics of background traffic flow, for example: average vehicle speed, average inter-vehicle distance, etc.
And 2, selecting a specific traffic scene from the driving experience scene library, and constructing a simulator of the specific traffic scene by utilizing a simulation technology.
The simulator building process uses common simulation platform software and a mainstream virtual reality engine to build a vehicle running simulation environment. Simulation platform software, including but not limited to: CarSim, PreScan, CarMake, VTD, Carla, aiSim, 51VR, and Simulink, among others.
Mainstream virtual reality engines, including but not limited to: UE, Unity, etc
In one possible embodiment, the process of constructing the simulator includes:
step 201, a virtual scene is constructed based on real road data of a place where a specific traffic scene is located, wherein the virtual scene comprises a road network, road marking lines, road facilities and other traffic participants.
In a possible embodiment, constructing a virtual scene for scene simulation includes:
(1) road network construction
The road network in the simulation scene is constructed by using the local high-precision map of the place where the scene is located, the map at least comprises longitude and latitude coordinates, altitude, transverse gradient, longitudinal gradient and curvature of roads and lanes, and the real road and lane conditions can be truly reflected.
(2) Road marking
Road markings that are consistent with actual roads are identified in the scene. The road marking is a road surface printed matter and at least comprises lane boundary lines, driving direction markings, flow guide belts, deceleration markings, stop lines, zebra stripes and the like. These may be derived from high precision maps or field acquisitions, facilitating simulated driving.
(3) Road installation
Road facilities consistent with actual roads are simulated in the scene. The road facilities are equipment and facilities closely related to traffic on the road, at least comprise guardrails, isolation belts, traffic lights and signboards, and serve for simulated driving.
(4) Other traffic participants
Other traffic participants refer to subjects influencing vehicle driving except for vehicle research objects, including pedestrians, livestock and non-motor vehicles. This item acts as a simulation option.
In step 202, a virtual vehicle and traffic flow are constructed based on the occurrence conditions described in the specific traffic scene.
And constructing a virtual vehicle to perform vehicle simulation, wherein the vehicle simulation means that the subject uses the virtual vehicle to replace a real vehicle, and then simulated driving can be realized in the virtual scene constructed in step 201. At least the following is required:
(1) vehicle model
The vehicle model comprises a motion model and a dynamic model, is constructed by using the completely consistent vehicle parameters of the real vehicle, and can be used for constructing a virtual vehicle in a virtual world so that the virtual vehicle has the static and dynamic properties basically consistent with those of the real vehicle.
(2) Sensor model
The sensor is a device for sensing the position and the surrounding environment of the vehicle, and helps a driver or an automatic driving device to make a decision and control on vehicle driving. For example, the position of the vehicle, the distance between the vehicle and the vehicle, whether the vehicle is traveling with a pressed line, and the like are confirmed. And constructing a sensor model, and using the sensor model to construct a virtual sensor in a virtual world, so that the virtual sensor provides data related to the self vehicle and the surrounding environment of the self vehicle.
(3) Driving simulator (or automatic driving algorithm)
The driving simulator is a device for operating a vehicle in the virtual world, and includes at least an accelerator pedal, a steering wheel, a brake pedal, a gear, and the like. Let the simulation drive more have the sense of immersing, can be equipped with VR glasses. An automatic driving algorithm module can be arranged in consideration of the large-scale walking occasions.
The virtual traffic flow is constructed to carry out background traffic flow simulation, the background traffic flow which is an important factor influencing vehicle running needs to be created in a virtual world, and a traffic environment is built for simulating running.
In one possible embodiment, after the simulator is built, the simulator needs to be evaluated. And evaluating whether the simulator performs in the virtual world equivalently to the real world. Evaluation was mainly carried out from two aspects: availability, namely the simulator can be normally used and can output walking data. Consistency, namely under the same condition, all parameters of the walking data output by the vehicle in the virtual world are basically consistent or equivalent to those of the walking data output in the real world. When the two conditions are met, the data for simulating the running by using the simulator can be verified to replace the data for real-vehicle running on the spot.
Step 203, setting the running conditions according with the actual traffic rules under the specific traffic scene, and making a simulated running plan.
Step 3, simulating walking is implemented in a simulator according to set conditions and plans in a mode of multi-person simulated driving or automatic driving, and the simulator outputs a large number of walking data files; the travel data file records information of the own vehicle and the preceding and following vehicles at each time.
The process controls the vehicle to complete a set traveling task by simulating driving or automatic driving in a simulator simulation environment according to requirements under a specific scene. In a possible embodiment, the following is specified:
step 301, determining a walking target.
And indicating the determined walking scene, the running indexes, the qualified values, the characteristics required to be met by the walking data and the like, and expecting the effect achieved by the simulated walking.
Step 302, configuring a running environment.
Various parameters of the simulator are set to meet the condition of scene definition. The method comprises the steps of defining time, place, weather in a scene, driving speed of a vehicle, average speed of a background vehicle, average inter-vehicle distance and the like, and outputting fields of traveling data.
Step 303, making a walking plan.
The walking plan comprises arrangement of drivers, walking rules, walking times, a walking data file naming mode, a storage position and the like.
And step 304, implementing a running scheme.
And (5) implementing walking according to the walking plan and storing walking data.
In a possible embodiment, the information of the own vehicle and the preceding and following vehicles at each time at least comprises the following contents:
the vehicle control information, the vehicle position information, the vehicle dynamics information, the vehicle posture information, the correlation information of the front and rear vehicles and the running index information, wherein the running index information comprises: whether the driving purpose, the safety index, the comfort index and the economic index are achieved.
Vehicle control information, for example: accelerator pedal, steering wheel angle, braking status, etc.
Vehicle position information, for example: longitude, latitude, altitude, lane of presence, etc.
Vehicle dynamics information, for example: speed, acceleration, angular velocity, torque, etc.
Vehicle attitude information, for example: pitch, Roll, Yaw, etc.
The related information of the preceding and following vehicles, for example: relative velocity, etc.
Whether or not the purpose of travel is achieved, for example: whether the running is successful, and the like.
Safety indicators, for example: the inter-vehicle distance, the inter-vehicle time THW, etc.
Comfort indicators, for example: lateral G (lateral acceleration), longitudinal G, etc.
Economic indicators, for example: hundred kilometers of oil consumption, power consumption, etc.
And 4, constructing a driving experience data mining model of the specific traffic scene based on the walking data.
The process of constructing the driving experience data mining model of the specific traffic scene in the step 4 comprises the following steps:
(1) running result calibration
And calibrating the result of each walking manually or by using a program, and storing the calibration result as the running index information. The calibration standards are: comprehensively judging whether the purpose is achieved, whether the safety is ensured, whether the comfort is ensured and the like according to the running of each time.
For example, in a confluent scenario:
is the vehicle normally merging from the ramp to the main road? Running process is complete
Is there a safe driving without or with rear-end collision? Time between plants (THW) > 3 seconds
(iii) comfortable or not? The horizontal G value and the vertical G value are both within 0.25G (G is a gravity acceleration constant)
If the above are satisfied, the running result is marked as 1, otherwise, the running result is marked as 0.
(2) Analysis of major influencing factors
The method comprises the steps of sequentially extracting parameter values of all influence factors described by vehicle dynamics information, association information of front and rear vehicles and the like from traveling data, establishing a scatter diagram by respectively using the parameter values of all the influence factors and the parameter values of determined traveling indexes, analyzing the correlation between all the parameters and the traveling index parameters by using the scatter diagram, and finding out strongly correlated (for example, approximately linear correlation or exponential correlation and the like) parameters to be used as main influence factors influencing the traveling indexes in all the influence factors.
For example, in a confluence scenario:
the difference value between the speed of the vehicle reaching the confluence point on the ramp and the speed of the vehicle on the main road is larger, the confluence failure rate is higher, and the judgment can be carried out: the speed difference is a main factor affecting the merge driving index.
Secondly, after entering the confluence point, an additional section of acceleration buffer area is arranged beside the main road, even if the speed difference described in the first step is larger, if the vehicle enters the acceleration buffer area, acceleration and deceleration can be carried out at proper acceleration to reduce the speed difference, and the confluence success rate can be improved. The judgment can be carried out as follows: acceleration of the vehicle after entering the confluence point is another major contributor to the confluence driving index.
(3) Estimating driving experience data
And after the main influence factors are determined, carrying out segmented cumulative statistics on the walking results of the walking records with all walking results marked as 1 according to the parameter values of the main influence factors, forming a statistical distribution map, and taking the segmented numerical value with the largest occupation ratio as the driving experience data of the main influence factors.
For example, in a confluence scenario:
after the speed difference and the acceleration are determined as main influence factors, the formed segmented statistical distribution graph is shown in fig. 2 and fig. 3, fig. 2 is a distribution graph of the successful accumulated times of confluence under different speed differences provided by the embodiment of the invention, and fig. 3 is a distribution graph of the successful accumulated times of confluence under different accelerations provided by the embodiment of the invention. It can be presumed that:
the speed difference is in the range of [ -5, 5], and the control is optimal.
And the acceleration is in the interval of 0.09g and 0.15g, so that the optimal control is realized.
The estimation result may be recommended as driving experience data.
Other methods for mining driving experience data have also been chosen, such as: and (4) AI.
(4) Construction of driving experience data model
And constructing a driving experience data mining model of the specific traffic scene based on the traveling data, the analysis method and the corresponding result.
In a possible embodiment, step 4 is further followed by:
and 5, collecting vehicle traveling data in the specific traffic scene, and excavating actual available driving experience data according to the driving experience data mining model.
It can be understood that, in step 5, the actual available driving experience data is mined according to the analysis and mining method described by the model.
And judging whether the traveling data output by the simulator is representative according to whether the specific traffic scene is a simple scene.
If the driving simulation method is oriented to a simple scene, the data of the simulated traveling is evaluated to be representative, and the driving experience data can be directly generated through the data of the simulated traveling.
Secondly, if the walking data output by the simulator is not representative, after the construction of the driving experience data mining model is completed, erecting observation equipment on the automatic driving vehicle or the road, collecting parameter data in the driving experience data mining model, and mining the driving experience data in real time on line according to a method described in the driving experience data mining model.
Example 2
The embodiment 2 provided by the invention is an embodiment of a driving experience data mining model construction system based on simulation provided by the invention, and the embodiment comprises the following steps: the system comprises a scene library generating module, a simulator generating module, a walking data simulation generating module and a model building module.
And the scene library generating module is used for extracting and defining the traffic scene needing to be supported by the driving experience data, recording the location and the occurrence condition described by the traffic scene and forming a driving experience scene library.
And the simulator generation module is used for selecting a specific traffic scene from the driving experience scene library and constructing a simulator of the specific traffic scene by utilizing a simulation technology.
The walking data simulation generation module is used for implementing simulated walking in the simulator according to set conditions and plans in a multi-person simulated driving or automatic driving mode, and the simulator outputs a walking data file; the travel data file records information of the own vehicle and the preceding and following vehicles at each time.
And the model building module is used for building a driving experience data mining model of the specific traffic scene based on the walking data.
It can be understood that the simulation-based driving experience data mining model construction system provided by the invention corresponds to the simulation-based driving experience data mining model construction method provided by each of the foregoing embodiments, and the relevant technical features of the simulation-based driving experience data mining model construction system can refer to the relevant technical features of the simulation-based driving experience data mining model construction method, and are not described herein again.
Referring to fig. 4, fig. 4 is a schematic view of an embodiment of an electronic device according to an embodiment of the invention. As shown in fig. 4, an embodiment of the present invention provides an electronic device, which includes a memory 1310, a processor 1320, and a computer program 1311 stored in the memory 1310 and executable on the processor 1320, where the processor 1320 executes the computer program 1311 to implement the following steps: step 1, extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene, and forming a driving experience scene library; step 2, selecting a specific traffic scene from the driving experience scene library, and constructing a simulator of the specific traffic scene by using a simulation technology; step 3, simulating walking is implemented in a simulator according to set conditions and plans in a mode of multi-person simulated driving or automatic driving, and the simulator outputs a walking data file; the walking data file records the information of the self vehicle and each time of the front vehicle and the rear vehicle; and 4, constructing a driving experience data mining model of the specific traffic scene based on the walking data.
Referring to fig. 5, fig. 5 is a schematic diagram of an embodiment of a computer-readable storage medium according to the present invention. As shown in fig. 5, the present embodiment provides a computer-readable storage medium 1400, on which a computer program 1411 is stored, which computer program 1411, when executed by a processor, implements the steps of: step 1, extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene, and forming a driving experience scene library; step 2, selecting a specific traffic scene from the driving experience scene library, and constructing a simulator of the specific traffic scene by using a simulation technology; step 3, simulating walking is implemented in a simulator according to set conditions and plans in a mode of multi-person simulated driving or automatic driving, and the simulator outputs a walking data file; the walking data file records the information of the self vehicle and each time of the front vehicle and the rear vehicle; and 4, constructing a driving experience data mining model of the specific traffic scene based on the walking data.
The simulation-based driving experience data mining model construction method, system and storage medium provided by the embodiment of the invention break through the limitation that the traveling data is protected by a private protocol in the traditional method, and can quickly and conveniently obtain the vehicle traveling data; the walking scene, walking conditions and simulation parameters can be customized at will; the actual vehicle is not needed to travel on the spot, so that the cost for acquiring the driving experience data is greatly reduced; the contents of the walking data can be abundant enough, and the dimensionality related to the driving experience data can be expanded; data under an extreme dangerous scene that a real vehicle cannot run on the spot can be acquired; if the simple scene is oriented, the data of the simulated walking is evaluated to be representative, and the driving experience data can be directly generated through the data of the simulated walking. In other cases, after the construction of the driving experience data mining model is completed, observation equipment is erected on the automatic driving vehicle or the road, parameter data in the model are collected, and the driving experience data is mined on line in real time according to the method described in the model, so that the development period of the driving experience data is effectively shortened.
It should be noted that, in the foregoing embodiments, the descriptions of the respective embodiments have respective emphasis, and reference may be made to relevant descriptions of other embodiments for parts that are not described in detail in a certain embodiment.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. A driving experience data mining model construction method based on simulation is characterized by comprising the following steps:
step 1, extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene, and forming a driving experience scene library;
step 2, selecting a specific traffic scene from the driving experience scene library, and constructing a simulator of the specific traffic scene by using a simulation technology;
step 3, simulating walking is implemented in the simulator according to set conditions and plans in a mode of multi-person simulated driving or automatic driving, and the simulator outputs walking data files; the walking data file records information of each time of the self vehicle and the front and rear vehicles;
and 4, constructing a driving experience data mining model of the specific traffic scene based on the walking data.
2. The building method according to claim 1, wherein the traffic scene requiring the support of driving experience data in the step 1 comprises:
the traffic scene that the autopilot function is low, the traffic scene that autopilot is difficult, the traffic scene that accident is frequent and the traffic scene of the complicated place of traffic.
3. The building method according to claim 1, wherein the step 2 comprises:
step 201, constructing a virtual scene based on real road data of a place where the specific traffic scene is located, wherein the virtual scene comprises a road network, road marked lines, road facilities and other traffic participants;
step 202, constructing virtual vehicles and traffic flows based on occurrence conditions recorded in the specific traffic scene;
step 203, setting the running conditions according with the actual traffic rules under the specific traffic scene, and making a simulated running plan.
4. The building method according to claim 1, wherein the information of each time of the own vehicle and the preceding and following vehicles includes:
the vehicle control information, the vehicle position information, the vehicle dynamics information, the vehicle posture information, the correlation information of the front and rear vehicles and the running index information, wherein the running index information comprises: whether the driving purpose, the safety index, the comfort index and the economic index are achieved.
5. The construction method according to claim 4, wherein the process of constructing the driving experience data mining model of the specific traffic scene in the step 4 comprises:
step 401, sequentially extracting parameter values of each influence factor from the walking data, respectively establishing a scatter diagram between the parameter values of each influence factor and the parameter values of the driving index, and analyzing to obtain main influence factors influencing the driving index in each influence factor;
step 402, carrying out segmented cumulative statistics on all the walking results according to the parameter values of the main influence factors, and taking the segmented numerical value with the largest ratio as the driving experience data of the main influence factors;
and 403, constructing a driving experience data mining model of the specific traffic scene based on the traveling data, the analysis method and the corresponding result.
6. The building method according to claim 1, wherein the step 4 is further followed by:
and 5, collecting vehicle traveling data in the specific traffic scene, and excavating driving experience data according to the driving experience data mining model.
7. The building method according to claim 6, the step 5 further comprising:
judging whether the walking data output by the simulator is representative or not according to whether the specific traffic scene is a simple scene or not;
and when the traveling data are not representative, erecting observation equipment on an automatic driving vehicle or a road after the construction of the driving experience data mining model is completed, collecting parameter data in the driving experience data mining model, and mining the driving experience data in real time on line according to a method described in the driving experience data mining model.
8. A simulation-based driving experience data mining model building system, characterized in that the building system comprises: the system comprises a scene library generating module, a simulator generating module, a walking data simulation generating module and a model building module;
the scene library generation module is used for extracting and defining a traffic scene needing to be supported by driving experience data, recording the location and the occurrence condition described by the traffic scene and forming a driving experience scene library;
the simulator generation module is used for selecting a specific traffic scene from the driving experience scene library and constructing a simulator of the specific traffic scene by utilizing a simulation technology;
the walking data simulation generation module is used for implementing simulated walking in the simulator according to set conditions and plans in a multi-person simulated driving or automatic driving mode, and the simulator outputs walking data files; the walking data file records information of each time of the self vehicle and the front and rear vehicles;
and the model building module is used for building a driving experience data mining model of the specific traffic scene based on the walking data.
9. An electronic device comprising a memory, a processor for implementing the steps of the simulation-based driving experience data mining model construction method according to any one of claims 1 to 7 when executing a computer management class program stored in the memory.
10. A computer-readable storage medium, having stored thereon a computer management-like program which, when executed by a processor, implements the steps of the simulation-based driving experience data mining model construction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110993321.6A CN113918615A (en) | 2021-08-26 | 2021-08-26 | Simulation-based driving experience data mining model construction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110993321.6A CN113918615A (en) | 2021-08-26 | 2021-08-26 | Simulation-based driving experience data mining model construction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113918615A true CN113918615A (en) | 2022-01-11 |
Family
ID=79233272
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110993321.6A Withdrawn CN113918615A (en) | 2021-08-26 | 2021-08-26 | Simulation-based driving experience data mining model construction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113918615A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115240409A (en) * | 2022-06-17 | 2022-10-25 | 上海智能网联汽车技术中心有限公司 | Method for extracting dangerous scene based on driver model and traffic flow model |
| CN115524996A (en) * | 2022-09-13 | 2022-12-27 | 工业和信息化部装备工业发展中心 | Edge scene supplement method and device of analog simulation scene library |
-
2021
- 2021-08-26 CN CN202110993321.6A patent/CN113918615A/en not_active Withdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115240409A (en) * | 2022-06-17 | 2022-10-25 | 上海智能网联汽车技术中心有限公司 | Method for extracting dangerous scene based on driver model and traffic flow model |
| CN115240409B (en) * | 2022-06-17 | 2024-02-06 | 上智联(上海)智能科技有限公司 | Method for extracting dangerous scene based on driver model and traffic flow model |
| CN115524996A (en) * | 2022-09-13 | 2022-12-27 | 工业和信息化部装备工业发展中心 | Edge scene supplement method and device of analog simulation scene library |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111897305B (en) | Data processing method, device, equipment and medium based on automatic driving | |
| CN108508881B (en) | Automatic driving control strategy adjusting method, device, equipment and storage medium | |
| US10943414B1 (en) | Simulating virtual objects | |
| CN113032285B (en) | High-precision map testing method and device, electronic equipment and storage medium | |
| US10713148B2 (en) | Using divergence to conduct log-based simulations | |
| CN109100155A (en) | A kind of unmanned vehicle is in ring high-speed simulation test macro and method | |
| CN110103983A (en) | System and method for the verifying of end-to-end autonomous vehicle | |
| CN101676689B (en) | Driving evaluation system and driving evaluation method | |
| Zhou et al. | A framework for virtual testing of ADAS | |
| CN108528458A (en) | System and method for vehicle dimension prediction | |
| US11385991B1 (en) | Collision evaluation for log-based simulations | |
| CN111795832B (en) | Intelligent driving vehicle testing method, device and equipment | |
| CN114879631A (en) | Automatic driving test system and method based on digital twin cloud control platform | |
| CN112819968B (en) | Test method and device for automatic driving vehicle based on mixed reality | |
| CN114077541A (en) | Method and system for validating automatic control software for an autonomous vehicle | |
| Sahlholm | Distributed road grade estimation for heavy duty vehicles | |
| US20220198107A1 (en) | Simulations for evaluating driving behaviors of autonomous vehicles | |
| CN113918615A (en) | Simulation-based driving experience data mining model construction method and system | |
| US20220204009A1 (en) | Simulations of sensor behavior in an autonomous vehicle | |
| WO2021147071A1 (en) | Cross-platform control profiling for autonomous vehicle control | |
| US20230150549A1 (en) | Hybrid log simulated driving | |
| US20230031130A1 (en) | Open door reconstruction for sensor simulation | |
| US11126763B1 (en) | Realism metric for testing software for controlling autonomous vehicles | |
| US20210387628A1 (en) | Extracting agent intent from log data for running log-based simulations for evaluating autonomous vehicle software | |
| CN110426215B (en) | Model establishing method for vehicle ride comfort test and intelligent driving system |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20220111 |