CN115061385A - Real vehicle in-loop simulation test platform based on vehicle road cloud cooperation - Google Patents
Real vehicle in-loop simulation test platform based on vehicle road cloud cooperation Download PDFInfo
- Publication number
- CN115061385A CN115061385A CN202210651440.8A CN202210651440A CN115061385A CN 115061385 A CN115061385 A CN 115061385A CN 202210651440 A CN202210651440 A CN 202210651440A CN 115061385 A CN115061385 A CN 115061385A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- traffic flow
- road
- information
- module
- 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
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 real vehicle in-loop simulation test platform based on vehicle-road cloud cooperation, and belongs to the technical field of automatic driving. The intelligent network vehicle twin-generation system comprises an intelligent road end, an intelligent network vehicle end, a simulation cloud end and a communication server end, wherein the intelligent road end uploads traffic flow data to the simulation cloud end through the communication server end, the intelligent network vehicle end uploads positioning information to the simulation cloud end through the communication server end, the simulation cloud end rapidly simulates interference factors such as different virtual traffic flow elements based on real scenes, the simulation cloud end applies an intelligent network vehicle twin-generation algorithm to make a correct decision according to real-time traffic flow data and transmits the decision to the intelligent network vehicle end in real time so as to drive the intelligent network vehicle end to make corresponding movement, and the intelligent network vehicle end updates posture and positioning information and uploads the decision to the simulation cloud end in real time. The invention constructs a closed loop test system combining the intelligent networked automobile and the mixed environment. The invention can be used as a safer and more effective test technology before the road side test of the automatic driving automobile.
Description
Technical Field
The invention belongs to the technical field of automatic driving, and particularly relates to a real vehicle in-loop simulation test platform based on vehicle-road cloud cooperation.
Background
The testing and verification method of autonomous driving is of great importance in the future. The current test of the automatic driving system is mainly based on real drive test and simulation test, the traditional road test has the advantages that the traffic environment is real, but the diversity of test scenes is limited, and the edge test scenes cannot be safely and efficiently customized; the simulation test is flexible and efficient, but lacks a real traffic flow test environment, and breaks the strong coupling relationship between the vehicle and the environment in an actual application scene. In order to solve the problems, a real-Vehicle in-the-Loop (ViL) test is taken as an effective verification method before a road test. However, current ViL systems are complex, rely heavily on expensive real-time truck systems or vehicles with full autopilot capability, and cannot apply real-time traffic flow data on roads to a simulation test environment. Therefore, the development of the automatic driving system testing method with low cost, high efficiency and high reliability has practical significance.
Disclosure of Invention
The invention provides a real vehicle in-loop simulation test platform based on vehicle road cloud cooperation so as to improve the safety and effectiveness of an automatic driving vehicle road test.
The technical scheme adopted by the invention is as follows:
a real vehicle-in-loop simulation test platform based on vehicle-road cloud cooperation comprises: the system comprises an intelligent road end, an intelligent network connection end, a simulation cloud end and a communication server end;
the intelligent network vehicle-connecting end comprises a combined inertial navigation module, a control module and a communication module, wherein the combined inertial navigation module is used for positioning the intelligent network vehicle-connecting end to obtain a vehicle-end positioning message; the control module is used for controlling a drive-by-wire chassis at the intelligent network vehicle connecting end so as to drive the vehicle to move; the communication module is used for carrying out data interaction with a communication service terminal, and comprises: the intelligent network vehicle-connecting end sends a vehicle-end positioning message to the intelligent road end through the communication server end, and receives decision information from the simulation cloud end through the communication server end;
the intelligent road end comprises a sensing unit, a positioning module and a communication module, wherein the sensing unit comprises a road side camera, an identification module, a screening module and a tracking module, the road side camera is used for collecting road image data and sending the road image data to the identification module, the identification module is used for identifying an appointed traffic element object, and sends the recognition result to a screening module, the screening module selects traffic element objects in the lane range and sends the screening result to a tracking module, the tracking module performs multi-target tracking on the screened traffic element objects, the positioning module converts the representative pixel point of each target into physical coordinate information based on a configured perspective transformation matrix, and obtains perception information of each target, namely first traffic flow information, based on the category information and the physical coordinate information of the targets; deleting perception information of a corresponding vehicle end in the first traffic information to obtain second traffic information; the first traffic information and the second traffic information are sent to a communication server through a communication module of the intelligent road side, and the communication server forwards the first traffic information and the second traffic information to the simulation cloud side;
the simulation cloud end comprises a traffic flow online simulation module, a real vehicle twin decision module and a communication module, wherein the traffic flow online simulation module is used for visualizing first traffic flow information from the intelligent road end; the real vehicle twin decision module obtains decision information in the current traffic scene based on second traffic flow information from the intelligent road end according to a configured decision algorithm to be tested, and sends the decision information to the communication server end through the communication module of the simulation cloud end; the communication server side forwards the decision information to the intelligent network vehicle-connecting side, and a communication module of the intelligent network vehicle-connecting side analyzes the decision information and sends the decision information to a control module of the intelligent network vehicle-connecting side so as to drive the intelligent network vehicle-connecting side to make corresponding movement.
Furthermore, the specific setting mode of the perspective transformation matrix of the positioning module of the intelligent road end is as follows:
determining a traffic flow identification positioning area;
and collecting pixel information and coordinate information of four corner points of the traffic flow identification positioning area, and solving a conversion matrix from the pixel information to the coordinate information through perspective transformation to obtain a perspective transformation matrix.
Further, the traffic flow online simulation module of simulation high in the clouds still is used for: adding virtual interference elements based on real traffic information, visualizing the virtual interference elements, and sending the added virtual interference elements to a real vehicle twin decision module.
Further, the traffic flow online simulation module at simulation high in the clouds is to coming from the first traffic flow information of wisdom road-side and is visualized including: simulating and visualizing a static road, a static building and a dynamic traffic flow scene;
wherein the modeling of the static road comprises: collecting GPS (global positioning system) geographic position information of a road, and converting the geographic position information of the road into plane coordinates through a GPS data conversion tool to obtain the road modeled based on real GPS data;
the modeling of static buildings includes: acquiring real image textures of a building in a driving visual range, and performing 1:1, modeling, and mapping the constructed building model at a corresponding position;
the dynamic traffic flow scene modeling comprises the following steps: and generating a traffic flow model based on the first traffic flow information and controlling the movement of the traffic flow model.
The technical scheme provided by the invention at least has the following beneficial effects:
compared with the existing real vehicle rack in-loop and complete automatic driving real vehicle in-loop simulation test scheme, the invention has the advantages that:
1) real-time traffic flow data is introduced to make up for the difference between the simulation test environment and the actual environment.
2) And adding the real vehicle in the test ring to ensure the strong coupling relation between the real vehicle and the environment.
3) By constructing a closed loop of a real vehicle and a virtual and real combined test environment, the test capability of the extreme scene is improved while the validity of a test result is ensured.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic diagram of a real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation according to an embodiment of the present invention;
FIG. 2 is a schematic processing diagram of a sensing unit of a strong real-time sensing server at a road edge according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a simulation processing process of the simulation cloud in the embodiment of the present invention;
FIG. 4 is a schematic diagram of static scene modeling in an embodiment of the present invention;
fig. 5 is a process of dynamic traffic flow online simulation processing in the simulation cloud according to the embodiment of the present invention.
FIG. 6 is a diagram of rendering effects in online simulation according to an embodiment of the present invention.
Fig. 7 is a schematic structural diagram of a real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation according to an embodiment of the present invention.
FIG. 8 is a schematic illustration of vehicle alignment correction, in accordance with an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
The embodiment of the invention aims at the problem of difference between an automatic driving simulation test environment and a real traffic environment. The reality deficiency of a simulation test scene is made up by introducing real traffic flow information, and a real vehicle-in-loop simulation test platform based on vehicle-road cloud cooperation is provided. On the road side, the embodiment of the invention extracts real-time traffic flow elements of the road and applies real traffic flow behavior information to an automatic driving simulation test. In the cloud platform, the software simulation platform is set based on OpenSceneGraph, and can rapidly simulate different interference factors such as weather, illumination and virtual traffic flow elements based on a real scene. On the intelligent networking real vehicle, the embodiment of the invention constructs a closed loop test system combining the intelligent networking real vehicle and a mixed environment. The real vehicle-in-loop simulation test platform based on vehicle road cloud cooperation provided by the embodiment of the invention can be used as a safer and more effective test technology before the test of the automatic driving vehicle road.
The real vehicle-in-loop simulation test platform based on vehicle-road cloud cooperation provided by the embodiment of the invention is a real vehicle-in-loop twin simulation system based on real-time traffic flow, and comprises a strong real-time perception server (intelligent road end for short), an intelligent network vehicle-connecting end, a simulation cloud end and a communication service end, wherein the intelligent road end mainly carries out beyond-the-horizon perception through road-end equipment and uploads perceived traffic flow data to the simulation cloud end in real time, and on one hand, the intelligent road end can liberate a perception system of an intelligent network vehicle and reduce the calculation requirement in the vehicle end, and on the other hand, the road perception range of the intelligent network vehicle can be increased, so that the decision result is more reasonable. The intelligent network vehicle-connecting end is a main terminal device of a road and carries out real-time data interaction with the intelligent road end and the simulation cloud end through a high-speed network according to corresponding data interface specifications. The simulation cloud carries out fusion processing according to data of the intelligent network vehicle-connecting end and the intelligent road end, on one hand, visual simulation is completed, and on the other hand, decision planning is carried out on the intelligent network vehicle-connecting end, as shown in fig. 1.
(1) And the strong real-time perception server at the edge of the road end.
Because the perception distance of vehicle is limited, wisdom road-end can the decoupling zero perception system of autopilot car on the one hand, and on the other hand can provide the perception effect of beyond the visual range for autopilot car. Specifically, the embodiment of the invention is realized based on a real-time traffic flow perception mode of a road side monocular camera: and (3) completing multi-target recognition and tracking through target recognition and tracking based on deep learning, such as through YOLOv5 and deep sort, by utilizing the advantage of fixed pose of the road end camera and based on the traffic flow picture acquired by the road end camera. Meanwhile, the positioning of the traffic flow element is completed by utilizing the perspective transformation which converts the pixel coordinate into the physical coordinate, as shown in fig. 2, that is, firstly, a road end image is obtained through a road end camera, the road end image is input into a target recognition network after image preprocessing (such as size adjustment, so as to be matched with the input of a target recognition network (such as YOLOv5 target recognition)), the recognition result is screened, a traffic element object in a lane range is selected, a tracking module tracks the screened traffic element object, so that multi-target tracking is realized, and the positioning of the traffic flow element is completed by utilizing the perspective transformation which converts the pixel coordinate into the physical coordinate, so that the multi-target category and the track are obtained.
(2) And (5) simulating a cloud.
In the embodiment of the invention, the provided simulation cloud end adopts a real-time simulation system based on traffic flow twinning and is used for realizing traffic flow online simulation and real vehicle twinning decision. The consistency of the automatic driving simulation test environment and the real environment determines the reliability of the simulation test result to a great extent, and the behavior mapping of virtual traffic elements and actual road traffic participants in the simulation test is a key.
Specifically, as shown in fig. 3, a specific processing process of the simulation cloud provided by the embodiment of the invention is mainly divided into static road, building and dynamic traffic flow scene simulation in a virtual scene construction process. The construction of the static road comprises the following steps: actual road data (geographical location information of the road) is collected and processed to obtain the road modeled by the real GPS data. In the embodiment, a static road drives once on a real road by a vehicle carrying high-precision positioning, the route geographical position information (longitude and latitude altitude points) is recorded, the recorded route geographical position information is converted into a plane coordinate through GPS data conversion software and finally stored in a specified file. Static building modeling based on CAD drawings performs a complete 1:1, modeling, and completing building model mapping by acquiring real image textures of a main area (in a driving visual range) of a building, wherein the overall modeling schematic diagram is shown in FIG. 4.
The dynamic traffic flow scene modeling data is derived from real-time full-factor perception of road traffic flow information by intelligent perception equipment, and comprises vehicles, pedestrians and the like. The dynamic traffic flow online simulation processing process of the simulation cloud is shown in fig. 5, and the final mixed traffic flow simulation rendering effect is shown in fig. 6.
The dynamic traffic flow online simulation treatment specifically comprises the following steps:
(1) initializing a model pool of a traffic flow model;
(2) receiving traffic flow data;
(3) judging whether the currently received traffic flow data is a first frame or not, if so, executing the step (4); otherwise, executing the step (5);
(4) after calling the traffic flow model from the model pool and allocating the model ID, continuing to execute the step (12); reducing the traffic flow models in the model pool by one every time the traffic flow models are called;
(5) if the set of the model IDs is the same as the set of the model IDs of the previous frame, executing the step (12); otherwise, executing the step (6);
(6) if the newly added model ID exists, executing the step (7), otherwise, directly executing the step (11);
(7) whether an available traffic flow model exists in the model pool or not, if so, executing the step (8), otherwise, executing the step (9);
(8) assigning a model ID to the current model, and continuing to execute the step (10);
(9) initializing a traffic flow model and assigning a model ID, and continuing to execute the step (10);
(10) detecting whether a model ID disappears, if so, executing the step (11), otherwise, directly executing the step (12);
(11) recovering the traffic flow model with the disappeared model ID into a model pool (the number of the traffic flow models in the model pool is added with 1), and continuing to execute the step (12);
(12) analyzing traffic flow data (including category information and physical coordinate information) according to the model ID;
(13) and controlling the motion of the current traffic flow model based on the analyzed traffic flow data.
As shown in fig. 7, as a possible implementation manner, the real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation according to the embodiment of the present invention includes: strong real-time perception server in way end edge, intelligent net connection car end, emulation high in the clouds and communication service end, wherein, intelligent net connection car end is used to the navigation module including the combination, control module and communication module, the combination is used to the navigation module and is used for the location (high accuracy location) of intelligent net connection car end, obtain car end positioning message, control module is used for controlling the drive-by-wire chassis of intelligent net connection car end, in order to drive the vehicle motion, communication module is used for the data interaction with communication service end, interactive data include: the intelligent network vehicle-connecting end sends a vehicle-end positioning message to the intelligent road end through the communication server end, and receives decision information from the simulation cloud end through the communication server end;
the strong real-time perception server of the road end edge comprises a perception unit, a positioning module and a communication module, wherein the perception unit comprises a road side camera, an identification module, a screening module and a tracking module, the road side camera is used for collecting road image data and sending the road image data to the identification module, the identification module is used for identifying appointed traffic element objects (such as vehicles and pedestrians) and sending identification results to the screening module, the screening module selects the traffic element objects in a road range and sends the screening results to the tracking module, the tracking module carries out multi-target tracking on the received traffic element objects and sends pixel point information of a tracked target to the positioning module in real time; the positioning module converts a representative pixel (which can be a central point or a point specified by a middle point at the bottom of a target frame and the like) of each target into physical coordinate information based on the configured perspective transformation matrix, and obtains perception information of each target, namely first traffic flow information, based on the category information and the physical coordinate information of the target; deleting perception information of a corresponding vehicle end in the first traffic information to obtain second traffic information, namely obstacle information; the first traffic information and the second traffic information are sent to a communication server through a communication module of the intelligent road side, and the communication server forwards the first traffic information and the second traffic information to the simulation cloud end;
the simulation cloud comprises a traffic flow online simulation module, a real vehicle twin decision module and a communication module, wherein the traffic flow online simulation module is used for visualizing first traffic flow information from the intelligent road end, and is also used for adding a virtual interference element based on real traffic information and visualizing the virtual interference element, and sending the added virtual interference element to the real vehicle twin decision planning module; the real vehicle twin decision module obtains decision information in the current traffic scene based on second traffic flow information from the intelligent road end according to a configured decision algorithm to be tested, and sends the decision information to the communication server end through the communication module of the simulation cloud end; the communication server side forwards the decision information to the intelligent network vehicle-connecting side, and a communication module of the intelligent network vehicle-connecting side analyzes the decision information and sends the decision information to a control module of the intelligent network vehicle-connecting side so as to drive the intelligent network vehicle-connecting side to make corresponding movement.
In the embodiment of the invention, the communication system of the real-vehicle in-loop simulation test platform based on vehicle-road cloud cooperation is formed by communication modules deployed at an intelligent internet connection end, a smart road end and a simulation cloud end, and the communication system selects an open source MQTT-Mosquito message platform scheme widely used in the field of the internet of things as a communication frame and combines a high-speed communication technology (such as mobile 5G/Wifi) to realize a real-time data interaction function between the intelligent internet connection end, the smart road end and the simulation cloud end. In this embodiment, the on-line traffic flow simulation module at the simulation cloud end can subscribe to the first traffic flow information at the intelligent road end in a topic publishing/subscribing manner, and the real vehicle twin decision module at the simulation cloud end subscribes to the second traffic flow information at the intelligent road end. The intelligent network vehicle-connecting end also acquires decision information issued by the simulation cloud end in a subscription mode, and the intelligent road end also acquires vehicle-end positioning information issued by the intelligent network vehicle-connecting end in a subscription mode.
In this example, the smart internet terminal runs a high performance PC. The software environment is a Robot Operating System (ROS), wherein all functional modules, such as a positioning module (combined inertial navigation module), a control module, a communication module and the like, implement internal communication by an ROS topic mechanism. The communication module mainly completes the task of data interaction between the intelligent network vehicle-connecting end and the simulation cloud end/intelligent road end. The information reported by the intelligent network vehicle-connecting end can adopt Json format data, and the related description of the specific data content is shown in table 1. The format and description of the reported data information of the strong real-time sensing server at the edge of the road end are shown in table 2.
Table 1 intelligent network vehicle-connected end reported data format
Table 2 format of reported data of edge server at path end
The simulation cloud end (also called as a simulation cloud server) fuses self data of the intelligent vehicle network connection end and road end traffic flow data of the road end edge strong real-time perception server to perform scene twinning, obtains decision information by applying an intelligent network vehicle twinning algorithm (namely an algorithm to be tested) according to the real-time traffic flow data, transmits the decision information to the intelligent network connection end communication module in real time through the communication service end, and issues corresponding control information to a control module (control module) of the intelligent network connection end. The control module controls the real vehicle chassis to make corresponding movement through a Controller Area Network (CAN), meanwhile updates self posture and positioning information, and uploads the self posture and the positioning information to the simulation cloud end in real time through the communication server end, so that the closed loop of the whole system is completed, and the bidirectional mapping from the physical world to the digital world is realized. The format of the data sent by the simulation cloud is shown in table 3.
Table 3 cloud simulation platform issued data format
As a possible implementation manner, the construction process of the real vehicle-in-loop simulation test platform based on vehicle-road cloud cooperation provided by the embodiment of the invention is as follows:
step S1, the intelligent road terminal obtains real-time traffic flow data based on vehicle-road cooperation:
the method comprises the following steps: obtaining roadside video image data, determining a traffic flow identification positioning area, identifying and tracking multiple targets, setting a perspective transformation matrix, correcting vehicle positioning, and calculating traffic flow element position information;
wherein, step S1 specifically includes:
setting a perspective transformation matrix: determining a traffic flow identification positioning area, collecting pixel information and coordinate information of four corner points of the area, and obtaining a conversion matrix from the pixel information to the coordinate information through perspective conversion, namely a perspective conversion matrix.
Acquiring road-end traffic flow data, acquiring image data of a road-side camera of a sensing unit, identifying traffic flow elements of the acquired image data through YOLOv5, selecting traffic lanes as positioning and tracking areas, screening identification results, and putting the screened identification results into a Deepsort multi-target tracker to realize identification and tracking of the traffic elements in the lanes. And based on the set perspective transformation matrix, the pixels of each object can be converted into physical coordinate information.
The vehicle positioning correction, namely the target vehicle positioning estimation correction is realized by using the geometric position relation between the target vehicle and the road, specifically comprising the following steps:
1. finding out a cross central point of a target vehicle tracking frame from vehicle tracking frame pixel information extracted from a road end, and estimating the actual position of a target vehicle by using local physical coordinates after perspective transformation corresponding to the cross central point pixel of the target vehicle tracking frame;
2. and finishing the positioning estimation correction of the vehicle according to the geometric relationship between the local physical coordinate corresponding to the cross center point of the target vehicle tracking frame and the actual center point coordinate of the target vehicle.
In actual use, if the actual positioning of the vehicle is estimated by directly using the physical coordinates after perspective transformation corresponding to the cross center point pixel of the vehicle tracking frame, under the condition that the height of the vehicle is fixed according to the geometric relationship between the physical coordinates corresponding to the center of the recognition tracking frame and the actual center coordinates of the vehicle, the error between the two will increase along with the increase of the distance between the vehicle and the camera. To reduce the error, the present embodiment adds a correction process of positioning in which the correction amount is calculated while the target vehicle is still regarded as a rectangular parallelepiped. As shown in fig. 8, the center of the vehicle recognition frame corresponds to the geometric center of the cuboid of the target vehicle, and the center point is determined to be located in the front left or right of the actual position of the vehicle by calculating the physical coordinates of the pixels of the center point of the tracking frame; and calculating parameters delta X and delta Y of the X (abscissa) value and the Y (ordinate) value of the central point to be corrected according to the judgment result and the height of the camera and the height of the target vehicle, and correcting on the basis of the previously obtained central point (X, Y) of the target vehicle. The specific calculation formula is as follows:
Δy/y＝h/2H，Δx＝Δy * tanβ；
updating X ═ X +/- Δ X, and updating Y ═ Y- Δ Y;
wherein H represents the height of the vehicle, H represents the height of the camera, and beta represents the included angle between the connecting line from the vehicle to the camera and the optical axis of the camera.
In fig. 8, points a and B indicate two vertices of a rectangle whose sides are Δ x and Δ y.
Step S2, the simulation cloud end is based on the on-line simulation of the real-time traffic flow:
modeling a static three-dimensional model based on real road and building data; the dynamic information mainly comes from the real-time traffic flow data acquired in the step S1, and the dynamic traffic flow information data is rendered in a real-time model after being analyzed; the combination of dynamic traffic flow and static roads restores the real traffic scene.
Wherein, step S2 specifically includes:
step S21: and constructing a virtual static scene. The method is mainly divided into static roads and building construction. And the static road acquires longitude and latitude coordinate points on the real road through a vehicle carrying high-precision positioning to obtain a road curve of the real GPS data. And carrying out complete 1:1 modeling based on a CAD drawing, collecting the real image texture of the main area of the building, and finishing mapping.
Step S22: and constructing a dynamic traffic flow scene. The dynamic traffic flow scene modeling data is derived from real-time full-factor perception of road traffic flow information by intelligent perception equipment, and comprises vehicles, pedestrians and the like.
Step S3, automatic driving simulation based on the real vehicle in the ring:
the intelligent network vehicle connecting end sends positioning state information (real vehicle data) to the simulation cloud end through the combined inertial navigation module; the simulation cloud end screens and makes decisions on real vehicle data and roadside traffic flow data; the simulation cloud end feeds back decision planning steering wheel and accelerator results to the intelligent network vehicle connecting end through the communication server end; and the communication module of the intelligent network vehicle-connecting end issues corresponding control information to the control module so as to drive the real vehicle to change the corresponding posture.
Wherein, step S3 specifically includes:
s31: a communication system. Preferably deploying a Mosquitto MQTT server; a communication module MQTT Client1 at the intelligent network connection end is deployed; deploying a communication module MQTT Client2 of the strong real-time perception server at the edge of the road end; deploying a communication module MQTT Client3 of the simulation cloud; the cross-network transmission is achieved through router port mapping; real-time data interaction between the intelligent network vehicle-connecting end, the intelligent road end and the simulation cloud end.
S32: deploying an intelligent network vehicle-connected end system, and deploying an ROS system at the intelligent network vehicle-connected end; deploying a combined inertial navigation positioning module; deploying a control module; deploying a communication module; the communication module mainly completes the task of data interaction between the intelligent network vehicle connecting end and the simulation cloud end.
S32: deployment of the real vehicle in a loop system, and uploading a road end sensing result (traffic flow data) of the intelligent road end to the simulation cloud end; the data of the intelligent network vehicle connecting end is uploaded to the simulation cloud end in real time; a traffic flow online simulation module of the simulation cloud end fuses self data of the intelligent vehicle network connection end and road end traffic flow data to perform scene twinning, and dynamic traffic flow online simulation processing is achieved; a real vehicle twin decision module of the simulation cloud end obtains decision information in the current traffic scene according to real-time traffic flow data by applying an intelligent networked vehicle twin algorithm; the simulation cloud transmits the decision information to a communication module of the intelligent network vehicle-connecting end in real time; a communication module of the intelligent network vehicle-connecting end issues a corresponding control message to a control module; the control module controls the real vehicle chassis to make corresponding movement through a Controller Area Network (CAN); and the intelligent network vehicle-connecting end updates the posture and the positioning information of the intelligent network vehicle-connecting end and uploads the updated posture and the positioning information to the simulation cloud end in real time.
The invention constructs a set of closed loop system consisting of an intelligent network connection end and a mixed test environment, and can quickly generate a test scene fitting the real traffic condition by introducing real-time traffic flow information; the customizable function of the complex scene is realized by combining virtual and real, the stability and the robustness of the automatic driving algorithm are tested by utilizing the random injection of scene elements, and powerful support is provided for the transition of simulation test to road test.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
What has been described above are merely some embodiments of the present invention. It will be apparent to those skilled in the art that various changes and modifications can be made without departing from the inventive concept thereof, and these changes and modifications can be made without departing from the spirit and scope of the invention.
Claims (6)
1. The utility model provides a real car is at ring simulation test platform based on vehicle road cloud is collaborative which characterized in that includes: the system comprises an intelligent road end, an intelligent network connection end, a simulation cloud end and a communication server end;
the intelligent network vehicle-connecting end comprises a combined inertial navigation module, a control module and a communication module, wherein the combined inertial navigation module is used for positioning the intelligent network vehicle-connecting end to obtain a vehicle-end positioning message; the control module is used for controlling a drive-by-wire chassis at the intelligent network vehicle connecting end so as to drive the vehicle to move; the communication module is used for carrying out data interaction with a communication service end and comprises: the intelligent network vehicle-connecting end sends a vehicle-end positioning message to the intelligent road end through the communication server end, and receives decision information from the simulation cloud end through the communication server end;
the intelligent road end comprises a sensing unit, a positioning module and a communication module, wherein the sensing unit comprises a road side camera, an identification module, a screening module and a tracking module, the road side camera is used for collecting road image data and sending the road image data to the identification module, the identification module is used for identifying a specified traffic element object, and sends the recognition result to a screening module, the screening module selects traffic element objects in the lane range and sends the screening result to a tracking module, the tracking module performs multi-target tracking on the screened traffic element objects, the positioning module converts the representative pixel point of each target into physical coordinate information based on a configured perspective transformation matrix, and obtains perception information of each target, namely first traffic flow information, based on the category information and the physical coordinate information of the targets; deleting the perception information of the corresponding vehicle end in the first traffic information to obtain second traffic information; the first traffic information and the second traffic information are sent to a communication server through a communication module of the intelligent road side, and the communication server forwards the first traffic information and the second traffic information to the simulation cloud side;
the simulation cloud end comprises a traffic flow online simulation module, a real vehicle twin decision module and a communication module, wherein the traffic flow online simulation module is used for visualizing first traffic flow information from the intelligent road end; the real vehicle twin decision module obtains decision information in the current traffic scene based on second traffic flow information from the intelligent road end according to a configured decision algorithm to be tested, and sends the decision information to the communication server end through the communication module of the simulation cloud end; the communication server side forwards the decision information to the intelligent network vehicle-connecting side, and a communication module of the intelligent network vehicle-connecting side analyzes the decision information and sends the decision information to a control module of the intelligent network vehicle-connecting side so as to drive the intelligent network vehicle-connecting side to make corresponding movement.
2. The real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation as claimed in claim 1, wherein the perspective transformation matrix of the positioning module of the smart road-end is specifically configured as follows:
determining a traffic flow identification positioning area;
and collecting pixel information and coordinate information of four corner points of the traffic flow identification positioning area, and solving a conversion matrix from the pixel information to the coordinate information through perspective transformation to obtain a perspective transformation matrix.
3. The real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation of claim 1, wherein the traffic flow online simulation module of the simulation cloud is further configured to: adding virtual interference elements based on real traffic information, visualizing the virtual interference elements, and sending the added virtual interference elements to a real vehicle twin decision module.
4. The real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation of claim 1, wherein the visualization of the first traffic flow information from the smart road-end by the traffic flow on-line simulation module at the simulation cloud end comprises: simulating and visualizing a static road, a static building and a dynamic traffic flow scene;
wherein the modeling of the static road comprises: collecting GPS geographical position information of a road, and converting the GPS geographical position information of the road into plane coordinates through a GPS data conversion tool to obtain a road modeled based on real GPS data;
the modeling of static buildings includes: acquiring real image textures of a building in a driving visual range, and performing 1:1, modeling, and mapping the constructed building model at a corresponding position;
the dynamic traffic flow scene modeling comprises the following steps: and generating a traffic flow model based on the first traffic flow information and controlling the movement of the traffic flow model.
5. The real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation according to claim 4, wherein the processing of dynamic traffic flow scene modeling comprises:
(1) initializing a model pool of a traffic flow model;
(2) judging whether the received first traffic flow information is a first frame, if so, executing the step (3); otherwise, executing the step (4);
(3) calling a traffic flow model from the model pool and allocating a model ID, and then continuing to execute the step (11); reducing the traffic flow models in the model pool by one every time the model pool is called;
(4) whether the model ID set is the same as the model ID set of the previous frame or not is judged, if yes, the step (11) is executed; otherwise, executing the step (5);
(5) if the newly added model ID exists, executing the step (6), otherwise, directly executing the step (10);
(6) whether an available traffic flow model exists in the model pool or not, if so, executing the step (7), otherwise, executing the step (8);
(7) assigning a model ID to the current model and continuing to execute the step (9);
(8) initializing a traffic flow model and assigning a model ID, and continuing to perform step (9);
(9) detecting whether a model ID disappears, if so, executing the step (10), otherwise, directly executing the step (11);
(10) recovering the lost model ID traffic flow model into a model pool, adding 1 to the number of the traffic flow models in the model pool,
and continuing to execute the step (11);
(11) analyzing the first traffic flow information according to the model ID, and acquiring the category information and the physical coordinate information of the traffic flow model;
(12) and controlling the current traffic flow model to move based on the category information and the physical coordinate information of the traffic flow model.
6. The real vehicle-on-loop simulation test platform based on vehicle-road cloud cooperation as claimed in any one of claims 1 to 5, wherein interaction data is realized between the intelligent road end and the communication service end, between the intelligent network vehicle-connecting end and the communication service end, and between the simulation cloud end and the communication service end in a topic publishing/subscribing manner.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210651440.8A CN115061385A (en) | 2022-06-09 | 2022-06-09 | Real vehicle in-loop simulation test platform based on vehicle road cloud cooperation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210651440.8A CN115061385A (en) | 2022-06-09 | 2022-06-09 | Real vehicle in-loop simulation test platform based on vehicle road cloud cooperation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115061385A true CN115061385A (en) | 2022-09-16 |
Family
ID=83200373
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210651440.8A Pending CN115061385A (en) | 2022-06-09 | 2022-06-09 | Real vehicle in-loop simulation test platform based on vehicle road cloud cooperation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115061385A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115688484A (en) * | 2022-11-30 | 2023-02-03 | 西部科学城智能网联汽车创新中心(重庆)有限公司 | WebGL-based V2X simulation method and system |
| CN115691111A (en) * | 2022-09-22 | 2023-02-03 | 连云港杰瑞电子有限公司 | Internet vehicle minimum permeability calculation method suitable for traffic flow data acquisition |
| CN116486612A (en) * | 2023-04-21 | 2023-07-25 | 清华大学 | Mixed traffic queue stability evaluation method and device based on vehicle Lu Yun cooperation |
| CN117074049A (en) * | 2023-10-19 | 2023-11-17 | 中汽传媒(天津)有限公司 | Intelligent driving vehicle road cloud testing system based on virtual-real combination |
-
2022
- 2022-06-09 CN CN202210651440.8A patent/CN115061385A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115691111A (en) * | 2022-09-22 | 2023-02-03 | 连云港杰瑞电子有限公司 | Internet vehicle minimum permeability calculation method suitable for traffic flow data acquisition |
| CN115691111B (en) * | 2022-09-22 | 2024-01-23 | 连云港杰瑞电子有限公司 | Network-connected vehicle minimum permeability calculation method suitable for traffic flow data acquisition |
| CN115688484A (en) * | 2022-11-30 | 2023-02-03 | 西部科学城智能网联汽车创新中心(重庆)有限公司 | WebGL-based V2X simulation method and system |
| CN115688484B (en) * | 2022-11-30 | 2023-07-25 | 西部科学城智能网联汽车创新中心(重庆)有限公司 | V2X simulation method and system based on WebGL |
| CN116486612A (en) * | 2023-04-21 | 2023-07-25 | 清华大学 | Mixed traffic queue stability evaluation method and device based on vehicle Lu Yun cooperation |
| CN116486612B (en) * | 2023-04-21 | 2023-12-26 | 清华大学 | Mixed traffic queue stability evaluation method and device based on vehicle Lu Yun cooperation |
| CN117074049A (en) * | 2023-10-19 | 2023-11-17 | 中汽传媒(天津)有限公司 | Intelligent driving vehicle road cloud testing system based on virtual-real combination |
| CN117074049B (en) * | 2023-10-19 | 2024-01-02 | 中汽传媒(天津)有限公司 | Intelligent driving vehicle road cloud testing system based on virtual-real combination |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109781431B (en) | automatic driving test method and system based on mixed reality | |
| CN115061385A (en) | Real vehicle in-loop simulation test platform based on vehicle road cloud cooperation | |
| CN108574929B (en) | Method and apparatus for networked scene rendering and enhancement in an onboard environment in an autonomous driving system | |
| US11551429B2 (en) | Photorealistic image simulation with geometry-aware composition | |
| Alonso et al. | Accurate global localization using visual odometry and digital maps on urban environments | |
| US11370422B2 (en) | Method and system in a vehicle for improving prediction results of an advantageous driver assistant system | |
| Furgale et al. | Toward automated driving in cities using close-to-market sensors: An overview of the v-charge project | |
| US20220036579A1 (en) | Systems and Methods for Simulating Dynamic Objects Based on Real World Data | |
| US20220137636A1 (en) | Systems and Methods for Simultaneous Localization and Mapping Using Asynchronous Multi-View Cameras | |
| US20220153298A1 (en) | Generating Motion Scenarios for Self-Driving Vehicles | |
| WO2023123837A1 (en) | Map generation method and apparatus, electronic device, and storage medium | |
| Zhao et al. | Autonomous driving simulation for unmanned vehicles | |
| CN117576652B (en) | Road object identification method and device, storage medium and electronic equipment | |
| Mallik et al. | Real-time Detection and Avoidance of Obstacles in the Path of Autonomous Vehicles Using Monocular RGB Camera | |
| Bai et al. | Cyber mobility mirror for enabling cooperative driving automation in mixed traffic: A co-simulation platform | |
| WO2022173647A1 (en) | Synthesizing three-dimensional visualizations from perspectives of onboard sensors of autonomous vehicles | |
| US20210142096A1 (en) | Landscape video stream compression using computer vision techniques | |
| Golovnin et al. | Video processing method for high-definition maps generation | |
| CN113436239A (en) | Monocular image three-dimensional target detection method based on depth information estimation | |
| CN111197986B (en) | Real-time early warning and obstacle avoidance method for three-dimensional path of unmanned aerial vehicle | |
| CN117036607A (en) | Automatic driving scene data generation method and system based on implicit neural rendering | |
| CN115131407B (en) | Robot target tracking method, device and equipment oriented to digital simulation environment | |
| Bai et al. | Cyber mobility mirror for enabling cooperative driving automation: A co-simulation platform | |
| Velat et al. | Vision based vehicle localization for autonomous navigation | |
| CN114565669A (en) | Method for fusion positioning of field-end multi-camera |
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 |