CN114882698A - Miniature vehicle dynamic path planning vehicle-road coordination system based on V2X - Google Patents

Abstract

The invention provides a micro vehicle dynamic path planning vehicle-road coordination system based on V2X, which relates to the technical field of intelligent networked automobiles and intelligent traffic, and comprises the following steps: a sand table model and a miniature vehicle model; the sand table model comprises: roads, road infrastructure, electronic components, and sensors; wherein the electronics are responsible for processing the asset data collected by the sensors. The invention can complete the test of the unmanned miniature vehicle in the simulation environment.

Description

Miniature vehicle dynamic path planning vehicle-road coordination system based on V2X

Technical Field

The invention relates to the technical field of intelligent networked automobiles and intelligent transportation, in particular to a micro-vehicle dynamic path planning vehicle-road coordination system based on V2X.

Background

V2X is a communication method with high reliability and low time delay between vehicles and other elements in the field of car networking, and mainly includes V2V (Vehicle-to-Vehicle), V2I (Vehicle-to-Infrastructure), V2P (Vehicle-to-peer) and V2N (Vehicle-to-Network). With the modern development of society, the demand of industrial production process, transportation and service industry for automatic transportation is higher and higher. In addition to these demands, the social environment also provides a good basic condition for the development of automatic driving, and the modern automobile industry is increasingly developing towards the direction of electromotion, intellectualization and networking. On the basis, the development of the AI and Internet industries also creates good conditions for more conveniently introducing intelligent and networking related functions.

In the face of emerging needs, the related art framework of autopilot has also been developed. The driving assistance functions such as blind area monitoring, lane change assistance, rear transverse warning, emergency braking, night vision system, fatigue monitoring, etc. are also developed toward full automation.

At present, the policy related to domestic automatic driving is continuously appeared and continuously developed and perfected. China Intelligent network automobile technology development route map indicates the specific development direction and approach of planning domestic automatic driving; the intelligent networking automobile road test management specifications (trial) issued by the Ministry of industry and communications confirm that test environment conditions of test subjects including drivers, vehicles and other detection targets are satisfied, and relevant contents of test application, management and accident handling. Meanwhile, unmanned vehicles are independently developed in various colleges and universities in academic circles, and some unmanned related college student competitions gradually appear in the visual field.

With the rapid development of the automatic driving technology and the gradual maturity of the automatic driving technology in recent years, the application and research based on automatic driving are widely concerned. Methods of planning and controlling vehicles, for example, are also constantly changing. Besides the control of a single vehicle, the research on the integral scheduling of the homogeneous vehicle queue also has social and economic values, and can improve the road throughput and the utilization rate. In order to achieve better effect and collect more data in the physical vehicle test, a simulation environment needs to be built for testing.

The existing sand table models have simpler scenes and functions and can only complete limited scene tasks, such as simple intersection driving, waiting for traffic lights and other scenes. The sand table model has various scenes, including parking lots, viaducts, crossroads, loops and the like. Besides, various equipment components and sensors such as laser radar, millimeter wave radar, cameras, computing platforms and the like are also built. Hardware and software support is provided for the function test of the unmanned automatic driving miniature vehicle, and the relevant test of the unmanned automatic driving of single vehicle and multiple vehicles can be supported.

The invention patent with publication number CN113453263A discloses a vehicle-road cooperation V2I simulation test system and method thereof, the system includes: the system comprises a scene virtual simulation system, a driving simulator, RSU simulation equipment, an OBU to be tested and a data processing unit, wherein the scene virtual simulation system comprises a scene database and a parameter setting unit; the driving simulator comprises a scene calculation unit, a path planning unit, a scene display unit, a signal transmission unit and a GPS simulator; the RSU simulation equipment comprises a host and a radio frequency unit; the OBU to be tested comprises a GPS component, a vehicle data memory, a WLAN card and an Ethernet interface; the data processing unit includes a data storage and a data analysis system.

Disclosure of Invention

Aiming at the defects in the prior art, the invention provides a micro vehicle dynamic path planning vehicle-road coordination system based on V2X.

According to the V2X-based dynamic path planning vehicle-road coordination system for the miniature vehicle, the scheme is as follows:

a micro vehicle dynamic path planning vehicle-road coordination system based on V2X, the system comprises: a sand table model and a miniature vehicle model;

the sand table model comprises: roads, road infrastructure, electronic components, and sensors; wherein the electronics are responsible for processing the asset data collected by the sensors.

Preferably, the road comprises a crossroad, a loop, an overpass, a U-shaped curve and an S-shaped curve;

the road equipment comprises: parking lot, parking space, gantry hanger and signal lamp;

the electronic component includes: a computing platform;

the sensor includes: laser radar, millimeter wave radar and camera;

each gantry hanging frame is provided with a signal lamp, a camera and a millimeter wave radar; the signal lamp is used for indicating the miniature vehicle to stop and advance; the camera is used for providing a visual captured image; the millimeter wave radar is used for acquiring the vehicle position information of the miniature vehicle; the computing platform is used for processing the data collected by the sensor; the laser radar is used for scanning scene information and generating cloud point images.

Preferably, the signal lamp, the camera and the millimeter wave radar form a traffic signal detection processing system;

the laser radar, the computing platform and the gantry hanging bracket form an intersection target detection and analysis system.

Preferably, a stand is installed on the gantry crane and used for installing a signal lamp, a camera and a computing platform.

Preferably, the virtual simulation test adopts a Gazebo simulator, and when the miniature vehicle simulation environment test is performed, the sand table model, each sensor, the computing platform and the miniature vehicle model are all led into the Gazebo, and the virtual simulation test is performed on the Gazebo.

Preferably, the laser radar adopts sixteen-line laser radar, and sixteen groups of laser components are arranged in the laser radar, and the laser radar simultaneously transmits and receives high-frequency laser beams, and performs real-time 3D imaging through 360-degree rotation.

Preferably, the millimeter wave radar adopts a novel radar sensor provided by continental aviation, can simultaneously measure the remote speed of 250 meters relative distance and the angle relationship between two objects, can also scan and finish the distance between a detection target and a monitoring point in real time, identifies the running speed of a detection target vehicle, and further analyzes the risk of collision.

Preferably, a spatial positioning function is added into the sand table model, a visual reference system AprilTag is set up to complete spatial positioning in sand table simulation, and three positioning modes of laser 2-D SLAM, WIFI positioning and tag code positioning are combined;

when the AprilTag is used for detecting and identifying the intelligent miniature vehicle, the labels are uniformly attached to a miniature vehicle fixing area, the label code id is calculated through an AprilTag detection program to confirm the label of the miniature vehicle, and meanwhile, the 3D position and the direction of the intelligent miniature vehicle in a sand table are identified.

Preferably, the label two-dimensional codes are placed at the upper corner, the lower corner, the long straight road and some curves of the viaduct, and the label two-dimensional codes contain information of the whole map and the current position.

Preferably, the 2-D SLAM positioning only uses a single line laser radar sensor and carries out two-dimensional positioning in a laser radar scanning plane;

the WIFI positioning method comprises the following steps that a plurality of AP base stations are placed around a field, a miniature vehicle receives WIFI signals, and the base stations position the vehicle by measuring and calculating the strength of the signals;

the tag codes are positioned in a road, a plurality of characteristic two-dimensional codes are placed in the road, and after a camera of the vehicle shoots the characteristic two-dimensional codes, the characteristic two-dimensional codes are matched with map information obtained by the 2-D SLAM, so that the vehicle is positioned.

Compared with the prior art, the invention has the following beneficial effects:

1. the sand table model built by the SolidWorks is provided with scenes such as crossroads, loops, viaducts, parking lots and the like, and is provided with components such as a camera, a signal lamp, a millimeter wave radar, a laser radar, a computing platform and the like, so that the functions of the unmanned miniature vehicle in different scenes are realized, and a virtual simulation test environment is provided;

2. according to the method, through data processing of a gantry computing platform and calculation of a sand table server, positioning service and path optimization service are provided for the unmanned miniature vehicle, the path planning of the networked vehicle is helped to avoid road conditions possibly jammed in advance, and meanwhile, on-site vehicle drainage is carried out on the vehicles which are not networked on the urban road according to real-time data, so that the commuting efficiency is improved;

3. according to the invention, by combining the respective positioning function advantages of the 2-D SLAM, the WIFI and the Tag code, an AprilTag detection system is built, and when the vehicle is positioned in environments with different heights, different signal strengths and detection equipment, accurate positioning can be carried out by combining map information. When the unmanned miniature vehicle confirms the destination, the corresponding destination periphery can reserve parking spaces in corresponding time periods, and the effect of intelligent parking guidance is achieved by using more accurate Tag codes;

4. the invention can analyze the collision risk for the networked miniature vehicles by the millimeter wave radar and the computing platform in the areas such as intersections and the like where the millimeter wave radar is installed, and send out signal prompts similar to warning lights and buzzing sounds when the distance is close.

Drawings

Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:

FIG. 1 is a view showing a miniature vehicle;

FIG. 2 is a top plan view of the sand table;

FIG. 3 is a 3D perspective view of a sand table;

FIG. 4 shows a sand table ground road size label;

FIG. 5 shows a sand table overhead dimension label;

FIG. 6 shows a sand table overhead dimension label;

FIG. 7 is a sand table crossroad size label;

FIG. 8 shows a cross road size marking of a sand table;

FIG. 9 shows a sand table ring size label;

fig. 10 shows the sand table ring size labels.

FIG. 11 shows the size of sand table parking lot and parking space;

FIG. 12 is a schematic view of a gantry crane bridge and components at the exit of a viaduct;

FIG. 13 is a schematic view of a gantry detection model of a "radar-computing platform" intersection;

FIG. 14 is a schematic view of a gantry crane bridge and components at the entrance of a viaduct;

FIG. 15 is a side view of a viaduct;

FIG. 16 is a schematic diagram of a lidar model in a sand table;

FIG. 17 is a schematic view of a scan angle;

FIG. 18 is a schematic diagram of a millimeter wave radar 3D model;

FIG. 19 is a geometric bicycle model;

FIG. 20 is a schematic diagram of a pure tracking algorithm;

FIG. 21 is a driving route of a viaduct of the miniature vehicle;

FIG. 22 is a side view of an intersection scene;

FIG. 23 is a partial driving route of a miniature vehicle at a crossroad;

FIG. 24 is a parking lot scene lane line;

FIG. 25 is a partial route of the miniature vehicle according to the lane line;

FIG. 26 is a partial route of the miniature vehicle stopping according to the lane line;

FIG. 27 is a two-dimensional code for a parking lot;

FIG. 28 is a diagram of the AprilTag layout in a gazebo;

fig. 29 is a real-time map of a 2D lidar.

Detailed Description

The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.

The embodiment of the invention provides a micro vehicle dynamic path planning vehicle-road cooperation system based on V2X, which comprises: a sand table model and a miniature car model. The sand table model comprises: roads, road infrastructure, electronic components, and sensors; wherein the electronics are responsible for processing the asset data collected by the sensors. The road comprises a crossroad, a return loop, a viaduct, a U-shaped bend and an S-shaped bend; the road equipment comprises: parking lot, parking space, gantry hanger and signal lamp; the electronic component includes: a computing platform; the sensor includes: laser radar, millimeter wave radar and camera; each gantry hanging bracket is provided with a signal lamp, a camera and a millimeter wave radar; the signal lamp is used for indicating the miniature vehicle to stop and advance; the camera is used for providing a visual captured image; the millimeter wave radar is used for acquiring the vehicle position information of the miniature vehicle; the computing platform is used for processing the data collected by the sensor; the laser radar is used for scanning scene information and generating cloud point images.

Firstly, a miniature vehicle model:

(1) introduction of a miniature vehicle:

the vehicle models adopted in the simulation experiment and the actual experiment in the design are miniature vehicle models, as shown in fig. 1, the miniature vehicle is a miniature vehicle, and the used sensors comprise a laser radar with a 360-degree scanning radius of 12 meters, two high-definition video cameras and an integrated accelerometer and gyroscope. In addition, the vehicle itself is equipped with an Intel atom (tm) processor's computing platform, WIFI module and high capacity battery are sufficient to accomplish the testing task.

(2) And (3) controlling the miniature vehicle:

the chassis of the miniature vehicle is provided with a steering engine for controlling the direction of the front wheels and a motor for controlling the rotation of the rear wheels, wherein the control modes are PWM control. It is worth pointing out that the control of the motor is open loop control since the wheels of the miniature vehicle are not provided with speed sensors.

II, Gazebo:

gazebo is a 3D dynamic simulator that can accurately and efficiently simulate robot populations in complex indoor and outdoor environments. Although similar to the game engine, Gazebo provides a higher fidelity physical simulation, a set of sensors, and a user and program interface. Typical uses of Gazebo include testing the robot's algorithms, designing the robot, performing regression tests with real scenes, and the like. The robot system is mainly characterized by comprising a plurality of physics engines, abundant robot models and environment libraries, covering various sensors, and more convenient programming functions, graphical interfaces and the like.

Thirdly, sand table model:

(1) and (3) overall model configuration:

the sand table integral model is mainly composed of roads, road facilities, electronic components and sensors as shown in fig. 2 and 3. The road characteristics comprise crossroads, loops, viaducts, U-shaped curves and S-shaped curves. The road facilities include parking lot, parking space, gantry hanger and signal lamp. The main body of the electronic element for detection is a computing platform, and the rest part of the electronic element is composed of sensors such as a laser radar, a millimeter wave radar and a camera for image processing. FIG. 4 shows a top view of the sand table and its component size labels; as shown in figures 5, 6, 7, 8, 9, 10 and 11, the dimensions of the various parts of the sand table are marked. A traffic signal detection processing system model composed of a signal lamp, a camera and a millimeter wave radar is shown in fig. 12; fig. 13 and 14 show a model of an intersection target detection and analysis system composed of a laser radar, a computing platform and a gantry crane.

Referring to fig. 15, the road surface of the whole sand table can be divided into five parts, i.e., a parking lot, an intersection, a loop, a viaduct and a general road, and each part will be described one by one. The parking lot part comprises 8 parking spaces, wherein each parking space is 600mm long and 480mm wide. The length of the whole parking lot is 4000mm, and the width is 2000 mm. Crossroad part has contained 1 longmen gallows, a square pallet, 4 signal lamps, 4 cameras, 1 computing platform and 1 laser radar. The length of the gantry hanger is 1240mm, and the height is 500 mm. A square stand is arranged on the gantry hanger and used for placing a signal lamp, a camera and a computing platform. The side length of square stand is 600mm, has all installed 1 signal lamp and 1 camera on its every limit. In the middle of the square stand 1 computing platform was installed. The laser radar is positioned at the upper right of the crossroad, and the measuring range of the laser radar can cover the whole crossroad. The return ring part contains 1 gantry hanger, 4 cameras, 1 computing platform and 1 laser radar. The length of the gantry hanger is 2340mm, and the height of the gantry hanger is 500 mm. The method comprises the following steps that 1 computing platform, 4 cameras and 1 laser radar are mounted on a gantry hanger, the computing platform is arranged on the uppermost portion of the gantry hanger, the cameras are mounted on the periphery of a support which is arranged at an included angle of 90 degrees to detect road information in each direction, and 4 cameras are arranged in each direction; 1 lidar has been settled to the camera below, can carry out 360 rotations, and its distance ground height is 145mm, and monitoring range is whole ring and 3 entrances. The part of overpass has contained 2 longmen gallows, 2 signal lamps, 2 cameras, 2 millimeter wave radars. The height of the road is 350mm, the width of the road is 250mm, the angle of the up-down ramp is 20 degrees, the length of the ramp is 1400mm, and the radius of the curve is 1300 mm. 1 gantry hanger is arranged at the entrance and exit of the elevated frame, and the height of the gantry hanger is 360 mm. 1 signal lamp, 1 millimeter wave radar and 1 camera have been arranged on gantry crane. The general road part contains 4 curves, is 2 ordinary curves respectively, 1U type curve and 1S type curve. The radius of 2 ordinary bends is 325mm, the radius of U type bend is 870mm, the radius of S type bend is 445 mm.

(2) Laser radar:

the sand table model is provided with 2 lidar units, and a 3D perspective view of the lidar unit is shown in FIG. 16. The lidar is the most commonly used sensor for detecting an object, has the advantages of high precision, high resolution, long distance, wide scanning area and the like, can quickly acquire three-dimensional space information of the detected object in real time and further acquire a physical model three-dimensional point cloud, and as shown in fig. 17, the lidar is used for scanning an angle and a region which can be scanned and identified. Compared with a visual camera, the laser radar can rapidly acquire large-scale three-dimensional point cloud data, provides accurate geometric information of the environment, has strong robustness to illumination intensity change and weather conditions, and plays an important role in a target detection task by virtue of stable and reliable performance. The laser radar is mainly used for target detection and obstacle detection under complex conditions, so that safety guarantee is provided for units such as pedestrian personal safety, automobile drivers and automatic driving automobiles, and theoretical support is provided for relieving urban road traffic operation pressure and reducing time wasted by congestion of travelers in the process.

The laser radar in the introduction is a 16-line laser radar. The radar has the biggest characteristics of small size, low energy consumption, and domestic and world-wide leading advantages, and is mainly used in the fields of robot environment perception, peripheral environment perception of automatic driving automobiles, intelligent security detection of mobile equipment, unmanned aerial vehicle surveying and mapping and the like at the present stage.

16 groups of laser components are arranged in the three-dimensional space point cloud three-dimensional object reflectivity three-dimensional space point cloud three-based laser radar system has the indispensable role in positioning, obstacle avoidance navigation scheduling and navigation scheduling, the sensor parameter functions are as shown in the following table 1, and the laser radar output port introduction is shown in the following table 2.

Table 116 line lidar sensor parameters

TABLE 2 introduction of multiline lidar output modules

(3) Millimeter wave radar:

2 millimeter wave radars are configured in the sand table model, and a millimeter wave radar sensor has robustness on environmental influences caused by severe light and weather conditions and can identify targets in complex scenes; the millimeter wave radar has smaller volume, low energy consumption and higher speed when processing data. With the development and improvement of 77GHz millimeter wave radar technology, the industry gradually upgrades from 24GHz radar to more advantageous 77GHz radar, the frequency is higher and the wavelength is shorter when the millimeter wave radar is used for target detection, and the detection precision and the measurement range reach the traffic detection requirement.

The millimeter wave radar and the vision camera and other sensors are fused, and the complex traffic environment is identified and processed by combining the laser radar three-dimensional point cloud data, so that the detection capability and detection accuracy of dangerous targets can be greatly improved, and the urban traffic safety and operation efficiency are improved.

A new radar sensor provided by continental aviation is adopted in the sand table, as shown in fig. 18. The sensor can independently measure the distance and speed to an object (doppler principle), and its real-time scanning speed is 17/s because FMCW (frequency modulated continuous wave) has a very fast slope and no reflector in one measuring period. One feature of the apparatus is that it is possible to measure both the distance velocity at a relative distance of 250 meters and the angular relationship between two objects. In addition, the millimeter wave radar can also scan in real time and finish the distance between the detection target and the monitoring point, identify the running speed of the detection target vehicle and further analyze the possible collision risk.

In the rapid development process of the supply of the automobile industry, the millimeter wave radar has a more simplified measurement technical principle and a smaller volume, the sensor also has a self-detection function, can automatically sense and monitor the environment of the sensor, identify the fault of the sensor and feed back fault information, and has a more reliable technology and a wider application in scenes such as remote course control, remote area detection, vehicle collision prevention, target detection in chaos and the like. The millimeter wave radar measurement performance table is shown in table 3 below:

TABLE 3 millimeter wave radar measurement performance table

(4) The design principle of the bend is as follows:

the design principle of the curve in the sand table model is based on a lane line detection and tracking module of a miniature vehicle, and the corresponding realization functions of the lane line detection and tracking module are two parts of U-shaped turning and S-shaped running. The algorithm for judging the vehicle position based on the unilateral lane line is as follows:

firstly, for visual processing, a picture read by a camera is converted and adjusted into a gray scale image, a rectangle of 10x10 is used for expansion (the expansion of white is equivalent to the corrosion of a black part including a lane line) to remove black noise points, then reverse threshold transformation is carried out, perspective transformation is carried out to convert the image into a bird's-eye view image, and the lane line is extracted from the bird's-eye view image.

Counting the number distribution of white dots in the x direction of half size, regarding the x coordinate of the maximum value as the starting point of the lane line, then counting the pixel point coordinate distribution of the lane line by a sliding window method, fitting a quadratic curve to process the coordinate distribution, and judging whether the identified lane line is a left lane or a right lane by using the intersection point position of the quadratic curve and a parallel line of a vehicle head under a real coordinate system (a coordinate system which is established by taking a camera as the origin of coordinates and is parallel to the ground), thereby obtaining all geometrical information of the required lane line.

After a quadratic curve equation of the lane and information of the left lane or the right lane are obtained, more important trajectory planning and motion control are carried out, a pure tracking method is used for analysis, and a method based on geometric tracking is simple and visual in derivation.

FIG. 19 shows a geometric bicycle model. A curve is designed by applying Ackerman steering, generally, a left-right steering angle relation is found out by using a conversion matrix, then a steering error between a wheel base and a wheel base is obtained, and finally an objective function of a minimum error is obtained by Ackerman steering geometry, so that the steering error is reduced to the minimum. The bicycle model in the figure can simply and intuitively simulate the model in Ackerman steering geometry.

The intelligent miniature vehicle of practical application in the sand table is a four-wheel vehicle, the four-wheel vehicle is simplified into a two-wheel vehicle model by utilizing a bicycle model during design, and the vehicle is assumed to only run on a plane, so that the design advantage is that the geometric relationship between the steering angle of the front wheel and the running curvature of the rear axle meets the following requirements:

where δ represents the steering angle of the front wheel, L represents the Wheelbase (Wheelbase) between the front and rear wheels of the bicycle, and R is the radius of the circle the rear axle follows at a given steering angle. The bicycle geometric relational expression can be used for estimating the running of the vehicle during design, and particularly, the estimation result is more accurate in a scene that the vehicle runs at a lower speed during turning. By using a pure tracking algorithm, the rear axle of the geometric bicycle model is taken as a tangent point, the longitudinal body of the vehicle is taken as a tangent line, and the vehicle can run along an arc passing through a target road point (goal point) by controlling the corner of the front wheel, as shown in fig. 20.

Through simple geometric derivation, the final expression of the pure tracking algorithm control quantity in the turning design of the geometric bicycle model can be obtained:

wherein l _d Representing the wheelbases of the front axle and the rear axle of the vehicle; alpha represents the vehicle attitude heading angle; under a coordinate system established by taking the center of the vehicle as an origin and taking the advancing direction as the forward direction of the y axis, the pure tracking model is in the form of:

the positive and negative values of x and δ are ignored, where I is the distance between the axle centers of the front and rear wheels, x is the abscissa of the target point, and L is the distance from the target point to the origin (the center point of the vehicle).

(5) And (4) scene functions:

in order to enable the unmanned miniature vehicle to better complete the test under different road scenes, viaducts, loops, crossroads and parking lot scenes are added into the sand table model.

1) And (3) viaduct:

there are 1 entry and 1 export in the scene of overpass, and the equipment that contains has 2 longmen gallows, 2 signal lamps, 2 cameras, 2 millimeter wave radars. The width of the road of the viaduct is 250mm, and 1 miniature vehicle can be accommodated for running; the radius of the U-shaped bend on the viaduct is 1300 mm; the up-down slope angles of the inlet and the outlet of the elevated are both 20 degrees, and the road lengths of the up-down slope are both 1400 mm; the height from the viaduct to the road surface is 350 mm. 2 gantry hangers are arranged at the inlet and the outlet of the viaduct, and the heights of the gantry hangers are 360 mm; each gantry is provided with 1 signal lamp, 1 camera and 1 millimeter wave radar. The signal lamp is used for indicating the miniature vehicle to stop and advance; the camera is used for providing visual captured images; the millimeter wave radar is used for collecting the vehicle position information of the miniature vehicle; the computing platform is used for processing data collected by the sensors.

The scene operation mode is as follows: the millimeter wave radar is used for collecting the vehicle position information of the miniature vehicle and the picture information captured by the camera, and the information is fused and processed through the computing platform to perform the operation of the whole scene. The default of the signal lamp at the entrance of the overhead uphill is red, and whether a miniature vehicle waits to enter the overhead is detected through a millimeter wave radar. The miniature vehicle can be parked at an overhead entrance, the millimeter wave radar and the camera detect that the miniature vehicle waits for passing through, and the signal lamp is controlled to be changed into a green lamp. And after detecting that the signal lamp at the entrance is green, the miniature vehicle starts to run. And the miniature vehicle is driven to the high-rise downhill exit all the time, detects that the signal lamp at the exit is a red lamp, and then stops. At the moment, the millimeter wave radar and the camera at the exit can detect that the miniature vehicle waits for exiting the overhead, and control the signal lamp to turn into a green lamp. And after detecting that the signal lamp at the exit is a green lamp, the miniature car starts to run out of the overhead. The driving route of the entire miniature vehicle is shown in fig. 21.

2) Looping:

there are 3 entrances and exits in the scene of returning the ring, and the equipment that contains has 1 longmen gallows, 4 cameras, 1 computing platform, 1 lidar. The width of the road surface is 500 mm; the 3 entrances and exits are respectively positioned at the upper left, right and lower left of the return ring, and the miniature vehicle can complete the steering of 90 degrees and 120 degrees through the return ring. Pedestrian crossing lines are arranged at the entrances and exits right and left below. The gantry hanger is located at the center of the circular ring, and is 2340mm in length and 500mm in height. The method comprises the following steps that 1 computing platform, 4 cameras and 1 laser radar are mounted on a gantry hanger, the computing platform is arranged on the uppermost portion of the gantry hanger, the cameras are mounted on the periphery of a support which is arranged at an included angle of 90 degrees to detect road information in each direction, and 4 cameras are arranged in each direction; 1 lidar has been settled to the camera below, can carry out 360 rotations, and its distance ground height is 145mm, and detection range is whole ring and 3 entrances. The laser radar is used for scanning scene information and generating cloud point images; the camera is used for capturing visual images; the computing platform is used for processing data collected by the sensors.

The scene operation mode is as follows: scanning current environmental information by using a 360-degree laser radar to generate point cloud information and picture information captured by a camera, and fusing and processing the information through a computing platform to operate the whole scene. The support of traffic flow planning is provided by using a V2X vehicle-road cooperation system, so that a single miniature vehicle or a whole vehicle team smoothly passes through a loop scene. Pedestrian crossing lines are arranged right and left below the miniature vehicle, and if the laser radar detects that pedestrians pass through the pedestrian crossing lines, the miniature vehicle is controlled through the computing platform, so that the movement of the miniature vehicle is stopped. And after the pedestrians pass through the pedestrian crossing line, controlling the miniature vehicle to drive in or out of the loop.

3) At the crossroad:

the sand table model comprises 1 crossroad and 2T-shaped crossroads. The part at crossroad has contained 1 gantry hanger, 1 square stand, 4 signal lamps, 4 cameras, 1 computing platform and 1 laser radar. The length of the gantry hanger is 1240mm, and the height is 500 mm. A square stand is arranged on the gantry hanger and used for placing a signal lamp, a camera and a computing platform. The side length of square stand is 600mm, has all installed 1 signal lamp and 1 camera on its every limit. In the middle of the square stand 1 computing platform was installed. The laser radar is positioned at the upper right of the crossroad, and the measuring range of the laser radar can cover the whole crossroad. The laser radar is used for scanning scene information and generating cloud point images; the camera is used for capturing visual images; the computing platform is used for processing data collected by the sensors. Pedestrian crossing lines are arranged on 4 road surfaces of the crossroads, as shown in figure 22.

The scene operation mode is as follows: scanning current environmental information by using a 360-degree laser radar to generate point cloud information and picture information captured by a camera, and fusing and processing the information through a computing platform to operate the whole scene. The support of traffic flow planning is provided by using a V2X vehicle-road cooperation system, so that a single miniature vehicle or a whole vehicle team smoothly passes through a crossroad scene. The signal lamp can light red light or green light according to current road conditions, if the horizontal road has the vehicle to travel, then 2 signal lamps of horizontal road can light green light, and 2 signal lamps of vertical road can light red light. On the contrary, if the vehicle runs on the longitudinal road, the 2 signal lamps on the longitudinal road can be turned on by green light, and the 2 signal lamps on the transverse road can be turned on by red light. When the miniature vehicle runs to a crossroad scene, the miniature vehicle can be selected to go straight, turn left or turn right. If the miniature vehicle selects to run straight, whether the miniature vehicle can be used for running continuously is judged according to the signal lamp. If the signal lamp is green, continuing driving; if the signal lamp is red, the vehicle stops and waits until the signal lamp is changed into green, and then the vehicle restarts to run. If the miniature vehicle selects left turning, when the signal lamp is green, the laser radar monitors whether the opposite side of the miniature vehicle passes by the coming vehicle. If no vehicle comes from the opposite side, starting to turn left; if the opposite side of the vehicle is ready for the coming vehicle or is passing through the opposite side, the left turn is carried out after the coming vehicle on the opposite side passes through the opposite side. Meanwhile, the laser radar can monitor whether pedestrians pass through the left pedestrian crossing, if the pedestrians pass through the left pedestrian crossing, the miniature vehicle stops before the pedestrian crossing line, and the miniature vehicle drives again after waiting for the pedestrians to pass through the miniature vehicle. If the miniature vehicle selects right turning, when the signal lamp is turned green, the miniature vehicle directly turns right, meanwhile, the laser radar monitors whether pedestrians pass through the right pedestrian crossing, if so, the miniature vehicle stops before the pedestrian crossing line, and continues to turn right after the pedestrians pass through; when the signal lamp is the red light, then judge whether its left side has the car that comes to pass through laser radar, if there is the car that comes to the left side, then wait for the left side vehicle and pass through after accomplishing, turn right again. Meanwhile, the laser radar can monitor whether pedestrians pass through the right-side pedestrian crossing, if the pedestrians pass through the right-side pedestrian crossing, the miniature vehicle stops in front of the pedestrian crossing line, and the miniature vehicle continues to turn right after waiting for the pedestrians to pass through the right-side pedestrian crossing.

The partial driving route of the miniature vehicle in the intersection scene is shown as 23.

4) Parking lot scene:

the parking lot part comprises 8 parking spaces, wherein each parking space is 600mm long and 480mm wide. The length of the whole parking lot is 4000mm, and the width is 2000 mm. In front of each parking space, a T-shaped white lane line is marked for identifying the two cameras of the miniature car, so as to realize the scene function of automatic parking, and the lane line is shown in fig. 24.

When the miniature vehicle enters the parking lot, the miniature vehicle can advance to the front of the parking space according to the guidance of the lane line, as shown in fig. 25.

After the miniature vehicle runs to the front of the parking space, the miniature vehicle can be used for positioning the center line of the parking space by virtue of the functional characteristics that 2 cameras behind the miniature vehicle can capture road surface information, as shown in fig. 26.

The miniature vehicle can complete the realization of the scene function of automatic parking by identifying the lane line of the scene of the parking lot. In addition, the two-dimensional code that is marked with spatial position information still on the road surface in parking area, the camera of miniature car can gather current positional information through scanning this two-dimensional code, fuses with the lane line information of discerning mutually and handles, makes more accurate parking action. The two-dimensional code identification of the parking lot is shown in fig. 27:

(6) space positioning:

in order to enable the unmanned miniature vehicle to accurately complete the driving of the road, a space positioning function is added into the sand table model. The space positioning of the vehicle generally comprises methods such as laser 2-D SLAM, WIFI positioning and tag code positioning. The 2-D SLAM positioning method only uses a single line laser radar sensor and carries out two-dimensional positioning in a plane scanned by the laser radar, but in actual measurement, the laser head cannot be always kept on the same plane, the ground is bumpy, the traveling speed and even the tire is insufficient, the positioning can be misaligned, and therefore the 2-D SLAM is limited to be used only. In WIFI positioning, a plurality of AP base stations are placed around a field, a miniature vehicle receives WIFI signals, the base stations position the vehicle by measuring and calculating the strength of the signals, and the positioning accuracy is also limited by the strength of the WIFI signals. the tag codes are positioned in a road, a plurality of characteristic two-dimensional codes are placed in the road, and the vehicle is positioned by matching the two-dimensional codes with map information obtained by 2-D SLAM after a camera of the vehicle shoots the two-dimensional codes.

In conclusion, the three positioning modes are realized, and no matter which positioning mode is adopted, the advantages of the three positioning modes are respectively superior and deficient, and the aprilTag test mode is set up to complete the spatial positioning in the sand table simulation. Aprilatag is a visual reference system and the marker labels resemble a bar code or two-dimensional code. When the AprilTag is used for detecting and identifying the intelligent miniature vehicle, the labels are uniformly attached to the fixed area of the miniature vehicle, the label code id can be calculated through an AprilTag detection program to confirm the label of the miniature vehicle, and meanwhile, the 3D position and direction of the intelligent miniature vehicle in a sand table can be identified. The method has important significance for number identification, space positioning and direction identification of the intelligent miniature vehicle in the sand table and subsequent detailed research of all directions.

As shown in fig. 28, some tag two-dimensional codes are placed on the upper and lower corners of the overpass as well as the long straight roads and some curved roads so that the car can be positioned more accurately.

These two-dimensional codes contain information of the entire map and the current position. When the two-dimensional code is scanned by the miniature vehicle, the current position information can be positioned according to the information obtained by reading the two-dimensional code, and planning for a subsequent driving route is facilitated.

The objects are mapped using slam _ mapping in the mapping packet in the ROS, and the movement of the vehicle is controlled using the keypad during mapping so that the lidar can sweep across the entire sand table to complete a 2-D mapping, where the schematic diagrams in gazebo and RVIZ are shown in fig. 29.

The embodiment of the invention provides a V2X-based miniature vehicle dynamic path planning vehicle-road coordination system, which provides a sand table model capable of testing for miniature vehicles of an automatic driving cloud platform, and comprises the following key point characteristics required by an automatic driving test: the method comprises the following steps of matching intersection traffic lights with crossroads, loop loops, viaducts and parking spaces, and completing the test of the unmanned miniature vehicle in a simulation environment by utilizing image detection technology, a camera, a millimeter wave radar, a laser radar and other sensors.

In the description of the present application, it is to be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", and the like indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience in describing the present application and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present application.

The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.

Claims (10)

1. A dynamic path planning vehicle-road coordination system of miniature vehicles based on V2X is characterized by comprising: a sand table model and a miniature vehicle model;

the sand table model comprises: roads, road infrastructure, electronic components, and sensors; wherein the electronics are responsible for processing the asset data collected by the sensors.

2. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 1, wherein said road comprises crossroads, loops, viaducts, U-shaped curves and S-shaped curves;

the road equipment comprises: parking lot, parking space, gantry hanger and signal lamp;

the electronic component includes: a computing platform;

the sensor includes: laser radar, millimeter wave radar and camera;

each gantry hanging frame is provided with a signal lamp, a camera and a millimeter wave radar; the signal lamp is used for indicating the miniature vehicle to stop and advance; the camera is used for providing a visual captured image; the millimeter wave radar is used for acquiring the vehicle position information of the miniature vehicle; the computing platform is used for processing the data collected by the sensor; the laser radar is used for scanning scene information and generating cloud point images.

3. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 2, wherein the signal lamp, the camera and the millimeter wave radar constitute a traffic signal detection processing system;

the laser radar, the computing platform and the gantry hanging bracket form an intersection target detection and analysis system.

4. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 2, wherein said gantry is provided with a stand for positioning signal lights, cameras and computing platforms.

5. The V2X-based dynamic path planning vehicle-road coordination system for miniature vehicles according to claim 2, wherein a Gazebo simulator is used for virtual simulation test, and when the miniature vehicle simulation environment test is performed, the sand table model, each sensor, the computing platform and the miniature vehicle model are all imported into the Gazebo, and the virtual simulation test is performed on the Gazebo.

6. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 2, wherein the lidar employs sixteen-line lidar which has sixteen sets of laser components, and transmits and receives high-frequency laser beams simultaneously, and performs real-time 3D imaging through 360 rotations.

7. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 2, wherein the millimeter wave radar adopts a novel radar sensor provided by continental aviation, can simultaneously measure the long-distance speed with a relative distance of 250m and the angle relationship between two objects, can also scan and complete the distance between a detection target and a monitoring point in real time, identify the running speed of the detection target vehicle, and further analyze the risk of collision.

8. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 1, wherein a spatial localization function is added to the sand table model, a visual reference system aprilTag is built to complete spatial localization in sand table simulation, and three localization modes of laser 2-D SLAM, WIFI localization and tag code localization are combined;

when the AprilTag is used for detecting and identifying the intelligent miniature vehicle, the labels are uniformly attached to a miniature vehicle fixing area, the label code id is calculated through an AprilTag detection program to confirm the label of the miniature vehicle, and meanwhile, the 3D position and the direction of the intelligent miniature vehicle in a sand table are identified.

9. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system, according to claim 8, wherein said tag two-dimensional code is placed on the upper and lower corners of the viaduct, as well as the long straight road and some curves, said tag two-dimensional code contains the information of the whole map and the current position.

10. The V2X-based miniature vehicle dynamic path planning vehicle-road coordination system according to claim 8, wherein said 2-D SLAM positioning uses only a single line lidar sensor and performs two-dimensional positioning in the plane scanned by the lidar;

the WIFI positioning method comprises the following steps that a plurality of AP base stations are placed around a field, a miniature vehicle receives WIFI signals, and the base stations position the vehicle by measuring and calculating the strength of the signals;

the tag codes are positioned in a road, a plurality of characteristic two-dimensional codes are placed in the road, and after a camera of the vehicle shoots the characteristic two-dimensional codes, the characteristic two-dimensional codes are matched with map information obtained by the 2-D SLAM, so that the vehicle is positioned.

Images