CN110807922A

CN110807922A - Unmanned vehicle intersection traffic capacity test system and test method

Info

Publication number: CN110807922A
Application number: CN201911040213.6A
Authority: CN
Inventors: 赵祥模; 王润民; 周文帅; 朱宇; 徐志刚; 刘占文; 吴霞; 王文威
Original assignee: Changan University
Current assignee: Changan University
Priority date: 2019-10-29
Filing date: 2019-10-29
Publication date: 2020-02-18

Abstract

The invention discloses a system and a method for testing the traffic capacity of an unmanned vehicle intersection, which utilize a cross test road provided with traffic lights as an unmanned vehicle test site, arrange roadside equipment and a control center computer beside the cross test road, utilize the roadside equipment and the control center computer to realize information interaction with a vehicle to be tested, obtain the running information of the vehicle to be tested, utilize a reference vehicle as an influencing vehicle in the test process of the unmanned vehicle to be tested, utilize the reference vehicle and the cross test road to test the unmanned vehicle to be tested, and can effectively simulate the road condition met by the unmanned vehicle in the actual road running process Compared with virtual simulation test, the test environment is closer to the real traffic environment, and the test result is more real and reliable.

Description

Unmanned vehicle intersection traffic capacity test system and test method

Technical Field

The invention relates to the technical field of unmanned vehicle performance testing, in particular to a system and a method for testing the traffic capacity of an unmanned vehicle intersection.

Background

With the continuous and deep research of unmanned vehicles at home and abroad, key technologies in the fields of unmanned vehicle perception, decision making, execution and the like are gradually broken through, the intelligent level of the unmanned vehicles is continuously improved, and the commercial application is increasingly closed. In each stage of unmanned vehicle research and development and application detection, the safety and the running performance of the unmanned vehicle need to be tested and evaluated, so that the technical iteration of the unmanned vehicle is promoted.

Unlike traditional automobile test, the unmanned vehicle test object is converted from a human-vehicle binary structure into a human-vehicle-environment-task strong coupling system; the test method is changed from driver test and vehicle mechanical performance test into unmanned vehicle operation scene-based test. In actual traffic, the running scenes of the unmanned vehicles are complex and changeable, and typical scenes in traffic running, such as crossing, obstacle avoidance, traffic flow influx and the like, can be extracted through analysis of real traffic scenes. Wherein, passing through the intersection is an important environment for unmanned testing. The unmanned vehicle passing through the intersection is an important task for the actual road operation of the unmanned vehicle and is also an important object for the unmanned vehicle operation capability test. The unmanned vehicle passing intersection test is a comprehensive test of the perception, decision and execution capabilities of the unmanned vehicle. The test aiming at the capability of the unmanned vehicle passing through the intersection is carried out, and the method has important significance for ensuring the running safety of the unmanned vehicle and detecting the running performance of the unmanned vehicle. However, a test method and a test field close to a real road environment are lacked at present, vehicle running task optimization is performed only by means of running scene simulation, and actual traffic running state information of an unmanned vehicle cannot be accurately acquired. Therefore, a method and a test field capable of acquiring operation data when an unmanned vehicle passes through an intersection in a real road environment are needed.

Disclosure of Invention

The invention aims to provide a system and a method for testing the traffic capacity of an unmanned vehicle intersection so as to overcome the defects of the prior art.

In order to achieve the purpose, the invention adopts the following technical scheme:

a system for testing the traffic capacity of an unmanned vehicle intersection comprises road side equipment, a control center computer, a reference vehicle and a cross test road provided with traffic signal lamps, wherein the reference vehicle can run on the cross test road according to a set running route, and the unmanned vehicle to be tested starts to enter the cross test road according to a control signal of the control center computer; the reference vehicle and the unmanned vehicle to be detected are both provided with a signal transmission module for wireless transmission with roadside equipment and a data acquisition module for acquiring the running information of the reference vehicle and the running information of the unmanned vehicle to be detected; the roadside equipment is arranged beside the cross test road and used for acquiring reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information and sending the acquired reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information to the control center computer, the control center computer receives and stores the reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information, and the control center computer obtains the traffic capacity of the unmanned vehicle to be tested at the intersection according to the received reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information.

Further, the signal transmission modules on the reference vehicle and the unmanned vehicle to be tested adopt V2X wireless communication.

Further, the reference vehicle travel information includes a reference vehicle position, a reference vehicle motion trajectory, and reference vehicle travel speed information.

Further, the running information of the unmanned vehicle to be detected comprises the position of the unmanned vehicle to be detected, the motion track of the unmanned vehicle to be detected and the running speed information of the unmanned vehicle to be detected.

Further, still including setting up the reference pedestrian on cross test road pedestrian passageway, the reference pedestrian can be passed on cross test road pedestrian passageway.

Furthermore, all reference pedestrians adopt a test mobile platform, and a human model for testing is carried on the test mobile platform.

Further, the road side equipment comprises an ENodeB base station, a road side test unit, an LTE-V core network and a local server group; the LTE-V core network is connected with the local server group through the aggregation switch, the aggregation switch is connected with the ENodeB base station, the ENodeB base station provides a wireless signal covering a cross test road, a data acquisition module on the unmanned vehicle to be tested realizes information interaction with the ENodeB base station in a wireless mode, the roadside test unit is installed beside the cross test road, and the roadside test unit realizes information interaction with the ENodeB base station and the control center computer in a wireless mode.

A method for testing the traffic capacity of an unmanned vehicle intersection comprises the following steps:

step 1), enabling a reference vehicle to run on a cross test road according to a set route;

step 2), when a reference vehicle runs to the crossroad test road intersection, starting the unmanned vehicle to be tested, and enabling the unmanned vehicle to be tested to run according to a set route;

step 3), acquiring reference vehicle running information, unmanned vehicle running information to be detected and traffic signal lamp information in real time through roadside equipment, and transmitting the acquired reference vehicle running information, unmanned vehicle running information to be detected and traffic signal lamp information to a control center computer;

step 4), if the to-be-tested unmanned vehicle enters the intersection in a normal green light passing state and simultaneously enters the intersection of the crossroad test road, the reference vehicle in the test road is braked to avoid collision under full force, the risk of collision between the to-be-tested unmanned vehicle and the internal reference vehicle of the test road is higher, and the capability of the to-be-tested unmanned vehicle entering the crossroad is judged to be unqualified; if the unmanned vehicle to be tested enters the crossroad in a normal green light passing state and simultaneously enters the crossroad of the crossroad test road, the reference vehicle in the test road has no full-force braking collision avoidance, and the capability of the unmanned vehicle to be tested entering the crossroad is judged to be qualified; and if the to-be-detected unmanned vehicle enters the crossroad in a red light or yellow light non-passing state, judging that the capability of the to-be-detected unmanned vehicle entering the crossroad is unqualified.

Further, in step 1), the reference vehicle is caused to run on the intersection test road according to the set route, and the reference vehicle runs through the intersection of the intersection test road according to the normal vehicle, specifically including left-turn running, straight running and right-turn running.

Further, when a reference vehicle runs to the intersection of the crossroad test road, the reference pedestrian is started to pass through the crossroad test road pedestrian passage at the same time.

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

the invention relates to an unmanned vehicle intersection traffic capacity testing system, which utilizes a cross testing road provided with traffic lights as an unmanned vehicle testing field, arranges roadside equipment and a control center computer beside the cross testing road, utilizes the roadside equipment and the control center computer to realize information interaction with a vehicle to be tested, acquires running information of the vehicle to be tested, utilizes a reference vehicle as an influencing vehicle in the testing process of the unmanned vehicle to be tested, utilizes the reference vehicle and the cross testing road to test the unmanned vehicle to be tested, can effectively simulate road conditions encountered by the unmanned vehicle in the actual road running process, quickly detects and obtains intersection traffic data and other vehicle traffic data of the unmanned vehicle to be tested, can accurately analyze and obtain the intersection traffic capacity of the unmanned vehicle to be tested according to the intersection traffic data and other vehicle traffic data of the unmanned vehicle to be tested, the invention is closer to the real driving environment, so that the test result can more accurately express the capability of the vehicle to be tested to pass through the intersection.

A crossroad unmanned vehicle traffic capacity test method, through making the reference vehicle go on the cross test road according to presuming the route, provide the real traffic environment, when there is reference vehicle to go to the cross test road crossing, start the unmanned vehicle to be measured, make the unmanned vehicle to be measured go according to presuming the route; then, the control center computer is used for acquiring reference vehicle running information and unmanned vehicle running information to be tested in real time through roadside equipment, so that the actual running condition of the unmanned vehicle to be tested in the test process is obtained, meanwhile, the reference vehicle running information can be acquired and used as a reference factor for influencing other vehicles in the running process of the unmanned vehicle to be tested, if the unmanned vehicle to be tested enters the intersection in a normal green light passing state and simultaneously enters the intersection of the crossroad test, the reference vehicle in the test road is braked and avoided in a full force mode, the risk of collision between the unmanned vehicle to be tested and the internal reference vehicle of the test road is higher, and the capability of the unmanned vehicle to be tested entering the intersection is judged to be unqualified; if the unmanned vehicle to be tested enters the crossroad in a normal green light passing state and simultaneously enters the crossroad of the crossroad test road, the reference vehicle in the test road has no full-force braking collision avoidance, and the capability of the unmanned vehicle to be tested entering the crossroad is judged to be qualified; if the unmanned vehicle to be tested enters the crossroad in a red light or yellow light non-passing state, judging that the unmanned vehicle to be tested cannot enter the crossroad, and therefore effectively judging the passing capacity of the unmanned vehicle to be tested at the crossroad.

Drawings

FIG. 1 shows a schematic view of a test yard with an unmanned vehicle traveling straight through an intersection with traffic lights;

FIG. 2 shows a schematic view of a test field in which an unmanned vehicle travels straight through an intersection with traffic lights to avoid a vehicle with a conflicting path;

FIG. 3 shows a schematic view of a test field in which an unmanned vehicle passes straight through a crossroad with traffic lights to avoid red-light running reference pedestrians;

FIG. 4 shows a schematic view of a test yard in which an unmanned vehicle travels straight through an intersection having traffic lights to avoid a vehicle running a red light;

FIG. 5 shows a schematic view of a test field with an unmanned vehicle turning left/right through an intersection with traffic lights;

FIG. 6 shows a schematic view of a test field where an unmanned vehicle turns right through an intersection with traffic lights, reference pedestrians and vehicles for avoiding path conflicts;

in the figure, 1, an unmanned vehicle to be tested; 2. a first reference vehicle; 3. a second reference vehicle; 4. a traffic light; 5. a roadside apparatus; 6. and a control center computer.

Detailed Description

The invention is described in further detail below with reference to the accompanying drawings:

the invention provides a traffic capacity test system for an unmanned vehicle intersection, which is used for monitoring the position of an unmanned vehicle to be tested, the track of the unmanned vehicle to be tested and the speed of the unmanned vehicle to be tested when the unmanned vehicle to be tested passes through the intersection so as to comprehensively test the traffic capacity of the unmanned vehicle at the intersection; and quantitatively evaluating the ability of the unmanned vehicle to be tested to pass through the intersection through the safety test of the unmanned vehicle to be tested to pass through the intersection. The test result can more accurately represent the capability of the unmanned vehicle to be tested to pass through the intersection. Compared with virtual simulation test, the test environment is closer to the real traffic environment, and the test result is more real and reliable.

The unmanned vehicle intersection traffic capacity test system comprises road side equipment 5, a control center computer 6, a reference vehicle and a cross test road provided with traffic signal lamps 4, wherein the reference vehicle runs on the cross test road according to a set running route, and the unmanned vehicle to be tested starts to enter the cross test road according to a control signal of the control center computer 6; the reference vehicle and the unmanned vehicle to be detected are both provided with a signal transmission module for wireless transmission with the roadside device 5 and a data acquisition module for acquiring the running information of the reference vehicle and the running information of the unmanned vehicle to be detected; the roadside device 5 is arranged beside the cross test road and used for acquiring reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information and sending the acquired reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information to the control center computer 6, and the control center computer 6 stores the reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information. And the signal transmission modules on the reference vehicle and the unmanned vehicle to be tested adopt V2X wireless communication. The reference vehicle travel information includes a reference vehicle position, a reference vehicle motion trajectory, and reference vehicle travel speed information. The running information of the unmanned vehicle to be detected comprises the position of the unmanned vehicle to be detected, the motion track of the unmanned vehicle to be detected and the running speed information of the unmanned vehicle to be detected. The road side equipment comprises an ENodeB base station, a road side test unit, an LTE-V core network and a local server group; the LTE-V core network is connected with the local server group through the aggregation switch, the aggregation switch is connected with the ENodeB base station, the ENodeB base station provides a wireless signal covering a cross test road, a data acquisition module on the unmanned vehicle to be tested realizes information interaction with the ENodeB base station in a wireless mode, the roadside test unit is installed beside the cross test road, and the roadside test unit realizes information interaction with the ENodeB base station and the control center computer in a wireless mode.

In the step 1), the reference vehicle is driven on the crossroad test road according to the set route, and the reference vehicle passes through the crossroad of the crossroad test road according to the normal vehicle driving, specifically including left-turn driving, straight driving and right-turn driving.

When a reference vehicle runs to the crossroad test road intersection, starting a reference pedestrian to pass through the crossroad test pedestrian passage, if the to-be-tested unmanned vehicle enters the crossroad in a normal green light passing state, and simultaneously enters the crossroad test road intersection, avoiding or colliding with the to-be-tested pedestrian, and judging that the capability of the to-be-tested unmanned vehicle entering the crossroad is unqualified; if the unmanned vehicle to be tested enters the crossroad in a normal green light passing state and enters the crossroad test road intersection at the same time, normal passing of reference pedestrians is not influenced, and the capability of the unmanned vehicle to be tested entering the crossroad is judged to be qualified; and if the to-be-detected unmanned vehicle enters the crossroad in a red light or yellow light non-passing state, judging that the capability of the to-be-detected unmanned vehicle entering the crossroad is unqualified.

Example 1:

as shown in fig. 1, in the present embodiment, it is used to simulate a scene in which an unmanned vehicle travels straight through an intersection having traffic lights, and to test the ability of the unmanned vehicle to pass through the intersection in the scene.

The test system comprises: the system comprises a crossroad test road provided with traffic signal lamps 4, roadside equipment 5 and a control center computer 6, wherein the crossroad test road is provided with a straight lane a with the length of not less than 200m, and the unmanned vehicle 1 to be tested can drive to the crossroad along the straight lane a; the roadside equipment 5 is arranged beside the crossroad, can acquire the running information and traffic signal lamp information of the unmanned vehicle to be detected and sends the running information and the traffic signal lamp information to the control center computer.

The test procedure of test case 1 is: starting the unmanned vehicle 1 to be tested, accelerating to 30km/h (typical speed of straight-going at an intersection in an urban environment), then driving at a constant speed, when the unmanned vehicle 1 to be tested drives to a position 150m away from a stop line of the crossroad, sending a straight-going crossroad passing command to the unmanned vehicle 1 to be tested through a control center computer 6 through wireless communication equipment, simultaneously controlling a signal to be red along the driving direction phase of the unmanned vehicle 1 to be tested, obtaining the driving information of the unmanned vehicle to be tested in real time through the control center computer 6, observing whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad, and evaluating a test result.

The response of the unmanned vehicle to be tested in the test case 1 is as follows: after receiving the command of straight-going through the intersection sent by the control center computer 6, the unmanned vehicle 1 to be detected identifies the phase information of the signal lamp by using the sensing equipment carried by the unmanned vehicle 1 or identifies the state of the signal lamp by receiving the phase information of the signal lamp by using the V2X communication equipment carried by the unmanned vehicle 1 to be detected. And because the phase of the signal lamp is red, the unmanned vehicle to be detected should stop before the stop line. The unmanned vehicle to be tested needs to stop in front of the stop line, if the front wheels cross the stop line, the unmanned vehicle to be tested is shown not to stop in front of the stop line, the sensing and executing capability is unqualified in the test case, and 0 is recorded; if the distance between the vehicle head and the stop line is not more than 2.5m (a half vehicle length typical value), the result shows that the unmanned vehicle to be tested has excellent perception and execution capability in the test case, and 2 points are counted; if the distance between the vehicle head and the stop line is more than 2.5m and less than 5m, the sensing and executing capability of the unmanned vehicle to be tested in the test case is qualified, and 1 point is counted; if the distance between the vehicle head and the stop line is more than 5 meters (one vehicle length typical value), the result shows that the perception and execution capacity of the unmanned vehicle to be tested in the test case is unqualified, and the score is 0. And carrying out at least 50 times of tests on the same type of unmanned vehicles by using the same test case, calculating the total score into the test case, and taking the evaluation score of the ability test of the unmanned vehicles passing the crossroads or normalizing the total score as the evaluation score of the ability test of the unmanned vehicles passing the crossroads in the test case.

The test procedure of test case 2 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h (typical speed of straight-going at an intersection in an urban environment), then driving at a constant speed, when the unmanned vehicle 1 to be detected drives to a position 150m away from a stop line of the crossroad, sending a straight-going command to the unmanned vehicle 1 to be detected through a control center computer 6 by wireless communication equipment, simultaneously controlling a phase of a signal to be green along the driving direction of the unmanned vehicle 1 to be detected, identifying the signal lamp information in the straight-going direction by the unmanned vehicle 1 to be detected by utilizing a carried sensing device or according to the signal lamp phase information received by V2X communication equipment, and executing the signal lamp command; and the control center computer 6 acquires the running information of the unmanned vehicle to be tested in real time, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result. After receiving the command of straight-going through the intersection sent by the control center computer 6, the unmanned vehicle 1 to be detected identifies the phase information of the signal lamp by using the carried sensing equipment or receives the phase information of the signal lamp by using the carried V2X communication equipment, and identifies the state of the signal lamp. And because the phase of the signal lamp is a green lamp, the unmanned vehicle to be detected can keep running at the uniform speed and pass through the intersection.

The unmanned vehicle to be tested needs to keep running straight at a constant speed, if the speed deviation is not more than 5km/h, the unmanned vehicle to be tested is excellent in sensing and executing capacity in the test case, and 2 points are counted; if the speed deviates from 5km/h-10km/h, the sensing and executing capacity of the unmanned vehicle to be tested in the test case is qualified, and the score is 1; and if the speed deviation is more than 10km/h, the sensing and executing capacity of the unmanned vehicle to be tested in the test case is unqualified, and the score of 0 is counted. Preferably, the response of the unmanned vehicle 1 to be tested may further include that the unmanned vehicle 1 to be tested travels to a correct lane after passing through the intersection. And carrying out at least 50 times of tests on the same type of unmanned vehicles by using the same test case, calculating the total score into the test case, and taking the evaluation score of the ability test of the unmanned vehicles passing the crossroads or normalizing the total score as the evaluation score of the ability test of the unmanned vehicles passing the crossroads in the test case.

The test procedure of test case 3 is: and keeping the signal lamp of the intersection as a red lamp, and stopping the unmanned vehicle 1 to be detected outside the stop line of the intersection to wait for the change of the signal lamp. The control center computer 6 sends a straight-going command to the unmanned vehicle 1 to be detected through the wireless communication equipment, and controls the phase of the signal lamp to be changed into green after waiting for not less than 2 s. The unmanned vehicle 1 to be tested identifies the signal lamp phase by using the sensing equipment carried on the unmanned vehicle or identifies the signal lamp information according to the signal lamp phase information received by the V2X communication equipment, and executes the signal lamp instruction. And the control center computer 6 acquires the running information of the unmanned vehicle to be tested in real time, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The response of the unmanned vehicle to be tested in the test case 3 is as follows: and the unmanned vehicle to be detected stops at the outer side of the stop line of the intersection. After receiving a straight-going crossroad passing command sent by the control center computer 6, the unmanned vehicle 1 to be detected continuously identifies the phase information of the signal lamp by using the carried sensing equipment or receives the phase information of the signal lamp by using the carried V2X communication equipment, and identifies the state of the signal lamp. When the phase of the signal lamp is changed into green, the unmanned vehicle to be detected is started and accelerated to run straight and pass through the intersection.

The evaluation method of the test case 3 is as follows: when the vehicle stops and the like, the red light is turned to the green light, the vehicle is started to run straight, and the vehicle is started within 2s, which shows that the unmanned vehicle to be tested has excellent perception and execution capability in the current test case and counts 2 points; starting within 2-4 seconds, and counting 1 minute, wherein the sensing and executing capabilities of the unmanned vehicle to be tested in the current test case are qualified; and 4s of external starting, which indicates that the sensing and executing capabilities of the unmanned vehicle to be tested in the current test case are unqualified, and the score is 0. When the vehicle approaches the intersection, the red light turns to the green light, and there is no essential difference between the perception and the stop and the like in the straight-going direction, and the vehicle is not tested as an independent test case.

The test procedure of test case 4 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h (typical straight-going speed of an intersection in an urban environment), then driving at a constant speed, sending a straight-going command to the unmanned vehicle 1 to be detected through the control center computer 6 by wireless communication equipment, and keeping a signal lamp of the intersection as a green light to enable the unmanned vehicle to be detected to approach the intersection at a constant speed. When the unmanned vehicle to be detected enters the range within 10 meters from the stop line of the intersection, the control center computer 6 randomly controls the phase of the signal lamp to be changed into yellow lamp and then to be changed into red lamp after 2s, and the unmanned vehicle 1 to be detected identifies the phase of the signal lamp by using the carried sensing equipment or identifies the signal lamp information according to the signal lamp phase information received by the V2X communication equipment and executes the signal lamp instruction. And the control center computer 6 acquires the running information of the unmanned vehicle to be tested in real time, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result. After receiving a command of going straight through the crossroad sent by the control center computer 6, the unmanned vehicle 1 to be detected accelerates to 30km/h and then drives at a constant speed to approach the crossroad, and meanwhile, the phase information of the signal lamp is continuously identified by using the carried sensing equipment or the phase information of the signal lamp is received by using the carried V2X communication equipment, so that the state of the signal lamp is identified. When the phase of the signal lamp is changed into yellow, the unmanned vehicle to be detected should judge the position of the unmanned vehicle and select to stop or pass through the intersection.

The evaluation method of the test case 4 is as follows: when the yellow lamp is on, the distance between the vehicle head and the stop line is not more than 5m (a vehicle length typical value), the vehicle can pass through the crossroad, at the moment, the vehicle does not have a speed reducer for 2 minutes, the speed reducer passes through the meter for 1 minute, and the parking meter is 0 minute; when the yellow lamp is on, the distance between the vehicle head and the stop line is more than 5m (one vehicle length typical value), the vehicle can not pass through the crossroad, at the moment, the wheel crosses the stop line and is marked with 0 minute, and the distance between the vehicle head and the stop line is not more than 2.5m (half vehicle length typical value) and is counted with 2 minutes; more than 2.5m, less than 5m and 1 min; more than 5 meters (one vehicle length typical value) and 0 minutes.

Example 2:

as shown in fig. 2, in the present embodiment, the method is used for simulating that an unmanned vehicle travels straight through an intersection with traffic lights, avoiding a scene of other vehicles with route conflicts, and testing the capability of the unmanned vehicle to pass through the intersection in the scene.

The test system comprises a cross test road provided with traffic signal lamps 4, an unmanned vehicle 1 to be tested, a plurality of reference vehicles, roadside equipment 5 and a control center computer 6; the cross test road provided with the traffic signal lamps 4 comprises a straight lane a, a lateral right-turn lane b and an opposite left-turn lane c.

The unmanned vehicle 1 to be detected can drive to the crossroad along the straight lane a; the roadside equipment 5 is arranged beside the crossroad, and can acquire reference vehicle running information, unmanned vehicle running information to be detected and traffic signal lamp information and send the reference vehicle running information, the unmanned vehicle running information to be detected and the traffic signal lamp information to the control center computer. The control center computer 6 can send a command of passing through the crossroad to the unmanned vehicle 1 to be detected through the wireless communication equipment. The test scenario used in this embodiment contains 2 test cases.

The test procedure of test case 1 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h and then driving at a constant speed, when the unmanned vehicle 1 to be detected drives to a position 150m away from a stop line of the crossroad, sending a straight-going crossroad passing command to the unmanned vehicle 1 to be detected through the control center computer 6 through the wireless communication equipment, and keeping the phase of the unmanned vehicle 1 to be detected as a green light along the driving direction of the unmanned vehicle 1 to be detected. The unmanned vehicle 1 to be detected senses the phase of the signal lamp and detects the vehicle in the driving path by using the carried sensing equipment, and the unmanned vehicle moves straight through the crossroad. When the unmanned vehicle passes the center of the intersection by about 10m (two vehicle length typical values), the first reference vehicle 2 is controlled to make a left turn over the center of the intersection from the lane c for making a left turn, and the left turn speed is 20km/h typical value for performing a turning action under urban environment. And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 1 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the opposite left-turning vehicle or not, and evaluates the test result. When the unmanned vehicle 1 to be tested crosses the stop line, detecting the vehicles in the driving path range, when the first reference vehicle 2 is detected and approaches, if the unmanned vehicle does not have the deceleration collision avoidance action, the collision risk is high, and the result shows that the sensing and executing capability of the unmanned vehicle is unqualified under the test case, and the score is 0; if the distance between the two vehicles is not less than 2.5m (a typical value of half vehicle length), the sensing and executing capability of the unmanned vehicle under the test case is excellent and 2 points are counted; if the distance between the two vehicles is less than 2.5m, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

The test procedure of test case 2 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h and then driving at a constant speed, when the unmanned vehicle 1 to be detected drives to a position 150m away from a stop line of the crossroad, sending a straight-going crossroad passing command to the unmanned vehicle 1 to be detected through the control center computer 6 through the wireless communication equipment, and keeping the phase of the unmanned vehicle 1 to be detected as a green light along the driving direction of the unmanned vehicle 1 to be detected. The unmanned vehicle 1 to be detected senses the phase of the signal lamp and detects the vehicle in the driving path by using the carried sensing equipment, and the unmanned vehicle moves straight through the crossroad. When the unmanned vehicle passes the center of the intersection by about 10m (two typical values of the vehicle length), the first reference vehicle 2 is controlled to cross the stop line from the side direction right-turn lane b, and the right-turn speed is 20km/h which is a typical value of the steering action performed in the urban environment. And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 2 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the side-turning right vehicle or not, and evaluates the test result. When the unmanned vehicle 1 to be tested crosses the stop line, detecting the vehicles in the driving path range, when the reference vehicle 2 is detected and approaches, if the unmanned vehicle does not have the action of speed reduction and collision avoidance, the collision risk is high, and the result shows that the sensing and execution capacity of the unmanned vehicle is unqualified under the test case, and the score is 0; if the distance between the two vehicles is not less than 2.5m (a typical value of half vehicle length), the sensing and executing capability of the unmanned vehicle under the test case is excellent and 2 points are counted; if the distance between the two vehicles is less than 2.5m, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

And sequentially testing 2 test cases of the same type of unmanned vehicle in the same scene, and calculating the total score of the 2 test cases as the evaluation score of the capability test of the unmanned vehicle passing through the intersection in the scene, or normalizing the total score to be used as the evaluation score of the capability test of the unmanned vehicle passing through the intersection in the scene.

Example 3:

as shown in fig. 3, in the present embodiment, the method is used for simulating that the unmanned vehicle passes through the intersection with the traffic light straight, avoiding the scene of the reference pedestrian running the red light, and testing the capability of the unmanned vehicle passing through the intersection in the scene. The test system comprises: the cross test road is provided with a traffic signal lamp 4, roadside equipment 5, a control center computer 6 and a plurality of reference pedestrians, a straight lane a, a first pedestrian f1 and a second pedestrian f2 are arranged on the cross test road, the plurality of reference pedestrians all adopt a test mobile platform, and a human body model for testing is carried on the test mobile platform.

The unmanned vehicle 1 to be detected can drive to the crossroad along the straight lane a; the roadside equipment 5 is arranged beside the crossroad, and can acquire reference pedestrian running information, running information of the unmanned vehicle to be detected and traffic signal lamp information and send the reference pedestrian running information, the running information of the unmanned vehicle to be detected and the traffic signal lamp information to the control center computer. The control center computer 6 can send a command of passing through the crossroad to the unmanned vehicle 1 to be detected through the wireless communication equipment. The test scenario used in this embodiment contains 2 test cases. And different test cases are used for testing the response condition of the unmanned vehicle under the condition of different reference pedestrian walking track conflicts. The 2 test cases should not be deployed in the test field at the same time.

The test procedure of test case 1 is: starting the unmanned vehicle 1 to be tested, accelerating to 30km/h and then driving at a constant speed, when the unmanned vehicle 1 to be tested drives to a position 150m away from a stop line of the crossroad, sending a straight-going command to the unmanned vehicle 1 to be tested through the control center computer 6 by the wireless communication equipment, and keeping the phase of a control signal along the driving direction of the unmanned vehicle 1 to be tested as a green light. The unmanned vehicle 1 to be detected senses the phase of a signal lamp and detects vehicles or reference pedestrians in a driving path by using the carried sensing equipment (including but not limited to a camera, a laser radar and a millimeter wave radar), and the unmanned vehicle moves straight through the crossroad. When the unmanned vehicle 1 to be detected drives to the outer side of the stop line, the first reference pedestrian 2 is controlled to enter the pedestrian crossing f1 to pass through the crossroad in a straight line according to the stop line 10m (two typical values of the vehicle length), and the speed is 4km/h (the typical value of the pedestrian walking speed). And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method in test case 1 is: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the reference pedestrian, and evaluates the test result. When the unmanned vehicle to be tested is detected to be close to the pedestrian crossing, no deceleration collision avoidance action is performed on the unmanned vehicle, the collision risk is high, the sensing and executing capability of the unmanned vehicle is unqualified under the test case, and the score is 0; when the unmanned vehicle decelerates and avoids collision, the collision risk is reduced, the distance between the vehicle and the human is not less than 2.5m (a typical value of half the vehicle length), which indicates that the sensing and executing capability of the unmanned vehicle under the test case is excellent and 2 points are counted; the distance between the vehicle and the person is less than 2.5m, which indicates that the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

The test procedure of test case 2 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h and then driving at a constant speed, when the unmanned vehicle 1 to be detected drives to a distance of 150m from a stop line of the crossroad, the control center computer 6 sends a command of going straight through the crossroad to the unmanned vehicle 1 to be detected through the wireless communication equipment, and the signal lamp is kept to be a green lamp. The unmanned vehicle 1 to be detected is started to move straight through the crossroad. When the unmanned vehicle 1 to be detected crosses the center of the intersection, the second reference pedestrian 3 is controlled to enter the pedestrian crossing f2 to pass through the intersection in a straight line, and the speed is 4km/h (the typical value of the walking speed of the reference pedestrian). And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The response of the unmanned vehicle 1 to be tested in the test case 2 in the test process is as follows: after receiving a straight-going crossroad passing command sent by the control center computer 6, the unmanned vehicle 1 to be detected detects a crosswalk and senses position and speed running information of a reference pedestrian by using the carried sensing equipment, decelerates and avoids collision with the reference pedestrian.

The evaluation method in the test case 2 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the reference pedestrian, and evaluates the test result. When the unmanned vehicle to be tested is detected to be close to the pedestrian crossing, no deceleration collision avoidance action is performed on the unmanned vehicle, the collision risk is high, the sensing and executing capability of the unmanned vehicle is unqualified under the test case, and the score is 0; when the unmanned vehicle decelerates and avoids collision, the collision risk is reduced, the distance between the vehicle and the human is not less than 2.5m (a typical value of half the vehicle length), which indicates that the sensing and executing capability of the unmanned vehicle under the test case is excellent and 2 points are counted; the distance between the vehicle and the person is less than 2.5m, which indicates that the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

Example 4:

as shown in fig. 4, in the present embodiment, the method is used for simulating that the unmanned vehicle passes through the intersection with the traffic signal lamp in a straight line, avoiding the scene of the vehicle running the red light, and testing the capability of the unmanned vehicle passing through the intersection in the scene.

The test system comprises: the system comprises a cross test road provided with traffic signal lamps 4, unmanned vehicles 1 to be tested, a plurality of reference vehicles, roadside equipment 5 and a control center computer 6; the criss-cross test road includes a straight lane a, a left lateral lane e, and a right lateral lane d.

The unmanned vehicle 1 to be detected can drive to the crossroad along the straight lane a; the roadside equipment 5 is arranged beside the crossroad, and can acquire reference vehicle running information, unmanned vehicle running information to be detected and traffic signal lamp information and send the reference vehicle running information, the unmanned vehicle running information to be detected and the traffic signal lamp information to the control center computer.

The test scenario used in this embodiment contains 2 test cases. And different test cases are used for testing the response condition of the unmanned vehicle under the condition of different vehicle running track conflicts. The 2 test cases should not be deployed in the test field at the same time.

The test procedure of test case 1 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h and then driving at a constant speed, when the unmanned vehicle 1 to be detected drives to a distance of 150m from a stop line of the crossroad, the control center computer 6 sends a command of going straight through the crossroad to the unmanned vehicle 1 to be detected through the wireless communication equipment, and the signal lamp is kept to be a green lamp. The unmanned vehicle 1 to be detected is started to move straight through the crossroad. When the unmanned vehicle 1 to be detected runs to a distance of about 20m (four vehicle length typical values) from the center of the crossroad, the first reference vehicle 2 runs the red light and enters the unmanned vehicle running path from the right side across the stop line, and the speed is 30km/h which is the typical speed value in the urban environment. And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 1 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the vehicle running the red light, and evaluates the test result. When no deceleration collision avoidance action of the unmanned vehicle is detected, the collision risk is high, the perception and execution capacity of the unmanned vehicle under the test case is unqualified, and the score is 0. If the distance between the two vehicles is not less than 5m (a vehicle length typical value), the perception and execution capability of the unmanned vehicle under the test case is excellent and 2 points are counted; if the distance between the two vehicles is less than 5m, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

And carrying out at least 50 times of tests on the same type of unmanned vehicles by using the same test case, calculating the total score into the test case, and taking the evaluation score of the ability of the unmanned vehicles passing through the intersection or normalizing the total score as the evaluation score of the ability of the unmanned vehicles passing through the intersection in the test case.

The test procedure of test case 2 is: starting the unmanned vehicle 1 to be detected, accelerating to 30km/h and then driving at a constant speed, when the unmanned vehicle 1 to be detected drives to a distance of 150m from a stop line of the crossroad, the control center computer 6 sends a command of going straight through the crossroad to the unmanned vehicle 1 to be detected through the wireless communication equipment, and the signal lamp is kept to be a green lamp. When the unmanned vehicle 1 to be detected runs to a distance of about 20m (four vehicle length typical values) from the stop line, the second reference vehicle 3 runs the red light to cross the stop line from the left side and enter the unmanned vehicle running path, and the speed is 30km/h which is the typical speed value in the urban environment. The unmanned vehicle 1 to be detected decelerates and avoids collision with the red light running vehicle 3. Meanwhile, the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 2 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the vehicle running the red light, and evaluates the test result. When no deceleration collision avoidance action of the unmanned vehicle is detected, the collision risk is high, the perception and execution capacity of the unmanned vehicle under the test case is unqualified, and the score is 0. If the distance between the two vehicles is not less than 5m (a vehicle length typical value), the perception and execution capability of the unmanned vehicle under the test case is excellent and 2 points are counted; if the distance between the two vehicles is less than 5m, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

Example 5:

as shown in fig. 5, in the present embodiment, a simulation is made for a scene in which an unmanned vehicle turns right/left through an intersection having traffic lights, and the ability of the unmanned vehicle to pass through the intersection in the scene is tested.

The test system comprises: the cross test road is provided with a traffic signal lamp 4, road side equipment 5 and a control center computer 6. The unmanned vehicle 1 to be tested can drive to the crossroad along the right-turn lane a or the left-turn lane b; the roadside equipment 5 is arranged beside the crossroad, can monitor the running information of the unmanned vehicle to be detected passing through the crossroad and sends the running information to the control center computer. The control center computer 6 can send a command of passing through the crossroad to the unmanned vehicle 1 to be tested through the wireless communication equipment, receive the running information of the unmanned vehicle to be tested sent by the unmanned vehicle to be tested and evaluate the test result.

The test procedure of test case 1 is: starting the unmanned vehicle 1 to be tested to accelerate to 30km/h along the right-turn lane and then to run at a constant speed, when the unmanned vehicle 1 to be tested runs to a distance of 150m from a stop line of the crossroad, the control center computer 6 sends a right-turn command at the crossroad to the unmanned vehicle 1 to be tested through the wireless communication equipment, and a right-turn signal lamp is kept to be a green lamp. And the control center computer 6 receives the running information of the unmanned vehicle to be tested sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad and evaluates the test result.

The evaluation method of the test case 1 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly execute right turn of the crossroad or not, and evaluates the test result. At the moment, the unmanned vehicle 1 to be detected should enter the rightmost lane of the lateral lane, if the unmanned vehicle deviates from the preset lane seriously, the driving of the vehicle on the lateral lane is greatly influenced, but the lane deviation of the unmanned vehicle to a certain degree is allowed in consideration of various shapes and parameters of the intersection. The typical width of the unmanned vehicle is 2m, and the typical width of the lane of the intersection in the urban environment is 3.75 m. Therefore, if the center line of the unmanned vehicle crosses the outer side line of the preset lane, the influence on the vehicle running of the lateral lane is large, and the result shows that the perception and execution capacity of the unmanned vehicle is unqualified under the test case, and the score is 0; if the center line of the unmanned vehicle is not more than 0.8m according to the center line of the preset lane, the unmanned vehicle is excellent in perception and execution capacity under the test case and 2 points are counted; if the center line of the unmanned vehicle is larger than 0.8m and smaller than 1.85m according to the center line of the preset lane, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

The test procedure of test case 2 is: starting the unmanned vehicle 1 to be tested to accelerate to 30km/h along a right-turn lane and then to run at a constant speed, when the unmanned vehicle 1 to be tested runs to a distance of 150m from a stop line of the crossroad, the control center computer 6 sends a left-turn command at the crossroad to the unmanned vehicle 1 to be tested through the wireless communication equipment, and a left-turn signal lamp is kept to be a green lamp. The unmanned vehicle 1 to be detected senses the phase of a signal lamp and the road environment of an intersection by using the carried sensing equipment (including but not limited to a camera, a laser radar and a millimeter wave radar), and turns left at the intersection. Meanwhile, the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 1 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly execute the left turn of the crossroad or not, and evaluates the test result. At the moment, the unmanned vehicle 1 to be detected should enter the leftmost lane of the lateral lane, if the unmanned vehicle deviates from the preset lane seriously, the driving of the vehicle in the lateral lane is greatly influenced, but the lane deviation of the unmanned vehicle to a certain degree is allowed in consideration of various shapes and parameters of the intersection. The typical width of the unmanned vehicle is 2m, and the typical width of the lane of the intersection in the urban environment is 3.75 m. Therefore, if the center line of the unmanned vehicle crosses the outer side line of the preset lane, the influence on the vehicle running of the lateral lane is large, and the result shows that the perception and execution capacity of the unmanned vehicle is unqualified under the test case, and the score is 0; if the center line of the unmanned vehicle is not more than 0.8m according to the center line of the preset lane, the unmanned vehicle is excellent in perception and execution capacity under the test case and 2 points are counted; if the center line of the unmanned vehicle is larger than 0.8m and smaller than 1.85m according to the center line of the preset lane, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

Example 6:

as shown in fig. 6, in the present embodiment, the method is used for simulating that the unmanned vehicle turns right to pass through the intersection with the traffic signal lamp, avoiding a scene of vehicles and reference pedestrians which conflict with the driving path of the unmanned vehicle to be tested, and testing the capability of the unmanned vehicle to pass through the intersection in the scene.

The test system comprises: the system comprises road side equipment 5, a control center computer 6, a cross test road provided with traffic signal lamps 4, a plurality of reference vehicles and a plurality of reference pedestrians.

The unmanned vehicle 1 to be detected can drive to the crossroad along the right-turn lane a; the roadside equipment 5 is arranged beside the crossroad, can monitor the running information of the unmanned vehicle to be detected passing through the crossroad and sends the running information to the control center computer.

The test scenario used in this embodiment contains 3 test cases. And different test cases are used for testing the ability of the unmanned vehicle to turn right to pass through the crossroad with the traffic signal lamp, and avoid vehicles and reference pedestrians which conflict with the running path of the unmanned vehicle to be tested. The 3 test cases should not be deployed in the test field at the same time.

The test procedure of test case 1 is: starting the unmanned vehicle 1 to be tested to accelerate to 20km/h (typical steering vehicle speed in urban environment) along a right-turn lane and then driving at a constant speed, when the unmanned vehicle 1 to be tested drives to a position 150m away from a stop line of a crossroad, sending a right-turn command at the crossroad to the unmanned vehicle 1 to be tested through the control center computer 6 through wireless communication equipment, and keeping a right-turn signal lamp as a green lamp. When the unmanned vehicle 1 to be detected crosses the stop line and enters the crossroad, the reference pedestrian is controlled to enter the pedestrian crossing f3 and to move straight through the crossroad, and the speed is 4km/h (the reference pedestrian moves at a typical speed). The unmanned vehicle 1 to be detected decelerates and avoids collision with the reference pedestrian 3, and the right turn passes through the crossroad. And the control center computer 6 receives the running information of the unmanned vehicle to be tested sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad and evaluates the test result.

The evaluation method of the test case 1 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the reference pedestrian, and evaluates the test result. When the unmanned vehicle 1 to be tested crosses the stop line, the reference pedestrian in the driving path range is detected, and when the reference pedestrian 3 is detected to be close to the pedestrian crossing, the unmanned vehicle does not have the action of speed reduction and collision avoidance, the collision risk is high, which indicates that the sensing and execution capacity of the unmanned vehicle is unqualified under the test case, and the score is 0. When the unmanned vehicle decelerates and avoids collision, the collision risk is reduced, the distance between the vehicle and the human is not less than 2.5m (a typical value of half the vehicle length), which indicates that the sensing and executing capability of the unmanned vehicle under the test case is excellent and 2 points are counted; the distance between the vehicle and the person is less than 2.5m, which indicates that the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

The test procedure of test case 2 is: starting the unmanned vehicle 1 to be tested to accelerate to 20km/h (typical steering vehicle speed in urban environment) along a right-turn lane and then driving at a constant speed, when the unmanned vehicle 1 to be tested drives to a position 150m away from a stop line of a crossroad, sending a right-turn command at the crossroad to the unmanned vehicle 1 to be tested through the control center computer 6 through wireless communication equipment, and keeping a right-turn signal lamp as a green lamp. When the unmanned vehicle 1 to be detected is 50m away from the stop line (10 vehicle length typical values), the reference vehicle 2 is controlled to pass through the crossroad by passing through the stop line and turning left, the speed is 20km/h (the vehicle driving typical speed in urban environment), and the collision with the driving path of the unmanned vehicle to be detected occurs. The unmanned vehicle 1 to be detected decelerates and avoids collision with the reference vehicle 2, and turns right to pass through the crossroad. And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 2 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the opposite left-turning vehicle or not, and evaluates the test result. When no deceleration collision avoidance action of the unmanned vehicle to be tested is detected, the collision risk is high, the perception and execution capacity of the unmanned vehicle under the test case is unqualified, and the score is 0. If the distance between the two vehicles is not less than 5m (a vehicle length typical value), the perception and execution capability of the unmanned vehicle under the test case is excellent and 2 points are counted; if the distance between the two vehicles is less than 5m, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

The test procedure of test case 3 is: starting the unmanned vehicle 1 to be tested to accelerate to 20km/h (typical steering vehicle speed in urban environment) along a right-turn lane and then driving at a constant speed, when the unmanned vehicle 1 to be tested drives to a position 150m away from a stop line of a crossroad, sending a right-turn command at the crossroad to the unmanned vehicle 1 to be tested through wireless communication equipment by the control center computer 6, and keeping a straight-going signal lamp as a red lamp, a right-turn signal lamp as a green lamp and a lateral straight-going signal lamp as a green lamp. When the unmanned vehicle 1 to be detected is 30m away from the stop line (6 vehicle length typical values), the reference vehicle 2 is controlled to pass through the crossroad by crossing the stop line and go straight at a speed of 30km/h (typical speed of vehicle running in urban environment), and conflict with the running path of the unmanned vehicle to be detected. The unmanned vehicle 1 to be detected decelerates and avoids collision with the reference vehicle 2, and turns right to pass through the crossroad. And the control center computer 6 receives the running data sent by the unmanned vehicle to be tested, observes whether the unmanned vehicle 1 to be tested can correctly pass through the crossroad or not, and evaluates the test result.

The evaluation method of the test case 3 is as follows: and the control center computer 6 receives the running data sent by the unmanned vehicle 1 to be tested, observes whether the unmanned vehicle 1 to be tested can correctly avoid the lateral straight-going vehicle or not, and evaluates the test result. When no deceleration collision avoidance action of the unmanned vehicle to be tested is detected, the collision risk is high, the perception and execution capacity of the unmanned vehicle under the test case is unqualified, and the score is 0. If the distance between the two vehicles is not less than 5m (a vehicle length typical value), the perception and execution capability of the unmanned vehicle under the test case is excellent and 2 points are counted; if the distance between the two vehicles is less than 5m, the sensing and executing capability of the unmanned vehicle is qualified under the test case, and the score is 1.

And sequentially testing 3 test cases of the same type of unmanned vehicle in the same scene, and calculating the total score of the 3 test cases as the evaluation score of the capability test of the unmanned vehicle passing through the intersection in the scene, or normalizing the total score to be used as the evaluation score of the capability test of the unmanned vehicle passing through the intersection in the scene.

The above description is only a preferred embodiment of the invention and is illustrative of the technology used. It will be appreciated by those skilled in the art that the scope of the invention according to the present invention is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is made without departing from the inventive concept.

Claims

1. The system for testing the traffic capacity of the crossroad of the unmanned vehicle is characterized by comprising road side equipment (5), a control center computer (6), a reference vehicle and a cross test road provided with a traffic signal lamp (4), wherein the reference vehicle can run on the cross test road according to a set running route, and the unmanned vehicle to be tested starts to enter the cross test road according to a control signal of the control center computer (6); the reference vehicle and the unmanned vehicle to be detected are both provided with a signal transmission module for wireless transmission with roadside equipment (5) and a data acquisition module for acquiring the running information of the reference vehicle and the running information of the unmanned vehicle to be detected; the roadside device (5) is arranged beside the cross test road and used for acquiring reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information and sending the acquired reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information to the control center computer (6), the control center computer (6) receives and stores the reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information, and the control center computer (6) obtains the traffic capacity of the unmanned vehicle to be tested at the intersection according to the received reference vehicle running information, unmanned vehicle running information to be tested and traffic signal lamp information.

2. The system for testing the trafficability of the unmanned vehicle intersection as claimed in claim 1, wherein the signal transmission modules of the reference vehicle and the unmanned vehicle to be tested communicate wirelessly using V2X.

3. The unmanned vehicle intersection trafficability testing system of claim 1, wherein the reference vehicle travel information comprises a reference vehicle position, a reference vehicle motion trajectory, and a reference vehicle travel speed information.

4. The unmanned vehicle intersection trafficability testing system of claim 1, wherein the unmanned vehicle driving information to be tested comprises a position of the unmanned vehicle to be tested, a motion trajectory of the unmanned vehicle to be tested, and driving speed information of the unmanned vehicle to be tested.

5. The unmanned vehicle intersection trafficability test system of claim 1, further comprising a reference pedestrian disposed on the crossroad pedestrian passageway, the reference pedestrian being capable of traveling on the crossroad pedestrian passageway.

6. The system for testing the trafficability of the unmanned vehicle intersection as claimed in claim 5, wherein the reference pedestrians each use a test mobile platform, and a human body model for testing is carried on the test mobile platform.

7. The unmanned vehicle intersection traffic capacity test system according to claim 1, wherein the road side equipment (5) comprises an ENodeB base station, a road side test unit, an LTE-V core network and a local server group; the LTE-V core network is connected with the local server group through the aggregation switch, the aggregation switch is connected with the ENodeB base station, the ENodeB base station provides a wireless signal covering a cross test road, a data acquisition module on the unmanned vehicle to be tested realizes information interaction with the ENodeB base station in a wireless mode, the roadside test unit is installed beside the cross test road, and the roadside test unit realizes information interaction with the ENodeB base station and the control center computer in a wireless mode.

8. An unmanned vehicle intersection traffic capacity test method based on the unmanned vehicle intersection traffic capacity test system of claim 5, characterized by comprising the steps of:

9. The intersection unmanned vehicle traffic capacity test method according to claim 8, wherein in the step 1), the reference vehicle is caused to travel on the intersection test road according to the set route, and the reference vehicle travels through the intersection of the intersection test road according to the normal vehicle, specifically including left-turn travel, straight travel, and right-turn travel.

10. The intersection unmanned vehicle capacity test method of claim 8, wherein when a reference vehicle is driven to the intersection of the intersection test road, the reference pedestrian is simultaneously started to pass through the intersection test road pedestrian passageway.