CN114067610B - Simulation scene construction method and device for missing turnout accident - Google Patents
Simulation scene construction method and device for missing turnout accident Download PDFInfo
- Publication number
- CN114067610B CN114067610B CN202111108101.7A CN202111108101A CN114067610B CN 114067610 B CN114067610 B CN 114067610B CN 202111108101 A CN202111108101 A CN 202111108101A CN 114067610 B CN114067610 B CN 114067610B
- Authority
- CN
- China
- Prior art keywords
- lane
- vehicle
- turnout
- missed
- simulation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/16—Anti-collision systems
- G08G1/167—Driving aids for lane monitoring, lane changing, e.g. blind spot detection
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Train Traffic Observation, Control, And Security (AREA)
Abstract
The invention discloses a method and a device for constructing a simulation scene of a missed turnout accident, wherein the method comprises the following steps: acquiring drive test data of a vehicle running in a running area; the drive test data comprises vehicle lane change information and manual take-over information; when the fact that the vehicle meets the preset missed turnout junction judgment condition is judged according to the vehicle lane changing information and the manual takeover information, extracting first time when the vehicle starts to enter a lane changing state and second time when the vehicle drives into a target turnout junction after being manually taken over; intercepting the drive test data from the first time to the second time from the drive test data to obtain missed turnout drive test data corresponding to a missed turnout accident; and constructing a simulation scene of missing turnout accidents according to the missing turnout road test data. By implementing the method and the device, the efficiency of constructing the simulation scene missing the turnout accident can be improved.
Description
Technical Field
The invention relates to the technical field of automatic driving, in particular to a method and a device for constructing a simulation scene of a missed turnout accident.
Background
The simulation by using the drive test data is a common simulation mode in automatic driving, but the drive test data volume is huge and cannot be directly used for simulation. In the prior art, a marking person is often required to manually mark a complete drive test data according to the requirements of different simulation scenes, and then the drive test data at the corresponding marking position is intercepted according to the requirements of the different simulation scenes, so as to construct a corresponding simulation scene; for example, in the simulation process of automatic driving, the fact that a vehicle misses a crossing is usually required to be simulated, a simulation scene of the crossing-missing accident is constructed, based on the prior art, the crossing-missing accident in the complete road test data of a person is required to be labeled for identification, then the start time and the end time of the accident are manually labeled, and finally the simulation scene of the crossing-missing accident is constructed by intercepting the road test data in the time period from the start time to the end time according to the labeled time node.
Disclosure of Invention
The embodiment of the invention provides a simulation scene construction method for a missed crossing accident, which can automatically construct a simulation scene of the missed crossing accident and improve the efficiency of constructing the simulation scene of the missed crossing accident.
An embodiment of the present invention provides a method for constructing a simulation scene of a missed turnout junction accident, including: acquiring drive test data of a vehicle running in a running area; the drive test data comprises vehicle lane change information and manual take-over information;
when the vehicle is judged to meet a preset missed turnout junction judgment condition according to the vehicle lane change information and the manual takeover information, extracting first time when the vehicle starts to enter a lane change state and second time when the vehicle drives into a target turnout junction after being manually taken over;
intercepting the drive test data from the first time to the second time in the drive test data to obtain missed turnout drive test data corresponding to a missed turnout accident;
and constructing a simulation scene of the missed turnout accident according to the missed turnout road test data.
Further, the preset missed turnout determination condition includes: the vehicle is manually taken over, the vehicle is in a lane changing state before being manually taken over, a passable lane exists in the lane changing direction corresponding to the vehicle before being manually taken over, the vehicle completes at least one lane changing after being manually taken over, the lane changing direction corresponding to the vehicle after being manually taken over is consistent with the lane changing direction corresponding to the vehicle before being manually taken over, and the vehicle drives into the target turnout opening after being manually taken over.
Further, when a simulation scene of the missed turnout accident is constructed according to the missed turnout road test data, a position point in a target turnout, which is away from the target turnout by a preset distance, is used as a destination of a simulated vehicle in the simulation scene.
Further, when a simulation scene of the missed road junction accident is constructed according to the missed road junction road testing data, at least the lane changing times and the corresponding lane changing directions which need to be executed when the vehicle enters the target road junction from the current lane are counted according to the missed road junction road testing data, and the lane changing times and the lane changing directions are used as evaluation indexes of the simulation scene, so that the simulation scene determines whether the simulated vehicle can reproduce the missed road junction accident in the simulation scene according to the evaluation indexes.
Further, the method also comprises the following steps: operating the simulation scene and extracting a simulation operation track of a simulation vehicle in the simulation scene;
comparing the simulation running track with an actual running track of a vehicle in the running area, and then determining whether the simulation vehicle can reappear the missed turnout accident in the simulation scene according to a track comparison result to obtain a first simulation evaluation result;
determining whether the simulated vehicle can reappear the crossing missing accident in the simulation scene according to the evaluation index to obtain a second simulation evaluation result;
and when the first simulation evaluation result is consistent with the second simulation evaluation result, judging that the simulation scene is qualified.
Further, determining that the vehicle enters the target turnout after being manually taken over specifically includes:
acquiring a lane changing direction of each lane changing event of the vehicle and a first lane sequence in which the vehicle before lane changing is located from the lane changing state information of the vehicle;
judging whether a necessary lane change event meeting preset lane change conditions exists according to the lane change direction of each lane change event and the first lane sequence;
if the lane change direction is lane change to the right, the preset lane change condition includes: the first lane sequence is a rightmost lane sequence of lanes where vehicles are located before lane changing, the right lane sequence adjacent to the first lane sequence is not the rightmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing; if the lane changing direction is changing lanes to the left, the preset lane changing conditions include: the first lane sequence is a leftmost lane sequence of lanes where vehicles are located before lane changing, a left lane sequence adjacent to the first lane sequence is not a leftmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing;
if the lane change direction is the lane change to the right, taking the rightmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located; if the lane change direction is the lane change to the left, taking the leftmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located;
and extracting the position information of the target turnout according to the lane sequence of the target turnout, and judging that the vehicle drives into the target turnout after being manually taken over when the position of the vehicle is consistent with the position of the target turnout and the vehicle is in a taken over state.
On the basis of the above method item embodiments, the present invention correspondingly provides apparatus item embodiments;
another embodiment of the present invention provides a simulation scene constructing device for a missed crossing accident, including: the system comprises a data acquisition module, a key time point extraction module, a drive test data interception module and a simulation scene construction module;
the data acquisition module is used for acquiring the drive test data of the vehicle running in the running area; the drive test data comprises vehicle lane change information and manual take-over information;
the key time point extraction module is used for extracting first time when the vehicle starts to enter a lane change state and second time when the vehicle drives into a target turnout after being manually taken over when the fact that the vehicle meets a preset missed turnout judgment condition is judged according to the vehicle lane change information and the manual taking over information;
the drive test data intercepting module is used for intercepting the drive test data from the first time to the second time in the drive test data to obtain missed turnout drive test data corresponding to a missed turnout accident;
the simulation scene construction module is used for constructing the simulation scene of the missed turnout accident according to the missed turnout road test data.
Further, the simulation scene construction module comprises a simulation destination construction unit;
the simulation destination building unit is used for taking a position point in a target turnout, which is away from the target turnout by a preset distance, as a destination of a simulated vehicle in the simulation scene when the simulation scene of the missed turnout accident is built according to the missed turnout test data.
Further, the simulation scene construction module further comprises an evaluation index construction unit;
the evaluation index construction unit is used for counting at least the lane change times and the corresponding lane change directions which need to be executed when the vehicle enters the target turnout from the current lane according to the missed turnout road test data when the simulation scene of the missed turnout accident is constructed according to the missed turnout road test data, and taking the lane change times and the lane change directions as the evaluation indexes of the simulation scene, so that the simulation scene determines whether the simulated vehicle can reproduce the missed turnout accident in the simulation scene according to the evaluation indexes.
Further, the method also comprises the following steps: a simulation scene evaluation module;
the simulation scene evaluation module is used for operating the simulation scene and extracting a simulation operation track of a simulation vehicle in the simulation scene;
comparing the simulation running track with an actual running track of a vehicle in the running area, and then determining whether the simulation vehicle can reappear the missed turnout accident in the simulation scene according to a track comparison result to obtain a first simulation evaluation result;
determining whether the simulated vehicle can reappear the turnout accident in the simulation scene according to the evaluation index to obtain a second simulation evaluation result;
and when the first simulation evaluation result is consistent with the second simulation evaluation result, judging that the simulation scene is qualified.
The embodiment of the invention has the following beneficial effects:
the method comprises the steps of judging whether a driving situation of a missed turnout scene exists in the course of a vehicle in the course of a journey or not based on vehicle lane change information and manual takeover information in road measurement data, if so, extracting the first time when the vehicle starts to change the lane state and the second time when the vehicle drives into a target turnout after being manually taken over, automatically intercepting road measurement data meeting the scene of the missed turnout accident from the road measurement data according to the first time and the second time to obtain the missed turnout road measurement data, and then carrying out simulation scene construction of the missed turnout accident according to the missed turnout road measurement data.
Drawings
Fig. 1 is a schematic flow chart of a simulation scene construction method for a crossing accident according to an embodiment of the present invention.
Fig. 2 is another schematic flow chart of a simulation scenario construction method for a crossing accident according to an embodiment of the present invention.
Fig. 3 is a schematic flow chart of a method for determining that a vehicle enters the target turnout after being manually taken over according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of a lane sequence according to an embodiment of the present invention.
Fig. 5 is a schematic diagram of a target turnout provided in an embodiment of the present invention.
Fig. 6 is another schematic diagram of a target turnout junction according to an embodiment of the present invention.
Fig. 7 is a schematic structural diagram of a simulation scene constructing device for a crossing accident according to an embodiment of the present invention.
Fig. 8 is another schematic structural diagram of the simulation scene constructing apparatus for a crossing accident according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive step based on the embodiments of the present invention, are within the scope of protection of the present invention.
As shown in fig. 1, an embodiment of the present invention provides a method for constructing a simulation scene of a missed crossing accident, which at least includes the following steps:
s1: acquiring drive test data of a vehicle running in a running area; the drive test data comprises vehicle lane change information and manual take-over information.
S2: when the vehicle is judged to meet the preset missed turnout junction judgment condition according to the vehicle lane change information and the manual takeover information, extracting first time when the vehicle starts to enter a lane change state and second time when the vehicle drives into a target turnout junction after being manually taken over.
S3: and intercepting the road test data from the first time to the second time in the road test data to obtain missed turnout road test data corresponding to the missed turnout accident.
And S4, constructing a simulation scene of the missed turnout accident according to the missed turnout road test data.
For the step S1, enabling the vehicle to run in the running path area set in advance, a security officer sits in the vehicle to take over the vehicle at any time so as to obtain the drive test data collected by the vehicle; the drive test data can be acquired through the log in the vehicle control module in an actual scene; the drive test data includes vehicle lane change information, for example: whether the vehicle is in a lane changing state, the lane changing direction of the vehicle when changing the lane each time, the starting time of the vehicle when changing the lane each time, the ending time of the vehicle when changing the lane each time, the lane information before the vehicle changes the lane each time and the lane information after the vehicle changes the lane each time; in addition, the drive test data also includes manual takeover information of the vehicle, such as: whether the vehicle is in a taken over state, a start time each time it is taken over, and a time each time it is finished taking over.
For step S2, a preset missed turnout determination condition is first described, and in a preferred embodiment, the preset missed turnout determination condition includes: the vehicle is manually taken over, the vehicle is in a lane changing state before being manually taken over, a lane capable of passing exists in the lane changing direction corresponding to the vehicle before being manually taken over, the vehicle completes at least one lane changing after being manually taken over, the lane changing direction corresponding to the vehicle after being manually taken over is consistent with the lane changing direction corresponding to the vehicle before being manually taken over, and the vehicle drives into the target turnout junction after being manually taken over.
The invention adopts a decision tree algorithm to carry out condition verification according to the vehicle lane change information and the manual takeover information to identify whether the vehicle misses the turnout junction accident or not in the process of road test, and concretely, the invention can learn whether the vehicle has the lane change event or not, the lane change direction of each lane change event, the information of the lane where the lane is located before and after the lane change, the start time of the lane change and the end event from the vehicle lane change information in the process of road test. And the vehicle can be informed of whether the vehicle is manually taken over or not and the starting time and the ending time of the taking over in the process of the drive test from the manual taking over information. Whether the set missed turnout judgment condition is met by the vehicle in the road test process can be judged one by one according to the information, if the condition is met, the vehicle cannot drive to the target turnout by means of automatic driving in the road test process, the vehicle can drive to the target turnout only under the condition of manual taking over, and therefore the missed turnout accident occurs in the road test process.
And after the vehicle misses the turnout accident in the road test process, extracting the starting time of the vehicle entering the lane change state for the first time to obtain the first time, and extracting the time of the vehicle entering the target turnout after being manually taken over to obtain the second time.
It should be noted that, the target turnout may manually select a turnout of the turnout in advance according to a map corresponding to a vehicle driving area as a target turnout in the road test process, and then when it is determined whether the vehicle is driven into the target turnout after being manually taken over, the vehicle position information and the position information of the target turnout calibrated in advance are compared, and when the vehicle position is consistent with the position of the turnout calibrated in advance and the vehicle is in a taken over state, it is determined that the vehicle is driven into the target turnout after being manually taken over;
in other embodiments of the present invention, the target turnout may be determined according to the road test data, and it may be determined that the vehicle has entered the target turnout after being manually taken over.
In a preferred embodiment, as shown in fig. 3, the method for determining that the vehicle drives into the target turnout after being manually taken over at least comprises the following steps:
and S21, acquiring the lane change direction of each lane change event of the vehicle and a first lane sequence of the vehicle before lane change from the lane change state information of the vehicle.
S22, judging whether a necessary lane change event meeting preset lane change conditions exists according to the lane change direction of each lane change event and the first lane sequence; if the lane change direction is the lane change to the right, taking the rightmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located; if the lane change direction is the lane change to the left, taking the leftmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located; if the lane change direction is lane change to the right, the preset lane change condition includes: the first lane sequence is a rightmost lane sequence of lanes where vehicles are located before lane changing, the right lane sequence adjacent to the first lane sequence is not the rightmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing; if the lane changing direction is lane changing to the left, the preset lane changing condition includes: the first lane sequence is a leftmost lane sequence of lanes where vehicles are located before lane changing, the left lane sequence adjacent to the first lane sequence is not the leftmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing.
And S23, extracting the position information of the target turnout according to the lane sequence of the target turnout, and judging that the vehicle drives into the target turnout after being manually taken over when the position of the vehicle is consistent with the position of the target turnout and the vehicle is in a taken over state.
In step S21, a lane change event is generated every time the vehicle changes lanes, and the lane change directions of all the lane change events and the lane sequence (i.e., the above lane sequence) where the vehicle before lane change is located in the drive test process are obtained from the lane change state information.
First, a lane sequence is explained: in an actual road condition, in a section of driving area, a plurality of lanes exist, each lane is composed of a plurality of lane sequences formed by a lane trunk and a plurality of lane branches, as shown in fig. 4, two lanes, a lane a and a lane B exist in the figure; the lane A comprises a lane trunk 1, a lane branch 11, a lane branch 12 and a lane branch 13; the lane trunk 1 and the lane branches 11 form a lane sequence A1, the lane trunk 1 and the lane branches 12 form a lane sequence A2, the lane trunk 1 and the lane branches 13 form a lane sequence A3, and the lane sequences A1, A2 and A3 form a lane A; similarly, the lane B includes a lane trunk 2, a lane branch 21 and a lane branch 22, the lane trunk 2 and the lane branch 21 form a lane sequence B1, the lane trunk 2 and the lane branch 22 form a lane sequence B2, and the lane sequence B1 and the lane sequence B2 form the lane B.
If the vehicle is located at the position of the lane trunk 1 before lane changing, the first lane sequence may be any one of lane sequences A1, A2, and A3 at this time; assuming that the vehicle is located at the lane branch 13 before lane change, the first lane sequence is the lane sequence A3 at this time, and so on.
For step S22, in the present invention, the preset lane change condition is divided into two cases according to the lane change condition of the vehicle:
first, the lane change direction is a lane change to the right, and then the preset lane change condition includes: condition 1: the first lane sequence is a rightmost lane sequence of lanes where vehicles are located before lane changing; condition 2: the right lane sequence adjacent to the first lane sequence is not the rightmost lane sequence of the lanes where the vehicles are located after lane changing; and 3, the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing. If the above 3 conditions are satisfied simultaneously, the lane change event is determined to be a necessary lane change event. As shown in fig. 5, when the vehicle changes lane from point d to point e to complete a lane change event, and point d is located on the lane trunk 1, as can be seen from the description in step S21, the first lane sequence corresponding to the lane change event may be any one of the lane sequences A1, A2, and A3, and when the first lane sequence is the lane sequence A3, the first lane sequence is the rightmost lane sequence belonging to the lane a, so that the condition 1 can be satisfied. The right lane sequence adjacent to the lane sequence A3 is a lane sequence B1 composed of a lane trunk 2 and lane branches 21, and the lane sequence B1 does not belong to the rightmost lane sequence of the lane B after lane change (the rightmost lane sequence of the lane B is a lane sequence B2), so that the condition 2 is met. And the first lane sequence is not the lane sequence B1 or the lane sequence B2, and thus the condition 3 is satisfied. Thereby determining that the lane-change event when the vehicle changes lane from point d to point e is a necessary lane-change event.
Second, if the lane change direction is lane change to the left, the preset lane change condition includes: condition 4: the first lane sequence is a leftmost lane sequence of lanes where vehicles are located before lane changing; condition 5: the left lane sequence adjacent to the first lane sequence is not the leftmost lane sequence of the lanes where the vehicles are after lane change; condition 6 the first lane sequence does not belong to any lane sequence of lanes in which the vehicle is located after lane change. If the 3 conditions are met, the lane change event is judged to be a necessary lane change event. As schematically shown in fig. 6, there are two lanes, lane a and lane C; the lane A comprises a lane trunk 1, a lane branch 11, a lane branch 12 and a lane branch 13; the lane trunk 1 and the lane branches 11 form a lane sequence A1, the lane trunk 1 and the lane branches 12 form a lane sequence A2, the lane trunk 1 and the lane branches 13 form a lane sequence A3, and the lane sequences A1, A2 and A3 form a lane A; similarly, the lane C includes a lane trunk 3, a lane branch 31 and a lane branch 32, the lane trunk 3 and the lane branch 31 form a lane sequence C1, the lane trunk 3 and the lane branch 32 form a lane sequence C2, and the lane sequence C1 and the lane sequence C2 form the lane C.
When the vehicle changes the lane from point d to point f to complete a lane change event, and point d is located in the lane trunk 1, the first lane sequence corresponding to the lane change event may be any one of the lane sequences A1, A2, and A3, and when the first lane sequence is the lane sequence A1, the first lane sequence is the leftmost lane sequence belonging to the lane a, so that the condition 4 can be satisfied. The left lane sequence adjacent to the lane sequence A1 is a lane sequence C1 formed by a lane trunk 3 and lane branches 31, and the lane sequence C1 does not belong to the leftmost lane sequence of the lane C after lane change (the leftmost lane sequence of the lane C is a lane sequence C2), so that the condition 5 is met. And the first lane sequence is not the lane sequence C1 or the lane sequence C2, and thus the condition 3 is satisfied. Thereby determining that the lane change event when the vehicle changes lane from point d to point f is a necessary lane change event.
According to the above-mentioned illustration, when it is determined that the necessary lane change event exists, if the lane change direction of the necessary lane change event is a lane change to the right, the rightmost lane sequence of the lanes where the vehicles are located after the lane change is the lane sequence where the target turnout is located; as shown in fig. 5, at this time, the lane sequence B2 is the lane sequence in which the target turnout is located, and if the lane change direction of the necessary lane change event is a lane change to the left, the leftmost lane sequence of the lane in which the vehicle is located after the lane change is the lane sequence in which the target turnout is located; as shown in fig. 6, the lane sequence C2 is the lane sequence of the target turnout.
And if the necessary lane change event does not exist, directly judging that the vehicle does not drive into the target turnout.
In step S23, the lane sequence in which the target road junction is located is automatically identified according to the above steps, the intersection in which the lane branch is connected to the lane trunk in the lane sequence in which the target road junction is located is used as the target road junction, schematically shown in fig. 5, the intersection G in which the lane trunk 2 is connected to the lane branch 22 is used as the target road junction, the position information of the target road junction is extracted and compared with the position information of the vehicle, when it is detected that the position of the vehicle is consistent with the position of the target road junction at a certain moment and the vehicle is in a takeover state, it is determined that the vehicle is driven into the target road junction after being manually taken over, and otherwise, it is determined that the vehicle is not driven into the target road junction after being manually taken over.
By the method shown in fig. 3, the target turnout does not need to be marked in advance, the automation degree of constructing the simulation scene of the accident of passing by the turnout is further improved, and the construction efficiency is improved.
In a preferred embodiment, it is determined that the vehicle has completed a lane change by; assume that the position of the vehicle is (x) e ，y e ) Road center position (x) closest to the vehicle w ，y w ) An orientation angle θ between the vehicle position and a road center point closest to the vehicle; the offset distance offset (e, w) = | (x) of the vehicle relative to its nearest road center point e -x w ，y e -y w ) X (cos θ, sin θ) |; in this embodiment, points are taken from the center line of each lane sequence of the travel area at preset intervals as the respective road center points. Preferably, the points can be taken every 0.5 m;
if t 0 Time of day, offset (e) 0 ，w 0 )＜1m，w 0 .lane is l 0 ；t 1 Time of day, offset (e) 1 ，w 1 )＜1m，w 1 .lane is right(l 0 ) Then at t 0 ～t 1 In time period, the vehicle completes one lane change to the right; wherein, offset (e) 0 ，w 0 ) Is t 0 Time vehicleRelative to the road center point w nearest thereto 0 Offset distance of (w) 0 .lane is l 0 Time t of finger 0 Road center point w nearest to vehicle at all times 0 The lane is l 0 ，offset(e 1 ，w 1 ) Is t 0 At the moment the vehicle is relative to the road centre point w closest thereto 1 Offset distance of, w 1 .lane is right(l 0 ) Time t of finger 1 Road center point w nearest to vehicle at any moment 1 The lane is l 0 Right side lane of the vehicle.
If t 0 Time of day, offset (e) 0 ，w 0 )＜1m，w 0 .lane is l 0 ；t 1 Time of day, offset (e) 1 ，w 1 )＜1m，w 1 .lane is left(l 0 ) Then at t 0 ～t 1 In time period, the vehicle completes one lane change to the left; in the formula w 1 .lane is left(l 0 ) Time t of finger 1 Road center point w nearest to vehicle at any moment 1 The lane is l 0 Left lane of the vehicle.
And S3, intercepting the drive test data of the corresponding time period from the drive test data according to the extracted first time and second time, and obtaining the missed turnout drive test data corresponding to the missed turnout accident.
For step S4, after the corresponding missed turnout road test data is obtained, a simulation scene of the missed turnout accident can be constructed according to the missed turnout road test data.
In a preferred embodiment, when a simulation scene of the missed turnout accident is constructed according to the missed turnout road test data, a position point in a target turnout, which is away from the target turnout by a preset distance, is used as a destination of a simulated vehicle in the simulation scene. Schematically, as shown in fig. 5, a point H which is 100 meters away from the target turnout G may be set as the destination of the simulated vehicle during vehicle simulation in the target turnout 22, so as to ensure that the precondition of the missed turnout accident can be reproduced, and ensure that the driving track of the simulated vehicle in the simulation scene is substantially consistent with the actual driving track of the vehicle in the driving area, thereby better reproducing the missed turnout accident;
in a preferred embodiment, when a simulation scene of the missed road junction accident is constructed according to the missed road junction test data, at least the lane change times and the corresponding lane change directions which need to be executed when the vehicle enters the target road junction from the current lane are counted according to the missed road junction test data, and the lane change times and the lane change directions are used as evaluation indexes of the simulation scene, so that the simulation scene determines whether the simulated vehicle can reproduce the missed road junction accident in the simulation scene according to the evaluation indexes.
For example, as shown in fig. 5, before the vehicle enters the target turnout G, the vehicle changes from the lane trunk 1 to the lane trunk 2, then changes from the lane trunk 2 to the lane trunk 1, and then changes from the lane trunk 1 to the lane trunk 2, so that the lane change is performed for 3 times, but the lane change from the trunk 1 to the target turnout G substantially only needs to perform the lane change from the lane trunk 1 to the lane trunk 2 for 1 time, and at this time, only the lane change from the lane trunk 1 to the lane trunk 2 for the first time is reserved, and the rest of invalid lane changes are removed; and counting lane changing times and lane changing directions after all invalid lane changing is removed, taking the lane changing times and the lane changing directions as evaluation indexes of the constructed simulation scene, then extracting the lane changing times and the lane changing directions of the simulation vehicle in the simulation scene in the simulation process, and comparing the lane changing times and the lane changing directions in the configured evaluation indexes so as to judge whether the missed crossing accident of the simulation vehicle recurs in the simulation scene.
As shown in fig. 2, in a preferred embodiment, the method for constructing a simulation scene of a missed crossing accident further includes:
s5, operating the simulation scene and extracting a simulation operation track of a simulation vehicle in the simulation scene; comparing the simulation running track with an actual running track of a vehicle in the running area, and then determining whether the simulation vehicle can reappear the missed turnout accident in the simulation scene according to a track comparison result to obtain a first simulation evaluation result; determining whether the simulated vehicle can reappear the crossing missing accident in the simulation scene according to the evaluation index to obtain a second simulation evaluation result; and when the first simulation evaluation result is consistent with the second simulation evaluation result, judging that the simulation scene is qualified.
In the embodiment, a track route of the simulated vehicle in the simulation scene is detected through algorithms such as track similarity and the like, the detected track route is compared with an actual running track of an actual vehicle in an actual running area, whether the simulated vehicle repeatedly misses a turnout junction accident in the simulation is detected, and a first simulation evaluation result is obtained; then, further determining whether the simulated vehicle can reappear the missed crossing accident in the simulation scene according to the evaluation index configured in the simulation scene, and obtaining a second simulation evaluation result; if the first simulation evaluation result is consistent with the second simulation evaluation result, the constructed simulation scene is accurate and qualified, and therefore the accuracy of the constructed simulation scene is evaluated.
By implementing the embodiment of the simulation scene construction method for each missed crossing accident, the simulation scene of the missed crossing accident can be automatically constructed, and during the construction, a marker does not need to manually identify and mark the corresponding part of the road test data in the road test data aiming at the scene of the missed crossing accident, so that the human resources are saved, and the construction efficiency of the simulation scene is greatly improved.
On the basis of the embodiment of the method item, the invention correspondingly provides an embodiment of a device item;
as shown in fig. 7, an embodiment of the present invention provides a simulation scene constructing device for a missed crossing accident, including: the system comprises a data acquisition module, a key time point extraction module, a drive test data interception module and a simulation scene construction module;
the data acquisition module is used for acquiring the drive test data of the vehicle running in the running area; the drive test data comprises vehicle lane change information and manual take-over information;
the key time point extraction module is used for extracting first time when the vehicle starts to enter a lane change state and second time when the vehicle drives into a target turnout after being manually taken over when the vehicle is judged to meet a preset missed turnout judgment condition according to the vehicle lane change information and the manual taking-over information;
the road test data intercepting module is used for intercepting the road test data from the first time to the second time in the road test data to obtain missed turnout road test data corresponding to a missed turnout accident;
the simulation scene construction module is used for constructing the simulation scene of the missed turnout accident according to the missed turnout road test data.
Preferably, the preset missed turnout determination condition includes: the vehicle is manually taken over, the vehicle is in a lane changing state before being manually taken over, a lane capable of passing exists in the lane changing direction corresponding to the vehicle before being manually taken over, the vehicle completes at least one lane changing after being manually taken over, the lane changing direction corresponding to the vehicle after being manually taken over is consistent with the lane changing direction corresponding to the vehicle before being manually taken over, and the vehicle drives into the target turnout junction after being manually taken over.
Preferably, the determining that the vehicle enters the target turnout junction after being manually taken over specifically includes:
acquiring the lane change direction of each lane change event of the vehicle and a first lane sequence where the vehicle before lane change is located from the lane change state information of the vehicle;
judging whether a necessary lane change event meeting preset lane change conditions exists or not according to the lane change direction of each lane change event and the first lane sequence;
wherein, if the lane change direction is lane change to the right, the preset lane change condition includes: the first lane sequence is a rightmost lane sequence of a lane where a vehicle is located before lane changing, a right lane sequence adjacent to the first lane sequence is not a rightmost lane sequence of a lane where a vehicle is located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where the vehicle is located after lane changing; if the lane changing direction is lane changing to the left, the preset lane changing condition includes: the first lane sequence is a leftmost lane sequence of lanes where vehicles are located before lane changing, a left lane sequence adjacent to the first lane sequence is not a leftmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing;
if the lane change direction is the lane change to the right, taking the rightmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located; if the lane change direction is the lane change to the left, taking the leftmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located;
and extracting the position information of the target turnout according to the lane sequence of the target turnout, and judging that the vehicle drives into the target turnout after being manually taken over when the position of the vehicle is consistent with the position of the target turnout and the vehicle is in a taken over state.
In a preferred embodiment, the simulation scenario construction module comprises a simulation destination construction unit;
the simulation destination building unit is used for taking a position point in a target turnout, which is away from the target turnout by a preset distance, as a destination of a simulated vehicle in the simulation scene when the simulation scene of the missed turnout accident is built according to the missed turnout test data.
In a preferred embodiment, the simulation scenario construction module further includes an evaluation index construction unit; the evaluation index construction unit is used for counting at least the lane changing times and the corresponding lane changing directions which need to be executed when the vehicle enters the target turnout from the current lane according to the missed turnout road test data when the simulation scene of the missed turnout accident is constructed according to the missed turnout road test data, and taking the lane changing times and the lane changing directions as the evaluation indexes of the simulation scene, so that the simulation scene determines whether the simulated vehicle can reproduce the missed turnout accident in the simulation scene according to the evaluation indexes.
As shown in fig. 8, in a preferred embodiment, the simulation scenario construction apparatus for a missed crossing accident further includes: a simulation scene evaluation module;
the simulation scene evaluation module is used for operating the simulation scene and extracting a simulation operation track of a simulation vehicle in the simulation scene;
comparing the simulation running track with an actual running track of a vehicle in the running area, and then determining whether the simulation vehicle can reappear the missed turnout accident in the simulation scene according to a track comparison result to obtain a first simulation evaluation result;
determining whether the simulated vehicle can reappear the turnout accident in the simulation scene according to the evaluation index to obtain a second simulation evaluation result;
and when the first simulation evaluation result is consistent with the second simulation evaluation result, judging that the simulation scene is qualified.
It should be noted that the above-described apparatus embodiments correspond to the method embodiments of the present invention, and can implement any method for constructing a simulation scenario of a crossing accident, and the above-described apparatus embodiments are merely illustrative, where the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. In addition, in the drawings of the embodiment of the apparatus provided by the present invention, the connection relationship between the modules indicates that there is a communication connection therebetween, and may be specifically implemented as one or more communication buses or signal lines. One of ordinary skill in the art can understand and implement it without inventive effort.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
Claims (9)
1. A simulation scene construction method for a missed turnout accident is characterized by comprising the following steps:
acquiring drive test data of a vehicle running in a running area; the drive test data comprises vehicle lane change information and manual take-over information;
when the vehicle is judged to meet a preset missed turnout junction judgment condition according to the vehicle lane change information and the manual takeover information, extracting first time when the vehicle starts to enter a lane change state and second time when the vehicle drives into a target turnout junction after being manually taken over;
intercepting the road test data from the first time to the second time in the road test data to obtain missed turnout road test data corresponding to a missed turnout accident;
constructing a simulation scene of the missed turnout accident according to the missed turnout road test data;
wherein, the predetermined missed turnout determination condition includes: the vehicle is manually taken over, the vehicle is in a lane changing state before being manually taken over, a passable lane exists in the lane changing direction corresponding to the vehicle before being manually taken over, the vehicle completes at least one lane changing after being manually taken over, the lane changing direction corresponding to the vehicle after being manually taken over is consistent with the lane changing direction corresponding to the vehicle before being manually taken over, and the vehicle drives into the target turnout opening after being manually taken over.
2. The method for constructing a simulation scene of a missed turnout accident according to claim 1, wherein when constructing the simulation scene of the missed turnout accident according to the missed turnout road test data, a position point in a target turnout which is a preset distance away from the target turnout is used as a destination of a simulated vehicle in the simulation scene.
3. The method for constructing a simulation scene of a missed road junction accident according to claim 1, wherein when constructing the simulation scene of the missed road junction accident according to the missed road junction road testing data, the number of times of lane change and the corresponding direction of lane change that the vehicle needs to be performed at least when driving from the current lane to the target road junction are counted according to the missed road junction road testing data, and the number of times of lane change and the direction of lane change are used as evaluation indexes of the simulation scene, so that the simulation scene determines whether the simulated vehicle can reproduce the missed road junction accident in the simulation scene according to the evaluation indexes.
4. The method for constructing the simulation scene of the missed road junction accident of claim 3, further comprising:
operating the simulation scene and extracting a simulation operation track of a simulation vehicle in the simulation scene;
comparing the simulation running track with an actual running track of the vehicle in the running area, and then determining whether the simulation vehicle can reappear the missed crossing accident in the simulation scene according to a track comparison result to obtain a first simulation evaluation result;
determining whether the simulated vehicle can reappear the crossing missing accident in the simulation scene according to the evaluation index to obtain a second simulation evaluation result;
and when the first simulation evaluation result is consistent with the second simulation evaluation result, judging that the simulation scene is qualified.
5. The method for constructing a simulation scene of a missed road junction accident according to any one of claims 2 to 4, wherein the step of judging that the vehicle drives into the target road junction after being manually taken over specifically comprises the following steps:
acquiring a lane changing direction of each lane changing event of the vehicle and a first lane sequence in which the vehicle before lane changing is located from the lane changing state information of the vehicle;
judging whether a necessary lane change event meeting preset lane change conditions exists or not according to the lane change direction of each lane change event and the first lane sequence;
wherein, if the lane change direction is lane change to the right, the preset lane change condition includes: the first lane sequence is a rightmost lane sequence of lanes where vehicles are located before lane changing, the right lane sequence adjacent to the first lane sequence is not the rightmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing; if the lane changing direction is changing lanes to the left, the preset lane changing conditions include: the first lane sequence is a leftmost lane sequence of lanes where vehicles are located before lane changing, a left lane sequence adjacent to the first lane sequence is not a leftmost lane sequence of lanes where vehicles are located after lane changing, and the first lane sequence does not belong to any lane sequence of lanes where vehicles are located after lane changing;
if the lane change direction is the lane change to the right, taking the rightmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located; if the lane change direction is the lane change to the left, taking the leftmost lane sequence of the lanes where the vehicles are located after the lane change in the necessary lane change event as the lane sequence where the target turnout is located;
and extracting the position information of the target turnout according to the lane sequence of the target turnout, and judging that the vehicle drives into the target turnout after being manually taken over when the position of the vehicle is consistent with the position of the target turnout and the vehicle is in a taken over state.
6. A simulation scene construction device for a missed turnout accident is characterized by comprising the following steps: the system comprises a data acquisition module, a key time point extraction module, a drive test data interception module and a simulation scene construction module;
the data acquisition module is used for acquiring the drive test data of the vehicle running in the running area; the drive test data comprises vehicle lane change information and manual take-over information;
the key time point extraction module is used for extracting first time when the vehicle starts to enter a lane change state and second time when the vehicle drives into a target turnout after being manually taken over when the fact that the vehicle meets a preset missed turnout judgment condition is judged according to the vehicle lane change information and the manual taking over information; wherein, the default missed turnout judging condition comprises: the vehicle is manually taken over, the vehicle is in a lane changing state before being manually taken over, a passable lane exists in the corresponding lane changing direction of the vehicle before being manually taken over, the vehicle completes at least one lane changing after being manually taken over, the corresponding lane changing direction of the vehicle after being manually taken over is consistent with the corresponding lane changing direction before being manually taken over, and the vehicle drives into the target turnout opening after being manually taken over;
the road test data intercepting module is used for intercepting the road test data from the first time to the second time in the road test data to obtain missed turnout road test data corresponding to a missed turnout accident;
the simulation scene construction module is used for constructing the simulation scene of the missed turnout accident according to the missed turnout road test data.
7. The simulation scene constructing device for the missed road junction accident according to claim 6, wherein the simulation scene constructing module comprises a simulation destination constructing unit;
the simulation destination building unit is used for taking a position point in a target turnout, which is away from the target turnout by a preset distance, as a destination of a simulated vehicle in the simulation scene when the simulation scene of the missed turnout accident is built according to the missed turnout test data.
8. The simulation scene constructing device for the missed road junction accident of claim 6, wherein the simulation scene constructing module further comprises an evaluation index constructing unit;
the evaluation index construction unit is used for counting at least the lane changing times and the corresponding lane changing directions which need to be executed when the vehicle enters the target turnout from the current lane according to the missed turnout road test data when the simulation scene of the missed turnout accident is constructed according to the missed turnout road test data, and taking the lane changing times and the lane changing directions as the evaluation indexes of the simulation scene, so that the simulation scene determines whether the simulated vehicle can reproduce the missed turnout accident in the simulation scene according to the evaluation indexes.
9. The apparatus for constructing a simulation scene of a missed road junction accident according to claim 8, further comprising: a simulation scene evaluation module;
the simulation scene evaluation module is used for operating the simulation scene and extracting a simulation operation track of a simulation vehicle in the simulation scene;
comparing the simulation running track with an actual running track of the vehicle in the running area, and then determining whether the simulation vehicle can reappear the missed crossing accident in the simulation scene according to a track comparison result to obtain a first simulation evaluation result;
determining whether the simulated vehicle can reappear the crossing missing accident in the simulation scene according to the evaluation index to obtain a second simulation evaluation result;
and when the first simulation evaluation result is consistent with the second simulation evaluation result, judging that the simulation scene is qualified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111108101.7A CN114067610B (en) | 2021-09-22 | 2021-09-22 | Simulation scene construction method and device for missing turnout accident |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111108101.7A CN114067610B (en) | 2021-09-22 | 2021-09-22 | Simulation scene construction method and device for missing turnout accident |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114067610A CN114067610A (en) | 2022-02-18 |
| CN114067610B true CN114067610B (en) | 2023-03-17 |
Family
ID=80233961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111108101.7A Active CN114067610B (en) | 2021-09-22 | 2021-09-22 | Simulation scene construction method and device for missing turnout accident |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114067610B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114596470B (en) * | 2022-03-31 | 2024-06-25 | 广州文远知行科技有限公司 | Merging dangerous scene classification system, method, equipment and medium |
| CN114581623B (en) * | 2022-04-08 | 2024-03-05 | 广州文远知行科技有限公司 | Simulation scene generation method and device, storage medium and computer equipment |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107153363B (en) * | 2017-05-08 | 2020-11-03 | 百度在线网络技术（北京）有限公司 | Simulation test method, device, equipment and readable medium for unmanned vehicle |
| CN108931927B (en) * | 2018-07-24 | 2019-07-30 | 百度在线网络技术（北京）有限公司 | The creation method and device of unmanned simulating scenes |
| CN111199088B (en) * | 2018-10-31 | 2023-06-20 | 百度在线网络技术（北京）有限公司 | Method and device for reproducing scene data |
| CN111199083B (en) * | 2018-10-31 | 2023-12-15 | 阿波罗智能技术(北京)有限公司 | Method and device for reproducing accident data of unmanned vehicle |
| CN110160804B (en) * | 2019-05-31 | 2020-07-31 | 中国科学院深圳先进技术研究院 | Test method, device and system for automatically driving vehicle |
| CN111353221A (en) * | 2020-02-24 | 2020-06-30 | 上海商汤临港智能科技有限公司 | Automatic driving simulation method and device, electronic equipment and storage medium |
| CN111967163B (en) * | 2020-08-27 | 2021-12-31 | 腾讯科技（深圳）有限公司 | Vehicle simulation control method and device, computer equipment and storage medium |
| WO2022141506A1 (en) * | 2020-12-31 | 2022-07-07 | 华为技术有限公司 | Method for constructing simulation scene, simulation method and device |
| CN113071515B (en) * | 2021-03-31 | 2022-07-05 | 广州文远知行科技有限公司 | Movable carrier control method, device, movable carrier and storage medium |
-
2021
- 2021-09-22 CN CN202111108101.7A patent/CN114067610B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN114067610A (en) | 2022-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2017157119A1 (en) | Method and device for identifying abnormal behavior of vehicle | |
| CN114067610B (en) | Simulation scene construction method and device for missing turnout accident | |
| KR101995813B1 (en) | Apparatus for providing traffic information based on image and method thereof | |
| CN105279971B (en) | A kind of information of vehicles matching process, system and supervising device | |
| CN108734129A (en) | mobile phone and vehicle location analysis method and system | |
| CN109212513A (en) | Multiple target between radar data transmitting, data fusion and localization method is continuously tracked | |
| CN106548629B (en) | Traffic violation detection method and system based on data fusion | |
| CN107369335B (en) | A kind of reverse vehicle searching system and method based on ultrasonic listening technology | |
| CN111275975B (en) | Method and device for acquiring intersection turning flow data and storage medium | |
| CN105528894B (en) | The method and apparatus of O＆M detection is carried out to monitoring frontend based on trace information | |
| CN105741553A (en) | Method for identifying parking road segment in vehicle track based on dynamic threshold | |
| CN106571038A (en) | Method for fully automatically monitoring road | |
| CN110362557A (en) | A kind of miss path restorative procedure based on machine learning and license plate identification data | |
| CN107610511A (en) | A kind of reverse vehicle searching system and method for merging ultrasonic listening and IMAQ | |
| CN110175654A (en) | A kind of update method and system of track road sign | |
| CN106340190A (en) | Method, device and system for determining traffic light timing | |
| CN106128126A (en) | Plane cognition technology is utilized to reduce the method and system meeting red light number of times rush hour | |
| CN110174116A (en) | Generate method, apparatus, equipment and the computer storage medium of navigation casting content | |
| CN114973659A (en) | Method, device and system for detecting indirect event of expressway | |
| CN112937598A (en) | Method and device for evaluating and/or adjusting vehicle driving assistance or automatic driving functions | |
| CN112150831A (en) | Big data-based smart city street lamp information control processing method | |
| CN112071066A (en) | Automatic driving vehicle testing method, judging method and device | |
| CN110164137A (en) | Based on bayonet to the recognition methods of the fake license plate vehicle of running time and system, medium | |
| CN110675631A (en) | Traffic flow traceability analysis method and system | |
| CN109920253A (en) | Parking charge method and system based on vehicle regular visit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |