CN114495489A - Method for generating topological connection relation of road junction lanes - Google Patents
Method for generating topological connection relation of road junction lanes Download PDFInfo
- Publication number
- CN114495489A CN114495489A CN202111662498.4A CN202111662498A CN114495489A CN 114495489 A CN114495489 A CN 114495489A CN 202111662498 A CN202111662498 A CN 202111662498A CN 114495489 A CN114495489 A CN 114495489A
- Authority
- CN
- China
- Prior art keywords
- lane
- road section
- intersection
- calculating
- list
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
- G01C21/30—Map- or contour-matching
- G01C21/32—Structuring or formatting of map data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computational Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Analysis (AREA)
- Evolutionary Biology (AREA)
- Operations Research (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Probability & Statistics with Applications (AREA)
- Molecular Biology (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Automation & Control Theory (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a method for generating topological connection relation of crossing lanes, which relates to the technical field of road traffic.A vehicle track data is collected and subjected to statistical analysis on the data to automatically obtain all topological connections and occurrence probabilities thereof, the topological connections with higher occurrence probabilities can be selected as the basis for generating lane connection lines on a map, the driving habits of human drivers are fully considered, the method is more reasonable than the method for generating the lane topology by mechanically using simple fixed rules, other traffic participants can easily predict the intention and the route of the current vehicle, safe driving is facilitated, traffic accidents are reduced, and the crossing traffic efficiency is improved.
Description
Technical Field
The invention relates to the technical field of road traffic, in particular to a method for generating topological connection relation of crossing lanes.
Background
The high-precision map is a necessary link of the existing unmanned technology, the lane data is the core content of the high-precision map, and the vehicle usually needs to automatically drive along a series of coordinate points given by the lane center line data of the high-precision map as a preset track, so that the lane center line not only needs to be smooth in curvature, but also needs to be reasonable in topological relation. At the intersection, because there is no clear ground marking to determine the topological relation, it is usually necessary to manually judge and add and generate the topological relation of the lane, because the topological relation of the intersection is complex, the efficiency of manual generation is very low, and the judging rule is fuzzy, the judging methods of different people are inconsistent, which easily causes the data quality inconsistency, and if the fixed rule is uniformly set, the data quality consistency can be ensured, but many topological relations are not reasonable, which results in that when the vehicle actually runs, it is not beneficial to safely, stably and efficiently running the vehicle. In order to make the vehicle run more stably and safely at the intersection, the center line of the lane not only needs to meet the requirements of traffic regulations and continuous curvature, but also needs to meet the driving habits of human beings as much as possible, so that other traffic participants can predict the current intention and route of the vehicle better, unnecessary traffic accidents are reduced, and the traffic efficiency is improved.
At present, the intersection topological relations are mainly marked by ground arrows, and then some simple topological connection rules are designed for automatic calculation, or manually and manually adding the intersection topological relations one by one. Some technologies use vehicle track information for clustering, but are mainly used for generating common lanes, are not used for generating topological relations inside intersections, and only generate connection relations of lanes inside intersections where tracks are located. The road junction topological relation is determined through the ground arrow marks, and the defect that the lane topology cannot be determined when a plurality of small road junctions or special road junctions have no ground arrow marks or lack part of ground arrow marks is overcome. And the intersection lane topology is generated by artificially setting some fixed rules, some lane topologies are not reasonable, and the probability of occurrence in actual driving is low, so that after lane connecting lines are generated, when a vehicle drives along the connecting lines, other traffic participants are difficult to predict the intention of the current vehicle, and traffic accidents are easily caused.
Disclosure of Invention
Aiming at the technical problems, the invention overcomes the defects of the prior art and provides a method for generating topological connection relation of crossing lanes, which comprises the following steps:
first, intersection lane form parameterization
For each intersection, determining whether a lane enters or leaves the intersection according to the lane direction, dividing the lane into an outgoing lane and an incoming lane, forming a road section by the lanes which are adjacent to each other in the same lane direction and are arranged side by side, and then dividing the road section into an outgoing road section and an incoming road section;
when the topological relation of the lanes is generated at each intersection, the topological relation between each driven road section and all driven road sections is calculated in sequence, whether topological connection is generated between each lane is determined, and after the calculation of all driven road sections is completed, all topological connections in the whole intersection are determined;
calculating the relation between single driving-out road section and driving-in road section
For a certain outgoing road section, sequentially finding all incoming road sections along the anticlockwise direction or the clockwise direction in the intersection from the outgoing road section to obtain a list, and setting the list as an incoming road section list;
normalizing the parameters for the known driving-out road section and the driving-in road section list;
sequentially calculating the topological connection of the lanes between all the outgoing road sections and the corresponding incoming road section list in the intersection to obtain the topological connection of all the lanes in the whole intersection;
collecting vehicle track information at the intersection, converting the vehicle track information into lane topological connection information, counting the occurrence frequency of each lane topological connection, and calculating the probability of each lane topological connection
For a certain intersection, collecting vehicle track information, after obtaining the vehicle track information, calculating an entrance point and an exit point of each track, and respectively calculating which lane the 2 points belong to according to coordinates, thereby obtaining the topological connection of the lanes corresponding to the current track;
after a large number of tracks are collected at the same intersection, counting the number of times of topological connection of each lane, then calculating the number of times of topological connection of all lanes of the same exiting lane as the total number, and calculating the probability of the topological connection of each lane, namely dividing the number of times of topological connection of each lane by the corresponding total number;
fourthly, the method one: carrying out parameter discretization on the data set, directly inquiring a required result in the data set, and carrying out result calculation on any unknown intersection after processing and data collection;
the second method comprises the following steps: and training the neural network model by using the data set, calculating a required result by using the trained neural network model, and calculating the result of any unknown intersection after processing and data collection.
The technical scheme of the invention is further defined as follows:
the method for generating the topological connection relationship of the road junction lane comprises the following steps of:
taking the end point of the lane sidelines of all lanes in the current driven road section, and performing straight line fitting to obtain a straight line, wherein the angle of the straight line is set as the direction of the lanes from left to right, the included angle between the straight line and the positive direction of the x axis is taken as the direction of 0 degree, and the included angle is set as v0 and taken as the transverse angle of the driven road section;
calculating the intersection point between the straight line and the extension line of the leftmost lane boundary, and setting the intersection point as P0(x0, y0) as a reference point of the driving-out road;
calculating the direction angles of all the lane sidelines at the end point, namely calculating the angle formed between a point before the end point and the end point for each sideline, and then averaging all the obtained direction angles to obtain a longitudinal angle h0 of the outgoing road section;
calculating a lane direction list, and setting a lane direction value: using 0 to represent no direction, 1 to represent straight going, 2 to represent left turning, 4 to represent right turning, 8 to represent turning around, calculating a composite value for a plurality of directions, setting a lane direction value for each lane according to ground marking lines and road marking conditions, and then obtaining a lane direction list { t0, t1, t2, } of the current whole driving-out road section;
calculating a lane type list, and setting a lane type value: 0 represents an unknown type, 1 represents a motor vehicle lane, 2 represents a non-motor vehicle lane, 4 represents a bus lane, a composite value can be set, a lane type value is set for each lane according to the actual situation, and then a lane direction list { c0, c1, c2, } of the current whole driving-out road section is obtained;
and calculating a lane width list, calculating a lane width value at the end point of each lane, setting the lane width value as w in meters, and obtaining a lane width list { w0, w1, w2, } of the current whole outgoing road section.
The method for generating the topological connection relationship of the road junction lane comprises the following steps:
calculating the transverse angle v1 of the driving road section, wherein the method is similar to the method for calculating the transverse angle v0 of the driving road section, and the difference is that all lane starting points are taken to perform straight line fitting to obtain a straight line;
calculating a reference point P1 of the driving-in road section in the same way as the reference point P0 of the driving-out road section;
calculating the longitudinal angle h1 of the driving road section, wherein the method is similar to the calculation method of the h1 in the driving road section, but the difference is that the angle at the starting point is calculated, and then the average value is calculated;
calculating a lane direction list, wherein the lane directions of the driving road sections are uniformly set to be 0, namely, no direction exists;
calculating a lane type list which is the same as the driven road section;
the lane width list is calculated, similarly to the above-described driven-out section, except that the lane width at the starting point is calculated.
The method for generating the topological connection relationship of the intersection lane comprises the following steps: for the outgoing road segment, P0 is translated to the origin of coordinates (0, 0), v0 is rotated to 0 degrees, the same rotation is made for h0, then the same translation is made for all reference points in the incoming road segment list, the same rotation is made for all lateral and longitudinal angles, and all angles are set to 0 to 360 degrees, or-180 to 180 degrees.
The method for generating the topological connection relationship of the intersection lane comprises the following steps: and (3) calculating the intersection points between the track and the transverse straight lines obtained by fitting in the step 1, calculating the intersection points between all lane sidelines and the transverse straight lines, calculating the relation between the track intersection points and the lane sideline intersection points, and calculating which two lane sideline intersection points the track intersection points belong to, namely which lane.
The method for generating the topological connection relationship of the road junction lane comprises two methods for collecting vehicle track information: the method comprises the following steps that firstly, a vehicle is provided with a positioning device, and various sensors including a satellite positioning receiver, a gyroscope, an accelerometer, a wheel speed meter, a camera and a laser radar help to perform self positioning to obtain vehicle track information of the vehicle, tracks of different vehicles are collected together, then the coordinate range of a current intersection is known, and which sections of tracks are in the current intersection are calculated to obtain vehicle track information corresponding to each intersection; and secondly, performing mobile measurement through a sensor arranged on the vehicle or performing fixed measurement through a sensor arranged on a roadside unit, wherein the sensor comprises a camera, a laser radar and a millimeter wave radar, calculating the track information of other vehicles on the road, collecting the track information together, and then calculating the vehicle track information corresponding to each intersection.
The method for generating the topological connection relationship of the intersection lane comprises the following steps: carrying out parameter discretization on the data set, directly inquiring a required result in the data set, and carrying out result calculation on any unknown intersection after processing and data collection:
discretizing the parameterized result of the lane form of the intersection;
for an intersection data set formed by collecting data of a large number of different intersections, constructing all (an outgoing road section and an incoming road section list) by using corresponding discretization parameters, then counting the total times of all the corresponding road topology connections (the outgoing road section and the incoming road section list) which are completely the same, and calculating the total probability of each road topology connection;
setting a probability threshold value T, deleting the topological connection of the lane when the probability value of the topological connection of the lane in the data set is smaller than T, then updating the data set to obtain a new corresponding relation table between the discretization input parameters (the road section is driven out, the road section is driven into) and the output parameters;
calculating each road junction (an outgoing road section and an incoming road section list) in the same way for any unknown road junction, discretizing the same parameters of the road junctions, then, matching output parameters corresponding to the completely same input parameters in a data set to directly obtain a lane topological relation list, and then, automatically generating lane connecting lines inside the road junction directly according to the obtained lane topological relation when a map is made; when input parameters are matched, if the identical input parameters are not matched in the data set, the output parameters corresponding to the data with the minimum distance value of the current input parameters in the data set can be calculated as a result through the input parameter distance value calculating module;
inputting a parameter distance value calculation module, namely calculating the distance d between two (an outgoing road section and an incoming road section list), setting a fixed rule if the distance d is (the outgoing road section 1 and the incoming road section list 1) and (the outgoing road section 2 and the incoming road section list 2), respectively comparing the difference between the outgoing road section 1 and the outgoing road section 2 with the difference between the incoming road section list 1 and the incoming road section list 2, quantifying the difference, and adding the two values together to obtain a distance value;
the second method comprises the following steps: training a neural network model by using a data set, calculating a required result by using the trained neural network model, and calculating the result of any unknown intersection by using the following steps after processing and data collection:
according to the obtained (driving-out road section and driving-in road section list) as input parameters, a corresponding lane topological relation list { lane topological connection 1, lane topological connection 2, lane topological connection 3,. to. } and probability values { p1, p2, p3,. to. } serve as output parameters, a training data set is constructed, a neural network model is used, the neural network model is trained by the data set, and the trained neural network model is obtained;
carrying out the same parameterization calculation on any unknown intersection to obtain input parameters (a road section is driven out and a road section is driven in) and inputting the input parameters into the neural network model, so that output values can be obtained from the model, wherein the output values comprise a lane topological relation list { lane topological connection 1, lane topological connection 2, lane topological connection 3, } and probability values { p1, p2, p3, };
and setting a probability threshold T, deleting the topological connection of the lane in the output result when the probability value of the topological connection of the lane is less than T, wherein the rest result is a final topological relation list of the lane, and then automatically generating the lane connection line in the intersection directly according to the obtained topological relation of the lane when a map is made.
The method for generating the topological connection relationship of the intersection lane comprises the following steps: for each parameter (exit route, entry route list), all the transverse angles V, longitudinal angles H, reference points P (X, Y) and lane widths W in the parameter, discretizing the 4 parameters, setting a uniform discretization step size step _ V, step _ H, step _ xy and step _ W, and then obtaining 5 integer values, namely discretization transverse angle V ═ rounding (V/step _ V), discretization longitudinal angle H ═ rounding (H/step _ H), discretization reference point coordinate X ═ rounding (X/step _ xy), Y ═ rounding (Y/step _ xy), discretization lane width W ═ rounding (W/step _ W).
The method for generating the topological connection relationship of the road junction lanes comprises the following steps of firstly comparing the number of road sections in a driving-in road section list 1 and a driving-in road section list 2, if the number of the road sections is equal, carrying out one-to-one correspondence, then carrying out comparison between the two road sections one by one, calculating the difference, and carrying out quantification.
According to the method for generating the topological connection relation of the lanes at the intersection, when the width values of some lanes in the driving-out road section or the driving-in road section are too large, a certain rule is manually set for lane segmentation; or clustering is carried out through the collected historical track information of the current position, whether the current lane is divided or not is judged, and the dividing proportion is determined.
The method for generating the topological connection relationship of the road junction lanes comprises the steps that for some small road junctions or small fork junctions, no obvious marked line or mark exists to determine the direction of the current road section, at the moment, the historical track information of vehicles at the current position is called for judgment, if the current position does not have the historical track information of the vehicles, the judgment cannot be carried out, and the current road section can be considered to be bidirectional.
The invention has the beneficial effects that:
(1) according to the invention, by collecting vehicle track data and carrying out statistical analysis on the data, all topological connections and occurrence probabilities thereof are automatically obtained, the topological connection with higher occurrence probability can be selected as a basis for generating lane connection lines on a map, the driving habits of human drivers are fully considered, the method is more reasonable than the method of mechanically generating the lane topology by using a simple fixed rule, other traffic participants can easily predict the intention and the route of the current vehicle, safe driving is facilitated, traffic accidents are reduced, and the crossing traffic efficiency is improved;
(2) according to the method, vehicle track data are collected, and after a series of processing, an intersection data set is constructed, so that not only can the topological relation of the intersection where the track is located be generated, but also the topological relation of an unknown intersection be generated, the intersection lane connection relation can be automatically constructed, the manufacturing efficiency of map lane connection line data at the intersection is greatly improved, and the road at the intersection is manufactured more reasonably;
(3) the invention can obtain all topological relations and the occurrence probability of the topological relations for any unknown intersection, and when other vehicles appear near the intersection, the future track can be predicted by using the result obtained by the method to obtain a plurality of possible tracks and the probability of the possible tracks, so that the self vehicle can automatically predict the intention of other vehicles in advance, the self vehicle can smoothly run at the intersection, and the passing efficiency and the running safety of the self vehicle can be effectively improved.
Drawings
FIG. 1 is a graph illustrating the relationship between a single outbound road segment and all inbound road segments;
FIG. 2 is a parameterization of an outgoing road segment;
FIG. 3 is a parameterization of an incoming road segment;
fig. 4 shows two topological relationship generation methods after parameterization of the intersection lane form.
Detailed Description
The method for generating the topological connection relation of the lanes at the intersection comprises the following steps:
first, intersection lane form parameterization
For each intersection, determining whether a lane enters or leaves the intersection according to the lane direction, dividing the lane into an outgoing lane and an incoming lane, forming a road section by the lanes which are adjacent to each other in the same lane direction and are arranged side by side, and then dividing the road section into an outgoing road section and an incoming road section;
when the topological relation of the lanes is generated at each intersection, the topological relation between each driven road section and all driven road sections is calculated in sequence, whether topological connection is generated between each lane is determined, and after the calculation of all driven road sections is completed, all topological connections in the whole intersection are determined.
The method for parameterizing the driven road section comprises the following steps:
taking the end point of the lane sideline of all lanes in the currently driven-out road section, and performing straight line fitting (the straight line fitting algorithm can use a least square method and other fitting algorithms), wherein the angle of the obtained straight line is set as the direction of the lane from left to right, the included angle between the direction and the positive direction of the x axis as the 0-degree direction is set as v0, and the included angle is used as the transverse angle of the driven-out road section;
calculating the intersection point between the straight line and the extension line of the leftmost lane boundary, and setting the intersection point as P0(x0, y0) as a reference point of the driving-out road;
calculating the direction angles of all the lane sidelines at the end point, namely calculating the angle formed between a point before the end point and the end point for each sideline, and then averaging all the obtained direction angles to obtain a longitudinal angle h0 of the outgoing road section;
calculating a lane direction list, and setting a lane direction value: using 0 to represent no direction, 1 to represent straight going, 2 to represent left turning, 4 to represent right turning, 8 to represent turning around, a composite value may be calculated for multiple directions, for example, 10 to represent that the lane may turn left and turn around, for example, 11 to represent that the lane may turn straight, turn left and turn around, a lane direction value may be set for each lane according to the ground marking and the road marking, and then a lane direction list { t0, t1, t2, }, taking fig. 2 as an example, which is {10, 1, 4 };
calculating a lane type list, and setting a lane type value: the unknown type is represented by 0, the motor lane is represented by 1, the non-motor lane is represented by 2, the bus lane is represented by 4 (any other type can be added later, and the power n of 2 is used as a value), and similarly, a composite value can be set, for example, 3 represents a common lane for the motor vehicle and the non-motor vehicle. Setting a lane type value for each lane according to actual conditions, and then obtaining a lane direction list { c0, c1, c2, } of the current whole outgoing road section, which is {1, 1, 1} by taking fig. 2 as an example;
and calculating a lane width list, calculating a lane width value at the end point of each lane, setting the lane width value as w in meters, and then obtaining a lane width list { w0, w1, w2,. }, taking fig. 2 as an example, namely {3.30, 3.40, 3.30}, of the current whole outgoing road section.
The method for parameterizing the driving road section comprises the following steps:
calculating the transverse angle v1 of the driving road section, wherein the method is similar to the method for calculating the transverse angle v0 of the driving road section, and the difference is that all lane starting points are taken to perform straight line fitting to obtain a straight line;
calculating a reference point P1 of the driving-in road section in the same way as the reference point P0 of the driving-out road section;
calculating the longitudinal angle h1 of the driving road section, wherein the method is similar to the calculation method of the h1 in the driving road section, but the difference is that the angle at the starting point is calculated, and then the average value is calculated;
calculating a lane direction list, wherein the lane directions of the driving road sections are uniformly set to be 0, namely no direction, and are {0, 0, 0} by taking fig. 3 as an example;
calculating a lane type list which is the same as the driven road section;
the lane width list is calculated, similarly to the above-described driven-out section, except that the lane width at the starting point is calculated.
Calculating the relation between single driving-out road section and driving-in road section
For a certain outgoing road section, sequentially finding all incoming road sections along the anticlockwise direction or the clockwise direction in the intersection from the outgoing road section to obtain a list, and setting the list as an incoming road section list;
for the known outgoing road section and the incoming road section list, carrying out normalization processing on parameters, wherein the normalization method comprises the steps of translating P0 to a coordinate origin (0, 0) for the outgoing road section, rotating v0 to 0 degree, similarly rotating h0, similarly translating all reference points in the incoming road section list, similarly rotating all transverse angles and longitudinal angles, and setting the range of all angles to be 0-360 degrees or-180 degrees;
sequentially calculating the topological connection of the lanes between all the outgoing road sections and the corresponding incoming road section list in the intersection to obtain the topological connection of all the lanes in the whole intersection; assuming that the lane topological connection calculation result of the single outgoing road section and the incoming road section list is a lane topological connection list { lane topological connection 0, lane topological connection 1, lane topological connection 2,. }, the content of one lane topological connection is (outgoing road serial number, incoming road serial number), the outgoing road serial number is the first few lanes from the leftmost side in the outgoing road section, the value is from 0, the incoming road serial number represents the first few incoming road sections in the incoming road section list, the value is from 0, the incoming road serial number represents the first few lanes from the leftmost side in the incoming road section, and the value is from 0;
collecting vehicle track information at the intersection, converting the vehicle track information into lane topological connection information, counting the occurrence frequency of each lane topological connection, and calculating the probability of each lane topological connection
For a certain intersection, two methods are available for collecting vehicle track information; firstly, a vehicle is provided with a positioning device, and various sensors including a satellite positioning receiver, a gyroscope, an accelerometer, a wheel speed meter, a camera and a laser radar help to perform self positioning to obtain self vehicle track information, tracks of different vehicles are collected together, then the coordinate range of the current intersection is known, and which sections of tracks are in the current intersection are calculated to obtain vehicle track information corresponding to each intersection; secondly, performing mobile measurement through a sensor arranged on a vehicle, or performing fixed measurement through a sensor arranged on a roadside unit, wherein the sensor comprises a camera, a laser radar and a millimeter wave radar, calculating the track information of other vehicles on the road, collecting the track information together, and then calculating the vehicle track information corresponding to each intersection;
after obtaining the vehicle track information, calculating an entrance point and an exit point of each track, and respectively calculating which lane the 2 points belong to according to coordinates so as to obtain the topological connection of the lanes corresponding to the current track; the calculation method of the entrance intersection point and the exit intersection point comprises the steps of calculating the intersection point between the track and the transverse straight line obtained by fitting in the step 1, calculating the intersection points between all lane sidelines and the transverse straight line, calculating the relationship between the track intersection point and the lane sideline intersection point, and calculating which two lane sideline intersection points the track intersection point belongs to, namely which lane;
after a large number of tracks are collected at the same intersection, counting the number of times of topological connection of each lane, then calculating the number of times of topological connection of all the lanes of the same outgoing lane as the total number, and calculating the probability of the topological connection of each lane, namely dividing the number of times of topological connection of each lane by the corresponding total number;
fourthly, the method one: carrying out parameter discretization on the data set, directly inquiring a required result in the data set, and carrying out result calculation on any unknown intersection after processing and data collection:
discretizing the intersection lane form parameterization result, wherein the discretizing method comprises the following steps: discretizing 4 parameters, namely, discretizing all transverse angles V, longitudinal angles H, reference points P (X, Y) and lane widths W in the parameters, all transverse angles V, longitudinal angles H, reference points P (X, Y) and lane widths W in each (the outgoing road section and the incoming road section list), respectively setting unified discretization step _ V, step _ H, step _ xy and step _ W, and then respectively obtaining 5 integer values, namely, discretization transverse angle V (X) and discretization longitudinal angle H (H/step _ V), discretization reference point coordinate X (X) and discretization reference point coordinate Y (Y/step _ xy) and discretization lane width W (W/step _ W);
for an intersection data set formed by collecting data of a large number of different intersections, constructing all (an outgoing road section and an incoming road section list) by using corresponding discretization parameters, then counting the total times of all the corresponding road topology connections (the outgoing road section and the incoming road section list) which are completely the same, and calculating the total probability of each road topology connection;
setting a probability threshold value T, for example, T is 0.05, that is, 5%, the probability value of the lane topological connection in the data set is less than T, deleting the lane topological connection, then updating the data set, and obtaining a corresponding relation table between new discretization input parameters (a road section is driven out, a road section is driven in list) and output parameters (a road section topological relation list);
calculating each road junction (an outgoing road section and an incoming road section list) in the same way for any unknown road junction, discretizing the same parameters of the road junctions, then, matching output parameters corresponding to the completely same input parameters in a data set to directly obtain a lane topological relation list, and then, automatically generating lane connecting lines inside the road junction directly according to the obtained lane topological relation when a map is made; when input parameters are matched, if the identical input parameters are not matched in the data set, the output parameters corresponding to the data with the minimum distance value of the current input parameters in the data set can be calculated as a result through the input parameter distance value calculating module;
inputting a parameter distance value calculating module, namely calculating the distance d between two (outgoing road section and incoming road section list), assuming that (outgoing road section 1 and incoming road section list 1) and (outgoing road section 2 and incoming road section list 2), setting a fixed rule, for example, if the number of the incoming road sections in the incoming road section list 1 and the incoming road section list 2 is not equal, d is 1000, if the number is equal, then sequentially and one by one pair comparing the distance between the incoming road section 1 and the incoming road section 2 inside the two, the minimum value is 0, the maximum value is 100, for example, the number of lanes is not equal, if the number of lanes is equal, sequentially and one by one comparing the distance between the internal lane 1 and the lane 2 inside the two, the minimum value is 0, the maximum value is 10, then performing distance calculation on the attribute values inside the lanes, and similarly setting the minimum value and the maximum value of the distance for each attribute value, then, respectively calculating to finally obtain the total distance d; the distance between the outgoing link 1 and the outgoing link 2 is calculated in the same manner as the distance between the two incoming links.
(2) The second method comprises the following steps: training a neural network model by using a data set, calculating a required result by using the trained neural network model, and calculating the result of any unknown intersection by using the following steps after processing and data collection:
according to the obtained (driving-out road section and driving-in road section list) as input parameters, a corresponding lane topological relation list { lane topological connection 1, lane topological connection 2, lane topological connection 3,. to. } and probability values { p1, p2, p3,. to. } serve as output parameters, a training data set is constructed, a neural network model is used, the neural network model is trained by the data set, and the trained neural network model is obtained;
carrying out the same parameterization calculation on any unknown intersection to obtain input parameters (a road section is driven out and a road section is driven in) and inputting the input parameters into the neural network model, so that output values can be obtained from the model, wherein the output values comprise a lane topological relation list { lane topological connection 1, lane topological connection 2, lane topological connection 3, } and probability values { p1, p2, p3, };
setting a probability threshold value T, for example, T is 0.05, namely 5%, deleting the lane topological connection when the probability value of the lane topological connection in the output result is smaller than T, and the rest result is a final lane topological relation list, and then automatically generating the lane connection line in the intersection directly according to the obtained lane topological relation when a map is made.
When the width values of some lanes in the driving-out road section or the driving-in road section are too large, a certain rule is manually set for lane segmentation, for example, if the lane width exceeds 6 meters, the lane is automatically and evenly segmented into two lanes with the width of 3 meters; or clustering is carried out through the collected historical track information of the current position, whether the current lane is divided or not is judged, and the dividing proportion is determined.
For some small intersections or small fork openings, such as entrances and exits of shopping malls or doorways of residential areas, no obvious marked lines or marks are provided to determine the direction of the current road section, at this time, the historical track information of the vehicles at the current position is called for judgment, if the current position has no historical track information of the vehicles, the judgment cannot be carried out, and the current road section can be considered to be bidirectional.
The method designs a set of intersection lane form parameterization method, converts the real track information of a large number of vehicles at different intersections into topological relation data by collecting the track information, associates the intersection lane form with the topological relation data to establish a corresponding relation, then designs two methods for automatically generating the topological relation for any intersection, one is to discretize the intersection lane form parameter, then establishes a mapping relation table between the intersection form and the topological relation data, parameterizes the intersection lane form for any intersection in the same way, then inquires the obtained mapping relation table to obtain the topological relation data, the other is to use a neural network model, take the intersection form data as input data, take the corresponding topological relation data as output data, train the neural network model, and then carry out training on any intersection, after the intersection lane form parameters are calculated, the intersection lane form parameters are input to the neural network model, and then output topological relation data are obtained.
The method can automatically generate the topological relation of the road junction lane, the generation speed is high, the effect is good, the data making efficiency of the map lane data at the road junction is greatly improved, the data quality is greatly improved, the generated map is used for automatically driving the vehicle, the route is more reasonable, the method is beneficial to accurately pre-judging the intentions and the routes of the self vehicle and other vehicles when the self vehicle and other vehicles pass at the road junction, and the passing efficiency and the stability of the road junction are improved.
In addition to the above embodiments, the present invention may have other embodiments. All technical solutions formed by adopting equivalent substitutions or equivalent transformations fall within the protection scope of the claims of the present invention.
Claims (10)
1. A method for generating topological connection relation of crossing lanes is characterized in that: the method comprises the following steps:
first, intersection lane form parameterization
For each intersection, determining whether a lane enters or leaves the intersection according to the lane direction, dividing the lane into an outgoing lane and an incoming lane, forming a road section by the lanes which are adjacent to each other in the same lane direction and are arranged side by side, and then dividing the road section into an outgoing road section and an incoming road section;
when the topological relation of the lanes is generated at each intersection, the topological relation between each driven road section and all driven road sections is calculated in sequence, whether topological connection is generated between each lane is determined, and after the calculation of all driven road sections is completed, all topological connections in the whole intersection are determined;
calculating the relation between single driving-out road section and driving-in road section
For a certain outgoing road section, sequentially finding all incoming road sections in the anticlockwise direction or the clockwise direction from the outgoing road section in the intersection to obtain a list, and setting the list as an incoming road section list;
normalizing the parameters of the known driving-out road section list and the known driving-in road section list;
sequentially calculating the topological connection of the lanes between all the outgoing road sections and the corresponding incoming road section list in the intersection to obtain the topological connection of all the lanes in the whole intersection;
collecting vehicle track information at the intersection, converting the vehicle track information into lane topological connection information, counting the occurrence frequency of each lane topological connection, and calculating the probability of each lane topological connection
For a certain intersection, collecting vehicle track information, after obtaining the vehicle track information, calculating an entrance point and an exit point of each track, and respectively calculating which lane the 2 points belong to according to coordinates, thereby obtaining the topological connection of the lanes corresponding to the current track;
after a large number of tracks are collected at the same intersection, counting the number of times of topological connection of each lane, then calculating the number of times of topological connection of all the lanes of the same outgoing lane as the total number, and calculating the probability of the topological connection of each lane, namely dividing the number of times of topological connection of each lane by the corresponding total number;
fourthly, the method one: carrying out parameter discretization on the data set, directly inquiring a required result in the data set, and carrying out result calculation on any unknown intersection after processing and data collection;
the second method comprises the following steps: and training the neural network model by using the data set, calculating a required result by using the trained neural network model, and calculating the result of any unknown intersection after processing and data collection.
2. The method for generating topological connection relationship of intersection lanes according to claim 1, wherein: the method for parameterizing the driven road section comprises the following steps:
taking the end point of the lane sidelines of all lanes in the current driven road section, and performing straight line fitting to obtain a straight line, wherein the angle of the straight line is set as the direction of the lanes from left to right, the included angle between the straight line and the positive direction of the x axis is taken as the direction of 0 degree, and the included angle is set as v0 and taken as the transverse angle of the driven road section;
calculating the intersection point between the straight line and the extension line of the leftmost lane boundary, and setting the intersection point as P0(x0, y0) as a reference point of the driving-out road;
calculating the direction angles of all the lane sidelines at the end point, namely calculating the angle formed between a point before the end point and the end point for each sideline, and then averaging all the obtained direction angles to obtain a longitudinal angle h0 of the outgoing road section;
calculating a lane direction list, and setting a lane direction value: using 0 to represent no direction, 1 to represent straight going, 2 to represent left turning, 4 to represent right turning, 8 to represent turning around, calculating a composite value for a plurality of directions, setting a lane direction value for each lane according to ground marking lines and road marking conditions, and then obtaining a lane direction list { t0, t1, t2, } of the current whole driving-out road section;
calculating a lane type list, and setting a lane type value: 0 represents an unknown type, 1 represents a motor vehicle lane, 2 represents a non-motor vehicle lane, 4 represents a bus lane, a composite value can be set, a lane type value is set for each lane according to the actual situation, and then a lane direction list { c0, c1, c2, } of the current whole driving-out road section is obtained;
and calculating a lane width list, calculating a lane width value at the end point of each lane, setting the lane width value as w in meters, and then obtaining a lane width list { w0, w1, w2, } of the current whole outgoing road section.
3. The method for generating topological connection relation of intersection lanes as claimed in claim 2, wherein: the method for parameterizing the driving road section comprises the following steps:
calculating the transverse angle v1 of the driving road section, wherein the method is similar to the method for calculating the transverse angle v0 of the driving road section, and the difference is that all lane starting points are taken to perform straight line fitting to obtain a straight line;
calculating a reference point P1 of the driving-in road section in the same way as the reference point P0 of the driving-out road section;
calculating the longitudinal angle h1 of the driving road section, wherein the method is similar to the calculation method of the h1 in the driving road section, but the difference is that the angle at the starting point is calculated, and then the average value is calculated;
calculating a lane direction list, wherein the lane directions of the driving road sections are uniformly set to be 0, namely, no direction exists;
calculating a lane type list which is the same as the driven road section;
the lane width list is calculated, similarly to the above-described driven-out section, except that the lane width at the starting point is calculated.
4. The method for generating topological connection relation of intersection lanes according to claim 1, characterized in that: the normalization method comprises the following steps: for the outgoing road segment, P0 is translated to the origin of coordinates (0, 0), v0 is rotated to 0 degrees, the same rotation is made for h0, then the same translation is made for all reference points in the incoming road segment list, the same rotation is made for all lateral and longitudinal angles, and all angles are set to 0 to 360 degrees, or-180 to 180 degrees.
5. The method for generating topological connection relation of intersection lanes according to claim 1, characterized in that: the calculation method of the entrance point and the exit point comprises the following steps: and (3) calculating the intersection points between the track and the transverse straight lines obtained by fitting in the step 1, calculating the intersection points between all lane sidelines and the transverse straight lines, calculating the relation between the track intersection points and the lane sideline intersection points, and calculating which two lane sideline intersection points the track intersection points belong to, namely which lane.
6. The method for generating topological connection relationship of intersection lanes according to claim 1, wherein: there are two collection methods for vehicle trajectory information: the method comprises the following steps that firstly, a vehicle is provided with a positioning device, and various sensors including a satellite positioning receiver, a gyroscope, an accelerometer, a wheel speed meter, a camera and a laser radar help to perform self positioning to obtain vehicle track information of the vehicle, tracks of different vehicles are collected together, then the coordinate range of a current intersection is known, and which sections of tracks are in the current intersection are calculated to obtain vehicle track information corresponding to each intersection; and secondly, performing mobile measurement through a sensor arranged on the vehicle or performing fixed measurement through a sensor arranged on a roadside unit, wherein the sensor comprises a camera, a laser radar and a millimeter wave radar, calculating the track information of other vehicles on the road, collecting the track information together, and then calculating the vehicle track information corresponding to each intersection.
7. The method for generating topological connection relation of intersection lanes according to claim 1, characterized in that:
the method comprises the following steps: carrying out parameter discretization on the data set, directly inquiring a required result in the data set, and carrying out result calculation on any unknown intersection after processing and data collection:
discretizing the parameterized result of the lane form of the intersection;
for an intersection data set formed by collecting data of a large number of different intersections, constructing all (an outgoing road section and an incoming road section list) by using corresponding discretization parameters, then counting the total times of all the corresponding road topology connections (the outgoing road section and the incoming road section list) which are completely the same, and calculating the total probability of each road topology connection;
setting a probability threshold value T, deleting the topological connection of the lane when the probability value of the topological connection of the lane in the data set is smaller than T, then updating the data set to obtain a new corresponding relation table between the discretization input parameters (the road section is driven out, the road section is driven into) and the output parameters;
calculating each road junction (an outgoing road section and an incoming road section list) in the same way for any unknown road junction, discretizing the same parameters of the road junctions, then, matching output parameters corresponding to the completely same input parameters in a data set to directly obtain a lane topological relation list, and then, automatically generating lane connecting lines inside the road junction directly according to the obtained lane topological relation when a map is made; when input parameters are matched, if the identical input parameters are not matched in the data set, the output parameters corresponding to the data with the minimum distance value of the current input parameters in the data set can be calculated as a result through the input parameter distance value calculating module;
inputting a parameter distance value calculation module, namely calculating the distance d between two (an outgoing road section and an incoming road section list), setting a fixed rule on the assumption that the distance d is (an outgoing road section 1 and an incoming road section list 1) and (an outgoing road section 2 and an incoming road section list 2), respectively comparing the difference between the outgoing road section 1 and the outgoing road section list 2 and the difference between the incoming road section list 1 and the incoming road section list 2, quantizing the differences, and adding the two values together to obtain a distance value;
the second method comprises the following steps: training a neural network model by using a data set, calculating a required result by using the trained neural network model, and calculating the result of any unknown intersection by using the following steps after processing and data collection:
according to the obtained (driving-out road section and driving-in road section list) as input parameters, a corresponding lane topological relation list { lane topological connection 1, lane topological connection 2, lane topological connection 3,. to. } and probability values { p1, p2, p3,. to. } serve as output parameters, a training data set is constructed, a neural network model is used, the neural network model is trained by the data set, and the trained neural network model is obtained;
carrying out the same parameterization calculation on any unknown intersection to obtain input parameters (a road section is driven out and a road section is driven in) and inputting the input parameters into the neural network model, so that output values can be obtained from the model, wherein the output values comprise a lane topological relation list { lane topological connection 1, lane topological connection 2, lane topological connection 3, } and probability values { p1, p2, p3, };
and setting a probability threshold T, deleting the topological connection of the lane in the output result when the probability value of the topological connection of the lane is less than T, wherein the rest result is a final topological relation list of the lane, and then automatically generating the lane connection line in the intersection directly according to the obtained topological relation of the lane when a map is made.
8. The method for generating topological connection relation of intersection lanes according to claim 7, characterized in that: the discretization method comprises the following steps: for each parameter (exit route, entry route list), all the transverse angles V, longitudinal angles H, reference points P (X, Y) and lane widths W in the parameter, discretizing the 4 parameters, setting a uniform discretization step size step _ V, step _ H, step _ xy and step _ W, and then obtaining 5 integer values, namely discretization transverse angle V ═ rounding (V/step _ V), discretization longitudinal angle H ═ rounding (H/step _ H), discretization reference point coordinate X ═ rounding (X/step _ xy), Y ═ rounding (Y/step _ xy), discretization lane width W ═ rounding (W/step _ W).
9. The method for generating topological connection relation of intersection lanes according to claim 1, characterized in that: when the width values of some lanes in the driving-out road section or the driving-in road section are too large, a certain rule is manually set for lane segmentation; or clustering is carried out through the collected historical track information of the current position, whether the current lane is divided or not is judged, and the dividing proportion is determined.
10. The method for generating topological connection relationship of intersection lanes according to claim 1, wherein: for some small intersections or small fork intersections, no obvious marked line or mark is provided to determine the direction of the current road section, at the moment, the historical track information of the vehicle at the current position is called for judgment, if the current position does not have the historical track information of the vehicle, the judgment cannot be carried out, and the current road section can be considered to be bidirectional.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111662498.4A CN114495489B (en) | 2021-12-30 | 2021-12-30 | Intersection lane topology connection relation generation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111662498.4A CN114495489B (en) | 2021-12-30 | 2021-12-30 | Intersection lane topology connection relation generation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114495489A true CN114495489A (en) | 2022-05-13 |
| CN114495489B CN114495489B (en) | 2023-07-25 |
Family
ID=81497123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111662498.4A Active CN114495489B (en) | 2021-12-30 | 2021-12-30 | Intersection lane topology connection relation generation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114495489B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115061479A (en) * | 2022-08-03 | 2022-09-16 | 国汽智控(北京)科技有限公司 | Lane relation determination method and device, electronic equipment and storage medium |
| CN115331425A (en) * | 2022-06-30 | 2022-11-11 | 银江技术股份有限公司 | Traffic early warning method, device and system |
| CN117576950A (en) * | 2024-01-16 | 2024-02-20 | 长沙行深智能科技有限公司 | Method and device for predicting vehicle to select crossing entrance and crossing exit |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105651295A (en) * | 2016-01-15 | 2016-06-08 | 武汉光庭信息技术股份有限公司 | Connection curve algorithm for constructing intersection entry and exit lane Links based on Bezier curve |
| DE102015211150A1 (en) * | 2015-06-17 | 2016-12-22 | Bayerische Motoren Werke Aktiengesellschaft | Method, driver assistance system and vehicle for learning a trajectory of a road section |
| CN109186617A (en) * | 2018-08-13 | 2019-01-11 | 武汉中海庭数据技术有限公司 | A kind of view-based access control model crowdsourcing data automatically generate method, system and the memory of lane grade topological relation |
| US20200064846A1 (en) * | 2018-08-21 | 2020-02-27 | GM Global Technology Operations LLC | Intelligent vehicle navigation systems, methods, and control logic for multi-lane separation and trajectory extraction of roadway segments |
| CN111065893A (en) * | 2017-06-01 | 2020-04-24 | 罗伯特·博世有限公司 | Method and device for creating a lane-accurate road map |
| CN111142525A (en) * | 2019-12-31 | 2020-05-12 | 武汉中海庭数据技术有限公司 | High-precision map lane topology construction method and system, server and medium |
| US20210001877A1 (en) * | 2019-07-02 | 2021-01-07 | DeepMap Inc. | Determination of lane connectivity at traffic intersections for high definition maps |
| CN112580179A (en) * | 2020-12-29 | 2021-03-30 | 武汉中海庭数据技术有限公司 | High-precision map intersection lane shape updating method and system, server and medium |
| US20210108936A1 (en) * | 2019-10-09 | 2021-04-15 | Argo AI, LLC | Methods and systems for topological planning in autonomous driving |
| CN112699708A (en) * | 2019-10-22 | 2021-04-23 | 北京初速度科技有限公司 | Method and device for generating lane-level topology network |
| CN113074744A (en) * | 2021-03-18 | 2021-07-06 | 重庆数字城市科技有限公司 | Method for generating topological connection line of map for intelligent network vehicle connection |
-
2021
- 2021-12-30 CN CN202111662498.4A patent/CN114495489B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102015211150A1 (en) * | 2015-06-17 | 2016-12-22 | Bayerische Motoren Werke Aktiengesellschaft | Method, driver assistance system and vehicle for learning a trajectory of a road section |
| CN105651295A (en) * | 2016-01-15 | 2016-06-08 | 武汉光庭信息技术股份有限公司 | Connection curve algorithm for constructing intersection entry and exit lane Links based on Bezier curve |
| CN111065893A (en) * | 2017-06-01 | 2020-04-24 | 罗伯特·博世有限公司 | Method and device for creating a lane-accurate road map |
| CN109186617A (en) * | 2018-08-13 | 2019-01-11 | 武汉中海庭数据技术有限公司 | A kind of view-based access control model crowdsourcing data automatically generate method, system and the memory of lane grade topological relation |
| US20200064846A1 (en) * | 2018-08-21 | 2020-02-27 | GM Global Technology Operations LLC | Intelligent vehicle navigation systems, methods, and control logic for multi-lane separation and trajectory extraction of roadway segments |
| US20210001877A1 (en) * | 2019-07-02 | 2021-01-07 | DeepMap Inc. | Determination of lane connectivity at traffic intersections for high definition maps |
| US20210108936A1 (en) * | 2019-10-09 | 2021-04-15 | Argo AI, LLC | Methods and systems for topological planning in autonomous driving |
| CN112699708A (en) * | 2019-10-22 | 2021-04-23 | 北京初速度科技有限公司 | Method and device for generating lane-level topology network |
| CN111142525A (en) * | 2019-12-31 | 2020-05-12 | 武汉中海庭数据技术有限公司 | High-precision map lane topology construction method and system, server and medium |
| CN112580179A (en) * | 2020-12-29 | 2021-03-30 | 武汉中海庭数据技术有限公司 | High-precision map intersection lane shape updating method and system, server and medium |
| CN113074744A (en) * | 2021-03-18 | 2021-07-06 | 重庆数字城市科技有限公司 | Method for generating topological connection line of map for intelligent network vehicle connection |
Non-Patent Citations (4)
| Title |
|---|
| 刘经南 等: "智能高精地图数据逻辑结构与关键技术", 《测绘学报》, vol. 48, no. 8, pages 939 - 953 * |
| 唐炉亮 等: "利用轨迹大数据进行城市道路交叉口识别及结构提取", 《测绘学报》, vol. 46, no. 6, pages 770 - 779 * |
| 张攀 等: "通用化高精地图数据模型", 《测绘学报》, vol. 50, no. 11, pages 1432 - 1446 * |
| 许广宏 等: "一种面向智能驾驶的电子地图模型", 《上海汽车》, pages 15 - 18 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115331425A (en) * | 2022-06-30 | 2022-11-11 | 银江技术股份有限公司 | Traffic early warning method, device and system |
| CN115331425B (en) * | 2022-06-30 | 2023-12-19 | 银江技术股份有限公司 | Traffic early warning method, device and system |
| CN115061479A (en) * | 2022-08-03 | 2022-09-16 | 国汽智控(北京)科技有限公司 | Lane relation determination method and device, electronic equipment and storage medium |
| CN115061479B (en) * | 2022-08-03 | 2022-11-04 | 国汽智控(北京)科技有限公司 | Lane relation determination method and device, electronic equipment and storage medium |
| CN117576950A (en) * | 2024-01-16 | 2024-02-20 | 长沙行深智能科技有限公司 | Method and device for predicting vehicle to select crossing entrance and crossing exit |
| CN117576950B (en) * | 2024-01-16 | 2024-04-09 | 长沙行深智能科技有限公司 | Method and device for predicting vehicle to select crossing entrance and crossing exit |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114495489B (en) | 2023-07-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114495489A (en) | Method for generating topological connection relation of road junction lanes | |
| CN112668153B (en) | Method, device and equipment for generating automatic driving simulation scene | |
| CN108920481B (en) | Road network reconstruction method and system based on mobile phone positioning data | |
| CN110553660B (en) | Unmanned vehicle trajectory planning method based on A-star algorithm and artificial potential field | |
| CN105118294B (en) | A kind of Short-time Traffic Flow Forecasting Methods based on state model | |
| CN107563566B (en) | Inter-bus-station operation time interval prediction method based on support vector machine | |
| WO2022141213A1 (en) | Gene prediction method and system for fault of autonomous rail rapid transit vehicle in smart city | |
| CN114005280B (en) | Vehicle track prediction method based on uncertainty estimation | |
| CN109558831B (en) | Cross-camera pedestrian positioning method fused with space-time model | |
| CN107703945A (en) | A kind of intelligent farm machinery paths planning method of multiple targets fusion | |
| CN109285348A (en) | A kind of vehicle behavior recognition methods and system based on two-way length memory network in short-term | |
| CN111710162B (en) | Urban road network traffic operation condition monitoring method and system | |
| CN112530158B (en) | Road network supplementing method based on historical track | |
| CN111314857B (en) | Vehicle real-time travel track acquisition method based on vehicle passing video data | |
| CN113033840B (en) | Method and device for judging highway maintenance | |
| CN115435798A (en) | Unmanned vehicle high-precision map road network generation system and method | |
| CN113537626A (en) | Neural network combined time sequence prediction method for aggregating information difference | |
| Meng et al. | Trajectory prediction for automated vehicles on roads with lanes partially covered by ice or snow | |
| CN111551185B (en) | Method for adding traffic lane | |
| Zhang et al. | Multi-modal virtual-real fusion based transformer for collaborative perception | |
| CN115610435B (en) | Method and device for predicting object driving intention, storage medium and electronic device | |
| CN112699575A (en) | Method and system for measuring and calculating relative position in virtual vehicle test platform | |
| Zhang et al. | Real time obstacle detection method based on lidar and wireless sensor | |
| CN116105717A (en) | Lane-level high-precision map construction method and system | |
| CN115188195A (en) | Method and system for extracting vehicle track of urban omnidirectional intersection in real time |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230323 Address after: 1 / F, building 28, 6055 Jinhai highway, Fengxian District, Shanghai 201400 Applicant after: Zhongzhixing (Shanghai) Transportation Technology Co.,Ltd. Applicant after: Tianyi Transportation Technology Co.,Ltd. Address before: 1 / F, building 28, 6055 Jinhai highway, Fengxian District, Shanghai 201400 Applicant before: Zhongzhixing (Shanghai) Transportation Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |