WO2024040499A1

WO2024040499A1 - Method and apparatus for determining high-precision navigation path, and device, medium and vehicle

Info

Publication number: WO2024040499A1
Application number: PCT/CN2022/114698
Authority: WO
Inventors: 张明星; 陈子晗
Original assignee: 北京初速度科技有限公司
Priority date: 2022-08-23
Filing date: 2022-08-25
Publication date: 2024-02-29
Also published as: CN117664157A

Abstract

Disclosed in the embodiments of the present invention are a method and apparatus for determining a high-precision navigation path, and a device, a medium and a vehicle. The method comprises: acquiring a traditional navigation path provided by a traditional navigation map, and matching the traditional navigation path with a high-precision map; during a matching process, determining high-precision road sections in the high-precision map which correspond to all the trajectory points of the traditional navigation path, wherein the corresponding high-precision road sections comprise first high-precision road sections that are determined as matching the trajectory points, and second high-precision road sections that possibly match the trajectory points; and screening, from among the first high-precision road sections and the second high-precision road sections, a plurality of consecutive target high-precision road sections with a topological relationship, and according to the plurality of consecutive target high-precision road sections, generating a high-precision navigation path matching the traditional navigation path, wherein the starting road section and the ending road section of the high-precision navigation path are both first high-precision road sections. By using the technical solution, the accuracy of a matching result between a traditional navigation path and a high-precision map is improved.

Description

Methods, devices, equipment, media and vehicles for determining high-precision navigation paths

Technical field

Embodiments of the present invention relate to the field of autonomous driving technology, and specifically to a method, device, equipment, medium and vehicle for determining a high-precision navigation path.

Background technique

Currently, in the process of autonomous driving, since real-time traffic information does not exist in high-precision maps, it is difficult to directly provide navigation functions. Therefore, in the actual application process of high-precision maps, it is necessary to match the traditional navigation paths obtained from traditional navigation maps. to high-precision maps to determine which road sections the car should turn on autonomous driving mode and which road sections require manual takeover.

In related technologies, hidden Markov models are usually used to match high-precision navigation maps with traditional navigation paths. During the matching process, it is necessary to find high-precision road segments that match each trajectory point in the traditional navigation map, and then connect the observed high-precision road segments through the topological connection relationships of the roads. However, due to different production processes of different maps, inconsistent labels, or differences in location accuracy, the matching solutions in related technologies will cause missed matching on some high-precision road sections. That is, the matching solutions in related technologies have poor compatibility, and the matching results have yet to be improve.

Contents of the invention

Embodiments of the present invention provide a method, device, equipment, medium and vehicle for determining a high-precision navigation path, which improves the accuracy of matching results between traditional navigation paths and high-precision maps.

The specific technical solutions are as follows:

In a first aspect, embodiments of the present invention provide a method for determining a high-precision navigation path, including:

Obtain the traditional navigation path provided by the traditional navigation map and match the traditional navigation path with the high-precision map;

During the matching process, a high-precision road section corresponding to each trajectory point of the unified navigation path is determined in the high-precision map. The corresponding high-precision road section includes the first high-precision road section that is determined to match each trajectory point, and the corresponding high-precision road section that is matched with each trajectory point. The second high-precision road segment that the track point may match;

Select multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, and generate a high-precision navigation path that matches the traditional navigation path based on the multiple continuous target high-precision road sections. Among them, the starting section and the ending section of the high-precision navigation path are both the first high-precision section;

Among them, the first high-precision road section is a high-precision road section that meets all the following matching conditions, and the second high-precision road section is a high-precision road section that meets any one or both of the following matching conditions, where the matching conditions include: for any current trajectory point, the label information of the traditional navigation path corresponding to the current trajectory point is consistent with the label information, position information, and angle information of the high-precision road section corresponding to the current trajectory point.

It can be seen from the above solution that for high-precision road sections that cannot meet all matching conditions, that is, for high-precision road sections that may match the trajectory points in the traditional navigation path, the embodiment of the present invention retains the observation information of this type of high-precision road section and combines the Observation information, as well as the topological connection relationship between the high-precision road section and other road sections, a high-precision navigation path that matches the traditional navigation path is obtained, which avoids the removal of track points in the traditional navigation path that may match the high-precision navigation path in related technologies. The phenomenon of mismatching or missed matching caused by the observation information of road sections makes the matching results of traditional navigation paths and high-precision maps compatible with the process errors, angle errors and precision errors between traditional navigation maps and high-precision maps. The matching results more precise.

Optionally, select multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, including:

Score the observation values of the first high-precision road section and the second high-precision road section respectively to obtain the first observation score corresponding to the first high-precision road section and the second observation score corresponding to the second high-precision road section, where the first The observation score is greater than the second observation score;

According to the topological relationship of the high-precision road section and the observation scores of different road sections, the transition probability scores of the first high-precision road section and the second high-precision road section are respectively obtained, and the first transition probability score corresponding to the first high-precision road section is obtained, as well as the first transition probability score corresponding to the first high-precision road section. The second transition probability score corresponding to the two high-precision road sections, wherein the second transition probability score is smaller than the first transition probability score;

For any current trajectory point, starting from the first first high-precision road segment that can be observed by the current trajectory point, the transition probability score of the first first high-precision road segment is used as the weight value, and based on the weight value, the The observation scores of the first first high-precision road section are weighted to obtain the Viterbi score of the first first high-precision road section;

The obtained Viterbi score is used as the observation score of the next candidate continuous high-precision road section in turn, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform Viterbi on the candidate continuous high-precision road section. The score calculation operation is performed until the Viterbi score of the last first high-precision road section that can be observed at the last trajectory point is calculated, in which the candidate continuous high-precision road section is the first high-precision road section or the second high-precision road section. road section;

Select the Viterbi score with the largest value from the multiple calculated Viterbi scores, and use the multiple continuous high-precision road sections corresponding to the largest Viterbi score as multiple target high-precision road sections with continuous topological relationships. .

It can be seen from the above solution that setting the second transition probability score to be smaller than the first transition probability score can increase the weight of the high-precision road section that is determined to match each trajectory point in the entire high-precision navigation path, and can reduce the weight of the high-precision road section that matches each trajectory point. The weight of high-precision road segments that may be matched by points in the entire high-precision navigation path to improve the accuracy of the matching results.

It can be seen from the above solution that the screening process can be simplified by using observation values and transition probability values to score and calculate the Viterbi score, and efficiently screen out high-precision road sections that are continuous with the current high-precision road section.

Optionally, score the observation values of the first high-precision road section and the second high-precision road section respectively to obtain the first observation score corresponding to the first high-precision road section and the second observation score corresponding to the second high-precision road section, include:

The observation value of each first high-precision road section is scored as 1, and the observation value of each second high-precision road section is scored as 0.5; correspondingly,

According to the topological relationship of the high-precision road section and the observation scores of different road sections, the transition probability scores of the first high-precision road section and the second high-precision road section are respectively obtained, and the first transition probability score corresponding to the first high-precision road section is obtained, as well as the first transition probability score corresponding to the first high-precision road section. The second transition probability scores corresponding to the two high-precision road sections include:

For any current high-precision road segment, if the observation score of a high-precision road segment that is continuous with the current high-precision road segment is 1, and when the candidate continuous high-precision road segment is the same road segment as the current high-precision road segment, the candidate high-precision road segment will be The transition probability score of a continuous high-precision road section is 1+E, where E is a minimum value greater than 0; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 1, and the candidate continuous high-precision road segment can be topologically connected to the current high-precision road segment, the transition probability of the candidate continuous high-precision road segment is divided into The value is scored as 1+2E; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 1, and the continuous high-precision road segment candidate cannot be topologically connected to the current high-precision road segment, the transition probability of the continuous high-precision road segment candidate is divided into The value is scored as 0; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 0.5, then the transition probability score of the high-precision road segment that is continuous with the candidate is scored as 1.

It can be seen from the above scheme that for a candidate continuous high-precision road section with an observation score of 1, if the candidate continuous high-precision road section is the same road section as the current high-precision road section, then the transition probability score of the candidate continuous high-precision road section is 1. +E, if the candidate continuous high-precision road segment and the current high-precision road segment are the next road segment that can be topologically connected, the transition probability score will be slightly greater than the transfer score of the same road segment 1+2E. This setting will be topologically connectable. The connected road segments are distinguished from the same road segments, so that the process of observing and finding the next road segment continues, thereby ensuring that the beginning and end of the road segment can be correctly matched and connected.

Optionally, when the observation score of the high-precision road section that is continuous with the current high-precision road section is 1, and the transition probability score of the candidate continuous high-precision road section is 0:

The current high-precision road segment is used as the ending road segment of the matched path, and the high-precision road segment that is continuous with the candidate is used as the starting road segment of the next matching path, and the execution continues according to the topological relationship of the high-precision road segment and the observation scores of different road segments. Scoring operation for transition probability scores.

It can be seen from the above solution that when the transition probability of a candidate continuous high-precision road section is 0, it means that the candidate continuous high-precision road section is disconnected from the current high-precision road section, that is, the matching result is discontinuous. This is because the high-precision map itself may not be complete, and the road network may be missing, so the real matching results will be discontinuous. In this embodiment, the transition probability score of the next candidate road segment that does not have a topological connection relationship with the current high-precision road segment is scored as 0, which is compatible with the lack of high-precision road network.

Optionally, the method provided by the embodiment of the present invention also includes:

According to the topological relationship of the road network, identify the intersecting road sections in the traditional navigation path, where the intersecting road sections are: there is no topological connection relationship between the target observation road sections that can be observed by at least two consecutive trajectory points in the traditional navigation path, and In the case where the target observation road section is not a road entrance or exit section, nor is it a starting section or an ending section of a road, the navigation section corresponding to the at least two continuous trajectory points;

The observation scores of the high-precision road sections that can be observed by the trajectory points corresponding to the transverse interpolation road sections are corrected to the second observation scores;

Correspondingly, according to the topological relationship of the high-precision road section and the observation scores of different road sections, the first high-precision road section and the second high-precision road section are respectively assigned transition probability scores, including:

According to the topological relationship of the high-precision road sections and the corrected observation scores of different road sections, the transition probability is scored for each first high-precision road section and each second high-precision road section.

It can be seen from the above scheme that the intersecting road sections in the traditional navigation path are identified through topological relationships, and the observation score of the target observation road section that can be observed in the intervening road sections is corrected to 0.5, eliminating the target observation road section as the starting point of the final matching path. and the possibility of ending points, solving the problem of mismatching of cross-cutting sections.

Optionally, before performing transition probability scoring on each first high-precision road section and each second high-precision road section according to the topological relationship of the high-precision road section and the corrected observation scores of different road sections, the method provided by the embodiment of the present invention Also includes:

In the opposite direction to the traditional navigation path, start from the target observation road section that can be observed by the trajectory point of the cross-intersection road section to find whether there is an entrance and exit section. If the entrance and exit section can be found, then connect the target observation section and the entrance and exit section. The observation scores of all high-precision road sections are corrected to the second observation scores, and the observation scores corresponding to the entrance and exit sections are corrected to the first observation scores.

It can be seen from the above scheme that during the observation process, the observation scores of the entrance and exit sections are measured through the road topology relationship, as well as the observation scores of all high-precision road sections between the target observation section and the entrance and exit sections that can be observed by the trajectory points of the intersecting road sections. The value is corrected to solve the problem of excessive accuracy of entrances and exits between high-precision maps and traditional navigation paths, further ensuring the accuracy of the matching results.

Optionally, consistent location information means that the trajectory points of the traditional navigation path intersect with the road range corresponding to the high-precision road section;

Among them, when the traditional navigation path passes through an urban road intersection, the road range corresponding to the urban road intersection is: all high-precision road sections within the urban road intersection share the same road range; or,

When the traditional navigation path passes through a non-two-way road that contains two road edge lines, the road range corresponding to the non-two-way road is: connecting the road boundaries of each high-precision road segment end to end; or,

When the traditional navigation path passes through a two-way road that only contains one road edge line, the road range corresponding to the two-way road is: formed by connecting the road edge line of the current two-way road and the edge line of the opposite two-way road end to end, where, The opposite two-way road is a two-way road adjacent to the current two-way road in the high-precision map and opposite to the road direction of the current two-way road.

Optionally, for the starting section of the high-precision navigation path, the proportion of the first high-precision section within the first distance threshold range from the starting section reaches the first proportion threshold;

For the end section of the high-precision navigation path, the proportion of the first high-precision section within the second distance threshold range from the end section reaches the second proportion threshold.

It can be seen from the above technical solution that by increasing the proportion of high-precision road sections that determine matching at the beginning section and end section of the high-precision navigation path, the accuracy of the matching results can be further ensured.

Optionally, the label information of the traditional navigation path corresponding to the current trajectory point and the label information of the high-precision road section corresponding to the current trajectory point are consistent: the label of the corresponding traditional navigation path within the third distance threshold range before and after the current trajectory point is consistent with the label information of the high-precision road section. The label information of the high-precision road section corresponding to the current trajectory point is consistent.

It can be seen from the above technical solution that the setting of the third distance threshold can be compatible with the problem that there is a certain gap in the location of the entrance and exit sections due to differences in production processes between high-precision maps and traditional navigation maps.

In a second aspect, embodiments of the present invention also provide a device for determining a high-precision navigation path, including:

The traditional navigation path acquisition module is configured to obtain the traditional navigation path provided by the traditional navigation map, and match the traditional navigation path with the high-precision map;

The observation module is configured to, during the matching process, determine a high-precision road section corresponding to each trajectory point of the traditional navigation path in the high-precision map, and the corresponding high-precision road section includes a path determined to match each trajectory point. The first high-precision road section, and the second high-precision road section that may match each trajectory point;

The matching module is configured to select multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, and generate a matching traditional navigation path based on the multiple continuous target high-precision road sections. A high-precision navigation path, in which the beginning section and the end section of the high-precision navigation path are both the first high-precision section;

Among them, the first high-precision road section is a high-precision road section that meets all the following matching conditions, and the second high-precision road section is a high-precision road section that meets any one or both of the following matching conditions, where the matching conditions include: for any one The current trajectory point. The label information of the traditional navigation path corresponding to the current trajectory point is consistent with the label information, position information, and angle information of the high-precision road section corresponding to the current trajectory point.

Optional matching modules include:

The observation value scoring unit is configured to score the observation values of the first high-precision road section and the second high-precision road section respectively, and obtain the first observation score corresponding to the first high-precision road section, and the second observation score corresponding to the second high-precision road section. Observation score, where the first observation score is greater than the second observation score;

The transition probability value scoring unit is configured to score the first high-precision road section and the second high-precision road section respectively according to the topological relationship of the high-precision road section and the observation scores of different road sections, and obtain the transition probability value corresponding to the first high-precision road section. The first transition probability score, and the second transition probability score corresponding to the second high-precision road section, wherein the second transition probability score is smaller than the first transition probability score;

The Viterbi score calculation unit is configured to calculate the transition probability score of the first high-precision road section for any current trajectory point starting from the first high-precision road section that can be observed by the current trajectory point. as the weight value, and weight the observation score of the first first high-precision road section based on the weight value to obtain the Viterbi score of the first first high-precision road section; in turn, the obtained Viterbi score will be used as the next The observation score of the candidate continuous high-precision road section, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform the Viterbi score calculation operation on the candidate continuous high-precision road section until the last one is calculated. The Viterbi value of the last first high-precision road section that can be observed by the trajectory point;

The target high-precision road section determination unit is configured to select the Viterbi score with the largest value from the multiple calculated Viterbi scores, and use the multiple continuous high-precision road sections corresponding to the maximum Viterbi score as Multiple target high-precision road sections with continuous topological relationships;

The high-precision navigation path generating unit is configured to generate a high-precision navigation path that matches the traditional navigation path based on multiple consecutive target high-precision road sections.

Optional, observation value scoring unit, specifically configured as:

The transition probability value scoring unit is specifically configured as:

Optionally, when the observation score of a continuous high-precision road section selected from the current high-precision road section is 1, and the transition probability score of the candidate continuous high-precision road section is 0:

Optionally, the device provided by the embodiment of the present invention also includes:

The transverse road segment identification module is configured to identify transverse road segments in the traditional navigation path according to the topological relationship of the road network, where the transverse road segment is: a target that can be observed by at least two consecutive trajectory points in the traditional navigation path. There is no topological connection relationship between the observed road sections, and when the target observed road section is not a road entrance or exit section, nor is it the starting section or the ending section of the road, the navigation section corresponding to at least two consecutive trajectory points;

The first observation score correction module is configured to correct the observation scores of high-precision road sections that can be observed by the trajectory points corresponding to the transverse road sections to the second observation scores;

Correspondingly, the transition probability value scoring unit includes:

The transition probability value scoring sub-unit is configured to score each first high-precision road section and each second high-precision road section with a transition probability according to the topological relationship of the high-precision road section and the corrected observation scores of different road sections.

The second observation score correction module is specifically configured to score the transition probability for each first high-precision road section and each second high-precision road section respectively according to the topological relationship of the high-precision road section and the corrected observation scores of different road sections. , in the opposite direction to the traditional navigation path, start from the target observation road section that can be observed by the trajectory point of the intersecting road section to find whether there is an entrance and exit section. If the entrance and exit section can be found, then combine the target observation section and the entrance and exit section. The observation scores of all high-precision road sections between them are corrected to the second observation scores, and the observation scores corresponding to the entrance and exit sections are corrected to the first observation scores.

When the traditional navigation path passes through a two-way road that only contains one road edge line, the road range corresponding to the two-way road is: formed by connecting the road edge line of the current two-way road and the edge line of the opposite two-way road end to end, where, The opposite two-way road is a two-way road adjacent to the current two-way road in the high-precision map and opposite to the direction of the current two-way road.

Optionally, the label information of the traditional navigation path corresponding to the current trajectory point and the label information of the high-precision road section corresponding to the current trajectory point are consistent means: the label of the corresponding traditional navigation path within the third distance threshold range before and after the current trajectory point. The label information of the high-precision road section corresponding to the current trajectory point is consistent.

In a third aspect, an embodiment of the present invention provides an electronic device, which includes:

one or more processors;

a storage device for storing one or more programs,

When the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method for determining a high-precision navigation path as provided by any embodiment of the present invention.

In a fourth aspect, embodiments of the present invention provide a storage medium on which a computer program is stored, characterized in that when the program is executed by a processor, the determination of a high-precision navigation path as provided by any embodiment of the present invention is achieved. method.

In a fifth aspect, embodiments of the present invention provide a vehicle, which includes a device for determining a high-precision navigation path provided by any embodiment of the present invention, or an electronic device provided by any embodiment of the present invention.

In a sixth aspect, an embodiment of the present invention provides a computer program. The computer program includes program instructions. When the program instructions are executed by a computer, the method for determining a high-precision navigation path provided by any embodiment of the present invention is implemented.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without exerting creative efforts.

Figure 1a is a flow chart of a method for determining a high-precision navigation path provided in Embodiment 1 of the present invention;

Figure 1b is a schematic diagram of matching between a traditional navigation path and a high-precision map in related technologies;

Figure 1c is a schematic diagram of matching between another traditional navigation path and a high-precision map in related technologies;

Figure 1d is a schematic diagram of a matching path provided in Embodiment 1 of the present invention;

Figure 2a is a flow chart of a method for determining a high-precision navigation path provided in Embodiment 2 of the present invention;

Figure 2b is a schematic diagram of a transverse road section provided in Embodiment 2 of the present invention;

Figure 2c is a schematic diagram of entrance and exit section matching provided by Embodiment 2 of the present invention;

Figure 3 is a structural block diagram of a device for determining a high-precision navigation path provided in Embodiment 3 of the present invention;

Figure 4 is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention;

Figure 5 is a schematic diagram of a vehicle provided in Embodiment 5 of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without exerting creative efforts fall within the scope of protection of the present invention.

It should be noted that the terms "including" and "having" and any variations thereof in the embodiments of the present invention and the drawings are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes Other steps or units inherent to such processes, methods, products or devices.

In the description of the embodiment of the present invention, the accuracy of the term "traditional navigation map" is generally meter level, in which only road-level data is recorded, such as road shape, slope, curvature, paving, direction, etc. The users of “traditional navigation maps” are mainly drivers.

In the description of the embodiment of the present invention, the accuracy of the term "high-precision map" is centimeter level. In addition to the road-level data in the "traditional navigation map" mentioned above, the "high-precision map" also adds data related to lane attributes (lane line type, lane width, etc.), such as overhead objects, guardrails, trees, roads, etc. A large amount of target data such as edge types and roadside landmarks. The main user of "high-precision map" is the automatic driving system of automobiles.

Embodiments of the present invention disclose a method, device, equipment, medium and vehicle for determining a high-precision navigation path. Each is explained in detail below.

Embodiment 1

Figure 1a is a flow chart of a method for determining a high-precision navigation path provided in Embodiment 1 of the present invention. This method can be applied to vehicle-mounted computers, vehicle-mounted industrial personal computers (IPC) and other vehicle-mounted terminals, and can also be applied to server, this embodiment of the present invention does not limit this. The method provided in this embodiment can be applied to the process of matching high-precision maps and traditional navigation maps. The method provided in this embodiment can be executed by a device for determining a high-precision navigation path, which can be implemented in software and/or hardware. As shown in Figure 1a, the method provided by this embodiment specifically includes:

S110. Obtain the traditional navigation path provided by the traditional navigation map, and match the traditional navigation path with the high-precision map.

In this embodiment, the autonomous driving system can obtain the traditional navigation path provided by the traditional navigation map based on the departure point and destination information. Among them, the departure place information and destination information can be set by the user according to the actual situation, and there are many setting methods. For example, the user can manually input the departure place and destination information through the human-computer interaction interface, or can also input it through voice and other methods. Departure place and destination information are not specifically limited in this embodiment. After receiving the departure point information and destination information, the autonomous driving system can obtain the traditional navigation path from the traditional navigation map based on the received information, and match the traditional navigation path with the high-precision map.

In this embodiment, the traditional navigation path is matched with the high-precision map by matching each track point of the traditional navigation path with the road range of each high-precision road section in the high-precision map. Among them, the road range of the high-precision road sections in the high-precision map can be obtained by connecting the road boundaries of the high-precision road sections end to end. Below, the methods for determining the road range of different high-precision road sections are introduced in detail:

(1) Non-two-way road

Among them, non-two-way roads refer to high-precision road sections containing two road edge lines. For this type of high-precision road section, the road range is obtained by connecting the road boundaries end-to-end during the matching process. Due to the technical problems of high-precision maps, the road range of the non-two-way road will appear as a non-convex polygon. The convex polygon refers to: if any one of all the sides of a road range is extended infinitely to both directions to become a straight line. When a line is drawn, all other sides are on the same side of the line. In the case of non-convex polygons, the road extent is not maximum. In this embodiment, by correcting the road range of the high-precision road section of the non-two-way road to the largest road range formed by connecting only the maximum range of boundary points of the road edge line end to end, that is, correcting it to the largest convex polygon, it can be ensured The accuracy of the road range is compatible with the process issues of high-precision maps.

(2) Two-way road

Among them, the two-way road in the high-precision map is a road that only contains one road edge line. In high-precision maps, the original road range of a two-way road is usually composed of its road edge line and another non-road edge line. If the non-road edge line is a traversable road line type, such as a white dotted line, or a virtual line, the original road range of the two-way road can be expanded, and the expanded two-way road range is determined by the road edge line of the current two-way road. It is formed by connecting end to end with the edge line of the two-way road on the opposite side. Wherein, the two-way road on the opposite side of the current two-way road is: a two-way road adjacent to the current two-way road in the high-precision map and opposite to the road direction of the current two-way road.

As an optional implementation, when the length of the current two-way road to be matched is consistent with the length of the adjacent opposite-side two-way road, by starting from the non-road edge line of the original road range of the current two-way road to be matched, Search for the road edge line in the direction of the opposite two-way road until the road edge line of the opposite two-way road is found; then compare the current road edge line of the current two-way road with the road edge line of the opposite two-way road with the same length. Connect end to end to get the road range of the two-way road.

As another optional implementation, when the length of the current two-way road to be matched is inconsistent with the length of the adjacent opposite-side two-way road, obtain the current road centerline of the current two-way road, and wait on the current road centerline. Insert multiple virtual points at intervals, and then draw normals to the multiple inserted virtual points. The normals of each virtual point will intersect with the road edge line of the opposite two-way road, thereby obtaining multiple intersection points; the boundary points of each intersection point Connected end-to-end with the boundary points of the road edge line of the current two-way road, the road range of the two-way road can be obtained.

In this embodiment, by expanding the original road range of the two-way road to a road range formed by connecting the edge line of the current two-way road and the edge line of the opposite two-way road end-to-end, the traditional navigation path can pass through the current two-way road. And also when passing through the two-way road on the opposite side, each track point in the traditional navigation path passing through the two-way road on both adjacent sides can be matched to the real road range of the two-way road to avoid missing matching.

(3) Urban road intersections

Among them, the routes within urban road intersections are relatively complex. For the road range of this type of road, all high-precision road sections within urban road intersections can be used in the same road range, that is, each of the urban road intersections can be divided into a rectangular frame. The road segments are included in a rectangular frame, and the rectangular frame is used as the road range of each high-precision road segment in the intersection, thereby avoiding the problem of complex routes in urban road intersections.

S120. During the matching process, determine the high-precision road section corresponding to each track point of the traditional navigation path in the high-precision map. The corresponding high-precision road section includes the first high-precision road section that matches each track point. and the second high-precision road segment that may match each trajectory point.

In this embodiment, the first high-precision road section is a high-precision road section that meets all the following matching conditions, and the second high-precision road section is a high-precision road section that meets any one or any two of the following matching conditions, where the matching conditions include: for any A current trajectory point. The label of the traditional navigation path corresponding to the current trajectory point is consistent with the label information, position information, and angle information of the high-precision road section corresponding to the current trajectory point.

The label information includes road grade information and road type information, for example, urban road label information, urban expressway label information, entrance and exit label information, ramp label information, etc. The consistent label information may be that the label of the traditional navigation path corresponding to the third distance threshold range before and after the current trajectory point is consistent with the label information of the high-precision road section corresponding to the current trajectory point. The range of the third distance threshold may be 80 to 120 meters. In this embodiment, the setting of the third distance threshold is compatible with the problem that there is a certain gap in the positions of entrance and exit sections due to differences in production processes between high-precision maps and traditional navigation maps.

In this embodiment, consistent location information means that the trajectory points of the traditional navigation path intersect with the road range corresponding to the high-precision road section. Consistent angle information means that the angle between the direction of the high-precision road section and the driving direction of the traditional navigation path is within a preset angle range, where the preset angle range can be 0 to 30°.

In the related technology, for high-precision road sections that may match the trajectory points that meet one or both of the above matching conditions, they are treated as unmatchable during the matching process, that is, in the subsequent matching process, this type of road section is There is no corresponding observation information. For this type of road section, it is necessary to correct the disconnected road section between the two successfully matched high-precision road sections through the topological relationship between the roads after the matching is completed. However, this method is prone to road connection errors.

For example, Figure 1b is a schematic diagram of matching between a traditional navigation path and a high-precision map in the related art. As shown in Figure 1b, the traditional navigation path S is in the process of matching with the high-precision road section a1-a2-a3 in the high-precision map. Among them, high-precision road section a1 and high-precision road section a3 satisfy all the above matching conditions and can be considered as high-precision road sections that are determined to match. As for the high-precision road section a2, as shown in the first picture in Figure 1b, the angle between the high-precision road section a2 and the traditional navigation path S exceeds the preset angle range. At this time, in the related technology, the high-precision road section is a2 is processed as a road section that does not match the traditional navigation path S, that is, as shown in the second picture in Figure 1b. In the subsequent matching process, there are no observations between high-precision road sections a1 and a3, and they are disconnected. The related technology is to connect the high-precision road section a1 and the high-precision road section a3 by finding topologically connectable road sections between the high-precision road sections a1 and a3. This method is highly random and can easily cause wrong topological connections. As shown in the third picture in Figure 1b, what is obtained is another road section a4 that is completely different from the high-precision road section a2. Moreover, since there is no observation information for road segment a4, it is difficult to provide automatic driving information corresponding to road segment a4.

Similarly, in the case where the accuracy of the high-precision map and the traditional navigation map is inconsistent with respect to the high-precision road section a2, for example, there is no intersection between the traditional navigation path and the high-precision road section a2 but the distance between them is within the preset distance range (for example, 3 -5 meters), or when the labels of high-precision road section a2 in the high-precision map and the traditional navigation map are inconsistent, for example, the high-precision road section a2 in the high-precision map is an urban expressway, but the corresponding label in the traditional navigation map is an urban road. , the technical solution of the related technology will also have the situation that the above-mentioned high-precision road section a2 cannot match the traditional navigation path S.

For another example, Figure 1c is a schematic diagram of matching between another traditional navigation path and a high-precision map in the related technology. As shown in Figure 1c, for the route from the starting point n to the end point m, due to possible car accidents or construction in route 4 When there is an abnormal situation, the traditional navigation path is to follow the route 5 shown by the arrow and avoid the abnormal section and drive from the starting point n to the end point m. Traditional navigation paths can only determine matching to high-precision road sections FA and high-precision road sections CE. High-precision road section CD and high-precision road section AB are the second high-precision road section that may be matched and do not participate in the subsequent path determination process. The related technology is to directly connect high-precision road section FA and high-precision road section CE through topological connection to obtain high-precision road section AC. In fact, there is no observation information in the high-precision road section AC. Direct topological connection will cause connection errors, and then an incorrect high-precision navigation path will be obtained. For example, if you drive along the high-precision road section 6 as shown in Figure 1c, you may be unable to drive due to road abnormalities.

If according to the matching scheme provided by the embodiment of the present invention, in the case where there are only straight roads AC, entrance road sections AB and exit road sections CD in the high-precision map, regarding Figure 1c, the final successfully matched path is the entrance road section FA-AB and the exit road section DC-CE (for the specific matching process, please refer to the matching instructions on cross-cutting sections and entrance and exit sections below). For other unmatched road sections other than the entrance and exit sections, you need to follow the traditional navigation path, which avoids the problem that related technologies directly perform topological connections without observation information and obtain incorrect matching paths.

In this embodiment, for the second high-precision road section that cannot meet all matching conditions, that is, the high-precision road section that may match the trajectory points in the traditional navigation path, compared to the related technology, which is treated as an unmatched road section and does not participate in subsequent Path matching method, and direct topological connection of high-precision road sections when observation information cannot be obtained. This embodiment retains the observation information of the high-precision road section, so that the observation information and the high-precision road section can be combined The topological connection relationship between the road segment and other road segments is used to obtain a high-precision navigation path that matches the traditional navigation path, avoiding the phenomenon of mismatching or missing matching, making the matching results of traditional navigation paths and high-precision maps compatible with traditional navigation There are no process errors, angular errors and accuracy errors between the map and the high-precision map, and the matching results are more accurate.

S130. Select multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, and generate high-precision navigation that matches the traditional navigation path based on the multiple continuous target high-precision road sections. Path, in which the starting section and the ending section of the high-precision navigation path are both the first high-precision section.

In this embodiment, the high-precision navigation path that matches the traditional navigation path starts from a certain first high-precision road section that is determined to match the trajectory points of the traditional navigation path, and ends with a certain first high-precision road segment that is determined to match the trajectory points of the traditional navigation path. A continuous road section that ends with the first high-precision road section. Each high-precision road section in the continuous road section may be the first high-precision road section or the second high-precision road section.

Optionally, when determining the high-precision navigation path, all topologically connectable high-precision road sections between the first high-precision road section at the beginning and the first high-precision road section at the end (can be the first high-precision road section, The path formed by the second high-precision road section may also be used as a high-precision navigation path.

Optionally, this embodiment can optimize the Viterbi algorithm to obtain multiple target high-precision road sections with continuous topological relationships. Specifically, this can be achieved through the following steps (1) to (4):

(1) Score the observation values of the first high-precision road section and the second high-precision road section respectively to obtain the first observation score corresponding to the first high-precision road section and the second observation score corresponding to the second high-precision road section, where , the first observation score is greater than the second observation score.

In this embodiment, a corresponding relationship between different high-precision road sections and observation scores can be established. Specifically, the observation value of each first high-precision road section that matches the trajectory point of the traditional navigation path can be scored as 1, for example , for a two-way road, if the trajectory points of the traditional navigation path and the corrected road range of the high-precision road section meet all the above matching conditions during the matching process, then the observation value of the high-precision road section corresponding to the two-way road will be scored as 1.

Specifically, the observation values of all second high-precision road segments that may match the trajectory points of the traditional navigation path can be scored as 0.5.

It should be noted that for road segments of the intersection type, since any navigation path cannot start or end within the intersection range, the observed values of all matching high-precision road segments within the intersection are calculated according to An observation score of 0.5 is used for calculation.

(2) According to the topological relationship of the high-precision road section and the observation scores of different road sections, the first high-precision road section and the second high-precision road section are respectively scored for the transition probability, and the first transition probability score corresponding to the first high-precision road section is obtained. , and the second transition probability score corresponding to the second high-precision road section, wherein the second transition probability score is smaller than the first transition probability score.

In this embodiment, setting the second transition probability score to be smaller than the first transition probability score can increase the weight of the high-precision road section that is determined to match each trajectory point in the entire high-precision navigation path, and can reduce the weight of the high-precision road section that matches each trajectory point. The weight of high-precision road segments that may be matched by points in the entire high-precision navigation path, thereby improving the accuracy of the matching results.

Specifically, for any current high-precision road section, if the observation score of the high-precision road section that is continuous with the current high-precision road section is 1, and when the candidate continuous high-precision road section is the same road section as the current high-precision road section, The transition probability score of the candidate continuous high-precision road section is scored as 1+E, where E is a minimum value greater than 0. Optional, the value range of E is 1e-1~1e-10; or,

If the observation score of the continuous high-precision road section candidate with the current high-precision road section is 1, and the candidate continuous high-precision road section cannot be topologically connected to the current high-precision road section, the transition probability score of the candidate continuous high-precision road section is scored. is 0; or,

In this embodiment, for a candidate continuous high-precision road section with an observation score of 1, if the candidate continuous high-precision road section is the same road section as the current high-precision road section, then the transition probability score of the candidate continuous high-precision road section is 1. +E, if the candidate continuous high-precision road segment and the current high-precision road segment are the next road segment that can be topologically connected, the transition probability score will be slightly greater than the transfer score of the same road segment 1+2E. This setting will be topologically connectable. The connected road segments are distinguished from the same road segments, so that the process of observing and finding the next road segment continues, thereby ensuring that the beginning and end of the road segment can be correctly matched and connected.

In addition, when the transition probability of a candidate continuous high-precision road section is 0, it means that the candidate continuous high-precision road section is disconnected from the current high-precision road section, that is, the matching result is discontinuous. This is because the high-precision map itself may not be complete, and the road network may be missing, so the real matching results will be discontinuous. In this embodiment, the transition probability score of the next candidate road segment that does not have a topological connection relationship with the current high-precision road segment is scored as 0, which is compatible with the lack of high-precision road network.

In this embodiment, for the discontinuity situation that occurs in the matching path, that is, the observation score of the candidate continuous high-precision road section in the current high-precision road section is 1, but the transition probability score of the candidate continuous high-precision road section is 0 In this case, the current high-precision road section can be used as the end section of the matched path, and the high-precision road section that is continuous with the current high-precision road section candidate can be used as the starting section of the next matching path, and the correction provided in this embodiment can be followed from this section. The subsequent Viterbi algorithm continues to determine the matching path. Specifically, FIG. 1d is a schematic diagram of a matching path provided in Embodiment 1 of the present invention. As shown in Figure 1d, high-precision road sections a, b, and c are determined matching road sections. There is no topological relationship between the current high-precision road section c and the candidate continuous high-precision road section d, that is, the lower part of the current high-precision road section c. The transition probability score of a candidate continuous high-precision road segment d is 0. At this time, the matching path is disconnected from c, and c is regarded as the end segment of the matching path a-b-c, and d is regarded as the beginning of the next matching path d-e-f. section.

(3) For any current trajectory point, starting from the first high-precision road section that can be observed by the current trajectory point, the transition probability score of the first high-precision road section is used as the weight value, and based on the weight The observation score of the first first high-precision road section is weighted by the value to obtain the Viterbi score of the first first high-precision road section. The obtained Viterbi score is used as the observation score of the next candidate continuous high-precision road section in turn, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform Viterbi on the candidate continuous high-precision road section. The score calculation operation is performed until the Viterbi score of the last first high-precision road section that can be observed at the last trajectory point is calculated, in which the candidate continuous high-precision road section is the first high-precision road section or the second high-precision road section. section.

In this embodiment, the transition probability score of a high-precision road section with an observation value of 0.5 is smaller than the transition probability score of a high-precision road section with an observation value of 1, which can increase the weight of the high-precision road section that matches each trajectory point in the entire high-precision navigation path. , and can reduce the weight of high-precision road segments that may match each trajectory point in the entire high-precision navigation path. By scoring the transition probability score of a high-precision road section with an observation score of 0.5 as 1, that is, setting the weight value to 1, the road section with an observation score of 0.5 will not be weighted in the Viterbi calculation process, so that it can Avoid the strong correlation between the final matching path score and the number of road segments.

(4) Select the Viterbi score with the largest value from the multiple Viterbi scores calculated, and use the multiple continuous high-precision road sections corresponding to the Viterbi score with the largest value as multiple targets with continuous topological relationships. High precision road section.

It should be noted that for the starting section of the high-precision navigation path, the proportion of the first high-precision section within the first distance threshold range before and after the starting section reaches the first proportion threshold. The first distance threshold range may be 100-300 meters, and the first proportion threshold may be 60%-80%. For the end section of the high-precision navigation path, the proportion of the first high-precision section within the second distance threshold range before and after the end section reaches the second proportion threshold. The second distance threshold range may be 100 to 300 meters, and the second proportion threshold may be 60% to 80%. In this embodiment, the accuracy of the matching results can be further ensured by increasing the proportion of high-precision road sections that are determined to match within the range before and after the start section and end section of the high-precision navigation path.

In this embodiment, for high-precision road sections that cannot meet all matching conditions, that is, for high-precision road sections that may match the trajectory points in the traditional navigation path, this embodiment retains the observation information of this type of high-precision road section and combines the observations Information, as well as the topological connection relationship between the high-precision road section and other road sections, a high-precision navigation path that matches the traditional navigation path is obtained, which avoids the removal of track points in the traditional navigation path that may match high-precision road sections in related technologies. The phenomenon of mismatching or missed matching caused by observation information makes the matching results of traditional navigation paths and high-precision maps compatible with the process errors, angle errors and precision errors between traditional navigation maps and high-precision maps, and the matching results are more precise. In addition, this embodiment can simplify the screening process by using observation values and transition probability values to score and calculate Viterbi scores, and efficiently screen out high-precision road sections that are continuous with the current high-precision road section.

Embodiment 2

Figure 2a is a flow chart of a method for determining a high-precision navigation path provided in Embodiment 2 of the present invention. In this embodiment, based on the above embodiment, the process of correcting the observed values of the intersecting road sections and the entrance and exit road sections is carried out. Refinement, as shown in Figure 2a, the method provided by this embodiment includes:

S200: Obtain the traditional navigation path provided by the traditional navigation map, and match the traditional navigation path with the high-precision map.

S210. During the matching process, determine the high-precision road section corresponding to each trajectory point of the traditional navigation path in the high-precision map, where the corresponding high-precision road section includes the first high-precision road section that is determined to match each trajectory point. , and the second high-precision road segment that may match each trajectory point.

Among them, the first high-precision road section is a high-precision road section that meets all the following matching conditions, and the second high-precision road section is a high-precision road section that meets any one or two of the following matching conditions, where the matching conditions include: for any current trajectory point , the label of the traditional navigation path corresponding to the current trajectory point is consistent with the label information, position information, and angle information of the high-precision road section corresponding to the current trajectory point.

S220: Score the observation values of the first high-precision road section and the second high-precision road section respectively to obtain the first observation score corresponding to the first high-precision road section and the second observation score corresponding to the second high-precision road section.

Specifically, the observation value of each first high-precision road section is scored as 1, and the observation value of each second high-precision road section is scored as 0.5.

S230. According to the topological relationship of the road network, identify the intersecting road sections in the traditional navigation path.

Among them, the cross-interpolation road section is: the target observation road section that can be observed by at least two consecutive trajectory points in the traditional navigation path does not have a topological connection relationship, and when the target observation road section is a non-road entrance and exit section, it is not a road. In the case of start road segment and end road segment, the navigation road segment corresponding to at least two consecutive trajectory points in the traditional navigation path.

Among them, there is no topological connection relationship between the target observation road sections that can be observed by at least two consecutive trajectory points in the traditional navigation path, including the following two situations:

(1) Among the two consecutive trajectory points on the traditional navigation path, one of the trajectory points can observe a certain high-precision road section. In the driving direction along the traditional navigation path, the trajectory point is traveling in front of the other The trajectory point cannot observe the high-precision road section, and the high-precision road section that is continuous with the high-precision road section cannot be observed.

(2) Among the two consecutive trajectory points on the traditional navigation path, one of the trajectory points can observe a certain high-precision road section. In the driving direction along the traditional navigation path, the trajectory point is traveling on the other trajectory in front of it. There is no topological connection between the high-precision road section that can be observed by the point and the high-precision road section that can be observed by the previous trajectory point.

Specifically, FIG. 2b is a schematic diagram of a transverse road section provided in Embodiment 2 of the present invention. As shown in Figure 2b, the traveling direction of the traditional navigation path 7 is the direction indicated by the arrow in Figure 2b. The trajectory point u in the traditional navigation path can observe the high-precision road section po in the high-precision map 8, that is, the trajectory point u has an intersection with the road range A of the high-precision road section po. In the case of non-entrance and exit road sections, the high-precision road sections that can be observed by trajectory point u and trajectory point v in the traditional navigation path should be continuous. As shown in Figure 2b, since the trajectory point v cannot observe the high-precision road section po, nor can it observe the high-precision road section that is continuous with the high-precision road section po, it can be explained that the road section uv in the traditional navigation path is a transverse road section.

S240. Correct the observation scores of the high-precision road sections that can be observed by the trajectory points corresponding to the transverse road sections to the second observation scores.

If there is a trajectory point that successfully matches the high-precision road segment in the transverse interpolation road segment, for example, as shown in Figure 2b, the trajectory point u intersects the road range of the target observation road segment po, and the driving direction of the traditional navigation path is consistent with the high-precision road segment. The angles are consistent, and if the labels of the two are also consistent, for example, the high-precision road section po is an urban expressway ramp, the traditional navigation path corresponds to the urban expressway, and the labels of both belong to the urban expressway. At this time, the high-precision road section can be The road section po is determined to be a high-precision road section that matches the trajectory point u in the transverse road section, and the high-precision road section po can be scored as 1.

It should be noted that for the transverse road section shown in Figure 2b, even if the trajectory point u matches the high-precision road section po successfully, this is a matching error. Because in the application scenario where the high-precision road section is a ramp section as shown in Figure 2b, it is impossible for the navigation section to cross the ramp section, and the normal vehicle driving path will not end on the ramp.

In this embodiment, the way to deal with the mismatching situation of the cross-interpolation road segment is to correct the observation scores of the high-precision road sections that can be observed by the trajectory points corresponding to the cross-intersection road segment to the second observation score. Specifically, The correction of the observation score of the high-precision road section that can be observed in the transverse road section is 0.5. Since the beginning section and the ending section of the high-precision navigation path are both the first high-precision section, that is, the observation score is 1. For the high-precision road section that can be observed in the cross-intersection road section, the observation score of the high-precision road section is corrected to 0.5, which eliminates the possibility of the high-precision road section being the starting point and end point of the final matching path, and solves the problem of cross-intersection road section The problem of matching errors, for example, as shown in Figure 2b, the high-precision road section po with an observation score of 0.5 will not be used as the end section of the matching path.

S250. In the opposite direction to the traditional navigation path, start from the target high-precision road section that can be observed from the trajectory point of the transverse road section to find whether there is an entrance and exit road section. If the entrance and exit road section can be found, then combine the target high-precision road section and the entrance and exit road section. The observation scores of all high-precision road sections are corrected to the second observation scores, and the observation scores corresponding to the entrance and exit sections are corrected to the first observation scores.

Those skilled in the art can understand that the reason why intersecting road sections appear in traditional navigation routes is usually due to the accuracy difference between the entrances and exits of high-precision road sections and the entrances and exits corresponding to traditional navigation maps. In this embodiment, by inserting road sections horizontally, it is possible to search whether there are entrance and exit road sections in the high-precision map along the opposite direction of the traditional navigation path within a set range, where the set range can be 700 to 2000 meters. If there are no entrance and exit sections, it means that the entrances and exits corresponding to the traditional navigation path are missing in the high-precision map. If there are entrance and exit sections, it can be stated that the intersecting road sections in the traditional navigation path should match the entrance and exit sections of the high-precision map. At this time, the observation scores of all high-precision road sections between the target high-precision road section and the entrance and exit sections can be It is corrected to the second observation score, which can be corrected to 0.5 specifically, that is, all high-precision road sections between the target high-precision road section and the entrance and exit road sections are corrected to the second high-precision road section that may match, and the observation scores corresponding to the entrance and exit road sections are corrected. is the first observation score, which can be modified to 1 specifically, that is, the entrance and exit road section is corrected to the first high-precision road section that is determined to match.

For example, FIG. 2c is a schematic diagram of entrance and exit section matching provided by Embodiment 2 of the present invention. Figure 2c is based on Figure 2b and adds the observation of querying the entrance and exit sections. As shown in Figure 2c, starting from the target observation section po that can be observed in the cross-intersection section, along the direction opposite to the direction indicated by the arrow, the exit section rt in the high-precision section can be found. This exit section rt is different from the traditional There is a difference in the position accuracy of exit xy in the navigation map. At this time, the observation score of the exit section is corrected to 1, that is, the exit section rt is regarded as a high-precision road section that matches the traditional navigation path, and all high-precision sections between the target observation section po and the entrance and exit section rt (including The observation scores of high-precision road section rq, high-precision road section qp and high-precision road section po) are corrected to the second observation score of 0.5, that is, these road sections are determined as high-precision road sections that may match the traditional navigation path. In this way, in the subsequent determination process of the high-precision navigation path, since the beginning section and the end section of the high-precision navigation path are both high-precision sections with an observation score of 1, after completing the above observation score correction, the exit section will be as At the end of the matching route, each possible matching road segment will not participate in the subsequent determination of the high-precision navigation path.

In this embodiment, during the observation process, the observation scores of the entrance and exit sections are measured through the road topology relationship, as well as the observation scores of all high-precision road sections between the target observation section and the entrance and exit sections that can be observed by the trajectory points of the intersecting road sections. The value is corrected, which solves the problem of excessive accuracy of entrances and exits between high-precision maps and traditional navigation paths, and ensures the accuracy of matching results.

S260. According to the topological relationship of the high-precision road sections and the corrected observation scores of different road sections, transfer probability scores are performed on each first high-precision road section and each second high-precision road section, and the first high-precision road section corresponding to each first high-precision road section is obtained. The transition probability score, and the second transition probability score corresponding to each second high-precision road section.

Among them, the observation score correction is to correct whether the high-precision road section is the first high-precision road section or the second high-precision road section. For example, correcting the observation score from 0.5 to 1 means correcting the high-precision road section from the second high-precision road section to the first high-precision road section. Correcting the observation score from 1 to 0.5 means changing the high-precision road section from the second high-precision road section to the first high-precision road section. The first high-precision road section is modified into the second high-precision road section. In this embodiment, for high-precision road sections whose observation scores have not changed, the transition probability scoring is performed based on the current observation score. For high-precision road sections with corrected observation scores, when scoring transition probabilities, the transition probability scores are based on the corrected observation scores. For the specific scoring process, please refer to the description of the above embodiment.

S270. For any current trajectory point, starting from the first first high-precision road section that can be observed by the current trajectory point, use the transition probability score of the first first high-precision road section as the weight value, and based on the weight The observation score of the first first high-precision road section is weighted by the value to obtain the Viterbi score of the first first high-precision road section.

S280. Use the obtained Viterbi score in turn as the observation score of the next candidate continuous high-precision road section, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform the execution on the candidate continuous high-precision road section. The calculation operation of the Viterbi score is until the Viterbi score of the last first high-precision road section that can be observed at the last trajectory point is calculated.

Among them, the observation score of each high-precision road section located between high-precision road sections with an observation score of 1 can be 1 or 0.5. In the calculation process of Viterbi score, if the observation score of the high-precision road section that can be observed by the trajectory point in the traditional navigation path is corrected, the calculation process of Viterbi score is based on the corrected observation score. , and calculated based on the transition probability score obtained from the corrected observation score.

Specifically, for any current trajectory point, starting from the high-precision road section with the first observation score of 1 that can be observed at the current trajectory point, the transition probability of the high-precision road section with the first observation score of 1 is divided into The value is used as the weight value, and the observation score of the high-precision road section with the first observation score of 1 is weighted based on the weight value to obtain the Viterbi score of the high-precision road section with the first observation score of 1. The obtained Viterbi score is used as the observation score of the next candidate continuous high-precision road section in turn, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform Viterbi on the candidate continuous high-precision road section. The score calculation operation is performed until the Viterbi score of the high-precision road section with the last observation score of 1 that can be observed at the last trajectory point is calculated.

S290. Select the Viterbi score with the largest value from the calculated Viterbi scores, and use the multiple continuous high-precision road sections corresponding to the Viterbi score with the largest value as multiple target high-precision road sections with continuous topological relationships. Precision road sections, and based on multiple continuous target high-precision road sections, generate a high-precision navigation path that matches the traditional navigation path.

For the specific implementation of steps S260 to S290, reference may be made to the description of the above embodiment, and details will not be described again here.

In this embodiment, during the observation process, the intersecting road sections in the traditional navigation path are identified, and the observation score of the target observation road section that can be observed by the intersecting road section is corrected to 0.5, thereby eliminating the target observation road section. As the possibility of finally matching the start and end points of the path, the problem of mismatching of transverse road segments is solved. In addition, during the observation process, the observation scores of the entrance and exit sections are corrected through the road topology relationship, as well as the observation scores of all high-precision road sections between the target observation section and the entrance and exit sections that can be observed by the trajectory points of the intersecting road sections. , which solves the problem of excessive accuracy difference between high-precision maps and traditional navigation path entrances and exits, and ensures the accuracy of matching results.

Embodiment 3

Figure 3 is a structural block diagram of a device for determining a high-precision navigation path provided in Embodiment 3 of the present invention. As shown in Figure 3, the device includes: a traditional navigation path acquisition module 310, an observation module 320 and a matching module 330, where,

The traditional navigation path acquisition module 310 is configured to acquire the traditional navigation path provided by the traditional navigation map, and match the traditional navigation path with the high-precision map;

The observation module 320 is configured to, during the matching process, determine a high-precision road section corresponding to each track point of the traditional navigation path in the high-precision map, and the corresponding high-precision road section includes a determined match with each track point. The first high-precision road section, and the second high-precision road section that may match each trajectory point;

The matching module 330 is configured to select multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, and generate a target high-precision road section that is consistent with the traditional navigation path based on the multiple continuous target high-precision road sections. Matching high-precision navigation path, in which the beginning section and the end section of the high-precision navigation path are both the first high-precision section;

Optional, matching module 330 includes:

Optional, observation value scoring unit, specifically configured as:

Score the observation value of each first high-precision road section as 1, and score the observation value of each second high-precision road section as 0.5;

Correspondingly, the transition probability value scoring unit is specifically configured as:

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 0.5, then the transition probability score of the continuous high-precision road segment candidate is scored as 1.

Correspondingly, the transition probability value scoring unit includes:

The transition probability value scoring sub-unit is configured to score the transition probability for each first high-precision road section and each second high-precision road section according to the topological relationship of the high-precision road section and the corrected observation scores of different road sections.

Optionally, the label information of the traditional navigation path corresponding to the current trajectory point and the label information of the high-precision road section corresponding to the current trajectory point are consistent: the label of the corresponding traditional navigation path within the third distance threshold range before and after the current trajectory point. The label information of the high-precision road section corresponding to the current trajectory point is consistent.

The high-precision navigation path device provided by the embodiment of the present invention can execute the high-precision navigation path determination method provided by any embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method. For technical details that are not described in detail in the above embodiments, please refer to the method for determining a high-precision navigation path provided by any embodiment of the present invention.

Embodiment 4

Figure 4 is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention. As shown in Figure 4, the electronic device includes:

Memory 510 storing executable program code;

processor 520 coupled to memory 510;

The processor 520 calls the executable program code stored in the memory 510 to execute the method for determining a high-precision navigation path provided by any embodiment of the present invention.

Based on the above embodiments, another embodiment of the present invention provides a vehicle, which includes the device as described in any of the above embodiments, or includes the electronic device as described above.

Embodiment 5

Figure 5 is a schematic diagram of a vehicle provided in Embodiment 5 of the present invention. As shown in Figure 5, the vehicle includes a speed sensor 61, an ECU (Electronic Control Unit) 62, a GPS (Global Positioning System) positioning device 63, and a T-Box (Telematics Box). 64. Among them, the speed sensor 61 is used to measure the vehicle speed and uses the vehicle speed as the empirical speed for model training; the GPS positioning device 63 is used to obtain the current geographical location of the vehicle; the T-Box64 can be used as a gateway to communicate with the server; the ECU62 can perform the above-mentioned high-level tasks. Method for determining precision navigation paths.

In addition, the vehicle may also include: V2X (Vehicle-to-Everything, Internet of Vehicles) module 65 , radar 66 and camera 67 . The V2X module 65 is used to communicate with other vehicles, roadside equipment, etc.; the radar 66 or the camera 67 is used to sense the road environment information ahead and/or in other directions to obtain original point cloud data; the radar 66 and/or the camera 67 can be configured At the front and/or at the rear of the car.

Based on the above method embodiments, another embodiment of the present invention provides a storage medium on which executable instructions are stored. When executed by the processor, the instructions enable the processor to implement high-precision navigation as described in any of the above embodiments. How to determine the path.

Those of ordinary skill in the art can understand that the accompanying drawing is only a schematic diagram of an embodiment, and the modules or processes in the accompanying drawing are not necessarily necessary for implementing the present invention.

Those of ordinary skill in the art can understand that the modules in the device in the embodiment may be distributed in the device in the embodiment according to the description of the embodiment, or may be correspondingly changed and located in one or more devices different from this embodiment. The modules of the above embodiments can be combined into one module, or further divided into multiple sub-modules.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be used Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

A method for determining a high-precision navigation path, which is characterized by including:

Obtain the traditional navigation path provided by the traditional navigation map, and match the traditional navigation path with the high-precision map;

During the matching process, a high-precision road section corresponding to each track point of the traditional navigation path in the high-precision map is determined, and the corresponding high-precision road section includes the first high-precision road section that is determined to match each track point. Precision road section, and the second high-precision road section that may match each trajectory point;

Select multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, and generate high-precision navigation that matches the traditional navigation path based on the multiple continuous target high-precision road sections. Path, wherein the starting section and the ending section of the high-precision navigation path are both the first high-precision section;

Wherein, the first high-precision road section is a high-precision road section that meets all the following matching conditions, and the second high-precision road section is a high-precision road section that meets any one or both of the following matching conditions, where the matching conditions include : For any current trajectory point, the label information of the traditional navigation path corresponding to the current trajectory point is consistent with the label information, position information, and angle information of the high-precision road section corresponding to the current trajectory point.
The method according to claim 1, characterized in that selecting a plurality of target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section includes:

Score the observation values of the first high-precision road section and the second high-precision road section respectively to obtain the first observation score corresponding to the first high-precision road section and the second observation score corresponding to the second high-precision road section, where the first The observation score is greater than the second observation score;

According to the topological relationship of the high-precision road section and the observation scores of different road sections, the transition probability scores of the first high-precision road section and the second high-precision road section are respectively obtained, and the first transition probability score corresponding to the first high-precision road section is obtained, as well as the first transition probability score corresponding to the first high-precision road section. The second transition probability score corresponding to the two high-precision road sections, wherein the second transition probability score is smaller than the first transition probability score;

For any current trajectory point, starting from the first first high-precision road section that can be observed by the current trajectory point, the transition probability score of the first first high-precision road section is used as the weight value, and based on the weight value The observation score of the first first high-precision road section is weighted to obtain the Viterbi score of the first first high-precision road section;

The obtained Viterbi score is used as the observation score of the next candidate continuous high-precision road section in turn, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform the execution on the candidate continuous high-precision road section. The calculation operation of the Viterbi score is until the Viterbi score of the last first high-precision road section that can be observed at the last trajectory point is calculated;

Select the Viterbi score with the largest value from the multiple calculated Viterbi scores, and use the multiple continuous high-precision road sections corresponding to the largest Viterbi score as multiple target high-precision road sections with continuous topological relationships. .
The method according to claim 2, characterized in that the first high-precision road section and the second high-precision road section are respectively scored with observation values to obtain the first observation score corresponding to the first high-precision road section, and the second high-precision road section. The second observation scores corresponding to high-precision road sections include:

The observation value of each first high-precision road section is scored as 1, and the observation value of each second high-precision road section is scored as 0.5; correspondingly,

According to the topological relationship of the high-precision road section and the observation scores of different road sections, the first high-precision road section and the second high-precision road section are respectively scored for transition probability to obtain the first transition probability score corresponding to the first high-precision road section, And the second transition probability score corresponding to the second high-precision road section, including:

For any current high-precision road segment, if the observation score of a high-precision road segment that is continuous with the current high-precision road segment is 1, and when the candidate continuous high-precision road segment is the same road segment as the current high-precision road segment, the candidate high-precision road segment will be The transition probability score of a continuous high-precision road section is 1+E, where E is a minimum value greater than 0; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 1, and the candidate continuous high-precision road segment can be topologically connected to the current high-precision road segment, the transition probability of the candidate continuous high-precision road segment is divided into The value is scored as 1+2E; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 1, and the continuous high-precision road segment candidate cannot be topologically connected to the current high-precision road segment, the transition probability of the continuous high-precision road segment candidate is divided into The value is scored as 0; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 0.5, then the transition probability score of the high-precision road segment that is continuous with the candidate is scored as 1.
The method according to claim 3, characterized in that when the observation score of the high-precision road section that is continuous with the current high-precision road section is 1, and the transition probability score of the candidate continuous high-precision road section is 0 :

The current high-precision road segment is used as the ending road segment of the matched path, and the high-precision road segment that is continuous with the candidate is used as the starting road segment of the next matching path, and the execution continues according to the topological relationship of the high-precision road segment and the observation scores of different road segments. Scoring operation for transition probability scores.
The method of claim 2, further comprising:

According to the topological relationship of the road network, identify the intersecting road sections in the traditional navigation path, wherein the intersecting road sections are: the target observation road section that can be observed by at least two consecutive trajectory points in the traditional navigation path does not have topology connection relationship, and in the case that the target observation road section is not a road entrance and exit section, nor is it a starting section or an ending section of the road, the navigation section corresponding to the at least two continuous trajectory points;

Correct the observation scores of the high-precision road sections that can be observed by the trajectory points corresponding to the transverse interpolation road sections to the second observation scores;

Correspondingly, according to the topological relationship of the high-precision road section and the observation scores of different road sections, the transition probability scores are respectively performed on the first high-precision road section and the second high-precision road section, including:

According to the topological relationship of the high-precision road sections and the corrected observation scores of different road sections, the transition probability is scored for each first high-precision road section and each second high-precision road section.
The method according to claim 5, characterized in that, according to the topological relationship of the high-precision road sections and the corrected observation scores of different road sections, a transition probability score is performed on each first high-precision road section and each second high-precision road section. Previously, the method also included:

According to the direction opposite to the traditional navigation path, start from the target observation road section that can be observed by the trajectory point of the intersecting road section to find whether there is an entrance and exit road section. If the entrance and exit road section can be found, then the target observation road section will be searched. The observation scores of all high-precision road sections between the road segment and the entrance and exit road sections are corrected to the second observation scores, and the observation scores corresponding to the entrance and exit road sections are corrected to the first observation scores.
The method according to claim 1, wherein the consistency of the location information means that the trajectory points of the traditional navigation path intersect with the road range corresponding to the high-precision road section;

Wherein, when the traditional navigation path passes through an urban road intersection, the road range corresponding to the urban road intersection is: all high-precision road sections within the urban road intersection share the same road range; or,

In the case where the traditional navigation path passes through a non-two-way road that contains two road edge lines, the road range corresponding to the non-two-way road is obtained by connecting the road boundaries of each high-precision road segment end to end; or,

In the case where the traditional navigation path passes through a two-way road that only contains one road edge line, the road range corresponding to the two-way road is: formed by connecting the road edge line of the current two-way road and the edge line of the opposite two-way road end to end, Wherein, the opposite two-way road is a two-way road adjacent to the current two-way road in the high-precision map and opposite to the road direction of the current two-way road.
The method according to claim 1, characterized in that:

For the starting section of the high-precision navigation path, the proportion of the first high-precision section within the first distance threshold range from the starting section reaches the first proportion threshold;

For the end section of the high-precision navigation path, the proportion of the first high-precision section within the second distance threshold range from the end section reaches the second proportion threshold.
The method according to claim 1, wherein the label information of the traditional navigation path corresponding to the current trajectory point is consistent with the label information of the high-precision road section corresponding to the current trajectory point means: the third time before and after the current trajectory point. The label of the corresponding traditional navigation path within the distance threshold range is consistent with the label information of the high-precision road section corresponding to the current trajectory point.
A device for determining a high-precision navigation path, which is characterized by including:

The traditional navigation path acquisition module is configured to obtain the traditional navigation path provided by the traditional navigation map, and match the traditional navigation path with the high-precision map;

The observation module is configured to determine, during the matching process, a high-precision road section corresponding to each track point of the traditional navigation path in the high-precision map, where the corresponding high-precision road section includes a link to each track point. Determine the matching first high-precision road segment and the second high-precision road segment that may match each trajectory point;

The matching module is configured to filter out multiple target high-precision road sections with continuous topological relationships from the first high-precision road section and the second high-precision road section, and generate the traditional navigation path based on the multiple continuous target high-precision road sections. A matching high-precision navigation path, wherein the starting section and the ending section of the high-precision navigation path are both the first high-precision section;

Among them, the first high-precision road section is a high-precision road section that meets all the following matching conditions, and the second high-precision road section is a high-precision road section that meets any one or both of the following matching conditions, where the matching conditions include: for any one The current trajectory point. The label information of the traditional navigation path corresponding to the current trajectory point is consistent with the label information, position information, and angle information of the high-precision road section corresponding to the current trajectory point.
The device according to claim 10, characterized in that the matching module includes:

The observation value scoring unit is configured to score the observation values of the first high-precision road section and the second high-precision road section respectively, and obtain the first observation score corresponding to the first high-precision road section, and the second observation score corresponding to the second high-precision road section. Observation score, where the first observation score is greater than the second observation score;

The transition probability value scoring unit is configured to score the first high-precision road section and the second high-precision road section respectively according to the topological relationship of the high-precision road section and the observation scores of different road sections, and obtain the transition probability value corresponding to the first high-precision road section. The first transition probability score, and the second transition probability score corresponding to the second high-precision road section, wherein the second transition probability score is smaller than the first transition probability score;

The Viterbi score calculation unit is configured to calculate the transition probability score of the first high-precision road section for any current trajectory point starting from the first high-precision road section that can be observed by the current trajectory point. As the weight value, and weighting the observation score of the first first high-precision road section based on the weight value, the Viterbi score of the first first high-precision road section is obtained; in turn, the obtained Viterbi score is The observation score of the next candidate continuous high-precision road section, and based on the transition probability score corresponding to the candidate continuous high-precision road section, continue to perform the calculation operation of the Viterbi score for the candidate continuous high-precision road section until Calculate the Viterbi value of the last first high-precision road section that can be observed at the last trajectory point;

The target high-precision road section determination unit is configured to select the Viterbi score with the largest value from the multiple calculated Viterbi scores, and use the multiple continuous high-precision road sections corresponding to the maximum Viterbi score as Multiple target high-precision road sections with continuous topological relationships;

The high-precision navigation path generating unit is configured to generate a high-precision navigation path that matches the traditional navigation path based on multiple consecutive target high-precision road sections.
The device according to claim 11, characterized in that the observation value scoring unit is specifically configured as:

The observation value of each first high-precision road section is scored as 1, and the observation value of each second high-precision road section is scored as 0.5; correspondingly,

The transition probability value scoring unit is specifically configured as:

For any current high-precision road segment, if the observation score of a high-precision road segment that is continuous with the current high-precision road segment is 1, and when the candidate continuous high-precision road segment is the same road segment as the current high-precision road segment, the candidate high-precision road segment will be The transition probability score of a continuous high-precision road section is 1+E, where E is a minimum value greater than 0; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 1, and the candidate continuous high-precision road segment can be topologically connected to the current high-precision road segment, the transition probability of the candidate continuous high-precision road segment is divided into The value is scored as 1+2E; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 1, and the continuous high-precision road segment candidate cannot be topologically connected to the current high-precision road segment, the transition probability of the continuous high-precision road segment candidate is divided into The value is scored as 0; or,

If the observation score of the high-precision road segment that is continuous with the current high-precision road segment candidate is 0.5, then the transition probability score of the high-precision road segment that is continuous with the candidate is scored as 1.
The device according to claim 12, characterized in that when the current high-precision road section is selected and the observation score of the continuous high-precision road section is 1, and the transition probability score of the candidate continuous high-precision road section is 0:

The current high-precision road segment is used as the ending road segment of the matched path, and the high-precision road segment that is continuous with the candidate is used as the starting road segment of the next matching path, and the execution continues according to the topological relationship of the high-precision road segment and the observation scores of different road segments. Scoring operation for transition probability scores.
The device according to claim 11, characterized in that the device further includes:

The transverse road segment identification module is configured to identify the transverse road segment in the traditional navigation path according to the topological relationship of the road network, wherein the transverse road segment is: a path that can be reached by at least two consecutive trajectory points in the traditional navigation path. The observed target observation road section does not have a topological connection relationship, and in the case that the target observation road section is not a road entrance or exit section, nor is it a starting section or an ending section of a road, the navigation section corresponding to the at least two continuous trajectory points;

The first observation score correction module is configured to correct the observation scores of the high-precision road sections that can be observed by the trajectory points corresponding to the transverse road sections to the second observation scores;

Correspondingly, the transition probability value scoring unit includes:

The transition probability value scoring sub-unit is configured to score each first high-precision road section and each second high-precision road section with a transition probability according to the topological relationship of the high-precision road section and the corrected observation scores of different road sections.
The device according to claim 14, characterized in that the device further includes:

The second observation score correction module is specifically configured to score the transition probability for each first high-precision road section and each second high-precision road section respectively according to the topological relationship of the high-precision road section and the corrected observation scores of different road sections. , in the opposite direction to the traditional navigation path, start from the target observation road section that can be observed by the trajectory point of the intersecting road section to find whether there is an entrance and exit road section. If the entrance and exit road section can be found, then the target The observation scores of all high-precision road sections between the observed road section and the entrance and exit road sections are corrected to the second observation scores, and the observation scores corresponding to the entrance and exit road sections are corrected to the first observation scores.
The device according to claim 10, wherein the consistency of the location information means that the trajectory points of the traditional navigation path intersect with the road range corresponding to the high-precision road section;

Wherein, when the traditional navigation path passes through an urban road intersection, the road range corresponding to the urban road intersection is: all high-precision road sections within the urban road intersection share the same road range; or,

In the case where the traditional navigation path passes through a non-two-way road that contains two road edge lines, the road range corresponding to the non-two-way road is obtained by connecting the road boundaries of each high-precision road segment end to end; or,

In the case where the traditional navigation path passes through a two-way road that only contains one road edge line, the road range corresponding to the two-way road is: formed by connecting the road edge line of the current two-way road and the edge line of the opposite two-way road end to end, Wherein, the opposite two-way road is a two-way road adjacent to the current two-way road in the high-precision map and opposite to the road direction of the current two-way road.
The device according to claim 10, characterized in that, for the starting section of the high-precision navigation path, the proportion of the first high-precision section within the first distance threshold range from the starting section reaches the first proportion threshold;

For the end section of the high-precision navigation path, the proportion of the first high-precision section within the second distance threshold range from the end section reaches the second proportion threshold.
The device according to claim 10, wherein the label information of the traditional navigation path corresponding to the current trajectory point is consistent with the label information of the high-precision road section corresponding to the current trajectory point means: the third time before and after the current trajectory point. The label of the corresponding traditional navigation path within the distance threshold range is consistent with the label information of the high-precision road section corresponding to the current trajectory point.
An electronic device, characterized by including:

one or more processors;

a storage device for storing one or more programs,

When the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method as described in any one of claims 1-9.
A storage medium on which a computer program is stored, characterized in that when the program is executed by a processor, the method according to any one of claims 1-9 is implemented.
A vehicle, characterized in that the vehicle includes the device according to any one of claims 10-18, or the electronic device according to claim 19.