Positioning initialization method and initialization device based on road features, electronic equipment and storage medium

Abstract

The invention provides a positioning initialization method and initialization device based on road features, electronic equipment and a storage medium, and relates to the technical field of inertial navigation. A road network map is obtained, the road network map comprises road network features of roads, and the road network features comprise at least one of geometric features, physical features and topological features; obtaining track features of a motion carrier, wherein feature parameters of the track features correspond to feature parameters of the road network features of the road; and performing similarity matching on the characteristic parameters of the track characteristics and the characteristic parameters of the road network characteristics of the roads in the road network map, and determining the positioning information of the motion carrier. According to the method, the road network map of the inertial navigation system is obtained, and the road network map is matched with the track characteristics of the motion carrier to determine the positioning information of the motion carrier, so that autonomous initialization positioning of the inertial navigation system is realized, and the applicability of the inertial navigation system is improved.

Description

Technical Field

[0001] The present invention relates to the field of inertial navigation technology, and in particular to a positioning initialization method, initialization device, electronic equipment and storage medium based on road features. Background Art

[0002] An inertial navigation system (INS) is a navigation system that is independent of external signals. It uses inertial sensors (accelerometers and gyroscopes) to measure the acceleration and angular velocity of a moving object. It then infers the object's position, velocity, and attitude using a strapdown inertial navigation update algorithm. It is widely used in various applications, particularly in military, aerospace, and autonomous driving, due to its high accuracy and anti-interference capabilities. However, INS cannot autonomously determine an initial position, and low-precision inertial measurement units cannot even sense the Earth's rotational angular velocity, making it impossible to determine an initial heading.

[0003] To address the problem of inertial navigation systems' inability to initialize themselves, they are often combined with other positioning technologies (such as satellite navigation systems, ground base stations, and visual or radar ranging) for integrated navigation initialization. However, the integrated navigation initialization method depends on the positioning accuracy and usage scenarios of the other navigation systems, which limits the use of inertial navigation systems. Summary of the Invention

[0004] The present invention provides a positioning initialization method, initialization device, electronic device and storage medium based on road features, which are used to solve the defect in the prior art that the use of the inertial navigation system is restricted by other navigation systems due to combined navigation initialization. By constructing the inertial navigation system's own road network map, matching the road network map with the trajectory characteristics of the moving carrier to determine the positioning information of the moving carrier, the autonomous initial positioning of the inertial navigation system is achieved, and the applicability of the inertial navigation system is improved.

[0005] The present invention provides a positioning initialization method based on road features, comprising the following steps: Acquire a road network map, the road network map including road network features, the road network features including at least one of geometric features, physical features, and topological features, the geometric features being used to characterize spatial attributes of roads, the physical features being used to characterize environmental attributes of roads, and the topological features being used to characterize connectivity between roads; Acquiring trajectory characteristics of the moving carrier, wherein characteristic parameters of the trajectory characteristics correspond to characteristic parameters of the road network characteristics of the road; The characteristic parameters of the trajectory feature are matched with the characteristic parameters of the road network feature of the road in the road network map for similarity, and the positioning information of the moving carrier is determined. The positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0006] According to a positioning initialization method based on road features provided by the present invention, obtaining a road network map includes: Acquire road characteristics based on open digital maps and / or sensors to obtain road network characteristics of the road; The road network features of the roads are randomly and evenly stored as a plurality of data blocks, and a dynamically updateable spatial index structure is established for each of the plurality of data blocks to obtain the road network map.

[0007] According to a positioning initialization method based on road features provided by the present invention, the geometric features include the length and slope of a first road, the length and slope of a second road, and the turning angle between the first road and the second road; the sensor includes an inertial measurement unit; and the road network features of each road are obtained based on an open digital map and / or sensors, including: Converting the coordinate system corresponding to the coordinates of the roads in the open digital map from the Earth-centered Earth-fixed coordinate system to the local horizontal coordinate system to obtain a converted open digital map; Extracting turning intersections between different roads based on the converted open digital map to obtain a plurality of turning intersections corresponding to the plurality of roads; Determining, according to the plurality of turning intersections corresponding to the plurality of roads, an intersection of a first road and a second road among the plurality of roads as a target intersection; Calculating the distance between a target endpoint of the first road and the target intersection, and the distance between a target endpoint of the second road and the target intersection in a local horizontal coordinate system to obtain the length of the first road and the length of the second road, wherein the target endpoint is used to represent a non-intersecting endpoint; Calculating the angle between the first road and the second road relative to the target intersection according to a vector dot product formula to obtain a turning angle between the first road and the second road; Acquiring acceleration data of the first road by using the inertial measurement unit, wherein the inertial measurement unit is disposed on the first road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the first road and a direction perpendicular to the direction of the first road; Calculating the acceleration data according to the accelerometer leveling to obtain a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, wherein the roll angle and the pitch angle represent the slope of the first road; Acquiring acceleration data of the second road by using the inertial measurement unit, wherein the inertial measurement unit is disposed on the second road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the second road and a direction perpendicular to the direction of the first road; The acceleration data is calculated according to the accelerometer leveling to obtain a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, where the roll angle and the pitch angle represent the slope of the second road.

[0008] According to a road feature-based positioning initialization method provided by the present invention, the road network features of the road are randomly and evenly stored as multiple data blocks, and a dynamically updateable spatial index structure is established for each of the multiple data blocks to obtain the road network map, including: Evenly distributing the road network features of the road into multiple storage buckets using a hash algorithm, each storage bucket storing corresponding road network feature data; An incremental kd tree is used to dynamically construct and manage the road network feature data in each storage bucket to obtain the road network map.

[0009] According to a road feature-based positioning initialization method provided by the present invention, the hash algorithm adopts a consistent hashing algorithm, and the road network features of the road are evenly distributed into multiple storage buckets by the hash algorithm, including: Mapping each storage bucket and the corresponding road network feature data to the same hash ring space through the consistent hashing algorithm; When the data in a bucket needs to be adjusted, the hash ring position of the newly added or deleted bucket is calculated, and only the road network feature data between adjacent positions on the hash ring is migrated to the newly added bucket, or the road network feature data of the deleted bucket is reallocated to the adjacent bucket, keeping the road network feature data storage position in non-adjacent areas unchanged.

[0010] According to a positioning initialization method based on road features provided by the present invention, the incremental kd tree is used to dynamically construct and manage the road network feature data of each storage bucket to obtain the road network map, including: Determining the data dimension for constructing the kd tree according to the feature type of the road network feature data in each storage bucket, and storing the road network feature data according to the data dimension; Select the feature data of one road in the road network feature data of each storage bucket as the root node to initialize the kd tree; For the road network features of the newly added roads, the following operations are recursively performed: a partition dimension is selected according to the current tree depth or data distribution, and the road network features of the newly added roads are recursively inserted into the left subtree or right subtree of the kd tree according to the partition dimension. If the insertion causes the subtree to be unbalanced, the corresponding subtree is locally rebalanced through node rotation or subtree reconstruction until all the newly added road network feature data are inserted, and an incremental kd tree structure that supports nearest neighbor queries and range queries is obtained as the road network map.

[0011] According to a positioning initialization method based on road features provided by the present invention, the step of obtaining trajectory features of a moving carrier includes: Acquiring continuous trajectory data of the moving carrier through an inertial navigation system; Dividing the continuous trajectory data into a plurality of trajectory segments, and extracting the length of each trajectory segment, the angle between adjacent trajectory segments, and the turning direction to obtain the trajectory features; The similarity matching of the characteristic parameters of the trajectory features with the characteristic parameters of the road network features of the roads in the road network map to determine the positioning information of the moving carrier includes: The characteristic parameters of the trajectory features are matched with the characteristic parameters of the road network features of the roads in the road network map for similarity, the initial position of the carrier is inferred based on the hidden Markov model and the Viterbi algorithm, and the initial position is converged to a unique solution through trajectory growth to obtain the positioning information of the moving carrier.

[0012] According to a positioning initialization method based on road features provided by the present invention, the method includes performing similarity matching on characteristic parameters of the trajectory features with characteristic parameters of the road network features of the roads in the road network map, inferring the initial position of the carrier based on the hidden Markov model and the Viterbi algorithm, and converging the initial position to a unique solution through trajectory growth to obtain the positioning information of the moving carrier, including: Constructing a hidden Markov model, wherein the hidden Markov model includes observation probability and transition probability, wherein the observation probability is determined based on a matching result of characteristic parameters of the trajectory characteristics and the road network characteristics, and the transition probability is determined based on a topological connection relationship of the roads; According to the hidden Markov model and the currently observed trajectory feature sequence, the optimal road sequence is calculated through the dynamic programming process of the Viterbi algorithm, and the center point coordinates of the first road in the state sequence obtained by backtracking are used as candidate initial position coordinates; As the motion trajectory of the moving vehicle grows, new trajectory segment features are added and the observation sequence is updated. The above two steps are repeated to dynamically adjust the candidate initial position. When all trajectory segment features correspond to only a single candidate road, the candidate initial position is determined to have converged to a unique solution, and the positioning information of the moving vehicle is output.

[0013] The present invention also provides a positioning initialization device based on road features, comprising the following modules: A road network map acquisition module is used to acquire a road network map, wherein the road network map includes a road network feature map extracted and / or acquired through sensors; A trajectory feature acquisition module, configured to acquire trajectory features of a moving carrier, wherein characteristic parameters of the trajectory features correspond to characteristic parameters of the road network features of the road; A positioning information determination module is used to perform similarity matching between the characteristic parameters of the trajectory characteristics and the characteristic parameters of the road network characteristics of the roads in the road network map to determine the positioning information of the moving carrier, wherein the positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the positioning initialization method based on road features as described above is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described positioning initialization methods based on road features.

[0016] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the positioning initialization method based on road features as described above is implemented.

[0017] The road feature-based positioning initialization method, initialization device, electronic device, and storage medium provided by the present invention construct the inertial navigation system's own road network map, match the road network map with the trajectory characteristics of the moving carrier to determine the moving carrier's positioning information, thereby realizing autonomous initial positioning of the inertial navigation system and improving the applicability of the inertial navigation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1It is a flowchart of the road feature-based positioning initialization method provided by the present invention.

[0020] Figure 2 It is a flow chart of the method for obtaining a road network map provided by the present invention.

[0021] Figure 3 This is a schematic diagram of the result after OSM map preprocessing provided by the present invention.

[0022] Figure 4 It is a schematic diagram of the road angle provided by the present invention.

[0023] Figure 5 It is a structural diagram of the IKD tree provided by the present invention.

[0024] Figure 6 It is a flow chart of the method for obtaining trajectory characteristics and positioning information provided by the present invention.

[0025] Figure 7 It is a schematic diagram of the process of obtaining trajectory features provided by the present invention.

[0026] Figure 8 It is a schematic diagram of the initial positioning of the three-segment trajectory provided by the present invention.

[0027] Figure 9 It is a schematic diagram of the initial positioning of the four-segment trajectory provided by the present invention.

[0028] Figure 10 It is a schematic diagram of the initial positioning of the five-segment trajectory provided by the present invention.

[0029] Figure 11 It is a structural schematic diagram of the positioning initialization device based on road features provided by the present invention.

[0030] Figure 12 It is a schematic diagram of the physical structure of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0031] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0032] An inertial navigation system (INS) is a navigation system that is independent of external signals. It uses inertial sensors (accelerometers and gyroscopes) to measure the acceleration and angular velocity of a moving object. It then infers the object's position, velocity, and attitude using a strapdown inertial navigation update algorithm. It is widely used in various applications, particularly in military, aerospace, and autonomous driving, due to its high accuracy and anti-interference capabilities. However, INS cannot autonomously determine an initial position, and low-precision inertial measurement units cannot even sense the Earth's rotational angular velocity, making it impossible to determine an initial heading.

[0033] To address the inability of inertial navigation systems to initialize autonomously, they are often combined with other positioning technologies (such as satellite navigation systems, ground-based base stations, and visual or radar ranging) for integrated navigation initialization. Specifically, because inertial navigation systems cannot initialize autonomously with high precision, they require assistance from external positioning sources (such as GNSS, vision systems, etc.) or manual initialization. For example, inertial navigation systems can be combined with real-time kinematic (RTK) technology, employing algorithms such as Kalman filtering to comprehensively consider data from multiple systems to determine the initial state of the inertial navigation system. Alternatively, they can be combined with a visual-inertial navigation system (VINS) to achieve initialization by matching feature points in the environment. However, the integrated navigation initialization method depends on the positioning accuracy and usage scenario of the other navigation systems, which limits the use of inertial navigation systems.

[0034] In view of this, an embodiment of the present invention provides a positioning initialization method based on road features, which obtains a road network map, wherein the road network map includes road network features of roads, wherein the road network features include at least one of geometric features, physical features, and topological features; obtains trajectory features of a moving carrier, wherein the characteristic parameters of the trajectory features correspond to the characteristic parameters of the road network features of the roads; and performs similarity matching between the characteristic parameters of the trajectory features and the characteristic parameters of the road network features of the roads in the road network map to determine the positioning information of the moving carrier. This method obtains a road network map of an inertial navigation system and matches the road network map with the trajectory features of the moving carrier to determine the positioning information of the moving carrier, thereby achieving autonomous initialization positioning of the inertial navigation system and improving the applicability of the inertial navigation system.

[0035] The technical solutions in the embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention.

[0036] Figure 1: is a flow chart of the positioning initialization method based on road features provided by the present invention. The positioning initialization method based on road features can be applied to electronic devices, which can be various types of devices with information processing capabilities during implementation. For example, the electronic device can include a personal computer, a laptop, a PDA or a server, etc.; the electronic device can also be a mobile terminal, for example, the mobile terminal can include a mobile phone, a car computer, a tablet computer or a projector, etc. Figure 1 As shown, the method includes the following: Step 101: Acquire a road network map, wherein the road network map includes road network features of roads, and the road network features include at least one of geometric features, physical features, and topological features. The geometric features are used to characterize spatial attributes of roads, the physical features are used to characterize environmental attributes of roads, and the topological features are used to characterize connection relationships of roads.

[0037] It should be noted that the road network map includes the road network features of the roads. The road network map can be obtained by extracting target features from an open data map or by collecting target features through sensors.

[0038] For example, the road network features may be extracted from the Open Street Map (OSM), or the target features of the road may be acquired through various high-precision sensors. The present invention does not limit the method for acquiring the road network map.

[0039] The road network features include at least one of geometric, physical, and topological features. The road network features may include only geometric features, both geometric and physical features, or all of them. The present invention does not limit the types of features corresponding to the road network features. For example, geometric features may include road length, slope, etc.; topological features may include turning angles, road connectivity, etc.; and physical features may include altitude, etc.

[0040] Step 102: Acquire trajectory characteristics of the moving carrier, wherein characteristic parameters of the trajectory characteristics correspond to characteristic parameters of the road network characteristics of the road.

[0041] It should be noted that there are many ways to obtain the trajectory characteristics of a moving vehicle, such as by inferring them through an inertial navigation system or extracting features through a model. The present invention does not limit the method for obtaining the trajectory characteristics of a moving vehicle. The characteristic parameters of the trajectory characteristics and the characteristic parameters of the road network characteristics can be the same. For example, the characteristic parameters of the trajectory characteristics include the length of the road, and the characteristic parameters of the road network characteristics also include the length of the road.

[0042] Step 103: Perform similarity matching on the characteristic parameters of the trajectory feature and the characteristic parameters of the road network feature of the road in the road network map to determine the positioning information of the moving carrier, wherein the positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0043] It should be noted that there are many ways to determine the positioning information of the moving vehicle by performing similarity matching on the characteristic parameters of the trajectory features with the characteristic parameters of the road network features of the roads in the road network map, such as matching via an algorithm or matching via a model. The present invention does not limit the method for determining the positioning information of the moving vehicle by performing similarity matching on the characteristic parameters of the trajectory features with the characteristic parameters of the road network features of the roads in the road network map.

[0044] Figure 2 This is a flow chart of the method for obtaining a road network map provided by the present invention. Figure 2 As shown, obtaining a road network map may include: Step 201: Obtain road features based on an open digital map and / or sensors to obtain road network features of the road.

[0045] For example, the road network characteristics of the road can be obtained based on the open digital map OSM and / or sensors. The target characteristics may include geometric characteristics, physical characteristics and / or topological characteristics. The geometric characteristics may include the length of the road, the size of the slope, etc. The topological characteristics may include the turning angle, the connection relationship of the road, etc. The physical characteristics may include the altitude, etc.

[0046] It's understandable that map-based services are now widely used in various fields. With the development of technology, the use cases of map services are constantly expanding, covering everything from daily life to complex business decisions. In recent years, with the development of technologies such as big data, artificial intelligence, and the Internet of Things, map-based services have gradually become the core component of many large-scale applications.

[0047] OSM Open Street Map is a free global map database that includes both spatial and attribute data. Spatial data primarily consists of three types: points, roads, and relationships, which together form the entire map. "Points" define the location of a point in space; "roads" define lines or areas; and "relationships" define the relationships between elements. OSM maps are highly scalable and support multiple data formats, so this application uses OSM data maps as the basic road network data source.

[0048] Based on the OSM digital map, high-precision sensors are used to extract the characteristic information of each road and after analysis and organization using mathematical tools, the positioning feature information of the digital map can be mined, such as the road width, altitude, type, length, corner size, slope size, magnetic field distribution, etc.

[0049] For example, OSM data can be preprocessed first. Check the OSM map for incorrect roads, missing roads, or unreasonable road sections. If so, update and maintain them. Then, extract the turning intersections between different roads to facilitate subsequent analysis and calculation of the road network topology.

[0050] Figure 3 This is a schematic diagram of the result after OSM map preprocessing provided by the present invention. Figure 3 As shown, the points between the line segments represent turning intersections between roads.

[0051] Next, high-precision sensors can be used to extract characteristic information of each road and mathematical tools can be used to analyze road characteristics. The road characteristics that can be obtained include but are not limited to road width, altitude, type, length, corner size, slope size, and magnetic field distribution.

[0052] Step 202: randomly and evenly storing the road network features of the road into a plurality of data blocks, and establishing a dynamically updateable spatial index structure for each of the plurality of data blocks to obtain the road network map.

[0053] It should be noted that the amount of road network feature data is very large. The road network features of the road can be randomly and evenly stored as multiple data blocks, and then a dynamically updateable spatial index structure can be established for each of the multiple data blocks to facilitate data search.

[0054] It is understandable that by obtaining road features through open digital maps and / or sensors, dispersing and storing the road features into multiple data blocks, and establishing a dynamically updateable spatial index structure for each data block, rich feature data can be obtained, map search can be facilitated, and the efficiency of feature matching can be improved.

[0055] In some embodiments, the physical feature includes a standard altitude of the first road, the sensor includes a barometer, and obtaining the road network characteristics of each road based on the open digital map and / or sensor may include: measuring the altitude of the first road using the barometer to obtain the measured altitude of the first road; and determining the standard altitude of the first road by performing a solution based on the measured altitude, the ideal gas hypothesis, and the International Standard Atmosphere Model.

[0056] For example, the altitude of the road can be measured using a high-precision barometer based on the ideal gas assumption and the International Standard Atmosphere (ISA) model: Solve to get the corresponding altitude. is at a height The air pressure at is the standard atmospheric pressure at sea level, is the temperature lapse rate, is the altitude, is the standard temperature at sea level, is the acceleration due to gravity at the Earth's surface, is the molar mass of air, is the gas constant.

[0057] In some embodiments, the geometric features include the length and slope of the first road, the length and slope of the second road, and the turning angle between the first road and the second road. The sensor includes an inertial measurement unit. The road network features of each road obtained based on the open digital map and / or the sensor may include: converting the coordinate system corresponding to the coordinates of the road in the open digital map from the Earth-centered Earth-fixed coordinate system to the local horizontal coordinate system to obtain the converted open digital map; extracting turning intersections between different roads based on the converted open digital map to obtain multiple turning intersections corresponding to multiple roads; determining the intersection of the first road and the second road in the multiple roads as the target intersection based on the multiple turning intersections corresponding to the multiple roads; calculating the distance between the target endpoint of the first road and the target intersection, and the distance between the target endpoint of the second road and the target intersection in the local horizontal coordinate system to obtain the length of the first road and the length of the second road, wherein the target endpoint is used to represent a non-intersecting endpoint; calculating the distance between the first road and the second road according to the vector dot product formula. The method further comprises: obtaining an angle between the first road and the second road relative to the target intersection, obtaining a turning angle between the first road and the second road; obtaining acceleration data of the first road by using the inertial measurement unit, wherein the inertial measurement unit is arranged on the first road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the first road and a direction perpendicular to the direction of the first road; calculating the acceleration data according to the accelerometer adjustment plane, obtaining a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, wherein the roll angle and the pitch angle are the magnitude of the slope of the first road; obtaining acceleration data of the second road by using the inertial measurement unit, wherein the inertial measurement unit is arranged on the second road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the second road and a direction perpendicular to the direction of the second road; calculating the acceleration data according to the accelerometer adjustment plane, obtaining a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, wherein the roll angle and the pitch angle are the magnitude of the slope of the second road.

[0058] It should be noted that for the length and angle of the road, the coordinates of the road can be converted from the Earth-centered Earth-fixed coordinate system (WGS-84 coordinate system) to the local horizontal coordinate system (Northeast Celestial or North East Earth). If two roads AB and BC intersect at point B, then using point B as the reference point, calculate the distance between point A and point C relative to point B in the local horizontal coordinate system (unit: meters), and the length of roads AB and BC can be solved. Then, using the dot product formula of the vectors , the angle between the roads can be obtained .

[0059] Figure 4 It is a schematic diagram of the road angle provided by the present invention. Figure 4 As shown, the angle of the road is , where AB represents road 1 and BC represents road 2, and the two intersect at point B, which is the actual turning intersection.

[0060] For example, to determine the slope of a road, a high-precision inertial measurement unit can be placed on the road, and its two orthogonal sensitive axes can be aligned in the direction parallel to and perpendicular to the direction of the road. The accelerometer leveling principle can then be used to calculate the roll and pitch angles of the IMU relative to the horizontal plane, thereby obtaining the slope of the road.

[0061] As can be understood, the proposed method for extracting road network features based on OSM maps and / or sensors exploits numerous positioning feature information from digital maps and sensors, resolving the issues of low stability and accuracy associated with inertial navigation systems (INS) when performing road network matching based on traditional digital maps. Furthermore, this method can exploit positioning feature information from digital maps, providing observational information for map-assisted inertial navigation system matching, positioning, and initialization.

[0062] The randomly and evenly storing the road network features of the road as a plurality of data blocks, and establishing a dynamically updateable spatial index structure for each of the plurality of data blocks to obtain the road network map may include: evenly distributing the road network features of the road into a plurality of storage buckets through a hash algorithm, each storage bucket storing corresponding road network feature data; and dynamically constructing and managing the road network feature data in each storage bucket using an incremental kd tree to obtain the road network map.

[0063] It should be noted that the road network characteristics are first hashed. To ensure uniform data distribution, a suitable hash function is selected to hash the road network data, and then a modulo operation is used to map the data into a predetermined number of buckets. Assuming that the input data is the unique OSM_ID number of each road, the OSM_ID will be processed by the hash function to generate an integer hash value. The corresponding bucket number is then calculated by taking the modulo of the predetermined total number of buckets, as shown in the following formula: ; in represents some kind of hash function, Representative road network The OSM_ID of the road, Representative road network The bucket number corresponding to the road, Indicates the amount by which you want to evenly distribute the data.

[0064] By hashing the OSM_ID of a road, all features corresponding to this road can be mapped to specific buckets, thereby achieving random and uniform distribution of large-scale data.

[0065] It is understood that the present invention discloses a method for hashing and managing large-scale road network features and IKD trees. After obtaining the features of the road network, all data is broken up into random and evenly distributed data using a hashing algorithm. Then, an incremental KD tree method is used to dynamically construct a KD tree of the data, thereby avoiding the need to reconstruct the KD tree after data updates, and realizing the function of efficient query and rapid search of features in large-scale road networks.

[0066] In some embodiments, the hash algorithm adopts a consistent hashing algorithm, and the road network features of the road are evenly distributed into multiple storage buckets through the hash algorithm, which can include: mapping each storage bucket and the corresponding road network feature data to the same hash ring space through the consistent hashing algorithm; when the data of the storage bucket needs to be adjusted, calculating the hash ring position of the newly added or deleted storage bucket, and only migrating the road network feature data between adjacent positions on the hash ring to the newly added storage bucket, or reallocating the road network feature data of the deleted storage bucket to the adjacent storage bucket, keeping the road network feature data storage position of the non-adjacent area unchanged.

[0067] It should be noted that as the amount of road network data increases significantly, the number of corresponding buckets may need to increase. To address the data redistribution problem caused by changes in the number of buckets, a consistent hashing algorithm can be used. This method allocates bucket nodes on a virtual hash ring, with each bucket node determining its position within the ring using a hash value. The corresponding road network data is then located at a point on the hash ring using the above hash calculation, and the corresponding bucket is found based on this location. When the number of buckets changes, data only needs to be allocated to the new buckets, without having to redistribute all data, reducing the cost of data migration.

[0068] Furthermore, the use of an incremental kd tree to dynamically construct and manage the road network feature data of each storage bucket to obtain the road network map may include: determining the data dimension for constructing the kd tree based on the feature type of the road network feature data of each storage bucket, and storing the road network feature data according to the data dimension; selecting the feature data of a road in the road network feature data of each storage bucket as the root node to initialize the kd tree; for the road network features of the newly added roads, recursively performing the following operations: selecting a partitioning dimension based on the current tree depth or data distribution, and recursively inserting the road network features of the newly added roads into the left subtree or right subtree of the kd tree according to the partitioning dimension; if the insertion causes the subtree to be unbalanced, locally rebalancing the corresponding subtree by node rotation or subtree reconstruction until all the newly added road network feature data are inserted, thereby obtaining an incremental kd tree structure that supports nearest neighbor queries and range queries as the road network map.

[0069] It is understandable that the incremental kd-tree is used to quickly retrieve specific road network features. The kd-tree (k-dimensional tree) is a tree-like data structure that stores instance points in k-dimensional space for fast retrieval. Its data structure is shown below: struct kdtree{ Node-data - A data point in a data vector dataset, which is a k-dimensional vector Range - The spatial vector representing the spatial range of the node. split - integer number of the axis perpendicular to the splitting hyperplane Left - kd tree consists of all data points in the left subspace of the node's split hyperplane. Right - The kd tree consists of all data points in the right subspace of the node's split hyperplane. parent - the kd-tree parent node } The IKD tree (Incremental KD-Tree) is a dynamic data structure that combines the traditional KD tree (k-dimensional tree) with the concept of incremental construction. Building on the traditional KD tree, it maintains tree balance through incremental insertion and dynamic adjustments to the tree structure. This design differs from conventional KD trees, which partition the data set all at once after it is constructed. In contrast, the IKD tree gradually builds and optimizes the tree structure as new data is inserted.

[0070] The construction process of the incremental kd tree can be summarized into the following steps: Determine the dimensions of the data. Assume that a road has k features, including OSM_ID, longitude and latitude of the road center point, longitude and latitude of the left end point of the road, longitude and latitude of the right end point of the road, road length, road segments connected to the left end point of the road, road segments connected to the right end point of the road, etc.

[0071] Initialize the root node of the tree. At the beginning, select any road data as the root node. The initial state of the tree is a single-node tree containing the root node.

[0072] Select a partitioning dimension. For each newly inserted data point, a partitioning dimension is selected. IKD trees typically select the optimal partitioning dimension dynamically, choosing the optimal partitioning dimension based on factors such as the current tree depth and data distribution. For example, for a tree depth of d, a dimension of d % k (where k is the dimension of the data point) may be selected. This ensures that the tree branches are as balanced as possible, avoiding excessive skewness in any particular dimension.

[0073] Inserting data. After selecting the appropriate partitioning dimension, the data points are partitioned according to that dimension and inserted into the appropriate locations in the tree. Unlike traditional KD trees, IKD trees insert data points into the appropriate locations in the tree by recursively stepping downward, rather than building the entire tree all at once.

[0074] Maintaining tree balance. Incrementally inserting data points may cause parts of the tree to become unbalanced, especially when the data points are unevenly distributed. In this case, the IKD tree needs to be locally rebalanced to maintain tree balance and ensure query performance.

[0075] Figure 5 This is a schematic diagram of the structure of the ikd tree provided by the present invention. Figure 5 As shown in the figure, the road network feature map data is incrementally inserted into the IKD tree to form nodes, gradually dividing the hyperspace into various parts. At this point, the IKD tree is constructed, and the target features of the road network can be quickly queried. There are usually two ways to query: Nearest neighbor query: Given a query point, the IKD tree can quickly find the nearest point. The query process recursively traverses the tree, deciding which subtree to search next based on the partitioning dimension of the current node. The IKD tree's incremental construction method enables efficient query operations on large datasets.

[0076] Range query: Given a query area (such as a rectangle or hypercube), the IKD tree can find all points within the area. Range query reduces unnecessary search areas by pruning different branches of the tree.

[0077] It is understood that the present invention discloses a method for hashing large-scale road network features and managing them using an incremental KD tree. While a large amount of road network location feature data enriches map details, it inevitably increases the burden of storage and retrieval. Hashing achieves a uniform distribution of large-scale road network feature data, evenly distributing this data into N parts. IKD tree management greatly improves the efficiency of searching for road network features, resolving the issue of inefficient map information retrieval during system operation. By managing large-scale road network features through hashing and IKD trees, the system's search efficiency for road network features is improved, with the advantages of low computing resource overhead and high information retrieval efficiency.

[0078] Figure 6 This is a flow chart of the method for obtaining trajectory features and positioning information provided by the present invention. Figure 6 As shown, the step 102 of obtaining the trajectory characteristics of the moving carrier may include: Step 301: Acquire continuous trajectory data of the moving carrier through an inertial navigation system; Step 302: Divide the continuous trajectory data into a number of trajectory segments, and extract the length of each trajectory segment, the angle between adjacent trajectory segments, and the turning direction to obtain the trajectory features.

[0079] The step 103 of performing similarity matching on the characteristic parameters of the trajectory characteristics and the characteristic parameters of the road network characteristics of the roads in the road network map to determine the positioning information of the moving carrier may include: Step 401: perform similarity matching on the characteristic parameters of the trajectory features and the characteristic parameters of the road network features of the roads in the road network map, infer the initial position of the carrier based on the hidden Markov model and the Viterbi algorithm, and converge the initial position to a unique solution through trajectory growth to obtain the positioning information of the moving carrier.

[0080] Furthermore, dividing the continuous trajectory data into a plurality of trajectory segments may include: projecting the continuous trajectory data onto a horizontal plane to form a two-dimensional trajectory; determining a turning feature point by analyzing a change in direction of trajectory points of the two-dimensional trajectory; and dividing the continuous trajectory data into a plurality of trajectory segments using the turning feature point as a segmentation point.

[0081] For example, let's take the three characteristics of a trajectory (the length of each segment, the angle between trajectories, and the left and right turns of each turn) as an example. Assuming that the target, or the moving vehicle, always walks along the road during its movement and does not cross the road, the length and angle of the road can be perceived based on the trajectory of the target. The trajectory formed by the inertial navigation results is projected onto a horizontal plane to obtain a two-dimensional trajectory. The turning points in the trajectory are determined by analyzing the changes in the front and back directions of the positioning points. The entire trajectory is then divided with the turning points as a reference, and adjacent turning points are connected into line segments. The length of the line segments and the angles between the line segments are calculated to obtain the corresponding characteristics of this segment of the trajectory.

[0082] Figure 7 This is a schematic diagram of the process of obtaining trajectory features provided by the present invention. Figure 7 As shown in the figure, the process of analyzing the trajectory to determine the turning points, then dividing the trajectory based on the turning points and obtaining the trajectory features is demonstrated.

[0083] Furthermore, a positioning initialization method based on hidden Markov and Viterbi algorithms is employed, based on road network characteristics. Due to the inherent characteristics of inertial navigation, positioning errors are bound to diverge over time. However, in the short term, inertial navigation systems are generally capable of accurate positioning. By comparing trajectory features calculated by the inertial navigation system with those in the road network map, all possible initial positions of the vehicle can be inferred.

[0084] Furthermore, the method of performing similarity matching between the characteristic parameters of the trajectory features and the characteristic parameters of the road network features of the roads in the road network map, inferring the initial position of the carrier based on the hidden Markov model and the Viterbi algorithm, and converging the initial position to a unique solution through trajectory growth to obtain the positioning information of the moving carrier may include: constructing a hidden Markov model, the hidden Markov model including an observation probability and a transition probability, the observation probability being determined based on the matching result of the characteristic parameters of the trajectory features and the road network features of the roads, and the transition probability being determined based on the topological connection relationship of the roads; calculating the optimal road sequence through the dynamic programming process of the Viterbi algorithm based on the hidden Markov model and the currently observed trajectory feature sequence, and using the center point coordinates of the first road in the state sequence obtained by backtracking as the candidate initial position coordinates; as the motion trajectory of the moving carrier grows, adding trajectory segment features and updating the observation sequence, repeating the above two steps to dynamically adjust the candidate initial position; when all trajectory segment features correspond to only a unique candidate road, determining that the candidate initial position converges to a unique solution, and outputting the positioning information of the moving carrier.

[0085] For example, the present invention matches trajectory features with road features. First, based on the first trajectory segment and the first corner, the initial probability matrix of the hidden Markov model is set by comparing the corresponding features of each road segment in the road network. If the errors in the length and angle features are within a set threshold, as shown in the following formula, the trajectory features and road segment features are considered similar.

[0086] ; in and Represents the length and angle of the first trajectory, and Representative road network The length and angle of each road segment, and Represents the set length error threshold and angle error threshold.

[0087] Assuming that there are m road segments that match the trajectory features, then The initial probability of a road segment is If the trajectory characteristics and If the characteristics of the road sections are similar, If not similar, then .

[0088] For the next k+1th trajectory, we can obtain the k+1th length feature, kth angle feature, and kth turn feature. By searching for road segments in the road network, if there is a road segment that satisfies all three of the above features within the allowable error range, as shown in the following formula, then this road segment is considered similar to the k+1th trajectory.

[0089] ; in and represents the length and angle of the k+1th trajectory, Indicates the turning situation of the i-th road segment at the endpoint, is the left or right turn of the k-th turn in the trajectory.

[0090] After obtaining the similarity relationship between the road section and the trajectory, the observation probability in the hidden Markov model is set based on this relationship. , determined by the following formula: ; Next, the state transition probability in the hidden Markov model is determined based on the topological relationship of the road The size of is determined by the following formula based on the road topology: ; If the moving target is The road at the time is , then at the next moment transfer to the road probability.

[0091] Finally, the Viterbi algorithm can be used to perform dynamic programming on the hidden Markov model to find the most likely hidden state sequence given a given observation sequence.

[0092] The Viterbi algorithm consists of four steps: Initialization: Calculate the optimal path probability for each state at the first moment (taking into account the initial state probability and the probability of the first observation).

[0093] Recursion: For each subsequent moment, calculate the optimal path probability for each state. This probability is to select the path with the highest probability from all possible states at the previous moment, and combine the state transition probability and observation probability at the current moment.

[0094] Backtracking: After the recursion is completed, backtrack to the optimal path at the last moment and gradually determine the most likely state at each moment.

[0095] Output result: The state sequence finally obtained by backtracking is the most likely hidden state sequence.

[0096] Finally, the Viterbi algorithm is used for decoding. Based on the observed trajectory features, the most likely corresponding state of the target in the road network is determined, thereby achieving initial positioning. As the observed trajectory lengthens, more trajectory features are obtained. This method of initializing positioning based on hidden Markov and Viterbi algorithms based on road network features allows the target's initial position to converge quickly to a unique solution. Once converged to a unique solution, the system's initial position and initial heading can be determined.

[0097] Figure 8 This is a schematic diagram of the initial positioning of the three-segment trajectory provided by the present invention. Figure 8 As shown in Figure 2, the initial positioning results of three experimental trajectories on the road network are obtained. It can be seen that the three trajectories correspond to 27 possible situations in the map.

[0098] Figure 9 Schematic diagram of the initial positioning of the four-segment trajectory provided by the present invention. Figure 9 As shown in Figure 1, the initial positioning results of four experimental trajectories on the road network are obtained. It can be seen that the four trajectories correspond to four possible situations in the map.

[0099] Figure 10 Schematic diagram of the initial positioning of the five-segment track provided by the present invention. Figure 10 As shown in Figure 1, the initial positioning results of five experimental trajectories on the road network are obtained. It can be seen that there is only one unique solution for the five trajectories in the map.

[0100] It is understandable that the present invention proposes a positioning initialization method based on the hidden Markov and Viterbi algorithms of the road network characteristics. By analyzing the trajectory calculated by the inertial navigation system and then statistically analyzing the probability characteristics of the positioning features in the road network, the possibility of all initial positions of the carrier can be inferred. As the trajectory grows, the hidden Markov and Viterbi algorithms can make the initial position probability of the carrier quickly converge to a unique solution, and finally achieve positioning initialization. This allows the inertial navigation system to be initialized without manually setting the initial position or using other high-precision equipment for positioning, solving the problem that the initialization process of the inertial navigation system is cumbersome and has many limitations. In addition, the hidden Markov and Viterbi algorithm positioning initialization method based on road network characteristics proposed in the present invention can realize the initialization of the inertial navigation system conveniently and at low cost without the assistance of other equipment. In the future, it can be applied to smart devices such as mobile phones and bracelets, with great economic benefits.

[0101] It can be understood that the road feature-based positioning initialization method provided by the present invention can eliminate the need to manually set the initial position or use other high-precision equipment for positioning when initializing the inertial navigation system, thereby improving the convenience of the inertial navigation system and expanding its application scope.

[0102] The innovations of the present invention include: (1) a method for extracting road network features based on an open digital map and a method for extracting and analyzing the positioning feature information of each road using high-precision sensors or mathematical tools, thereby obtaining a road network map with multi-dimensional positioning features. (2) a method for hashing and managing large-scale road network features and an IKD tree is proposed. By using a hash algorithm to evenly distribute a wide range of road network feature data and using an incremental KD tree to dynamically construct a KD tree for each piece of data, efficient management of road network maps and rapid data search are achieved. (3) a method for positioning initialization based on hidden Markov and Viterbi algorithms is proposed. The trajectory features calculated by the inertial navigation system are compared with the features in the road network map to determine the observation probability. The transfer probability is then determined based on the topological relationship of the roads in the road network, thereby inferring the possibility of all initial positions of the carrier. Finally, the initial position converges to a unique solution, achieving positioning initialization.

[0103] Based on the aforementioned embodiments, an embodiment of the present invention provides a positioning initialization device based on road features. The modules included in the device, as well as the units included in each module, can be implemented by a processor; of course, they can also be implemented by specific logic circuits. During implementation, the processor can be a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP), or a field programmable gate array (FPGA), etc.

[0104] The following describes a positioning initialization device based on road features provided by the present invention. The positioning initialization device based on road features described below and the positioning initialization method based on road features described above can refer to each other.

[0105] Figure 11 Schematic diagram of the structure of the positioning initialization device based on road features provided by the present invention. Figure 11 As shown, the device 500 includes a road network map acquisition module 501, a trajectory feature acquisition module 502 and a positioning information determination module 503, wherein: A road network map acquisition module 501 is used to acquire a road network map, wherein the road network map includes a road network feature map extracted and / or acquired through sensors; A trajectory feature acquisition module 502 is configured to acquire trajectory features of a moving vehicle, wherein characteristic parameters of the trajectory features correspond to characteristic parameters of the road network features of the road; The positioning information determination module 503 is used to perform similarity matching between the characteristic parameters of the trajectory characteristics and the characteristic parameters of the road network characteristics of the roads in the road network map to determine the positioning information of the moving carrier. The positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0106] In some embodiments, the road network map acquisition module 501 includes a feature acquisition unit and a feature storage unit, wherein: The feature acquisition unit is used to acquire road features based on an open digital map and / or sensors to obtain road network features of the road; The feature storage unit is used to randomly and evenly store the road network features of the road as multiple data blocks, and establish a dynamically updateable spatial index structure for each data block in the multiple data blocks to obtain the road network map.

[0107] In some embodiments, the geometric features include the length and slope of the first road, the length and slope of the second road, and the turning angle between the first road and the second road. The sensor includes an inertial measurement unit. The feature acquisition unit is specifically configured to: Converting the coordinate system corresponding to the coordinates of the roads in the open digital map from the Earth-centered Earth-fixed coordinate system to the local horizontal coordinate system to obtain a converted open digital map; Extracting turning intersections between different roads based on the converted open digital map to obtain a plurality of turning intersections corresponding to the plurality of roads; Determining, according to the plurality of turning intersections corresponding to the plurality of roads, an intersection of a first road and a second road among the plurality of roads as a target intersection; Calculating the distance between a target endpoint of the first road and the target intersection, and the distance between a target endpoint of the second road and the target intersection in a local horizontal coordinate system to obtain the length of the first road and the length of the second road, wherein the target endpoint is used to represent a non-intersecting endpoint; Calculating the angle between the first road and the second road relative to the target intersection according to a vector dot product formula to obtain a turning angle between the first road and the second road; Acquiring acceleration data of the first road by using the inertial measurement unit, wherein the inertial measurement unit is disposed on the first road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the first road and a direction perpendicular to the direction of the first road; Calculating the acceleration data according to the accelerometer leveling to obtain a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, wherein the roll angle and the pitch angle represent the slope of the first road; Acquiring acceleration data of the second road by using the inertial measurement unit, wherein the inertial measurement unit is disposed on the second road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the second road and a direction perpendicular to the direction of the first road; The acceleration data is calculated according to the accelerometer leveling to obtain a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, where the roll angle and the pitch angle represent the slope of the second road.

[0108] In some embodiments, the feature storage unit includes a feature storage component and a feature management component, wherein: The feature storage component is used to evenly distribute the road network features of the road into multiple storage buckets through a hash algorithm, and each storage bucket stores corresponding road network feature data; The feature management component is used to dynamically construct and manage the road network feature data in each storage bucket using an incremental kd tree to obtain the road network map.

[0109] In some embodiments, the hash algorithm adopts a consistent hashing algorithm, and the feature storage component is specifically used to: Mapping each storage bucket and the corresponding road network feature data to the same hash ring space through the consistent hashing algorithm; When the data in a bucket needs to be adjusted, the hash ring position of the newly added or deleted bucket is calculated, and only the road network feature data between adjacent positions on the hash ring is migrated to the newly added bucket, or the road network feature data of the deleted bucket is reallocated to the adjacent bucket, keeping the road network feature data storage position in non-adjacent areas unchanged.

[0110] In some embodiments, the feature management component is specifically configured to: Determining the data dimension for constructing the kd tree according to the feature type of the road network feature data in each storage bucket, and storing the road network feature data according to the data dimension; Select the feature data of one road in the road network feature data of each storage bucket as the root node to initialize the kd tree; For the road network features of the newly added roads, the following operations are recursively performed: a partition dimension is selected according to the current tree depth or data distribution, and the road network features of the newly added roads are recursively inserted into the left subtree or right subtree of the kd tree according to the partition dimension. If the insertion causes the subtree to be unbalanced, the corresponding subtree is locally rebalanced through node rotation or subtree reconstruction until all the newly added road network feature data are inserted, and an incremental kd tree structure that supports nearest neighbor queries and range queries is obtained as the road network map.

[0111] In some embodiments, the trajectory feature acquisition module 502 includes a continuous trajectory acquisition unit and a trajectory feature acquisition unit, wherein: The continuous trajectory acquisition unit is used to acquire the continuous trajectory data of the moving carrier through an inertial navigation system; The trajectory feature acquisition unit is used to divide the continuous trajectory data into a plurality of trajectory segments, and extract the length of each trajectory segment, the angle between adjacent trajectory segments, and the turning direction to obtain the trajectory feature; The positioning information determination module 503 includes an initial position determination unit, wherein: The initial position determination unit is used to perform similarity matching between the characteristic parameters of the trajectory features and the characteristic parameters of the road network features of the roads in the road network map, infer the initial position of the carrier based on the hidden Markov model and the Viterbi algorithm, and converge the initial position to a unique solution through trajectory growth to obtain the positioning information of the moving carrier.

[0112] In some embodiments, the trajectory feature acquisition unit is specifically configured to: project the continuous trajectory data onto a horizontal plane to form a two-dimensional trajectory; determine a turning feature point by analyzing a direction change of a trajectory point of the two-dimensional trajectory; and divide the continuous trajectory data into a plurality of trajectory segments using the turning feature point as a segmentation point.

[0113] In some embodiments, the initial position determination unit is specifically configured to: Constructing a hidden Markov model, wherein the hidden Markov model includes observation probability and transition probability, wherein the observation probability is determined based on a matching result of characteristic parameters of the trajectory characteristics and the road network characteristics, and the transition probability is determined based on a topological connection relationship of the roads; According to the hidden Markov model and the currently observed trajectory feature sequence, the optimal road sequence is calculated through the dynamic programming process of the Viterbi algorithm, and the center point coordinates of the first road in the state sequence obtained by backtracking are used as candidate initial position coordinates; As the motion trajectory of the moving vehicle grows, new trajectory segment features are added and the observation sequence is updated. The above two steps are repeated to dynamically adjust the candidate initial position. When all trajectory segment features correspond to only a single candidate road, the candidate initial position is determined to have converged to a unique solution, and the positioning information of the moving vehicle is output.

[0114] In an embodiment of the present invention, the inertial navigation system can construct its own road network map, match the road network map with the trajectory characteristics of the moving carrier to determine the positioning information of the moving carrier, realize the autonomous initial positioning of the inertial navigation system, and improve the applicability of the inertial navigation system.

[0115] Figure 12 Schematic diagram of the physical structure of the electronic device provided by the present invention. Figure 12 As shown, the electronic device may include: a processor 610, a communications interface 620, a memory 630, and a communications bus 640, wherein the processor 610, the communications interface 620, and the memory 630 communicate with each other via the communications bus 640. The processor 610 may call logic instructions in the memory 630 to execute a positioning initialization method based on road features, the method comprising: obtaining a road network map, the road network map including road network features of roads, the road network features including at least one of geometric features, physical features, and topological features, the geometric features being used to characterize spatial attributes of roads, the physical features being used to characterize environmental attributes of roads, and the topological features being used to characterize connectivity between roads; Acquiring trajectory characteristics of the moving carrier, wherein characteristic parameters of the trajectory characteristics correspond to characteristic parameters of the road network characteristics of the road; The characteristic parameters of the trajectory feature are matched with the characteristic parameters of the road network feature of the road in the road network map for similarity, and the positioning information of the moving carrier is determined. The positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0116] Furthermore, the logic instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0117] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being storable on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is capable of executing the road feature-based positioning initialization method provided by the above methods, the method including: obtaining a road network map, the road network map including road network features of roads, the road network features including at least one of geometric features, physical features, and topological features, the geometric features being used to characterize spatial attributes of roads, the physical features being used to characterize environmental attributes of roads, and the topological features being used to characterize connectivity between roads; Acquiring trajectory characteristics of the moving carrier, wherein characteristic parameters of the trajectory characteristics correspond to characteristic parameters of the road network characteristics of the road; The characteristic parameters of the trajectory feature are matched with the characteristic parameters of the road network feature of the road in the road network map for similarity, and the positioning information of the moving carrier is determined. The positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0118] The computer program product includes one or more computer instructions. When loaded and executed on a computer, the computer program instructions fully or partially perform the processes or functions described in accordance with the embodiments of the present invention. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium capable of computer storage or a data storage device such as a server or data center that integrates one or more available media. The available medium may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

[0119] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the road feature-based positioning initialization method provided by the above methods, the method comprising: obtaining a road network map, the road network map including road network features of roads, the road network features including at least one of geometric features, physical features, and topological features, the geometric features being used to characterize spatial attributes of roads, the physical features being used to characterize environmental attributes of roads, and the topological features being used to characterize connectivity between roads; Acquiring trajectory characteristics of the moving carrier, wherein characteristic parameters of the trajectory characteristics correspond to characteristic parameters of the road network characteristics of the road; The characteristic parameters of the trajectory feature are matched with the characteristic parameters of the road network feature of the road in the road network map for similarity, and the positioning information of the moving carrier is determined. The positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

[0120] The computer-readable storage medium may be any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0121] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0122] Program code embodied on a computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, radio frequency (RF), etc., or any suitable combination of the foregoing.

[0123] Computer program code for performing the operations of this specification may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, and C++, as well as conventional procedural programming languages such as "C" or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0124] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0125] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A positioning initialization method based on road features, characterized in that: include: Acquire a road network map, the road network map including road network features, the road network features including at least one of geometric features, physical features, and topological features, the geometric features being used to characterize spatial attributes of roads, the physical features being used to characterize environmental attributes of roads, and the topological features being used to characterize connectivity between roads; Acquiring trajectory characteristics of the moving carrier, wherein characteristic parameters of the trajectory characteristics correspond to characteristic parameters of the road network characteristics of the road; The characteristic parameters of the trajectory feature are matched with the characteristic parameters of the road network feature of the road in the road network map for similarity, and the positioning information of the moving carrier is determined. The positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the initial movement direction of the moving carrier in the road network map.

2. The positioning initialization method based on road features according to claim 1, characterized in that: The obtaining of the road network map includes: Acquire road characteristics based on open digital maps and / or sensors to obtain road network characteristics of the road; The road network features of the roads are randomly and evenly stored as a plurality of data blocks, and a dynamically updateable spatial index structure is established for each of the plurality of data blocks to obtain the road network map.

3. The positioning initialization method based on road features according to claim 2, characterized in that: The geometric features include the length and slope of the first road, the length and slope of the second road, and the turning angle between the first road and the second road. The sensor includes an inertial measurement unit. The road network features of each road obtained based on the open digital map and / or the sensor include: Converting the coordinate system corresponding to the coordinates of the roads in the open digital map from the Earth-centered Earth-fixed coordinate system to the local horizontal coordinate system to obtain a converted open digital map; Extracting turning intersections between different roads based on the converted open digital map to obtain a plurality of turning intersections corresponding to the plurality of roads; Determining, according to the plurality of turning intersections corresponding to the plurality of roads, an intersection of a first road and a second road among the plurality of roads as a target intersection; Calculating the distance between a target endpoint of the first road and the target intersection, and the distance between a target endpoint of the second road and the target intersection in a local horizontal coordinate system to obtain the length of the first road and the length of the second road, wherein the target endpoint is used to represent a non-intersecting endpoint; Calculating the angle between the first road and the second road relative to the target intersection according to a vector dot product formula to obtain a turning angle between the first road and the second road; Acquiring acceleration data of the first road by using the inertial measurement unit, wherein the inertial measurement unit is disposed on the first road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the first road and a direction perpendicular to the direction of the first road; Calculating the acceleration data according to the accelerometer leveling to obtain a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, wherein the roll angle and the pitch angle represent the slope of the first road; Acquiring acceleration data of the second road by using the inertial measurement unit, wherein the inertial measurement unit is disposed on the second road, and two orthogonal sensitive axes of the inertial measurement unit are respectively aligned with a direction parallel to the direction of the second road and a direction perpendicular to the direction of the first road; The acceleration data is calculated according to the accelerometer leveling to obtain a roll angle and a pitch angle of the inertial measurement unit relative to a horizontal plane, where the roll angle and the pitch angle represent the slope of the second road.

4. The positioning initialization method based on road features according to claim 2, characterized in that: The method of randomly and evenly storing the road network features of the road into a plurality of data blocks and establishing a dynamically updateable spatial index structure for each of the plurality of data blocks to obtain the road network map comprises: Evenly distributing the road network features of the road into multiple storage buckets using a hash algorithm, each storage bucket storing corresponding road network feature data; An incremental kd tree is used to dynamically construct and manage the road network feature data in each storage bucket to obtain the road network map.

5. The positioning initialization method based on road features according to claim 4, characterized in that: The hash algorithm adopts a consistent hash algorithm, and the road network features of the road are evenly distributed into multiple storage buckets through the hash algorithm, including: Mapping each storage bucket and the corresponding road network feature data to the same hash ring space through the consistent hashing algorithm; When the data in a bucket needs to be adjusted, the hash ring position of the newly added or deleted bucket is calculated, and only the road network feature data between adjacent positions on the hash ring is migrated to the newly added bucket, or the road network feature data of the deleted bucket is reallocated to the adjacent bucket, keeping the road network feature data storage position in non-adjacent areas unchanged.

6. The method for initializing positioning based on road features according to claim 4, characterized in that: The incremental kd tree is used to dynamically construct and manage the road network feature data of each storage bucket to obtain the road network map, including: Determining the data dimension for constructing the kd tree according to the feature type of the road network feature data in each storage bucket, and storing the road network feature data according to the data dimension; Select the feature data of one road in the road network feature data of each storage bucket as the root node to initialize the kd tree; For the road network features of the newly added roads, the following operations are recursively performed: a partition dimension is selected according to the current tree depth or data distribution, and the road network features of the newly added roads are recursively inserted into the left subtree or right subtree of the kd tree according to the partition dimension. If the insertion causes the subtree to be unbalanced, the corresponding subtree is locally rebalanced through node rotation or subtree reconstruction until all the newly added road network feature data are inserted, and an incremental kd tree structure that supports nearest neighbor queries and range queries is obtained as the road network map.

7. The positioning initialization method based on road features according to claim 1, characterized in that: The acquiring of the trajectory characteristics of the moving carrier includes: Acquiring continuous trajectory data of the moving carrier through an inertial navigation system; Dividing the continuous trajectory data into a plurality of trajectory segments, and extracting the length of each trajectory segment, the angle between adjacent trajectory segments, and the turning direction to obtain the trajectory features; The similarity matching of the characteristic parameters of the trajectory features with the characteristic parameters of the road network features of the roads in the road network map to determine the positioning information of the moving carrier includes: The characteristic parameters of the trajectory features are matched with the characteristic parameters of the road network features of the roads in the road network map for similarity, the initial position of the carrier is inferred based on the hidden Markov model and the Viterbi algorithm, and the initial position is converged to a unique solution through trajectory growth to obtain the positioning information of the moving carrier.

8. The method for positioning initialization based on road features according to claim 7, characterized in that: The method includes performing similarity matching on the characteristic parameters of the trajectory features and the characteristic parameters of the road network features of the roads in the road network map, inferring the initial position of the carrier based on the hidden Markov model and the Viterbi algorithm, and converging the initial position to a unique solution through trajectory growth to obtain the positioning information of the moving carrier, including: Constructing a hidden Markov model, wherein the hidden Markov model includes observation probability and transition probability, wherein the observation probability is determined based on a matching result of characteristic parameters of the trajectory characteristics and the road network characteristics, and the transition probability is determined based on a topological connection relationship of the roads; According to the currently observed trajectory feature sequence, the optimal road sequence is calculated through the dynamic programming process of the Viterbi algorithm, and the center point coordinates of the first road in the state sequence obtained by backtracking are used as candidate initial position coordinates; As the motion trajectory of the moving vehicle grows, new trajectory segment features are added and the observation sequence is updated. The above two steps are repeated to dynamically adjust the candidate initial position. When all trajectory segment features correspond to only a single candidate road, the candidate initial position is determined to have converged to a unique solution, and the positioning information of the moving vehicle is output.

9. A positioning initialization device based on road features, characterized in that: include: A road network map acquisition module is used to acquire a road network map, wherein the road network map includes a road network feature map extracted and / or acquired through sensors; A trajectory feature acquisition module, configured to acquire trajectory features of a moving carrier, wherein characteristic parameters of the trajectory features correspond to characteristic parameters of the road network features of the road; A positioning information determination module is used to perform similarity matching between the characteristic parameters of the trajectory characteristics and the characteristic parameters of the road network characteristics of the roads in the road network map to determine the positioning information of the moving carrier, wherein the positioning information includes an initial position and an initial heading. The initial position is used to represent the coordinates of the starting movement position of the moving carrier in the road network map, and the initial heading is used to represent the movement direction of the moving carrier in the road network map.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the positioning initialization method based on road features as described in any one of claims 1 to 9 is implemented.

Images