CN111721306B - Road matching method and device, electronic equipment and readable storage medium - Google Patents

Road matching method and device, electronic equipment and readable storage medium Download PDF

Info

Publication number
CN111721306B
CN111721306B CN201910216427.8A CN201910216427A CN111721306B CN 111721306 B CN111721306 B CN 111721306B CN 201910216427 A CN201910216427 A CN 201910216427A CN 111721306 B CN111721306 B CN 111721306B
Authority
CN
China
Prior art keywords
road
candidate
information
track information
running track
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910216427.8A
Other languages
Chinese (zh)
Other versions
CN111721306A (en
Inventor
李海波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Didi Infinity Technology and Development Co Ltd
Original Assignee
Beijing Didi Infinity Technology and Development Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Didi Infinity Technology and Development Co Ltd filed Critical Beijing Didi Infinity Technology and Development Co Ltd
Priority to CN201910216427.8A priority Critical patent/CN111721306B/en
Publication of CN111721306A publication Critical patent/CN111721306A/en
Application granted granted Critical
Publication of CN111721306B publication Critical patent/CN111721306B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
    • G01C21/30—Map- or contour-matching
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
    • G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
    • G01C21/30—Map- or contour-matching
    • G01C21/32—Structuring or formatting of map data

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Traffic Control Systems (AREA)

Abstract

The embodiment of the invention provides a road matching method, a road matching device, electronic equipment and a readable storage medium, and belongs to the technical field of navigation positioning. The method comprises the following steps: acquiring running track information of a vehicle, wherein the running track information comprises at least three position points; obtaining a plurality of candidate roads according to any one of the at least three position points; inputting the running track information and the candidate road characteristic information of the candidate roads into a long-short term memory network (LSTM) model, and matching the running track information with the candidate roads through the LSTM model so as to obtain a target road matched with the running track information from the candidate roads. Since the LSTM model can screen useful data from the input data, it can filter out useless data, so that the final output results are all obtained based on the useful data, and therefore, the LSTM model is used for matching, thereby improving the accuracy of road matching.

Description

Road matching method and device, electronic equipment and readable storage medium
Technical Field
The invention relates to the technical field of navigation and positioning, in particular to a road matching method, a road matching device, electronic equipment and a readable storage medium.
Background
Currently, Location Based Service (LBS) applications of the mobile internet have been deeply introduced into the lives of the public. People often plan a trip route (Road Planning) using these mobile internet maps, and estimate the time spent at the starting End (ETA), etc., and an important basis of these technologies is to sense the route of the current vehicle through data collected in real time (including satellite navigation System (GPS) data and sensor data), that is, map-matching in the map field.
In the prior art, the road matching is carried out by acquiring the GPS data of the vehicle, but the GPS data has a drift phenomenon, so that the road matching accuracy is low.
Disclosure of Invention
In view of the above, an object of the embodiments of the present invention is to provide a road matching method, apparatus, electronic device and readable storage medium.
In a first aspect, an embodiment of the present invention provides a road matching method, where the method includes: acquiring running track information of a vehicle, wherein the running track information comprises at least three position points, the at least three position points comprise a current position point and at least two historical position points, and the distance from each historical position point to the current position point is smaller than or equal to a preset distance; obtaining a plurality of candidate roads according to any one of the at least three position points, wherein each candidate road comprises candidate road characteristic information; inputting the running track information and the candidate road characteristic information of the candidate roads into a long-short term memory network (LSTM) model, and matching the running track information with the candidate roads through the LSTM model so as to obtain a target road matched with the running track information from the candidate roads.
In the implementation process, the driving track information is matched with the candidate roads through the LSTM model to obtain a target road matched with the driving track information from the candidate roads, and the LSTM model can screen useful data from input data, so that useless data can be filtered out, and final output results are obtained based on the useful data.
Optionally, acquiring a plurality of candidate roads according to any one of the at least three location points includes: acquiring any one of the at least three position points as corresponding road density information of a target position point in a target range; when the road density information is greater than or equal to a first preset value, acquiring a plurality of candidate roads in a first preset range including the target position point, wherein the number of the candidate roads in the first preset range is smaller than that of the candidate roads in the target range; and when the road density information is smaller than a second preset value, acquiring a plurality of candidate roads in a second preset range including the target position point, wherein the number of the candidate roads in the second preset range is larger than that of the candidate roads in the target range.
In the implementation process, the last candidate roads are obtained based on the obtained road density information, so that the number of the obtained candidate roads meets the matching requirement, and further, the matching precision can be improved and the matching difficulty can be reduced during matching.
Optionally, the LSTM model includes a forgetting gate attenuation formula, a cell state filtering formula, an input gate updating formula, a cell state updating formula, and an output updating formula, the driving track information and the candidate road feature information of the multiple candidate roads are input to a long-short term memory network LSTM model, and the driving track information is matched with the multiple candidate roads through the LSTM model to obtain a target road matched with the driving track information from the multiple candidate roads, including: determining a forgetting function by using the forgetting gate attenuation formula based on the running track information and the candidate road characteristic information of the plurality of candidate roads; based on the forgetting function, filtering useless information in the running track information and the candidate road characteristic information of the candidate roads by using the cell state filtering formula to obtain effective running track information and effective candidate road characteristic information; determining an update value by using the input gate update formula based on the travel track information and the candidate road characteristic information of the plurality of candidate roads; updating the effective driving track information and the effective candidate road characteristic information based on the updated value by using the cell state updating formula, and determining an initial output value, wherein the initial output value is an initial matching probability of the driving track information respectively matching with the candidate roads; and carrying out standardization processing and updating on the initial output value by using the output updating formula to obtain a model output result, wherein the model output result is the matching probability of the running track information respectively matching with the candidate roads.
Optionally, before acquiring the traveling track information of the vehicle including at least three location point information within a preset time period before the current time, the method further includes: acquiring training sample data, wherein the training sample data comprise sample running track information and a plurality of sample candidate roads acquired according to the sample running track information, and the sample running track information is running track information of a vehicle in a preset time period before the current moment and comprises a plurality of position point information; and taking the sample running track information and the plurality of sample candidate roads as the input of the LSTM model, taking the sample running track information and the matching information of each sample candidate road as the output, and training the LSTM model to obtain the trained LSTM model.
In the implementation process, a large amount of data is acquired in advance to train the LSTM model, so that parameters in the LSTM model can be optimal, and then the trained LSTM model is used for road matching, so that a better matching result can be obtained.
Optionally, before acquiring the driving track information of the vehicle including at least three position points, the method includes: acquiring a plurality of historical roads; and dividing each historical road according to the road attributes to obtain a plurality of sub-road sections corresponding to each historical road, wherein the sub-road sections corresponding to each historical road are used as candidate roads in the historical road.
In the implementation process, each historical road is divided according to the road attributes to obtain a plurality of sub-road segments, the sub-road segments are used as candidate roads, so that the obtained candidate roads are road segments with small distances, the driving track information of the vehicle can be matched with the road segments with short distances in the matching process, and compared with the road segments with long distances, the matching difficulty can be reduced and the matching accuracy can be improved.
In a second aspect, an embodiment of the present invention provides a road matching apparatus, where the apparatus includes:
the vehicle driving system comprises a position information acquisition module, a driving path information acquisition module and a driving path information acquisition module, wherein the driving path information acquisition module is used for acquiring driving path information of a vehicle, the driving path information comprises at least three position points, the at least three position points comprise a current position point and at least two historical position points, and the distance from each historical position point to the current position point is smaller than or equal to a preset distance;
the candidate road acquisition module is used for acquiring a plurality of candidate roads according to any one of the at least three position points, and each candidate road comprises candidate road characteristic information;
and the matching module is used for inputting the running track information and the candidate road characteristic information of the candidate roads into a long-short term memory network (LSTM) model, matching the running track information with the candidate roads through the LSTM model, and obtaining a target road matched with the running track information from the candidate roads.
Optionally, the candidate road obtaining module is specifically configured to obtain road density information corresponding to any one of the at least three location points in a target range as a target location point; when the road density information is greater than or equal to a first preset value, acquiring a plurality of candidate roads in a first preset range including the target position point, wherein the number of the candidate roads in the first preset range is smaller than that of the candidate roads in the target range; and when the road density information is smaller than a second preset value, acquiring a plurality of candidate roads in a second preset range including the target position point, wherein the number of the candidate roads in the second preset range is larger than that of the candidate roads in the target range.
Optionally, the LSTM model includes a forgetting gate attenuation formula, a cell state filtering formula, an input gate update formula, a cell state update formula, and an output update formula, and the matching module is specifically configured to:
determining a forgetting function by using the forgetting gate attenuation formula based on the running track information and the candidate road characteristic information of the candidate roads;
based on the forgetting function, filtering useless information in the running track information and the candidate road characteristic information of the candidate roads by using the cell state filtering formula to obtain effective running track information and effective candidate road characteristic information;
determining an update value by using the input gate update formula based on the travel track information and the candidate road characteristic information of the plurality of candidate roads;
updating the effective driving track information and the effective candidate road characteristic information based on the updated value by using the cell state updating formula, and determining an initial output value, wherein the initial output value is an initial matching probability of the driving track information respectively matching with the candidate roads;
and carrying out standardization processing and updating on the initial output value by using the output updating formula to obtain a model output result, wherein the model output result is the matching probability of the running track information respectively matching with the candidate roads.
Optionally, the apparatus further comprises:
the model training module is used for acquiring training sample data, wherein the training sample data comprises sample running track information and a plurality of sample candidate roads acquired according to the sample running track information, and the sample running track information is running track information of a vehicle in a preset time period before the current time and comprises a plurality of position point information; and taking the sample running track information and the plurality of sample candidate roads as the input of the LSTM model, taking the sample running track information and the matching information of each sample candidate road as the output, and training the LSTM model to obtain the trained LSTM model.
Optionally, the apparatus further comprises:
the road dividing module is used for acquiring a plurality of historical roads; and dividing each historical road according to the road attributes to obtain a plurality of sub-road sections corresponding to each historical road, wherein the sub-road sections corresponding to each historical road are used as candidate roads in the historical road.
In a third aspect, an embodiment of the present invention provides an electronic device, including: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating via the bus when the electronic device is running, the processor executing the machine-readable instructions to perform the steps of the road matching method as provided in the first aspect.
In a fourth aspect, embodiments of the present invention provide a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to perform the steps of the road matching method as provided in the first aspect.
In the method, the driving track information is matched with the candidate roads through the LSTM model so as to obtain a target road matched with the driving track information from the candidate roads, and the LSTM model can screen out useful data from input data and can filter out useless data so that the final output result is obtained based on the useful data.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of the embodiments of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a schematic diagram illustrating interaction between a terminal and a server according to an embodiment of the present invention;
FIG. 2 illustrates a schematic diagram of exemplary hardware and software components of an electronic device that may implement a server of the present concepts in accordance with some embodiments of the present inventions;
fig. 3 is a flowchart of a road matching method according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of obtaining candidate roads according to an embodiment of the present invention;
fig. 5 is a schematic view of a road sub-road segment according to an embodiment of the present invention;
FIG. 6 is a schematic structural diagram of an LSTM model according to an embodiment of the present invention;
FIG. 7 is a schematic structural diagram of the cell states in an LSTM model according to an embodiment of the present invention;
FIG. 8 is a schematic structural diagram of a control gate in an LSTM model according to an embodiment of the present invention;
FIG. 9 is a schematic structural diagram of an input gate in an LSTM model according to an embodiment of the present invention;
fig. 10 is a schematic structural diagram of a forgetting gate in an LSTM model according to an embodiment of the present invention;
FIG. 11 is a schematic diagram illustrating cell state updating in an LSTM model according to an embodiment of the present invention;
FIG. 12 is a schematic structural diagram of an output gate in an LSTM model according to an embodiment of the present invention;
fig. 13 is a flowchart illustrating step S130 of a road matching method according to an embodiment of the present invention;
fig. 14 is a schematic structural diagram of a road matching device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures. Meanwhile, in the description of the present invention, the terms "first", "second", and the like are used only for distinguishing the description, and are not construed as indicating or implying relative importance.
Fig. 1 is a schematic diagram of interaction between a terminal 10 and a server 20 according to an embodiment of the present invention, where the server 20 is communicatively connected to one or more terminals 10 through a network 30 for data communication or interaction. The terminal 10 may be a Personal Computer (PC), a tablet PC, a smart phone, a Personal Digital Assistant (PDA), a wearable device, a vehicle-mounted terminal, or the like.
In some embodiments, the server 20 may be a single server or a group of servers. The set of servers may be centralized or distributed (e.g., server 20 may be a distributed system). In some embodiments, the server 20 may be local or remote to the terminal. For example, the server 20 may access information and/or data stored in the terminal 10 via the network 30. As another example, the server 20 may be directly connected to the terminal 10 to access stored information and/or data. In some embodiments, the server 20 may be implemented on a cloud platform; by way of example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud (community cloud), a distributed cloud, an inter-cloud, a multi-cloud, and the like, or any combination thereof. In some embodiments, server 20 may be implemented on an electronic device 200 having one or more of the components shown in FIG. 2 in the present disclosure.
In some embodiments, the server 20 may include a processor. The processor may process information and/or data related to the service request to perform one or more of the functions described in this disclosure. In some embodiments, a processor may include one or more processing cores (e.g., a single-core processor (S) or a multi-core processor (S)). Merely by way of example, a Processor may include a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), an Application Specific Instruction Set Processor (ASIP), a Graphics Processing Unit (GPU), a Physical Processing Unit (PPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), a Programmable Logic Device (PLD), a controller, a microcontroller Unit, a Reduced Instruction Set computer (Reduced Instruction Set computer), a microprocessor, or the like, or any combination thereof.
The network 30 may be used for the exchange of information and/or data. In some embodiments, the server 20 and the terminal 10 may send information and/or data to other components. For example, the server 20 may obtain a service request from the terminal 10 via the network 30. In some embodiments, the network 30 may be any type of wired or wireless network, or combination thereof. Merely by way of example, the Network 30 may include a wired Network, a Wireless Network, a fiber optic Network, a telecommunications Network, an intranet, the internet, a Local Area Network (LAN), a Wide Area Network (WAN), a Wireless Local Area Network (WLAN), a WLAN, a Metropolitan Area Network (MAN), a Wide Area Network (WAN), a Public Switched Telephone Network (PSTN), a bluetooth Network, a ZigBee Network, a Near Field Communication (NFC) Network, or the like, or any combination thereof. In some embodiments, the network 30 may include one or more network access points. For example, the network 30 may include wired or wireless network access points, such as base stations and/or network switching nodes, through which one or more components of other systems may connect to the network 30 to exchange data and/or information.
Fig. 2 illustrates a schematic diagram of exemplary hardware and software components of an electronic device 200 that may implement a server of the present concepts according to some embodiments of the present inventions. For example, a processor may be used on the electronic device 200 and to perform the functions herein.
The electronic device 200 may be a general purpose computer or a special purpose computer, both of which may be used to implement the road matching method of the present application. Although only one computer is shown, the functionality described herein may be implemented in a distributed manner across multiple similar platforms to balance processing loads for convenience.
For example, the electronic device 200 may include a network port 210 connected to a network, one or more processors 220 for executing program instructions, a communication bus 230, and a different form of storage medium 240, such as a disk, ROM, or RAM, or any combination thereof. Illustratively, the computer platform may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The method of the present application may be implemented in accordance with these program instructions. The electronic device 200 also includes an Input/Output (I/O) interface 250 between the computer and other Input/Output devices (e.g., keyboard, display screen).
For ease of illustration, only one processor is depicted in the electronic device 200. However, it should be noted that the electronic device 200 in the present application may also comprise a plurality of processors, and thus the steps performed by one processor described in the present application may also be performed by a plurality of processors in combination or individually. For example, if the processor of the electronic device 200 executes steps a and B, it should be understood that steps a and B may also be executed by two different processors together or separately in one processor. For example, a first processor performs step a and a second processor performs step B, or the first processor and the second processor perform steps a and B together.
Referring to fig. 3, fig. 3 is a flowchart of a road matching method according to an embodiment of the present invention, the method includes the following steps:
step S110: travel track information including at least three position points of a vehicle is acquired.
In the driving process of the vehicle, the driving track information of the vehicle acquired by the intelligent equipment and the actual driving track information of the vehicle have deviation, and the deviation degree is different from several meters to dozens of meters, so that the position points in the acquired track information are random drift points and cannot completely reflect the real position of the vehicle. In the practical application process, for example, when map matching is performed, it is necessary to know what track the vehicle has traveled and on what road at what time, so that the vehicle can be matched to the correct road according to the travel track of the vehicle.
Therefore, in this embodiment, the driving track information of the vehicle is collected by an intelligent device (such as an intelligent device like a vehicle-mounted terminal or a mobile terminal carried by a driver), and then the driving track information of the vehicle is sent to the server through the network. At present, a vehicle-mounted terminal or a mobile terminal of a driver in a vehicle is generally integrated with a GPS chip, the GPS chip can collect GPS positioning information once at intervals, the information comprises information such as GPS horizontal and vertical coordinates, precision, speed, acceleration in each direction and the like, and the GPS chip has a short collection period and a long duration, so that the data scale is very large in application with a large user amount, and the data is the most direct description of the user in a physical world time space.
The driving track information of the vehicle collected in this embodiment is a series of GPS data points collected by a positioning system, and is referred to as location points in this embodiment, that is, the driving track information includes at least three pieces of location point information. The Positioning technology that can be used may be based on Global Positioning System (GPS), Global Navigation Satellite System (GLONASS), COMPASS Navigation System (COMPASS), galileo Positioning System, Quasi-Zenith Satellite System (QZSS), Wireless Fidelity (WiFi), or the like, or any combination thereof. One or more of the above-described positioning systems may be used interchangeably in the present invention.
It should be noted that, in order to ensure that the current position of the vehicle is correctly matched, a plurality of position points in the travel track information need to be obtained, that is, the at least three position points include a current position point and at least two historical position points, and a distance from each historical position point to the current position point is less than or equal to a preset distance.
That is to say, during the vehicle driving process, a series of GPS data points, that is, location points, in the vehicle driving track information may be acquired, and the acquired location points are locations where the vehicle is driven within a very short period of time before the current time, which may enable better matching to the road where the vehicle is currently located, because if the location points are location points before the current time, it is obvious that the location points cannot represent the location where the current vehicle is located.
The preset distance may be defined by itself according to practical applications, for example, the preset distance may be set to a shorter distance such as 10 meters or 5 meters, and the position of the vehicle within the 10 meters or 5 meters generally does not change too much. It should be noted that, since the location point is a GPS data point, the distance from the historical location point to the current location point may be a spherical distance between two GPS data points in a spherical coordinate system.
Step S120: and acquiring a plurality of candidate roads according to any one of the at least three position points.
In order to perform road matching, a plurality of candidate roads are also required to be obtained for road matching, and the candidate road feature information of each candidate road is included, and the candidate road feature information may be obtained in advance and stored in the server, or may be input by a driver through a terminal and transmitted to the server, or may be obtained by the server through obtaining a road image captured by a vehicle event data recorder on a vehicle and then identifying the road image.
The manner of acquiring the plurality of candidate roads may be: taking the current position point of the three position points as an example, the candidate roads may be candidate roads in a preset range including the current position point, for example, candidate roads with a radius of 30 meters around the current position point, as shown in fig. 4, where the roads (r1, r2, r3) enclosed by circles may be the candidate roads.
Of course, the preset range may be not only a circle, but also a rectangle, for example, any one position point is a central point of the rectangle, the area of the rectangle is a range of the preset area, or a polygon, an ellipse, or the like is used as the preset range.
In practical application, each road is a long road, but in the process of road matching, since the acquired driving track information of the vehicle is a GPS data point within a very small period of time, in order to accurately match the road where the current position of the vehicle is located, a plurality of historical roads, which may be all roads in a certain city, may also be acquired before the driving track information of the vehicle is acquired, and then each historical road is divided according to road attributes to acquire a plurality of sub-roads corresponding to each historical road, and the sub-road corresponding to each historical road is taken as a candidate road in the historical road.
It is to be understood that the road attribute may be understood as the number of lanes of a road, road signs beside the road, intersections, etc., for example, a road shown in fig. 5 may be divided into a sub-section of three lanes from the left side to a two lanes from the right side, a sub-section of two lanes may be divided into a three-lane section, a sub-section of two lanes may be divided into a two-lane section, if a prompt mark is beside a lane in two lanes, a sub-section before the prompt mark may be divided into a sub-section, a sub-section after the prompt mark may be divided into a sub-section, if an intersection is behind two lanes, a section between intersections may be divided into a sub-section, and a sub-section after the intersection may be used as a candidate road for the historical road.
Therefore, according to the mode, each historical road can be divided into a plurality of sub-road segments with relatively short distances, so that the driving track of the vehicle can be matched to the sub-road segments when the roads are matched, and more accurate matching of the roads can be realized.
Each candidate road comprises candidate road characteristic information, which may include, but is not limited to, the number of road lanes, road speed limit, road grade, road type, and average vehicle flow velocity of the road, road transition probability, and the like, where the road transition probability refers to the probability of the proximity of the distance between two GPS data points on the spherical surface and the distance of the two GPS data points projected onto a certain road.
In addition, the manner of acquiring the plurality of candidate roads according to any one of the at least three position points may be: acquiring any one of the at least three position points as road density information corresponding to a target position point in a target range, and acquiring a plurality of candidate roads in a first preset range including the target position point when the road density information is greater than or equal to a first preset value, wherein the number of the candidate roads in the first preset range is less than that of the candidate roads in the target range; and when the road density information is smaller than a second preset value, acquiring a plurality of candidate roads in a second preset range including the target position point, wherein the number of the candidate roads in the second preset range is larger than that of the candidate roads in the target range.
It is to be understood that the road density information corresponding to the target location point may be obtained as follows: the road density information may be characterized by the number of the acquired candidate roads, such as the target range with the target position point as the center, if the number of the acquired candidate roads is large, for example, ten candidate roads, if the value set by the first preset value is 10, it indicates that the road density information in the range of 30 meters is greater than or equal to 10, and this indicates that the number of the acquired candidate roads is too large, which increases the difficulty of matching, then a plurality of candidate roads within a first preset range including the target location point may be obtained, for example, the first preset range takes the target location point as the center of the circle, the radius is obtained as 20 meters, candidate roads within the first preset range are acquired so that the number of candidate roads acquired by the first range is smaller than the number of candidate roads within the target range. However, in some cases, if the roads around the position where the vehicle is located are rare, the number of candidate roads obtained in the target range may be small, for example, 2, at this time, the number of candidate roads is too small, and correct matching may not be performed, if the set second preset value is 5, at this time, it is indicated that the road density information is less than 5, a plurality of candidate roads in a second preset range including the target position point are obtained, the number of candidate roads in the second preset range is greater than the number of candidate roads in the target range, and if the second preset range is a range with the target position as a center and a radius of 50 meters, and the number of candidate roads is 5, 5 candidate roads may be obtained as candidate roads for matching.
Of course, by obtaining the candidate roads in a certain range, the range determination is not only the above-described illustrated embodiment, but also may be a range in a shape of a rectangle, a polygon, or the like, and the road density information may be characterized not only by the number of the candidate roads, but also by other manners, such as a ratio of the number of the candidate roads to the range.
Step S130: inputting the running track information and the candidate road characteristic information of the candidate roads into a long-short term memory network (LSTM) model, and matching the running track information with the candidate roads through the LSTM model so as to obtain a target road matched with the running track information from the candidate roads.
In order to correctly match the driving track information of the vehicle with the plurality of candidate roads, the driving track information and the candidate road feature information of the plurality of candidate roads may be input into a Long short-term memory network (LSTM) model, and the driving track information is matched with the plurality of candidate roads through the LSTM model, so as to obtain a target road matched with the driving track information from the plurality of candidate roads.
Among them, the LSTM model is a special unit called memory cell like accumulator and gated neuron that will have a weight and couple to itself at the next time step, copying the true value of its state and accumulated external signals, but this self-coupling is controlled by a multiplicative gate that another unit learns and decides when to clear memory. Also, the LSTM model is a time-recursive neural network suitable for processing and predicting significant events of relatively long intervals and delays in a time series. The LSTM is distinguished from other memory networks in that a processor for judging whether information is useful or not is added in an algorithm, the structure of the function of the processor is called a cell unit (cell), three doors, namely an input door, a forgetting door and an output door, are placed in one cell unit, one information enters the LSTM network, whether the information is useful or not can be judged according to rules, only the information which accords with the algorithm authentication is left, and the inconsistent information is left through the forgetting door. Therefore, in the LSTM model, only useful input data is processed and the corresponding result is output, so that more useful information can be screened from the input data, and the final output result is more accurate.
Therefore, in this embodiment, the LSTM model is used to match the driving track information with the candidate roads to obtain a target road matched with the driving track information from the candidate roads, and since the LSTM model can screen useful data from input data, which can filter out useless data, so that the final output result is obtained based on the useful data, compared with the prior art in which road matching is performed only by obtaining GPS data of a vehicle, the present solution obtains the driving track information of the vehicle and feature information of the candidate roads, and performs matching using the LSTM model, thereby improving the accuracy of road matching.
The following describes the specific principles of the LSTM model.
Referring to fig. 6, fig. 6 shows a schematic diagram of the structure of the LSTM model, in fig. 6, each black line carries an entire vector, from the output of one node to the input of the other node, circles represent operations of pointwise, such as the sum of vectors, while rectangles are learned neural network layers, lines together represent the concatenation of vectors, and separate lines represent content that is copied and then distributed to different locations.
As in fig. 7, the key to LSTM is the cellular state, with horizontal lines running across the top in fig. 7. The cell state is similar to a carousel, running directly over the entire chain, with only a few linear interactions, on which the information remains unchanged. Of course, the conveyor belt itself cannot control which information is memorized, and the control gate in fig. 8 is used for the control.
The control gates, which are a method of selectively passing information, contain a sigmoid neural network layer and pointwise multiplication operations. The sigmoid layer outputs a value between 0 and 1, describing how much of each part can pass, 0 representing "no amount is allowed to pass" and 1 representing "any amount is allowed to pass".
The LSTM model has three control gates for protecting and controlling the cell state, namely an input gate, a forgetting gate and an output gate, and the LSTM model changes the self-circulation weight through the three control gates, and meanwhile, under the condition that the model parameters are fixed, the integral scales at different moments can be dynamically changed, so that the gradient disappearance or gradient expansion in the deep learning training process is avoided.
As shown in figure 9 of the drawings,the input gate is responsible for processing the input of the current sequence position, and as can be seen from fig. 9, the input gate is composed of two parts, the first part uses sigmoid activation function, and the output is itThe second part uses the tanh activation function and the output is
Figure BDA0002001683820000161
The results of the two are multiplied later to update the cell state, and the mathematical expression is as follows:
it＝σ(Wi[ht-1,xt]+bi)
Figure BDA0002001683820000162
wherein, Wi,Wc,bi,bcCoefficients and biases that are linear relationships, like in RNN, σ is the sigmoid activation function.
As shown in FIG. 10, the forgetting gate controls whether to forget, controls whether to forget the hidden cell state of the previous layer in the LSTM model with a certain probability, and inputs the hidden state h with the previous sequence in FIG. 10t-1And the present sequence data xtObtaining the output f of the forgetting gate by an activation function, generally a sigmoid functiont. Due to the output f of the sigmoid functiontIn [0,1 ]]And hence the output f heretThe probability of the state of the hidden cell in the upper layer is represented, and the mathematical expression is as follows:
ft＝σ(Wf[ht-1,xt]+bf)
wherein, W and b are respectively corresponding weight system matrix and bias term, xtFor the input data sequence at time t, σ is the sigmoid function, ct－1As a filter function at time t-1, ht－1Update the formula for the output at time t-1, ftIs a forgetting function.
The results of both the forgetting gate and the entry gate contribute to the cell state CtRenewal of cell state, i.e. from cell state Ct-1Is updated toCtCell state C, as shown in FIG. 11tThe mathematical expression is as follows:
Figure BDA0002001683820000171
where, is the Hadamard product.
With the new hidden cell state, referring to the output gate structure in FIG. 12, the hidden state h can be seentThe update part of (2) is composed of two parts, the first part is otFrom the previous sequence of hidden states ht-1And the present sequence data xtAnd the activation function sigmoid, the second part being derived from the hidden state CtAnd tanh activation function, i.e.:
ot＝σ(Wo[ht-1,xt]+bo)
ht＝ot*tanh(Ct)
the following describes a road matching process in the embodiment of the present invention with reference to the principles of the embodiment of the present invention and the LSTM model described above: the LSTM model may include a forgetting gate attenuation formula, a cell state filtering formula, an input gate updating formula, a cell state updating formula, and an output updating formula, as shown in fig. 13, where step S130 includes:
step S131: and determining a forgetting function by using the forgetting gate attenuation formula based on the running track information and the candidate road characteristic information of the candidate roads.
Step S132: and filtering useless information in the running track information and the candidate road characteristic information of the candidate roads by using the cell state filtering formula based on the forgetting function to obtain effective running track information and effective candidate road characteristic information.
The forgetting gate attenuation formula is ft＝σ(Wf[ht-1,xt]+bf) Wherein W and b are respectively corresponding weight system matrix and bias term, xtσ is sigmoid function for input data sequence at time tNumber, ct－1Is a filter function at time t-1, ht－1Updating the formula for the output at time t-1, ftIs a forgetting function. The forgetting gate describes the process of automatic information attenuation according to the forgetting curve rule for the previously learned information, namely, the forgetting process of the learner is described and represented.
The cell state filter formula is ct＝ftct-1Wherein c istThe term "filtering function" is used to describe information that, for previously learned information, filters out information that conflicts with the attribute hierarchy or information that is not useful for prediction at time t.
Step S133: and determining an updating value by using the input door updating formula based on the running track information and the candidate road characteristic information of the candidate roads.
Step S134: and updating the effective driving track information and the effective candidate road characteristic information based on the updated value by using the cell state updating formula, and determining an initial output value, wherein the initial output value is the initial matching probability of the driving track information respectively matched with the candidate roads.
The input gate updates the formula to it＝σ(Wi[ht-1,xt]+bi) The cell state update formula is
Figure BDA0002001683820000181
Figure BDA0002001683820000182
It should be understood that the cell state update formula includes the cell state filtering formula described above, while adding information that is updated based on the update value. Optionally, the output gate further obtains an initial output value through a sigmoid function, that is: ot＝σ(Wo[ht-1,xt]+bo)。
Step S135: and carrying out standardization processing and updating on the initial output value by using the output updating formula to obtain a model output result, wherein the model output result is the matching probability of the running track information respectively matching with the candidate roads.
The output updating formula is ht＝ot*tanh(Ct)。
The matching probability of the travel track information with each candidate road can be output by the above calculation inside the LSTM model.
In addition, it should be understood that, when the LSTM is used to match the driving track information with the candidate road, the LSTM model is obtained after training in advance, so before the LSTM model is used to match the driving track information with the candidate road, the LSTM model needs to be trained as follows: firstly, acquiring training sample data, wherein the training sample data comprises sample running track information and a plurality of sample candidate roads acquired according to the sample running track information, and the sample running track information is running track information of a vehicle in a preset time period before the current time and comprises a plurality of position point information; and taking the sample running track information and the plurality of sample candidate roads as the input of the LSTM model, taking the sample running track information and the matching information of each sample candidate road as the output, and training the LSTM model to obtain the trained LSTM model.
The training process is similar to the matching process, and only in the training process, the acquired training data is a large amount of sample driving track information, such as a large amount of GPS data points, when training, each GPS data point is labeled in advance, and the labeling mode is as follows: manual marking; rule selection: the binding result of the middle part of the track can be judged by using rules for the complete track which occurs, for example, a certain section in the middle of the track can not determine whether the track is on the main road or the auxiliary road, but can be determined according to the later track direction of the vehicle, for example, the auxiliary road has a right turn but the main road goes straight, and the track has a right turn, so that the previous driving point is on the auxiliary road but not on the main road; and (3) image identification and annotation mode: and determining the road to which the GPS data point corresponding to the image belongs by using the image acquired by the formation recorder on the vehicle.
It can be understood that, during training, each GPS data point may be labeled with the corresponding candidate road and the corresponding candidate road feature information, and then these data are input to the LSTM model for training, so as to train parameters in the LSTM model to be optimal, thereby obtaining an optimal output result.
Referring to fig. 14, fig. 14 is a block diagram of a road matching device 300 according to an embodiment of the present invention, the device includes:
the position information acquiring module 310 is configured to acquire travel track information of a vehicle, where the travel track information includes at least three position points, where the at least three position points include a current position point and at least two historical position points, and a distance from each historical position point to the current position point is less than or equal to a preset distance;
a candidate road obtaining module 320, configured to obtain multiple candidate roads according to any one of the at least three location points, where each candidate road includes candidate road feature information;
the matching module 330 is configured to input the driving track information and the candidate road feature information of the multiple candidate roads into a long-term and short-term memory network LSTM model, and match the driving track information with the multiple candidate roads through the LSTM model, so as to obtain a target road matched with the driving track information from the multiple candidate roads.
Optionally, the candidate road obtaining module 320 is specifically configured to obtain road density information corresponding to any one of the at least three location points in a target range as a target location point; when the road density information is greater than or equal to a first preset value, acquiring a plurality of candidate roads in a first preset range including the target position point, wherein the number of the candidate roads in the first preset range is smaller than that of the candidate roads in the target range; and when the road density information is smaller than a second preset value, acquiring a plurality of candidate roads in a second preset range including the target position point, wherein the number of the candidate roads in the second preset range is larger than that of the candidate roads in the target range.
Optionally, the LSTM model includes a forgetting gate attenuation formula, a cell state filtering formula, an input gate update formula, a cell state update formula, and an output update formula, and the matching module 330 is specifically configured to:
determining a forgetting function by using the forgetting gate attenuation formula based on the running track information and the candidate road characteristic information of the candidate roads;
based on the forgetting function, filtering useless information in the running track information and the candidate road characteristic information of the candidate roads by using the cell state filtering formula to obtain effective running track information and effective candidate road characteristic information;
determining an update value by using the input gate update formula based on the travel track information and the candidate road characteristic information of the plurality of candidate roads;
updating the effective driving track information and the effective candidate road characteristic information based on the updated value by using the cell state updating formula, and determining an initial output value, wherein the initial output value is an initial matching probability of the driving track information respectively matching with the candidate roads;
and carrying out standardization processing and updating on the initial output value by using the output updating formula to obtain a model output result, wherein the model output result is the matching probability of the running track information respectively matching with the candidate roads.
Optionally, the apparatus further comprises:
the model training module is used for acquiring training sample data, wherein the training sample data comprises sample running track information and a plurality of sample candidate roads acquired according to the sample running track information, and the sample running track information is running track information of a vehicle in a preset time period before the current time and comprises a plurality of position point information; and taking the sample running track information and the plurality of sample candidate roads as the input of the LSTM model, taking the sample running track information and the matching information of each sample candidate road as the output, and training the LSTM model to obtain the trained LSTM model.
Optionally, the apparatus further comprises:
the road dividing module is used for acquiring a plurality of historical roads; and dividing each historical road according to the road attributes to obtain a plurality of sub-road sections corresponding to each historical road, wherein the sub-road sections corresponding to each historical road are used as candidate roads in the historical road.
An embodiment of the present invention provides a readable storage medium, and the computer program, when executed by a processor, performs the method processes performed by the electronic device in the method embodiment shown in fig. 3.
It can be clearly understood by those skilled in the art that, for convenience and simplicity of description, the specific working process of the apparatus described above may refer to the corresponding process in the foregoing method, and redundant description is not repeated here.
In summary, embodiments of the present invention provide a road matching method, an apparatus, an electronic device, and a readable storage medium, in which the LSTM model is used to match the driving track information with the candidate roads to obtain a target road matched with the driving track information from the candidate roads, and the LSTM model can filter out useful data from input data, so that the final output result is obtained based on the useful data.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, the functional modules in the embodiments of the present invention may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions may be stored in a computer-readable storage medium if they are implemented in the form of software functional modules and sold or used as separate products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

Claims (10)

1. A road matching method, characterized in that the method comprises:
acquiring running track information of a vehicle, wherein the running track information comprises at least three position points, the at least three position points comprise a current position point and at least two historical position points, and the distance from each historical position point to the current position point is smaller than or equal to a preset distance;
obtaining a plurality of candidate roads according to any one of the at least three position points, wherein each candidate road comprises candidate road characteristic information;
inputting the running track information and the candidate road characteristic information of the candidate roads into a long-short term memory network (LSTM) model, and matching the running track information with the candidate roads through the LSTM model to obtain a target road matched with the running track information from the candidate roads;
the LSTM model comprises a forgetting gate attenuation formula, a cell state filtering formula, an input gate updating formula, a cell state updating formula and an output updating formula, the running track information and the candidate road characteristic information of the candidate roads are input into a long-short term memory network LSTM model, the running track information is matched with the candidate roads through the LSTM model, and a target road matched with the running track information is obtained from the candidate roads, and the method comprises the following steps:
determining a forgetting function by using the forgetting gate attenuation formula based on the running track information and the candidate road characteristic information of the candidate roads;
based on the forgetting function, filtering useless information in the running track information and the candidate road characteristic information of the candidate roads by using the cell state filtering formula to obtain effective running track information and effective candidate road characteristic information;
determining an update value by using the input gate update formula based on the travel track information and the candidate road characteristic information of the plurality of candidate roads;
updating the effective driving track information and the effective candidate road characteristic information based on the updated value by using the cell state updating formula, and determining an initial output value, wherein the initial output value is an initial matching probability of the driving track information respectively matching with the candidate roads;
and carrying out standardization processing and updating on the initial output value by using the output updating formula to obtain a model output result, wherein the model output result is the matching probability of the running track information respectively matching with the candidate roads.
2. The method of claim 1, wherein obtaining a plurality of candidate roads from any one of the at least three location points comprises:
acquiring any one of the at least three position points as corresponding road density information of a target position point in a target range;
when the road density information is greater than or equal to a first preset value, acquiring a plurality of candidate roads in a first preset range including the target position point, wherein the number of the candidate roads in the first preset range is smaller than that of the candidate roads in the target range;
and when the road density information is smaller than a second preset value, acquiring a plurality of candidate roads in a second preset range including the target position point, wherein the number of the candidate roads in the second preset range is larger than that of the candidate roads in the target range.
3. The method according to claim 1, wherein before acquiring the travel track information including at least three position point information of the vehicle within a preset time period before the current time, the method further comprises:
acquiring training sample data, wherein the training sample data comprises sample running track information and a plurality of sample candidate roads acquired according to the sample running track information, and the sample running track information is running track information of a vehicle in a preset time period before the current time and comprises a plurality of position point information;
and taking the sample running track information and the plurality of sample candidate roads as the input of the LSTM model, taking the sample running track information and the matching information of each sample candidate road as the output, and training the LSTM model to obtain the trained LSTM model.
4. The method of claim 1, wherein prior to obtaining travel track information for a vehicle including at least three location points, comprising:
acquiring a plurality of historical roads;
and dividing each historical road according to the road attributes to obtain a plurality of sub-road sections corresponding to each historical road, wherein the sub-road sections corresponding to each historical road are used as candidate roads in the historical road.
5. A road matching device, said device comprising:
the vehicle driving system comprises a position information acquisition module, a driving path information acquisition module and a driving path information acquisition module, wherein the driving path information acquisition module is used for acquiring driving path information of a vehicle, the driving path information comprises at least three position points, the at least three position points comprise a current position point and at least two historical position points, and the distance from each historical position point to the current position point is smaller than or equal to a preset distance;
the candidate road obtaining module is used for obtaining a plurality of candidate roads according to any one of the at least three position points, and each candidate road comprises candidate road characteristic information;
the matching module is used for inputting the running track information and the candidate road characteristic information of the candidate roads into a long-short term memory network (LSTM) model, matching the running track information with the candidate roads through the LSTM model, and obtaining a target road matched with the running track information from the candidate roads;
the LSTM model includes a forgetting gate attenuation formula, a cell state filtering formula, an input gate update formula, a cell state update formula, and an output update formula, and the matching module is specifically configured to:
determining a forgetting function by using the forgetting gate attenuation formula based on the running track information and the candidate road characteristic information of the candidate roads;
based on the forgetting function, filtering useless information in the running track information and the candidate road characteristic information of the candidate roads by using the cell state filtering formula to obtain effective running track information and effective candidate road characteristic information;
determining an update value by using the input gate update formula based on the travel track information and the candidate road characteristic information of the plurality of candidate roads;
updating the effective driving track information and the effective candidate road characteristic information based on the updated value by using the cell state updating formula, and determining an initial output value, wherein the initial output value is an initial matching probability of the driving track information respectively matching with the candidate roads;
and carrying out standardization processing and updating on the initial output value by using the output updating formula to obtain a model output result, wherein the model output result is the matching probability of the running track information respectively matching with the candidate roads.
6. The apparatus according to claim 5, wherein the candidate road obtaining module is specifically configured to obtain road density information corresponding to any one of the at least three location points as a target location point within a target range; when the road density information is greater than or equal to a first preset value, acquiring a plurality of candidate roads in a first preset range including the target position point, wherein the number of the candidate roads in the first preset range is smaller than that of the candidate roads in the target range; and when the road density information is smaller than a second preset value, acquiring a plurality of candidate roads in a second preset range including the target position point, wherein the number of the candidate roads in the second preset range is larger than that of the candidate roads in the target range.
7. The apparatus of claim 5, further comprising:
the model training module is used for acquiring training sample data, wherein the training sample data comprises sample running track information and a plurality of sample candidate roads acquired according to the sample running track information, and the sample running track information is running track information of a vehicle in a preset time period before the current time and comprises a plurality of position point information; and taking the sample running track information and the plurality of sample candidate roads as the input of the LSTM model, taking the sample running track information and the matching information of each sample candidate road as the output, and training the LSTM model to obtain the trained LSTM model.
8. The apparatus of claim 5, further comprising:
the road dividing module is used for acquiring a plurality of historical roads; and dividing each historical road according to the road attributes to obtain a plurality of sub-road sections corresponding to each historical road, wherein the sub-road sections corresponding to each historical road are used as candidate roads in the historical road.
9. An electronic device, comprising: a processor, a storage medium and a bus, the storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating via the bus when the electronic device is running, the processor executing the machine-readable instructions to perform the steps of the road matching method according to any one of claims 1-4.
10. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, which computer program, when being executed by a processor, carries out the steps of the road matching method according to any one of claims 1-4.
CN201910216427.8A 2019-03-20 2019-03-20 Road matching method and device, electronic equipment and readable storage medium Active CN111721306B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910216427.8A CN111721306B (en) 2019-03-20 2019-03-20 Road matching method and device, electronic equipment and readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910216427.8A CN111721306B (en) 2019-03-20 2019-03-20 Road matching method and device, electronic equipment and readable storage medium

Publications (2)

Publication Number Publication Date
CN111721306A CN111721306A (en) 2020-09-29
CN111721306B true CN111721306B (en) 2022-07-05

Family

ID=72562823

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910216427.8A Active CN111721306B (en) 2019-03-20 2019-03-20 Road matching method and device, electronic equipment and readable storage medium

Country Status (1)

Country Link
CN (1) CN111721306B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112229412B (en) * 2020-10-21 2024-01-30 腾讯科技（深圳）有限公司 Lane positioning method and device, storage medium and server
CN112597401A (en) * 2020-11-26 2021-04-02 深圳市跨越新科技有限公司 Rest point recommendation method and system
CN112380314B (en) * 2020-12-04 2022-07-29 腾讯科技（深圳）有限公司 Road network information processing method and device, storage medium and electronic equipment
CN112559660B (en) * 2020-12-11 2022-06-17 腾讯科技（深圳）有限公司 Road data processing method and device, electronic equipment and storage medium
CN115144880A (en) * 2022-06-28 2022-10-04 平安银行股份有限公司 Road matching method, equipment and storage medium

Citations (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101571400A (en) * 2009-01-04 2009-11-04 四川川大智胜软件股份有限公司 Embedded onboard combined navigation system based on dynamic traffic information
CN102831768A (en) * 2012-08-15 2012-12-19 大连理工大学 Hybrid power bus driving condition forecasting method based on internet of vehicles
CN104318766A (en) * 2014-10-22 2015-01-28 北京建筑大学 Bus GPS track data road network matching method
CN104330089A (en) * 2014-11-17 2015-02-04 东北大学 Map matching method by use of historical GPS data
CN104781686A (en) * 2012-11-21 2015-07-15 苹果公司 Pathway matching
CN105628033A (en) * 2016-02-26 2016-06-01 广西鑫朗通信技术有限公司 Map matching method based on road connection relationship
CN105674995A (en) * 2015-12-31 2016-06-15 百度在线网络技术（北京）有限公司 Method for acquiring commuting route based on user's travel locus, and apparatus thereof
CN106441316A (en) * 2016-09-08 2017-02-22 复旦大学 Single-point road network matching method based on historical data
CN106525057A (en) * 2016-10-26 2017-03-22 陈曦 Generation system for high-precision road map
CN106855415A (en) * 2017-01-09 2017-06-16 北京京东尚科信息技术有限公司 Map-matching method and system
CN106969764A (en) * 2016-01-13 2017-07-21 北京四维图新科技股份有限公司 A kind of road matching method, device and vehicular map acquisition system
CN107421555A (en) * 2017-07-10 2017-12-01 奇瑞汽车股份有限公司 The method and apparatus for determining guidance path
CN107993437A (en) * 2017-12-08 2018-05-04 东莞中国科学院云计算产业技术创新与育成中心 Method, apparatus, computer equipment and the storage medium of newly-increased bus routes
CN207510306U (en) * 2017-03-30 2018-06-19 深圳市圣卓芯科技有限公司 A kind of on-vehicle navigation apparatus
CN108288096A (en) * 2017-01-10 2018-07-17 北京嘀嘀无限科技发展有限公司 Method and device for estimating journey time, model training
CN108446324A (en) * 2018-02-11 2018-08-24 重庆邮电大学 A kind of GPS data reconstructing method based on long memory network LSTM in short-term
CN108510741A (en) * 2018-05-24 2018-09-07 浙江工业大学 A kind of traffic flow forecasting method based on Conv1D-LSTM neural network structures
CN108780156A (en) * 2016-03-14 2018-11-09 北京嘀嘀无限科技发展有限公司 System and method for the path for determining mobile device
EP3410348A1 (en) * 2017-05-31 2018-12-05 HERE Global B.V. Method and apparatus for building a parking occupancy model
CN109035761A (en) * 2018-06-25 2018-12-18 复旦大学 Travel time estimation method based on back-up surveillance study
CN109300309A (en) * 2018-10-29 2019-02-01 讯飞智元信息科技有限公司 Road condition predicting method and device
WO2019028269A2 (en) * 2017-08-02 2019-02-07 Strong Force Iot Portfolio 2016, Llc Methods and systems for detection in an industrial internet of things data collection environment with large data sets
CN109429506A (en) * 2017-06-21 2019-03-05 北京嘀嘀无限科技发展有限公司 System and method for route planning
CN109425345A (en) * 2017-08-31 2019-03-05 北京嘀嘀无限科技发展有限公司 A kind of map route display methods, display system and computer installation
CN109444932A (en) * 2018-10-30 2019-03-08 百度在线网络技术（北京）有限公司 A kind of vehicle positioning method, device, electronic equipment and storage medium

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7657372B2 (en) * 2002-03-29 2010-02-02 Panasonic Corporation Map matching method, map matching device, database for shape matching, and shape matching device
US9459626B2 (en) * 2014-12-11 2016-10-04 Here Global B.V. Learning signs from vehicle probes
US10060751B1 (en) * 2017-05-17 2018-08-28 Here Global B.V. Method and apparatus for providing a machine learning approach for a point-based map matcher

Patent Citations (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101571400A (en) * 2009-01-04 2009-11-04 四川川大智胜软件股份有限公司 Embedded onboard combined navigation system based on dynamic traffic information
CN102831768A (en) * 2012-08-15 2012-12-19 大连理工大学 Hybrid power bus driving condition forecasting method based on internet of vehicles
CN104781686A (en) * 2012-11-21 2015-07-15 苹果公司 Pathway matching
CN104318766A (en) * 2014-10-22 2015-01-28 北京建筑大学 Bus GPS track data road network matching method
CN104330089A (en) * 2014-11-17 2015-02-04 东北大学 Map matching method by use of historical GPS data
CN105674995A (en) * 2015-12-31 2016-06-15 百度在线网络技术（北京）有限公司 Method for acquiring commuting route based on user's travel locus, and apparatus thereof
CN106969764A (en) * 2016-01-13 2017-07-21 北京四维图新科技股份有限公司 A kind of road matching method, device and vehicular map acquisition system
CN105628033A (en) * 2016-02-26 2016-06-01 广西鑫朗通信技术有限公司 Map matching method based on road connection relationship
CN108780156A (en) * 2016-03-14 2018-11-09 北京嘀嘀无限科技发展有限公司 System and method for the path for determining mobile device
CN106441316A (en) * 2016-09-08 2017-02-22 复旦大学 Single-point road network matching method based on historical data
CN106525057A (en) * 2016-10-26 2017-03-22 陈曦 Generation system for high-precision road map
CN106855415A (en) * 2017-01-09 2017-06-16 北京京东尚科信息技术有限公司 Map-matching method and system
CN108288096A (en) * 2017-01-10 2018-07-17 北京嘀嘀无限科技发展有限公司 Method and device for estimating journey time, model training
CN207510306U (en) * 2017-03-30 2018-06-19 深圳市圣卓芯科技有限公司 A kind of on-vehicle navigation apparatus
EP3410348A1 (en) * 2017-05-31 2018-12-05 HERE Global B.V. Method and apparatus for building a parking occupancy model
CN109429506A (en) * 2017-06-21 2019-03-05 北京嘀嘀无限科技发展有限公司 System and method for route planning
CN107421555A (en) * 2017-07-10 2017-12-01 奇瑞汽车股份有限公司 The method and apparatus for determining guidance path
WO2019028269A2 (en) * 2017-08-02 2019-02-07 Strong Force Iot Portfolio 2016, Llc Methods and systems for detection in an industrial internet of things data collection environment with large data sets
CN109425345A (en) * 2017-08-31 2019-03-05 北京嘀嘀无限科技发展有限公司 A kind of map route display methods, display system and computer installation
CN107993437A (en) * 2017-12-08 2018-05-04 东莞中国科学院云计算产业技术创新与育成中心 Method, apparatus, computer equipment and the storage medium of newly-increased bus routes
CN108446324A (en) * 2018-02-11 2018-08-24 重庆邮电大学 A kind of GPS data reconstructing method based on long memory network LSTM in short-term
CN108510741A (en) * 2018-05-24 2018-09-07 浙江工业大学 A kind of traffic flow forecasting method based on Conv1D-LSTM neural network structures
CN109035761A (en) * 2018-06-25 2018-12-18 复旦大学 Travel time estimation method based on back-up surveillance study
CN109300309A (en) * 2018-10-29 2019-02-01 讯飞智元信息科技有限公司 Road condition predicting method and device
CN109444932A (en) * 2018-10-30 2019-03-08 百度在线网络技术（北京）有限公司 A kind of vehicle positioning method, device, electronic equipment and storage medium

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Predicting Fine-Grained Traffic Conditions via Spatio-Temporal LSTM;XiaojuanWei等;《Wireless Communications and Mobile Computing》;20190114;第1-12页 *
基于出租车时空轨迹深度学习的城市道路行程时间预测方法;喻钢;《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》;20180915;第C034-303页 *
深度异步残差网络及在路网交通流预测中的应用;赵博;《中国优秀博硕士学位论文全文数据库(硕士) 工程科技Ⅱ辑》;20190115;C034-2525 *

Also Published As

Publication number Publication date
CN111721306A (en) 2020-09-29

Similar Documents

Publication Publication Date Title
CN111721306B (en) Road matching method and device, electronic equipment and readable storage medium
US10810419B2 (en) Method and apparatus for updating road map geometry based on received probe data
CN101275841B (en) Feature information collecting apparatus and feature information collecting method
CN110686686B (en) System and method for map matching
US10380503B2 (en) Distributed online learning for privacy-preserving personal predictive models
CN111854779B (en) Route planning method and device, electronic equipment and readable storage medium
Song et al. Enhancing GPS with lane-level navigation to facilitate highway driving
CN109937344A (en) For generating the method and system of the profile data of the section of electronic map
CN111198560B (en) Method and apparatus for predicting feature spatial attenuation using a variational self-encoder network
CN109781122A (en) High-precision map updating method and device
CN106767835B (en) Positioning method and device
JP2007140745A (en) Traffic congestion prediction system, traffic congestion factor estimation system, traffic congestion prediction method and traffic congestion factor estimation method
CN111859178A (en) Method and system for recommending boarding points
CN112163063A (en) Method, apparatus and computer-readable storage medium for generating high-precision map
US20230137263A1 (en) Method and apparatus for generating structured trajectories from geospatial observations
Ekpenyong et al. Extracting road information from recorded GPS data using snap-drift neural network
CN111385868A (en) Vehicle positioning method, system, device and storage medium
CN111899505B (en) Detection method and device for traffic restriction removal
US20210270629A1 (en) Method and apparatus for selecting a path to a destination
CN111539401A (en) Lane line detection method, device, terminal and storage medium based on artificial intelligence
Karimi et al. A methodology for predicting performances of map-matching algorithms
CN110793531A (en) Road matching method and device and readable storage medium
CN114396956A (en) Navigation method and apparatus, computing device, storage medium, and computer program product
CN117705141B (en) Yaw recognition method, yaw recognition device, computer readable medium and electronic equipment
CN112525213B (en) ETA prediction method, model training method, device and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant