WO2018145235A1

WO2018145235A1 - Distributed storage system for use with high-precision maps and application thereof

Info

Publication number: WO2018145235A1
Application number: PCT/CN2017/073022
Authority: WO
Inventors: 黄波; 张玉新
Original assignee: 驭势（上海）汽车科技有限公司
Priority date: 2017-02-07
Filing date: 2017-02-07
Publication date: 2018-08-16

Abstract

Provided in the present application is a storage node for use with high-precision maps, comprising: a distributed storage system for a high-precision map of the storage node, a method for deploying a high-precision map in the distributed storage system, a smart driving system which uses a high-precision map, and a vehicle in which the smart driving system is installed. The storage node for use with high-precision maps comprises: an information transceiving module, which is used for sending and/or receiving a sub-map of a high-precision map; and a storage equipment, which is used for storing a sub-map set of the high-precision map, wherein the sub-map set comprises map information of a driving direction which is directly relevant to a present storage node, and each sub-map corresponds to map information of a driving direction which is directly relevant to the present storage node. The technical solution of the present application saves the storage space of a smart driving system, and provides a feasible smart driving system which carries out smart driving by using high-precision maps.

Description

Distributed storage system with high precision map and its application

Technical field

The present invention relates to the field of intelligent driving technology, and in particular, to a storage node of a high-precision map, a high-precision map distributed storage system including the storage node, and a method for deploying a high-precision map in the distributed storage system, which is high in use. An intelligent driving system with an accuracy map and a vehicle in which the intelligent driving system is installed.

Background technique

Intelligent driving can be divided into automatic driving and driverless driving. Autopilot is a man-machine driving. It refers to driving on a certain section of a road and driving by a car on certain sections. The less the person is driving, the higher the degree of automatic driving. Driverless driving is not required for driverless driving. High-precision maps are one of the necessary technologies for large-scale deployment of unmanned vehicles. High-precision maps are an indispensable technology even for highly automated driving. Because high-precision maps contain rich road traffic information elements, high-precision maps can provide not only high-precision geographic coordinates, but also accurate road shapes, number of lanes, and lanes. Slope, curvature, heading, roll, etc. This combination of rich information and related positioning technology ensures smart driving safety and a good ride experience.

The general navigation map can be stored offline in the car and provide navigation services to the driver due to the small amount of information. Different from the general navigation map, the richness of the information described by the high-precision map determines that the high-precision map needs to occupy a relatively large storage space (for example, the high-resolution map formed by the map scanned by Google with lidar is occupied every mile. Approximately 1 gigabyte of storage space). Due to the limited capacity of storage devices that store high-precision maps on smart-drive vehicles, high-precision maps covering a large area cannot be stored entirely on a typical smart-driving car, which directly affects the range of exercise of intelligently-driving vehicles. However, the dynamic update of the global high-precision map through the cloud during the driving process of the smart car requires a large bandwidth and generates a long delay, so the feasibility of the dynamic update method is relatively small in practical application scenarios. The promotion of intelligent driving urgently requires a practical method to achieve effective dynamic update of high-precision maps on smart driving vehicles.

For a conventional navigation map, a large area-wide map can be stored on the vehicle. The dynamic update of the map is relatively simple. Generally, only the difference between the two map versions needs to be compared and the differentiated part is sent to the vehicle. can. However, for high-precision maps, the methods of storage, update, and use applied to conventional navigation maps cannot be directly applied.

Summary of the invention

For one or more problems existing in the prior art, the present application provides a storage node of a high-precision map, a high-precision map distributed storage system including the storage node, and a high-precision map deployed in the distributed storage system. The method, the intelligent driving system using a high-precision map, and the vehicle in which the intelligent driving system is installed.

The application specifically includes the following contents:

Embodiment 1. A storage node of a high precision map, comprising:

An information transceiving module for transmitting and/or receiving subgraphs of high precision maps; and

a storage device for storing a sub-atlas of a high-precision map, wherein the sub-atlas includes map information of a traveling direction directly related to the storage node, and each sub-picture corresponds to a traveling direction directly related to the storage node Map information.

Embodiment 2. The storage node of the high precision map according to embodiment 1, wherein the sub-atlas comprises a direct pre-order storage node of the current storage node (which is a pre-order storage node of the current storage node, and in the pre-order There is no other storage node on the shortest path of the storage node to the current storage node) through the current storage node to the direct subsequent storage node (which is the subsequent storage node of the current storage node, and not on the shortest path from the current storage node to the subsequent storage node) There is high-precision map information of all traveling directions between other storage nodes), and each sub-picture corresponds to high-precision map information of each traveling direction.

Embodiment 3. The storage node of embodiment 1, wherein the information transceiving module comprises a wireless communication module and optionally a communication module with a cloud system.

Embodiment 4. The storage node of embodiment 3, wherein the wireless communication module performs wireless communication using a V2X communication protocol (a communication protocol between the automobile and other objects).

The storage node according to embodiment 4, wherein the V2X communication protocol is selected from the group consisting of DSRC (Dedicated Short Range Communications), LTE-V (a 4G wireless broadband technology), and 5G ( At least one of the 5th generation mobile communication technologies).

The storage node according to any one of embodiments 1 to 5, wherein the storage node stores a high-precision map sub-picture from the storage node to a nearest emergency station route, the emergency site storage ratio A high-resolution map submap of a wider area of the storage node.

The storage node according to any one of embodiments 1 to 5, further comprising a CPU, a GPS system, and a memory.

Embodiment 8. A distributed storage system for high precision maps, including

a master node for storing global high precision maps; and

A plurality of storage nodes according to any one of embodiments 1 to 7, wherein each storage node is disposed at a specific location of the high precision map.

Embodiment 9. The distributed storage system of embodiment 8, wherein the storage node is installed on at least one of an existing road infrastructure and a roadside building or as a new road infrastructure.

Embodiment 10. The distributed storage system of embodiment 9, wherein the existing road infrastructure is selected from the group consisting of traffic lights, street lights, illegal cameras, street signs, roadside utility poles, and bridges.

Embodiment 11. The distributed storage system of embodiment 8, wherein the high precision map comprises: high precision geographic location coordinates, and road shape, number of lanes, slope of each lane, curvature, heading, and side One, many or all of the information.

Embodiment 12. The distributed storage system of embodiment 8, further comprising an emergency site, wherein each storage node stores a high-resolution map sub-map from the current storage node to the nearest emergency site route, the most recent The emergency site stores at least a high-precision map sub-picture stored by all direct subsequent storage nodes of the current storage node.

The distributed storage system of embodiment 12, wherein the emergency site has an interface for accessing and downloading a high-resolution map sub-picture in a network cable and/or USB communication manner. Passengers of smart driving vehicles can download more map sets than ordinary storage nodes through these interfaces and human-computer interaction interfaces, such as downloading to the local high-precision atlas that is needed to complete the driving.

The distributed storage system of embodiment 8, wherein the density of the plurality of storage nodes on the ground is set according to an information storage density of the high-precision map and a communication efficiency including a bandwidth and a delay. The storage capacity of each storage node can accommodate the size of the stored high-precision map sub-atlas, and each storage node can meet the bandwidth and delay requirements of high-precision map sub-picture transmission.

Embodiment 15. The distributed storage system of embodiment 12 or 13, wherein the emergency site is the storage node according to any one of embodiments 1 to 7, wherein the emergency site further stores a surrounding thereof A sub-atlas of a high-precision map stored in a storage node.

The distributed storage system of embodiment 8, wherein the primary node is a cloud system, wherein the cloud system comprises a cloud communication module.

Embodiment 17. A method of deploying a high-precision map in the distributed storage system according to any one of embodiments 8-16, comprising the steps of:

Store a global high-precision map in the primary node;

For each storage node (called the current storage node), perform the following process:

Searching all the direct subsequent nodes of the current storage node determined by all the direct preamble nodes and each direct preamble node, for each direct preamble node to the current storage node's traveling direction, constructing a pass from the direct preamble node a high-resolution map submap of the current storage node to all its immediate subsequent nodes,

All constructed high precision map subgraphs are stored in a storage node to form the sub-atlas.

Embodiment 18. The method of embodiment 17, further comprising:

Forming a location topology map of the storage node,

The step of "searching out all direct preamble nodes and all direct subsequent nodes of the current storage node determined by each direct preamble node" is performed according to the location topology map of the storage node.

Embodiment 19. The method of embodiment 17, wherein the distributed storage system further comprises an emergency site, the method further comprising: storing, in each storage node, a high precision map from the current storage node to the nearest emergency site route a submap, and a high precision map submap stored by at least all of the direct subsequent storage nodes of the current storage node in the nearest emergency site.

Embodiment 20. The method of embodiment 17, comprising:

Map update process:

Replace the original high-precision map in the area with the high-precision map of the update area in the master node;

For each storage node in the update area, perform the following process:

Searching all direct immediate nodes of the current storage node and all direct subsequent nodes of the current storage node determined by each direct preamble node, constructing a current storage from each direct preamble node to the current storage node The direct preamble of the node Pointing through the high-precision map submap of the current storage node to all direct subsequent nodes of the current storage node,

Embodiment 21. The method of embodiment 17, comprising:

Storage node deployment update process:

For each storage node in the new storage node set, re-search all direct forward nodes and all direct subsequent nodes of the current storage node determined by each direct pre-order node, and re-search the results with the pre-update The results are compared,

For a storage node that has a change in "all direct successor nodes and all direct subsequent nodes of the current storage node determined by each direct preamble node", the following process is performed:

For each direct preamble node to the current storage node's driving direction, construct a high-precision map sub-picture from the direct pre-order node through the current storage node to all its immediate subsequent nodes,

Embodiment 22. An intelligent driving system using a high-precision map, comprising:

Intelligent driving information transceiver module;

a control device for controlling vehicle motion; and

An intelligent driving storage device for storing a current high-precision map, the current high-precision map including a sub-atlas of a high-precision map, and optionally a frequency of use corresponding to each sub-graph of the high-precision map;

The intelligent driving system is configured to perform the following steps:

Receiving step: in the process of driving the vehicle with the intelligent driving system from the direct pre-order storage node to the current storage node, the smart driving information transceiver module receives a sub-graph of the high-precision map from the current storage node, which is a slave a subgraph of a high-precision map of the direct pre-order storage node heading for the direction of travel of the current storage node;

Map update step: update the subgraph of the received high precision map to the current high precision map; and

Execution step: The control device controls the vehicle to perform intelligent driving according to the current high-precision map.

Embodiment 23. The intelligent driving system of embodiment 22, wherein the receiving In the step, the smart driving information transceiver module first receives the version information of the submap of the high-precision map from the current storage node, and determines whether the sub-graph of the same high-precision map exists in the “smart driving storage device”, if the same does not exist. The subgraph of the high-precision map continues to receive the sub-picture of the high-precision map, and if there is a sub-picture of the same high-precision map, the sub-picture of the high-precision map is stopped;

The usage frequency information of the subgraph of the high precision map is updated.

The smart driving system of embodiment 23, wherein the version information comprises at least one, a plurality or all of: a subgraph of the high precision map from a direct preamble storage node to a current storage The direction of travel of the node, the date of the update, and the updated version number.

The intelligent driving system of embodiment 22, wherein the map updating step further comprises: determining whether a storage space of the smart driving storage device is sufficient, if sufficient, continuing to store, if not, deleting A subgraph of one or more high precision maps until the storage space is sufficient.

[Embodiment 26] The intelligent driving system according to Embodiment 25, wherein in the step of "deleting a sub-picture of one or more high-precision maps", priority is deleted according to usage frequency information of a sub-picture of the high-precision map Use subgraphs of high-precision maps with lower frequency.

The intelligent driving system of embodiment 22, wherein the receiving step further comprises: the smart driving information transceiver module receiving, from the current storage node, a high-precision map of the current storage node traveling to the direction of travel of the nearest emergency site Subgraph.

The intelligent driving system of embodiment 27, wherein the driving system is further configured to not receive information from a current storage node when the vehicle in which the intelligent driving system is installed is close to a current storage node At the time, the control vehicle is driven to the nearest emergency site based on a sub-graph of a high-precision map that travels from the immediate pre-order storage node to the direction of travel of the nearest emergency site.

The intelligent driving system of embodiment 22, wherein the intelligent driving system is further configured to perform the following steps:

The storage node update step: after the vehicle drives past the current storage node and heads to a direct subsequent storage node, updates the current storage node to a direct preamble node, and updates the direct subsequent storage node to the current node.

Embodiment 30. A vehicle in which the intelligent driving system according to any one of Embodiments 22 to 29 is installed.

The application solves the problem that the high-precision map has large storage requirements and dynamic update on the smart car according to the system of distributed storage of the global high-precision map according to the storage node. The method for deploying and updating the global high-precision map provided by the invention can conveniently determine the high-precision map sub-graph of each storage node according to the distribution of the storage nodes and effectively implement the high-accuracy map sub-graph of each storage node when updating the global high-precision map. Dynamic update. The intelligent driving system of the present application instantly updates the high-precision map required for driving during driving, thereby saving the storage space of the intelligent driving system, and caching the high-precision map sub-picture according to the frequency of use effectively avoids frequent sub-picture downloading. Thus, it provides a practical and intelligent driving system that uses high-precision maps for intelligent driving. Emergency site considerations also add to the safety of smart driving vehicles.

DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments will be briefly described below. It is obvious that the drawings in the following description relate only to some embodiments of the present disclosure, and are not to limit the disclosure. .

1 is an external schematic diagram of a distributed storage system of a high precision map of the present application.

2 is a schematic structural diagram of a system of a storage node of the present application.

3 is a schematic diagram of storage of a local high precision map in a storage node of the present application.

4 is an example of an algorithm flow diagram of a high precision map of the present application deployed in a storage node of a distributed storage system.

FIG. 5 is an example of an algorithm flow chart of a high-precision map dynamic update of the present application.

6 is an example of a flow chart for updating and using a local high-precision map of the intelligent driving system of the present application.

detailed description

The technical solutions of the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings of the embodiments of the present disclosure. It is apparent that the described embodiments are part of the embodiments of the present disclosure, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the described embodiments of the present disclosure without departing from the scope of the invention are within the scope of the disclosure.

The global high-precision map can actually be divided into many small blocks, each of which is a partial sub-map. In theory, no matter which location, as long as each smart car is stored inside With a number of high-resolution map subgraphs near the current location, the smart car can drive normally. Therefore, in the present application, the subgraph of the high precision map refers to a local submap of the global high precision map. In an ideal situation, as long as the smart driving car can continuously obtain high-precision map sub-pictures from the current position to the previous distance on the planned driving path, the intelligent driving car can also travel normally.

An aspect of the present application provides a storage node of a high-precision map, comprising: an information transceiving module for transmitting and/or receiving a sub-picture of a high-precision map; and a storage device for storing a sub-map of the high-precision map Atlas, wherein the sub-atlas includes map information of a driving direction directly related to the storage node, and each sub-picture corresponds to map information of one driving direction directly related to the storage node. The storage node of the present embodiment is an important component of a distributed storage system with high-precision maps. It is used to transmit sub-graphs of high-precision maps to a moving car in real time, which is the key to realizing the intelligent driving method of the present application. At each storage node, the car approaches the storage node from different directions, and it is considered that there are different driving directions. In order to facilitate the distinction among the stored high-precision map sub-pictures, the storage node and the car that the car has recently passed are The running storage node serves as a driving direction, and the driving direction determines a high-precision map sub-picture, and the high-precision map sub-picture should include high-precision map information reaching all subsequent storage nodes in the traveling direction.

In some embodiments, the sub-atlas includes a direct pre-order storage node of the current storage node (which is a pre-order storage node of the current storage node, and there is no other shortest path on the pre-order storage node reaching the current storage node) Storage node) high precision of all driving directions between the current storage node and the direct subsequent storage node (which is the subsequent storage node of the current storage node and no other storage nodes on the shortest path from the current storage node to the subsequent storage node) Map information, each sub-picture corresponds to high-precision map information for each driving direction. It can be seen that each direction of travel is determined by a direct pre-order storage node and a current storage node, and each high-precision map sub-picture contains high-precision map information that can reach all direct subsequent storage nodes from the direction of travel. The driving direction in the present application is mainly used for planning a reasonable driving route. Therefore, in general, the driving direction described in the present application only includes a reasonable driving direction, and does not include driving without obeying traffic rules. Directions, such as retrograde, reversing across solid lines are not a reasonable direction of travel.

In some embodiments, the information transceiving module includes a wireless communication module and an optional communication module with the cloud system. The wireless communication module exchanges information with the moving automobile, and mostly sends high-precision map sub-picture information to the intelligent driving system on the automobile.

In some embodiments, the wireless communication module uses a V2X communication protocol (car and Communication protocol between other objects) performs wireless communication. Other objects herein mainly refer to roadside equipment units, such as storage nodes installed on road infrastructure or on roadside buildings.

In some embodiments, the V2X communication protocol may be selected from the group consisting of DSRC (Dedicated Short Range Communications), LTE-V (a 4G wireless broadband technology), and 5G (5th generation mobile communication technology). At least one of them. However, those skilled in the art will appreciate that all other communication protocols that meet the communication requirements for bandwidth and latency can be used in this application.

In some embodiments, the storage node stores a high-precision map sub-map from the storage node to the nearest emergency site route, the emergency site storing a high-precision map sub-graph within a wider area than the storage node. The emergency site may be one of the storage nodes, except that the storage capacity of the emergency site should be larger than that of the ordinary storage node, and the latest emergency site stores at least the high storage of all direct subsequent storage nodes of the current storage node. Accurate map subgraphs, and also store a high-resolution map of a wide range of current administrative regions. When the intelligent driving vehicle runs to the current node, it is found that it cannot communicate with the current node. At this time, the intelligent driving vehicle runs to the emergency site using the recently downloaded high-precision map sub-picture from the storage node to the nearest emergency station route. And to obtain a wider range of stored maps at the emergency site, even in the case of a wide range of storage node downtime, the current driving requirements can be met.

In some embodiments, the storage node further includes a CPU, a GPS system, and a memory. The storage node can contain a complete computer and GPS system, so it can contain CPU and memory.

The distributed storage system of the high precision map provided by the present application includes a master node for storing a global high precision map; and a plurality of the above described storage nodes, wherein each storage node is disposed at a specific location of the high precision map. The master node stores a global high-precision map, and the storage node stores a high-resolution map subgraph. The global high-precision map refers to the entire high-precision map. In general, the map includes a map including all maps in a country such as China, or all maps including an administrative region such as a province. The size of the high-resolution map sub-picture is allocated by the master node from the global high-precision map according to the deployment situation of the storage node.

In some embodiments of the distributed storage system, the storage node is installed on at least one of an existing road infrastructure and a roadside building or as a new road infrastructure. The existing road infrastructure may be selected from one or more of traffic lights, street lights, illegal cameras, street signs, roadside poles, and bridges.

In some implementations of the distributed storage system, the high precision map includes: high precision Geographical coordinates, and one or more of the shape of the road, the number of lanes, the slope of each lane, the curvature, the heading, and the roll information.

In some embodiments, the distributed storage system further includes an emergency site, wherein each storage node stores a high-precision map sub-map from the current storage node to the nearest emergency site route, the latest emergency site storing at least A high-resolution map submap stored by all direct subsequent storage nodes of the current storage node. The emergency site has strong storage capacity, can store a large range of high-precision map sub-pictures, and can even store high-precision map sub-pictures of an administrative area such as a county or the entire city.

In some embodiments, the emergency site has an interface to access and download a high precision map submap in a network cable and/or USB communication mode. Passengers of intelligent driving vehicles can download more map sets than ordinary storage nodes through these interfaces and corresponding human-computer interaction interfaces (human-computer interaction interfaces can be used on smart driving cars or on the emergency site), such as in smart driving. In the case where the storage space of the car's intelligent driving system can be received, it can even be downloaded to all local high-precision maps required to complete the current driving.

In some embodiments, the density of the plurality of storage nodes on the ground is set according to the information storage density of the high-precision map and the communication efficiency including the bandwidth and the delay, so that the storage capacity of each storage node can be accommodated. The size of the stored high-resolution map sub-atlas, and each storage node can meet the bandwidth and delay requirements of high-precision map sub-picture transmission. For example, in the case of 4G wireless broadband communication protocol, storage nodes can be installed at each intersection. On the highway without intersections, storage nodes can be installed at the entrance and exit of the highway. Currently 4G wireless broadband technology It can achieve a download speed of about 100 megabits per second, and the speed will be reduced at high speeds. Therefore, under this design condition, even if the distance between every two intersections is one mile, the high-precision map occupies 1 Gbyte. It is also possible to download the map in about tens of seconds, and the high-precision map information of the expressway takes up less storage space. Therefore, this layout method can be completed by using the current common communication network. In the case of using high-precision maps of lower precision, a more sparse storage node layout can also be employed. With the improvement of the communication system, in the case of adopting the 5th generation mobile communication network in the future, it is also possible to adopt a more sparse storage node layout manner.

In some embodiments, the emergency site is any one of the storage nodes described in the manner described above, wherein the emergency site also stores a sub-atlas of high precision maps stored in storage nodes around it. In this embodiment, the emergency site is an enhanced storage node.

In a particularly specific embodiment, each storage node is configured as an emergency site, that is, each storage node stores a sub-atlas of high-precision maps stored in storage nodes around it. In this case, each storage node acts not only as a storage node but also as an emergency site, so that if there is a problem with one or two storage nodes in the network of the storage node, it does not affect the operation of the entire distributed storage system, thus The entire distributed storage system has high robustness.

In some implementations of the distributed storage system, the primary node is a cloud system, and the cloud system includes a cloud communication module. The cloud described in the present application has a meaning that is generally understood by those skilled in the art, and is actually a network. The term cloud system refers to a web server. The cloud system uses a cloud communication module, so that the storage node accesses the cloud system data (global high-precision map) in any place that can connect to the network.

The present application also provides a method for deploying a high-precision map in the distributed storage system described in any of the above, comprising the following steps:

Store a global high-precision map in the primary node;

For each storage node, perform the following process:

In the present application, the direct preamble node refers to the predecessor storage node of the current storage node, and there are no other storage nodes on the shortest path of the predecessor storage node to the current storage node. In the present application, a direct subsequent node refers to a subsequent storage node of the current storage node, and there are no other storage nodes on the shortest path from the current storage node to the subsequent storage node. In the present application, the term "all direct subsequent nodes determined by each direct preamble node" refers to all direct subsequent nodes that can be reached in the direction of travel determined from the immediate preamble node to the current node. In general, in the present application, each high precision map submap includes high precision map information from a direct preamble node of the current node to all direct subsequent nodes determined by the direct preamble node.

In some embodiments of the method, the method further comprises: forming a location topology map of the storage node, wherein "searching out all direct direct preamble nodes and all direct current nodes determined by each direct preamble node directly The steps of the subsequent nodes are performed according to the location topology map of the storage node. The direct preamble node and the direct subsequent node can be searched more conveniently by forming a topology map of the storage node, and the almost complete off-the-shelf graph-based data structure algorithm in the computer field can complete the search.

In some embodiments of the method, the distributed storage system further includes an emergency site, the method further comprising: storing, in each storage node, a high-precision map sub-picture from the current storage node to the nearest emergency site route And storing at least the high-precision map sub-picture stored by all direct subsequent storage nodes of the current storage node in the nearest emergency site.

In practical applications, high-precision maps may be updated at any time as the measurement technology is updated and the road conditions are updated. Therefore, the method of the present invention also includes a map update process. Therefore, in some implementations of the method, the method further includes:

Map update process:

For each storage node in the update area, perform the following process:

Searching all direct preamble nodes of the current storage node and all direct subsequent nodes determined by each direct preamble node, constructing the direct from the current storage node for each direct preamble node to the current storage node The high-precision map subgraph of the pre-order node through all current direct nodes of the current storage node to the current storage node,

In practical applications, the deployment of high-precision maps has problems that need to be updated at any time according to changes in the deployment of storage nodes. Therefore, the method of the present invention also includes a storage node deployment update process. Therefore, in some implementations of the method, the method further includes:

Storage node deployment update process:

For a storage node that has a change in "all direct preamble nodes and all direct subsequent nodes determined by each direct preamble node", the following process is performed:

For each direct preamble node to the current storage node's direction of travel, construct a high precision from the direct preamble node through the current storage node to all its immediate subsequent nodes Picture map,

The present application also provides an intelligent driving system using a high-precision map, comprising: an intelligent driving information transceiver module; a control device for controlling vehicle motion; and a smart driving storage device for storing a current high-precision map, The current high-precision map includes a sub-atlas of the high-precision map, and optionally the frequency of use corresponding to the sub-picture of each high-precision map; the intelligent driving system is configured to perform the following steps: receiving step: installing the In the process of driving the vehicle of the intelligent driving system from the direct pre-order storage node to the current storage node, the intelligent driving information transceiver module receives a sub-graph of the high-precision map from the current storage node, which is from the direct pre-order storage node to the current a sub-picture of a high-precision map of the direction of travel of the storage node; a map update step: updating the sub-picture of the received high-precision map to the current high-precision map; and performing steps: controlling the device to control the vehicle according to the current high-precision map Smart driving. In this embodiment, the storage node may be a storage node in a distributed storage system of the high precision map described in the present application.

In response to the smart driving system, the present application also provides an intelligent driving method of an intelligent driving system using a high-precision map, the smart driving reminder comprising: an intelligent driving information transceiver module; and a control device for controlling vehicle motion; And a smart driving storage device for storing a current high-precision map comprising a sub-atlas of a high-precision map, and optionally a frequency of use corresponding to a sub-picture of each high-precision map; The method includes the following steps: receiving a step of: receiving, in a process of driving a vehicle with the intelligent driving system from a direct pre-order storage node to a current storage node, the intelligent driving information transceiver module receives a high-precision map sub-sector from a current storage node a sub-picture of a high-precision map that travels from a direct pre-order storage node to a current storage node; a map update step: updating a sub-picture of the received high-precision map to a current high-precision map; and executing Step: The control device controls the vehicle for intelligent driving according to the current high-precision map.

In some embodiments, in the receiving step, the smart driving information transceiver module first receives version information of a submap of the high precision map from the current storage node, and determines whether the same high precision exists in the “smart driving storage device”. The sub-picture of the map, if there is no sub-picture of the same high-precision map, the sub-picture of the high-precision map is continued to be received, and if the sub-picture of the same high-precision map exists, the sub-picture of the high-precision map is stopped; The usage frequency information of the subgraph of the high precision map is updated. In this step, the smart driving storage device in the intelligent driving system is often capable of storing a plurality of high-precision map sub-pictures, and the frequently used map sub-pictures can be saved in the smart driving storage device. In the middle, when you need to use it, you only need to exchange the version number with the current storage node. If it is the latest one, you can download the high-resolution map sub-picture without using it repeatedly, and use it directly, which can save bandwidth. Save energy. When the same high-precision map sub-picture is used each time, it is updated with the frequency information. Generally speaking, if the frequency information is used, the frequency information is incremented by 1, indicating that the frequency of use is increased once.

In some embodiments, the version information includes at least one, a plurality or all of: a driving direction of the sub-picture of the high-precision map from a direct pre-order storage node to a current storage node, an update date, and an update Version number.

In some embodiments, the map updating step further comprises: determining whether the storage space of the smart driving storage device is sufficient, if sufficient, continuing to store, if not, deleting one or more submaps of the high precision map Until the storage space is sufficient.

In some embodiments, the smart driving system is further configured to be able to share a sub-picture of a high precision map with a surrounding smart driving vehicle.

In some embodiments, in the step of “deleting a sub-picture of one or more high-precision maps”, the sub-pictures of the high-precision map are used to preferentially delete the sub-high-precision maps with lower frequency. Figure.

In some embodiments, the receiving step further comprises: the smart driving information transceiver module receiving, from the current storage node, a sub-graph of the high-precision map of the current storage node heading toward the traveling direction of the nearest emergency station.

In some embodiments, the driving system is further configured to control the vehicle to be stored according to the direct pre-order when the location where the vehicle with the smart driving system is close to the current storage node still does not receive information from the current storage node. A subgraph of the high-precision map of the node heading for the direction of travel of the nearest emergency site heads for the emergency site. In some embodiments, the smart driving method further includes the step of: controlling the vehicle according to the slave directly when the location of the vehicle in which the smart driving system is installed is close to the current storage node and the information is still not received from the current storage node. The sub-picture of the high-precision map of the pre-order storage node heading for the direction of travel of the nearest emergency station heads to the emergency site.

In some embodiments, the intelligent driving system is further configured to perform the step of: storing a node update step: updating the current storage node to direct after the vehicle drives past the current storage node and heads to a direct subsequent storage node Preorder node and update the immediate subsequent storage node to the current node. In some embodiments, the smart driving method further includes the storage node update step described above.

The application also provides a vehicle equipped with the intelligent driving system of any of the above.

Example

The technical solutions of the present application are explained in more detail below with reference to the accompanying drawings. It should be understood that the embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the technical solutions of the present application.

1 is an external schematic diagram of a distributed storage system of a high precision map of the present application. In Figure 1, the storage node is mounted on a pole of a street light (2 in Figure 1). If a roadside infrastructure (such as traffic lights, illegal cameras, street lights, etc.) is installed with a storage node to store a high-precision map near the infrastructure (actually a sub-graph of a global high-precision map), the storage node is essentially one A small computer device that stores, receives, transmits (one-to-many), and updates functions such as high-precision map sub-pictures (see Figure 2). The intelligent driving car (1 in Figure 1) can download the stored high-precision map sub-picture to the smart driving car through communication with the corresponding current storage node when passing the corresponding infrastructure during driving (4 in Figure 1) . In a certain sense, the smart driving car is equipped with an intelligent driving system, and can also be “watched” as an enhanced version of the mobile “storage node” that can store, receive, update and share multiple high-precision map sub-pictures. Because the high-precision map stored by the storage nodes on the infrastructure is a partial sub-graph of the global high-precision map, it is designed and deployed to make V2X communication protocols (such as DSRC, LTE-V, and 5G, even at high speeds). It can also meet the bandwidth and delay requirements of local high-precision map transmission. In this way, each smart car only needs to store some small local maps. The smart car can continue to obtain the high-precision map needed for the current driving from the roadside storage infrastructure with high-precision map (similar to a partial window). Switch frequently in the global high-precision map).

In this embodiment, the nodes stored on the roadside infrastructure that store high-precision maps actually constitute a network map. On the roadside infrastructure, each storage node only needs to store the direct pre-order storage node during the driving process (there is no other storage node on the way that the pre-order node reaches the current node) through this storage node to the direct subsequent storage node. High-precision map information of all reasonable sections between (there are no other storage nodes on the way from the current storage node to the subsequent storage node). Since there are many different directions for driving to a storage node, each storage node may store multiple high-resolution map subgraphs. For example, in Figure 3, A, B, C, D, E, F, G, H, and K are storage nodes on the road, and the arrows drawn on each road indicate the direction in which the car can be used on the road. Storage Node B stores three high-resolution map sub-pictures in three different ways (from left to right, right to left, and top to bottom) that may pass through it. Each smart driving car follows B when it passes. Direction of travel Obtain the corresponding high-resolution map subgraph. In each subgraph, a subsection can be uniquely determined from a predecessor node to the current node. It should be noted that some of the road travel directions in Figure 3 are one-way roads, and some are two-way roads. The high-precision map sub-pictures directly ignore the travel paths that do not obey the traffic rules. In FIG. 3, when the vehicle approaches the storage node B from left to right, the high-resolution map sub-picture acquired by the vehicle from the storage node B should be reached in the direction of travel determined by the direct pre-order storage node A to the storage node B. All direct subsequent storage nodes F, D, E, G, and C. A high-precision map sub-picture corresponding to the other two driving directions in which the storage node B exists can be similarly explained according to FIG.

The information storage density of high-precision maps (the average number of bytes per kilometer) and the communication efficiency (bandwidth and latency) between the vehicle and the infrastructure will affect the denseness of the storage node deployment. Once the location of the storage node is determined, the cloud system (5 in FIG. 1) can calculate one or more local high-precision map sub-maps that each storage node needs to store according to the global high-precision map and the location topology of the storage node. (For details, see Figure 4), and then the corresponding high-resolution map submap will be deployed to the corresponding storage node through the network (3 in Figure 1). Whenever the storage node increases or decreases, the cloud system will re-execute the algorithm of Figure 3 to calculate the high-resolution map sub-graph that each storage node needs to deploy after the storage node increases or decreases. At this time, the cloud system only needs to change the sub-picture. The storage node can deploy a new high-resolution map submap.

In practical applications, the high-precision map may be updated at any time as the measurement technology is updated and the road conditions are updated. Therefore, the method of the present application also includes a dynamic update process of the high-precision map. Regardless of the method used to update the global high-precision map, when the global high-precision map is updated, the cloud system calculates which high-resolution map sub-graphs in the distributed storage nodes need to be updated, and then the storage nodes that need to be updated A new high-resolution map subgraph (see Figure 5).

With the distributed storage system described above, the corresponding intelligent driving system can be designed. In this embodiment, the intelligent driving system includes: an intelligent driving information transceiver module; a control device for controlling vehicle motion; and a smart driving storage device for storing a current high-precision map, the current high-precision map including high a sub-atlas of the accuracy map, and optionally a frequency of use corresponding to each of the sub-graphs of the high-precision map; the intelligent driving system is configured to perform the following steps: receiving the step: the vehicle from which the intelligent driving system is installed In the process of driving the direct pre-order storage node to the current storage node, the intelligent driving information transceiver module receives a sub-graph of the high-precision map from the current storage node, which is a driving direction from the direct pre-order storage node to the current storage node. Subgraph of high-precision map; map update step: update the subgraph of the received high-precision map to the current high-precision map And the execution step: the control device controls the vehicle to perform intelligent driving according to the current high-precision map.

After the above intelligent driving system is installed in the car, it is possible to use the high-precision map for intelligent driving.

The storage space on common smart cars should be enough to store some high-precision map sub-pictures, but there is often no global high-precision map. The high-precision map storage space on the smart car can be managed according to the frequency of use, that is, those high-precision map sub-pictures that retain high frequency of use if the storage space allows. When the smart car receives a new high-precision map sub-picture and the storage space on the car is not enough, then those sub-pictures with relatively low frequency used in the previously cached high-precision map sub-pictures will be replaced. The larger the storage space on the smart car, the more high-resolution map sub-pictures it can cache. The high-precision map sub-pictures can be reduced in the frequently-traveled area without the update of the high-precision map. The probability of the graph. With high-precision maps and corresponding positioning algorithms, smart driving cars can get rich information in high-precision maps for better planning and control, allowing passengers to enjoy a safe travel experience with good user experience (see Figure 6).

In the actual deployment process, you can consider setting up emergency sites in certain places (such as charging stations, parking lots, rest service areas, etc.) so that when there is communication problem between the smart driving car and the storage node (such as sudden network failure) There is an opportunity to update high-resolution maps. In the emergency site, the occupant of the intelligent driving car can connect to the high-precision map service system in the smart driving car and the emergency station by other communication methods (such as network cable, USB, etc.), and then download the data to the present according to the destination selection sub-picture data set. The car so that the smart car can continue to travel to the destination. That is to say, a smart driving vehicle supporting such a contingency plan must draw a specific storage area to store a high-precision map sub-picture from a storage node to some nearby emergency stations, so that the smart driving car can travel safely when a problem occurs. Go to the nearby emergency site. In fact, storing a high-precision map sub-picture from the current storage node to the nearest emergency site in each storage node can meet the needs of this scenario. At this time, the algorithms of Figure 4, Figure 5 and Figure 6 need to be corresponding. Expansion.

The algorithms of Figures 4, 5 and 6 are described below.

4 is an example of an algorithm flow diagram of a high precision map of the present application deployed in a storage node of a distributed storage system. The process of the algorithm is as follows. First, a global high-precision map is stored in the master node, and the master node acquires a location topology map of the storage node. Then all the storage nodes are processed. In the process of processing each storage node N, each driving direction of the storage node N is processed: in the current running direction, all direct fronts of the storage node N are calculated according to the location topology map of the storage node. Sequence node and all direct subsequent nodes (actually each straight According to the global high-precision map, a high-precision map sub-atlas containing N and all its direct pre-order nodes and all direct subsequent nodes in the current driving direction is constructed according to the global high-precision map.

FIG. 5 is an example of an algorithm flow chart of a high-precision map dynamic update of the present application. The process of the algorithm is as follows. First, the storage node set S included in the update area is found according to the update area of the global high-precision map and the topology map of all the storage nodes. A high precision map deployment algorithm similar to that shown in Figure 4 is then performed for all of the storage nodes in S. This algorithm is an example of a map update algorithm.

6 is an example of a flow chart for updating and using a local high-precision map of the intelligent driving system of the present application. The process of the algorithm is as follows. When the smart driving vehicle detects the presence of the front storage node CurNode (ie, the current storage node), it communicates with the current storage node CurNode, and sends the direct preamble node PreNode to the storage node CurNode. Then the smart driving car (that is, the intelligent driving system installed on the car) receives the version identifier of the high-resolution map sub-picture sent by the CurNode. If the high-precision map cache of the smart driving car contains the same version of the high-precision map sub-picture, there is no need to re-download, otherwise the high-resolution map sub-picture is re-downloaded. In the case of re-downloading the high-resolution map sub-picture, if the high-precision map buffer space on the smart-driving car is not enough, the high-precision map with a lower frequency of use in the cache is eliminated according to the size of the high-resolution map sub-picture to be received. Subgraph. Then, the high-precision map sub-picture sent by the CurNode as the first road segment (ie, the driving direction) sent by the CurNode is received, and the high-precision map sub-picture is stored in the high-precision map buffer. After the high-precision map sub-picture in the current traveling direction is prepared, the intelligently-driving vehicle (that is, the intelligent driving system) performs intelligent driving based on the current high-precision map.

The above description is only an exemplary embodiment of the present disclosure, and is not intended to limit the scope of the disclosure. The scope of the disclosure is determined by the appended claims.

Claims

A storage node of high precision map, comprising:

An information transceiving module for transmitting and/or receiving subgraphs of high precision maps; and

a storage device for storing a sub-atlas of a high-precision map, wherein the sub-atlas includes map information of a traveling direction directly related to the storage node, and each sub-picture corresponds to a traveling direction directly related to the storage node Map information.
The storage node of the high-precision map according to claim 1, wherein the sub-atlas comprises a direct pre-order storage node of the current storage node (which is a pre-order storage node of the current storage node, and the pre-order storage node arrives at the current There is no other storage node on the shortest path of the storage node) through the current storage node to the direct subsequent storage node (which is the subsequent storage node of the current storage node, and there is no other storage node on the shortest path from the current storage node to the subsequent storage node) High-precision map information of all driving directions between each, and each sub-picture corresponds to high-precision map information of each traveling direction.
The storage node of claim 1 wherein said information transceiving module comprises a wireless communication module and optionally a communication module with a cloud system.
The storage node according to claim 3, wherein said wireless communication module performs wireless communication using a V2X communication protocol (a communication protocol between the automobile and other objects).
The storage node according to claim 4, wherein said V2X communication protocol is selected from the group consisting of DSRC (Dedicated Short Range Communications), LTE-V (a 4G wireless broadband technology), and 5G (5th generation mobile) At least one of communication technologies).
The storage node according to any one of claims 1 to 5, wherein the storage node stores a high-precision map sub-picture from the storage node to a nearest emergency site route, the emergency site storage being wider than the storage node High-resolution map subgraphs within the region.
The storage node according to any one of claims 1 to 5, further comprising a CPU, a GPS system, and a memory.
High-resolution map distributed storage system, including

a master node for storing global high precision maps; and

A plurality of storage nodes according to any one of claims 1 to 7, wherein each storage node is disposed at a specific location of the high precision map.
The distributed storage system of claim 8 wherein said storage node is mounted to Installed on at least one of the existing road infrastructure and roadside buildings or as a new road infrastructure.
The distributed storage system of claim 9 wherein said existing road infrastructure is selected from the group consisting of traffic lights, street lights, illegal cameras, street signs, roadside utility poles, and bridges.
The distributed storage system according to claim 8, wherein said high-precision map comprises: high-precision geographic location coordinates, and one of a road shape, a number of lanes, a gradient of each lane, a curvature, a heading, and a roll information. Kind, multiple or all.
The distributed storage system of claim 8 further comprising an emergency site, wherein each storage node stores a high precision map submap from the current storage node to the nearest emergency site route, said recent emergency site being at least A high-precision map submap stored by all direct subsequent storage nodes of the current storage node.
The distributed storage system of claim 12 wherein said emergency site has an interface for accessing and downloading high precision map submaps in a network cable and/or USB communication mode.
The distributed storage system according to claim 8, wherein the density of the plurality of storage nodes in the ground is set according to the information storage density of the high-precision map and the communication efficiency including the bandwidth and the delay, so that: The storage capacity of the storage node can accommodate the size of the stored high-precision map sub-atlas, and each storage node can meet the bandwidth and delay requirements of high-precision map sub-picture transmission.
A distributed storage system according to claim 12 or 13, wherein said emergency site is a storage node according to any one of claims 1 to 7, wherein said emergency site further stores a storage node therearound A sub-atlas of stored high-resolution maps.
The distributed storage system of claim 8, wherein the primary node is a cloud system, and wherein the cloud system comprises a cloud communication module.
A method of deploying a high-precision map in the distributed storage system of any one of claims 8-16, comprising the steps of:

Store a global high-precision map in the primary node;

For each storage node (called the current storage node), perform the following process:

Searching all the direct subsequent nodes of the current storage node determined by all the direct preamble nodes and each direct preamble node, for each direct preamble node to the current storage node's traveling direction, constructing a pass from the direct preamble node Current storage node to all its straight a high-resolution map sub-picture of the subsequent node,

All constructed high precision map subgraphs are stored in a storage node to form the sub-atlas.
The method of claim 17 further comprising:

Forming a location topology map of the storage node,

The step of "searching out all direct preamble nodes and all direct subsequent nodes of this storage node determined by each direct preamble node" is performed according to the location topology map of the storage node.
The method of claim 17 wherein said distributed storage system further comprises an emergency site, said method further comprising: storing, in each storage node, a high precision map submap from the current storage node to the nearest emergency site route, and At least the high-precision map sub-picture stored by all direct subsequent storage nodes of the current storage node is stored in the nearest emergency site.
The method of claim 17 comprising:

Map update process:

Replace the original high-precision map in the area with the high-precision map of the update area in the master node;

For each storage node in the update area, perform the following process:

Searching all direct immediate nodes of the current storage node and all direct subsequent nodes of the current storage node determined by each direct preamble node, constructing a current storage from each direct preamble node to the current storage node The high-precision map sub-graph of the direct pre-node of the node through the current storage node to all direct subsequent nodes of the current storage node,

All constructed high precision map subgraphs are stored in a storage node to form the sub-atlas.
The method of claim 17 comprising:

Storage node deployment update process:

For each storage node in the new storage node set, re-search all its direct pre-order nodes and all direct subsequent nodes determined by each direct pre-order node, and re-search the results and the pre-update results. Compared,

For a storage node that has a change in "all direct preamble nodes and all direct subsequent nodes determined by each direct preamble node", the following process is performed:

For each direct preamble node to the current storage node's direction of travel, construct from The high-precision map subgraph of the direct pre-order node through the current storage node to all its immediate subsequent nodes,

All constructed high precision map subgraphs are stored in a storage node to form the sub-atlas.
An intelligent driving system that uses high-precision maps, including:

Intelligent driving information transceiver module;

a control device for controlling vehicle motion; and

An intelligent driving storage device for storing a current high-precision map, the current high-precision map including a sub-atlas of a high-precision map, and optionally a frequency of use corresponding to each sub-graph of the high-precision map;

The intelligent driving system is configured to perform the following steps:

Receiving step: in the process of driving the vehicle with the intelligent driving system from the direct pre-order storage node to the current storage node, the smart driving information transceiver module receives a sub-graph of the high-precision map from the current storage node, which is a slave a subgraph of a high-precision map of the direct pre-order storage node heading for the direction of travel of the current storage node;

Map update step: update the subgraph of the received high precision map to the current high precision map; and

Execution step: The control device controls the vehicle to perform intelligent driving according to the current high-precision map.
The intelligent driving system according to claim 22, wherein in said receiving step, said smart driving information transmitting and receiving module first receives version information of a subgraph of a high-precision map from a current storage node, and judges that "smart driving storage device" Whether there is a subgraph of the same high-precision map, if there is no subgraph of the same high-precision map, the subgraph of the high-precision map is continued to be received, and if there is a subgraph of the same high-precision map, the reception is stopped high. Subgraph of the accuracy map;

The usage frequency information of the subgraph of the high precision map is updated.
The intelligent driving system according to claim 23, wherein said version information comprises at least one, a plurality or all of: a traveling direction of a sub-picture of the high-precision map from a direct pre-order storage node to a current storage node , update date, and updated version number.
The intelligent driving system according to claim 22, wherein said map updating step further comprises: determining whether a storage space of said smart driving storage device is sufficient, if sufficient, continuing to store, if not, deleting one or more Subgraphs of high-precision maps until the storage space is sufficient.
The intelligent driving system according to claim 25, wherein in said step of "deleting a sub-picture of one or more high-precision maps", the frequency of priority deletion is lower according to the frequency of use of the sub-picture of the high-precision map Subgraph of a high precision map.
The intelligent driving system according to claim 22, wherein the receiving step further comprises: the smart driving information transceiving module receiving, from the current storage node, a sub-picture of the high-precision map of the current storage node traveling to the nearest emergency station.
The intelligent driving system according to claim 27, wherein said driving system is further configured to control the vehicle when said vehicle in which said intelligent driving system is installed close to a current storage node has not received information from a current storage node A subgraph of a high-precision map that travels from the immediate pre-order storage node to the direction of travel of the nearest emergency site to the emergency site.
The intelligent driving system of claim 22, wherein the intelligent driving system is further configured to perform the following steps:

The storage node update step: after the vehicle drives past the current storage node and heads to a direct subsequent storage node, updates the current storage node to a direct preamble node, and updates the direct subsequent storage node to the current node.
A vehicle equipped with the intelligent driving system according to any one of claims 22 to 29.