CN113724323A

CN113724323A - Map construction method, device and equipment

Info

Publication number: CN113724323A
Application number: CN202110975934.7A
Authority: CN
Inventors: 吕吉鑫; 孟超; 胡兵; 孙杰
Original assignee: Hangzhou Hikvision Digital Technology Co Ltd
Current assignee: Hangzhou Hikvision Digital Technology Co Ltd
Priority date: 2021-08-24
Filing date: 2021-08-24
Publication date: 2021-11-30
Anticipated expiration: 2041-08-24
Also published as: CN113724323B

Abstract

The application provides a map construction method, a map construction device and map construction equipment, wherein the method comprises the following steps: collecting backtracking data of a backtracking path and storing the backtracking data in the process that the teaching vehicle travels to the starting point of the task area along the backtracking path; when the teaching vehicle runs to the starting point of the task area, starting a map building data filling function, and filling backtracking data into the map building data; acquiring task data of a task path and filling the task data into map building data in the process that a teaching vehicle drives from a task area starting point to a task area end point along the task path; and when the teaching vehicle runs to the task area end point, stopping the map building data filling function, and building the target map based on all the map building data. By the technical scheme, the map building range of the target map is expanded, the operation interval of initial positioning is increased, and the success rate of the initial positioning is increased.

Description

Map construction method, device and equipment

Technical Field

The present application relates to the field of intelligent transportation technologies, and in particular, to a map construction method, apparatus, and device.

Background

The automatic parking means that the vehicle can be automatically parked without manual control, and the automatic parking can help a driver to automatically park. The automatic parking system may employ different methods to detect obstacles around the vehicle, for example, sensors are disposed around the vehicle, and the obstacles around the vehicle are detected by the sensors, or cameras are disposed around the vehicle, and the obstacles around the vehicle are detected by the cameras, or radars are disposed around the vehicle, and the obstacles around the vehicle are detected by the radars. Based on the data detected by these devices (e.g., sensors, or cameras, or radar), the automated parking system can determine the location of obstacles and then drive the vehicle into the parking space.

An Automatic Valet Parking (AVP) system is an application of automatic Parking, and in a Parking lot scene, the AVP system can help a vehicle to complete automatic driving, automatic garage searching and automatic Parking at a certain distance. For example, after a vehicle arrives at a parking lot, a driver parks the vehicle at a starting point of a task area, and starts an automatic parking operation by using an application program (such as an APP located in an intelligent terminal), and after receiving a start command, the AVP system can automatically drive the vehicle into the parking space (i.e., a destination of the task area).

In the related art, the vehicle positioning and path planning functions of the automatic parking scheme need to be manually set in advance, and the functions of real automatic vehicle positioning, automatic path planning, automatic passenger-replacing parking and the like cannot be realized.

Disclosure of Invention

The application provides a map construction method, which comprises the following steps:

collecting backtracking data of a backtracking path and storing the backtracking data in the process that a teaching vehicle travels to a starting point of a task area along the backtracking path; when the teaching vehicle runs to the starting point of the task area, starting a graph building data filling function, and filling the backtracking data into graph building data;

acquiring task data of a task path in the process that the teaching vehicle drives from a task area starting point to a task area end point along the task path, and filling the task data into map building data;

and when the teaching vehicle runs to the task area end point, stopping the map building data filling function, and building a target map based on all the map building data.

Illustratively, the collecting the trace-back data of the trace-back path and storing the trace-back data includes: acquiring backtracking data of the current position of the teaching vehicle when a data acquisition condition is met each time;

judging whether the quantity of backtracking data stored in the first-in first-out queue reaches n, wherein n is a positive integer;

if not, storing the backtracking data of the current position as the last backtracking data of the first-in first-out queue; if so, deleting the stored first backtracking data from the first-in first-out queue, and storing the backtracking data of the current position as the last backtracking data of the first-in first-out queue.

Illustratively, the determination that the data acquisition condition is satisfied includes:

counting the running time of the teaching vehicle from the time of storing backtracking data in the first-in first-out queue, and determining that a data acquisition condition is met when the running time is a preset time threshold; or the like, or, alternatively,

counting the travel distance of the teaching vehicle from the time of storing backtracking data in the first-in first-out queue, and determining that a data acquisition condition is met when the travel distance is a preset distance threshold.

Illustratively, the target map comprises a backtracking path sub-map constructed based on the backtracking data and a task path sub-map constructed based on the task data;

after the target map is constructed based on all the mapping data, the method further comprises the following steps:

positioning a target vehicle based on a backtracking path sub-map in the target map in the process that the target vehicle travels to the starting point of the task area along the backtracking path;

and if the target vehicle is successfully positioned before the target vehicle runs to the task area starting point, starting an automatic driving function when the target vehicle runs to the task area starting point, automatically driving the target vehicle based on a task path sub-map in the target map, and driving the target vehicle to the task area end point from the task area starting point along the task path.

After the target vehicle is located based on the traceback path sub-map in the target map, the method further includes: if the target vehicle is not positioned successfully when the target vehicle runs to the starting point of the task area, positioning the target vehicle based on the task path sub-map in the process that the target vehicle runs to the end point of the task area from the starting point of the task area along the task path; and starting an automatic driving function when the target vehicle is successfully positioned, automatically driving the target vehicle based on the task path sub-map, and driving the target vehicle to the task area end point.

For example, the locating the target vehicle based on the backtracking path sub-map in the target map includes: acquiring pose information of the target vehicle in the running process of the target vehicle;

and if the distance between the target vehicle and the coverage range of the backtracking path sub-map is determined to be smaller than a threshold value based on the pose information, positioning the target vehicle based on the backtracking path sub-map.

When the teaching vehicle runs to the starting point of the task area, acquiring and storing the reference data characteristics of the starting point of the task area; in the running process of a target vehicle, acquiring candidate data characteristics of each position, and positioning the target vehicle based on the candidate data characteristics; and if the candidate data feature of one position is matched with the reference data feature, positioning the position as the starting point of the task area, starting an automatic driving function, automatically driving the target vehicle based on a task path sub-map in the target map, and driving the target vehicle from the starting point of the task area to the end point of the task area along the task path.

In one possible embodiment, the backtracking data may include, but is not limited to, at least one of: image data, point cloud data, pose data and motion data; the task data may include, but is not limited to, at least one of: image data, point cloud data, pose data and motion data; the task area end point comprises a target parking space, and the target map is used for automatically parking a target vehicle to the target parking space.

The present application provides a map construction apparatus, the apparatus including:

the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring backtracking data of a backtracking path and storing the backtracking data in the process that a teaching vehicle runs to a starting point of a task area along the backtracking path;

the filling module is used for starting a drawing data filling function when the teaching vehicle runs to the starting point of the task area, and filling the backtracking data into drawing data;

the acquisition module is further used for acquiring task data of the task path in the process that the teaching vehicle runs from the starting point of the task area to the end point of the task area along the task path;

the filling module is further used for filling the task data into the mapping data;

and the construction module is used for stopping the map construction data filling function when the teaching vehicle runs to the task area end point and constructing a target map based on all the map construction data.

The application provides an intelligent driving device, include: a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor;

the processor is configured to execute machine executable instructions to perform the steps of:

The application provides a vehicle, wherein in the process that the vehicle runs to a starting point of a task area along a backtrack path, the vehicle acquires backtrack data of the backtrack path and stores the backtrack data;

when the vehicle runs to the starting point of the task area, the vehicle starts a map building data filling function and fills the backtracking data into the map building data;

in the process that the vehicle runs from the task area starting point to the task area end point along a task path, the vehicle collects task data of the task path and fills the task data into map building data;

and when the vehicle runs to the task area end point, stopping the map building data filling function of the vehicle, and building a target map based on all the map building data.

As can be seen from the above technical solutions, in the embodiments of the present application, when a target map is constructed, the target map includes a backtracking path sub-map constructed based on backtracking data and a task path sub-map constructed based on task data, the task path sub-map is a map of a task path between a task area starting point and a task area ending point, the backtracking path sub-map is a map of a backtracking path before the task area starting point, when a vehicle is positioned based on the target map, the vehicle can be positioned when the vehicle travels to the backtracking path, so that when the vehicle travels to the task area starting point, the vehicle is positioned successfully (instead of when the vehicle travels to the task area starting point, the vehicle starts to be positioned), thereby reducing a waiting time for positioning the vehicle, improving an accuracy of positioning the vehicle, and enabling the vehicle to travel from the task area starting point to the task area ending point based on the target map, the automatic passenger-replacing parking function is realized. By the method, the self-map building and non-inductive initial positioning functions of the AVP system can be realized, and the initial positioning success rate and the use convenience of the AVP system in a self-map building scene are greatly improved. When the AVP system utilizes the backtracking data and the task data to construct the target map, the map construction range of the target map is expanded, the operation interval of the initial positioning is increased, and the success rate of the initial positioning is increased.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments of the present application or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art according to the drawings of the embodiments of the present application.

FIG. 1 is a schematic flow chart diagram illustrating a mapping method according to an embodiment of the present application;

FIG. 2 is a schematic error ratio diagram of an odometer module in one embodiment of the present application;

FIG. 3 is a diagram illustrating backtracking data preservation according to an embodiment of the present application;

FIG. 4A is a schematic illustration of a teaching charting process in one embodiment of the present application;

FIG. 4B is a schematic flow chart diagram illustrating a mapping method according to an embodiment of the present application;

FIG. 4C is a schematic illustration of the composition of mapping data in an embodiment of the present application;

FIG. 4D is a schematic illustration of a path of a target map in one embodiment of the present application;

FIG. 5A is a schematic illustration of an automated valet parking process in one embodiment of the subject application;

FIG. 5B is a flowchart of a method for passenger parking based on a target map according to an embodiment of the present application;

FIG. 6 is a schematic diagram of a map construction apparatus according to an embodiment of the present application;

fig. 7 is a hardware configuration diagram of an intelligent driving apparatus according to an embodiment of the present application.

Detailed Description

The terminology used in the embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein is meant to encompass any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in the embodiments of the present application to describe various information, the information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present application. Depending on the context, moreover, the word "if" as used may be interpreted as "at … …" or "when … …" or "in response to a determination".

The embodiment of the present application provides a map construction method, which may be applied to an AVP system, and as shown in fig. 1, is a flow diagram of the map construction method, and the method may include:

step 101, collecting backtracking data of a backtracking path and storing the backtracking data in the process that the teaching vehicle travels to the starting point of the task area along the backtracking path. Illustratively, the backtracking data may include, but is not limited to, at least one of: image data, point cloud data, pose data and motion data.

Illustratively, in the process that the teaching vehicle travels to the starting point of the task area along the backtracking path, backtracking data of the current position of the teaching vehicle can be acquired every time a data acquisition condition is met; after obtaining the backtracking data of the current position, judging whether the number of the backtracking data stored in the first-in first-out queue reaches n, wherein n is a positive integer; if not, the backtracking data of the current position can be stored as the last backtracking data of the first-in first-out queue; if so, deleting the stored first backtracking data from the first-in first-out queue, and storing the backtracking data at the current position as the last backtracking data of the first-in first-out queue.

In one possible embodiment, the determination process of satisfying the data acquisition condition may include, but is not limited to: counting the running time of the taught vehicle from the time of storing the backtracking data in a first-in first-out queue, and determining that a data acquisition condition is met when the running time is a preset time threshold (which can be configured according to experience); or, counting the travel distance of the taught vehicle from the beginning of storing the backtracking data in the first-in first-out queue, and determining that the data acquisition condition is met when the travel distance is a preset distance threshold (which can be configured according to experience).

And 102, when the teaching vehicle runs to the starting point of the task area, starting a map building data filling function, and filling the backtracking data into the map building data. For example, after the mapping data filling function is started, all the saved trace data (i.e., n trace data) may be filled in the mapping data.

And 103, acquiring task data of the task path in the process that the teaching vehicle travels from the starting point of the task area to the end point of the task area along the task path, and filling the task data into the map building data.

For example, after the mapping data filling function is started, after task data of a task path is collected each time, the task data needs to be filled into the mapping data. Wherein the task data may include, but is not limited to, at least one of: image data, point cloud data, pose data and motion data.

And step 104, when the teaching vehicle runs to the end point of the task area, stopping the map building data filling function, and building a target map based on all the map building data, wherein the target map can comprise a backtracking path sub-map built based on the backtracking data and a task path sub-map built based on the task data.

For example, after the map data filling function is stopped, the target map may be constructed based on all the map data, and since all the map data includes the trace-back data and the task data, the target map may be constructed based on the trace-back data and the task data, which is not limited to this target map construction method. The map constructed based on the backtracking data is called a backtracking path sub-map, and the map constructed based on the task data is called a task path sub-map, i.e. the target map may comprise the backtracking path sub-map and the task path sub-map.

In one possible implementation, after the target map is constructed based on all the mapping data, the target vehicle is positioned based on the backtracking path sub-map in the target map in the process of driving the target vehicle to the starting point of the task area along the backtracking path. And if the target vehicle is successfully positioned before the target vehicle runs to the task area starting point, starting an automatic driving function when the target vehicle runs to the task area starting point, automatically driving the target vehicle based on the task path sub-map in the target map, and driving the target vehicle to the task area end point from the task area starting point along the task path. In summary, since the target map includes the traceback path sub-map, before the target vehicle travels to the starting point of the task area, the target vehicle can be positioned based on the traceback path sub-map, so that the target vehicle can be successfully positioned before traveling to the starting point of the task area, and when the target vehicle travels to the starting point of the task area, the automatic driving function can be directly started.

For example, the automatic driving function may include, but is not limited to, an automatic valet parking function, i.e., the task area end point may include a target parking space, the target map is used for automatically parking the target vehicle to the target parking space, for example, the automatic valet parking function is started when the target vehicle drives to the task area starting point, and the AVP system automatically drives the target vehicle to the target parking space based on the target map.

For example, after the target vehicle is located based on the backtracking path sub-map in the target map, if the target vehicle is not located successfully when the target vehicle travels to the task area starting point, the target vehicle is located based on the task path sub-map while the target vehicle travels from the task area starting point to the task area ending point along the task path. And starting an automatic driving function when the target vehicle is successfully positioned, automatically driving the target vehicle based on the task path sub-map, and driving the target vehicle to the destination of the task area.

For example, locating the target vehicle based on the traceback path sub-map in the target map may include, but is not limited to: in the running process of the target vehicle, the pose information of the target vehicle can be collected; if the distance between the target vehicle and the coverage range of the backtracking path sub-map is determined to be smaller than the threshold value based on the pose information, the target vehicle can be positioned based on the backtracking path sub-map in the target map.

In one possible embodiment, the reference data characteristic of the task area starting point may be collected and saved while the teaching vehicle travels to the task area starting point. On the basis, in the running process of the target vehicle, candidate data characteristics of all positions are collected, and the target vehicle is positioned based on the candidate data characteristics. In the process of positioning the target vehicle based on the candidate data features, if the candidate data features of a position are matched with the reference data features, the position can be positioned as a task area starting point (namely, the positioning of the target vehicle is completed), an automatic driving function is started, the target vehicle is automatically driven based on a task path sub-map in a target map, and the target vehicle is driven from the task area starting point to a task area end point along a task path.

The map construction method according to the embodiment of the present application will be described below with reference to specific embodiments.

Prior to describing the map construction method of the present application, technical terms related to the present application will be described.

AVP system: in a parking lot scene, the intelligent driving system helps a vehicle to complete automatic driving, automatic garage searching and automatic parking at a certain distance. For example, after the vehicle arrives at the parking lot, the driver parks the vehicle at the start point of the task area and starts the automatic parking operation using the application program, and the AVP system can automatically drive the vehicle into the parking space (i.e., the end point of the task area) after receiving the start command. In a parking lot without a high-precision map, the AVP system can realize automatic navigation and automatic parking functions of a vehicle. In order to realize the automatic parking function, the vehicle needs to be manually driven to complete teaching map building in a task area (from a task area starting point to a task area ending point) of a parking lot. After the map is built, when the vehicle approaches to a task area, the initial positioning in the built map needs to be quickly realized, the task path and scene information is obtained, and the automatic passenger-riding parking function is realized.

GNSS (Global Navigation Satellite System): GNSS is a space-based radio navigation positioning system that uses satellite observations to achieve the coordinates and velocity of a terrestrial receiver.

IMU (Inertial Measurement Unit): generally refers to a measuring device comprising a gyroscope and an accelerometer, and the IMU can measure motion data such as acceleration and angular velocity.

Wheel Speed Sensor (Wheel Speed Sensor): the sensor used for measuring the wheel rotating speed of the vehicle can obtain the information of the moving speed, the moving track and the like of the vehicle through the vehicle kinematic model.

Odometer module (Odometer Unit): the algorithm device for estimating the motion state of the carrier and recurrently estimating the motion track by using the sensor comprises an inertial odometer based on a wheel speed sensor and an IMU (inertial measurement Unit), a visual odometer based on a visual sensor, a laser odometer based on a laser radar and the like, and the type of the odometer module is not limited.

Tracking locator Unit (Tracking locator Unit): by using observation data (such as image data and radar point cloud data) of the perception sensor and relatively accurate prior pose and reference map, the tracking and positioning module can perform matching in a smaller local range to obtain the global pose relative to a map coordinate system.

Relocation Module (relocation Unit): by using the observation data of the perception sensor, the relatively inaccurate prior pose and the reference map, the repositioning module can perform matching in a smaller local range or a global range to obtain the global pose relative to a map coordinate system. The repositioning module is applied to initial positioning when the unmanned vehicle/robot enters a map range for the first time, and is also called as an initial positioning algorithm.

Integrated Localization System (Integrated Localization System): based on motion estimation and global position and orientation information provided by algorithm devices such as a GNSS, a relocation module, a tracking and positioning module, a milemeter module and the like, the combined positioning module can finally provide robust and accurate positioning information, namely the positioning information is accurate and reliable.

In order to realize the automatic passenger-replacing parking function, a map of a parking lot scene, for example, a map between a task area starting point and a task area ending point, needs to be constructed first, so that when a vehicle reaches the task area starting point, the vehicle can be positioned on the basis of the map, and after the vehicle is positioned successfully, the vehicle can be driven from the task area starting point to the task area ending point on the basis of the map, so that the automatic passenger-replacing parking function can be realized. However, if the vehicle is not successfully positioned when the vehicle reaches the task area starting point, the vehicle cannot be driven from the task area starting point to the task area ending point based on the map, that is, the automatic valet parking function cannot be realized.

In view of the above findings, in the embodiment of the present application, a self-map-building and non-inductive initial positioning method applied to an AVP system is provided, where a vehicle estimates sensory data (such as bit data and motion data) of the vehicle in real time during manual control driving, and simultaneously stores the sensory data in a latest preset driving distance in real time in a first-in first-out manner, where the sensory data is recorded as backtracking data, and a corresponding path is referred to as a backtracking path.

In the embodiment, a teaching map building process and an automatic passenger-replacing parking process are involved, and for convenience of distinguishing, a vehicle in the teaching map building process is recorded as a teaching vehicle, and a vehicle in the automatic passenger-replacing parking process is recorded as a target vehicle.

Aiming at the teaching and drawing building process, a teaching vehicle is driven to a starting point (such as an elevator port) of a task area of a parking lot, a drawing building command is sent, and the teaching vehicle is driven to travel to a target parking space along a task path to finish parking. After receiving the mapping command, the AVP system performs mapping by using the backtracking data stored before the start point of the task area and the sensing data (task data) acquired after the start point of the task area, so that the range of the finally constructed target map is the sum of the distances between the backtracking path and the task path. In the mapping process, all perception data are fused based on poses provided by combined positioning, and the self-mapping is aligned with a global coordinate system adopted by the combined positioning.

Aiming at the automatic passenger-replacing parking process, after a target vehicle approaches a certain established map (namely a target map) range in the manual driving process through combined positioning poses, an AVP system automatically enters an insensitive initial positioning state, and the current sensing data and the map are matched and positioned by using a repositioning algorithm. After the initial positioning is successful, the AVP system prompts readiness, and whether the automatic passenger-replacing parking function is operated or not can be manually selected.

In summary, the self-map-building and non-inductive initial positioning functions applied to the AVP system of the embodiment of the application greatly improve the initial positioning success rate and the use convenience of the AVP system in the self-map-building scene.

A combined positioning module: the vehicle utilizes the combined positioning module to estimate the global pose (such as position and attitude, the position represents longitude and latitude coordinates and the like, the attitude represents pitch angle, yaw angle, roll angle and the like) and motion data (also can be called as motion states such as acceleration, angular velocity and the like) of the vehicle in real time. Under the condition of map missing, the combined positioning module can fuse GNSS signals and data of the odometer module in an outdoor scene to obtain accurate and continuous global poses. After the vehicle enters an indoor scene with GNSS signals shielded, the combined positioning module can conjecture the global pose through the odometer module. In this case, the global pose accuracy provided by the combined positioning module gradually decreases with the increase of the travel distance, and as shown in fig. 2, although the global pose accuracy gradually decreases with the increase of the travel distance, the error rate of the odometer module is generally one percent to one thousandth, and in a scene where the travel distance is relatively limited, such as a parking lot, the global pose error estimated by the odometer module is limited, so that the pose accuracy requirement can be met. Referring to fig. 2, the combined positioning module can obtain an accurate global pose by using GNSS outdoors, and can estimate the global pose by using an odometer module indoors without a map. The global pose estimation based on the odometer module will decline with increasing distance traveled, and the true motion trajectory and the trajectory of pose estimates by the odometer module are shown in fig. 2.

Backtracking data storage: the AVP system may maintain a first-in-first-out queue that stores a maximum of n backtracking data, as shown in fig. 3, where each vertical line represents a location where the backtracking data is stored, and each rectangle represents a total travel distance where the backtracking data is stored. For example, during the driving process of the teaching vehicle, a certain number n of trace-back data are stored according to a fixed driving distance interval Δ d, the total driving distance of the stored data is about Δ d × n, that is, one trace-back data is stored every driving distance interval Δ d, when new trace-back data is stored, old trace-back data exceeding the limited number n is deleted, so that the size of the stored data volume is ensured to be unchanged, that is, n trace-back data are stored in a first-in first-out queue. Of course, in practical application, the method is not limited to the first-in first-out queue storage method, as long as n pieces of trace-back data can be stored.

For a teaching map building process, a map building method is provided in this embodiment of the present application, as shown in fig. 4A, for a schematic diagram of the teaching map building process, a task area starting point of a parking lot may be configured at will, which is not limited to this, for example, an elevator entrance, and a task area ending point of the parking lot may be a target parking space, that is, a teaching vehicle needs to be driven into the target parking space. And a path between the task area starting point and the task area end point is a task path. Teaching the process of the vehicle to travel from the starting point to the starting point of the task area, and calling the path to be traveled as a backtracking path.

Referring to fig. 4B, a schematic flow chart of a map building method is shown, where the method may include:

step 401, collecting the backtracking data of the backtracking path and storing the backtracking data in the process that the teaching vehicle travels to the starting point of the task area along the backtracking path. For example, the teaching vehicle needs to be driven from a starting point to a task area starting point of the parking lot, that is, the teaching vehicle needs to travel to the task area starting point along a backtracking path, and in the process, backtracking data of the backtracking path needs to be collected and stored.

Referring to fig. 3, when the teaching vehicle is at a starting point, the trace-back data a1 of the current position of the teaching vehicle is collected, and since the number of the trace-back data stored in the fifo queue is 0, if n is 3, the trace-back data a1 is stored as the last trace-back data of the fifo queue. After the teaching vehicle travels delta d (namely, the starting point + delta d) along the backtracking path, backtracking data a2 of the current position of the teaching vehicle is collected, and since the number of the backtracking data stored in the first-in first-out queue is 1, the backtracking data a2 is stored as the last backtracking data of the first-in first-out queue. After the teaching vehicle continues to travel for delta d (namely, the starting point +2 delta d) along the backtracking path, backtracking data a3 of the current position of the teaching vehicle is collected, and since the number of the backtracking data stored in the first-in first-out queue is 2, the backtracking data a3 is stored as the last backtracking data of the first-in first-out queue. After the teaching vehicle continues to travel for Δ d (i.e., the starting point +3 Δ d) along the backtracking path, backtracking data a4 of the current position of the teaching vehicle is collected, and since the number of the backtracking data stored in the fifo queue is 3, it is necessary to delete the stored first backtracking data a1 from the fifo queue and store the backtracking data a4 as the last backtracking data of the fifo queue. After the teaching vehicle continues to travel for Δ d (i.e., the starting point +4 Δ d) along the backtracking path, backtracking data a5 of the current position of the teaching vehicle is collected, and since the number of the backtracking data stored in the fifo queue is 3, it is necessary to delete the stored first backtracking data a2 from the fifo queue and store the backtracking data a5 as the last backtracking data of the fifo queue. By analogy, it is obvious that at most 3 pieces of backtracking data are stored in the fifo queue.

In the above embodiments, the backtracking data includes, but is not limited to, at least one of: image data, point cloud data, pose data and motion data. For example, cameras can be deployed around the teaching vehicle, and image data of the current position of the teaching vehicle is acquired through the cameras. Radars can be deployed around the teaching vehicle, and point cloud data of the current position of the teaching vehicle are collected through the radars. The combined positioning module can be used for collecting pose data (namely global pose) of the current position of the teaching vehicle and motion data, wherein the pose data comprises longitude and latitude coordinates, a pitch angle, a yaw angle, a roll angle and the like, and the motion data comprises acceleration, angular velocity and the like.

Step 402, when the teaching vehicle runs to the start point of the task area, a map building command is sent out, the map building command is used for starting a map building data filling function, and after the AVP system receives the map building command, the trace back data can be filled into the map building data, that is, all the stored trace back data (namely n trace back data) can be filled into the map building data. And after receiving the mapping command, the AVP system may further acquire a reference data feature (i.e., observation data) of the start point of the task area, and store the reference data feature of the start point of the task area, which may include, but is not limited to, image data and/or point cloud data.

For example, referring to fig. 4C, after receiving the mapping command, the AVP system fills all the trace-back data into the mapping data in a first-in first-out manner. For example, the first trace data in the first-in first-out queue is sent to the graph building module, the graph building module takes the first trace data as graph building data (which needs to participate in the graph building process subsequently), then the second trace data in the first-in first-out queue is sent to the graph building module, the graph building module takes the second trace data as the graph building data, and so on, the nth trace data in the first-in first-out queue is sent to the graph building module, and the graph building module takes the nth trace data as the graph building data, so that the graph building data may include n pieces of trace data. After all the n pieces of backtracking data are used as mapping data, the success of the teaching mapping initialization can be fed back through a display interface.

For example, after receiving the mapping command, the AVP system may further acquire image data of a starting point of the task area and point cloud data of the starting point of the task area, for example, cameras may be disposed around the teaching vehicle, and image data of the current position of the teaching vehicle is acquired through the cameras. Radars can be deployed around the teaching vehicle, and point cloud data of the current position of the teaching vehicle are collected through the radars. On the basis, the image data and the point cloud data can be stored to serve as reference data characteristics.

And step 403, after the driver obtains feedback of successful initialization of teaching map building of the AVP system, driving the teaching vehicle to perform teaching driving along an expected task path, namely driving the teaching vehicle to drive from the task area starting point to the task area end point along the task path. In the process that the teaching vehicle travels from the starting point of the task area to the end point of the task area along the task path, task data of the task path are collected and filled into the mapping data, namely, the AVP system fills the task data obtained in real time into the mapping data.

Referring to fig. 4C, for composition of mapping data and a use mode of the mapping data, after the mapping data filling function is started, the mapping module can keep the mapping task running in real time, and each time the AVP system collects task data of a task path, the task data can be sent to the mapping module, and the mapping module uses the task data as mapping data, so that the task data is filled in the mapping data.

In the above embodiments, the task data includes, but is not limited to, at least one of: image data, point cloud data, pose data and motion data. For example, cameras can be deployed around the teaching vehicle, and image data of the current position of the teaching vehicle is acquired through the cameras. Radars can be deployed around the teaching vehicle, and point cloud data of the current position of the teaching vehicle are collected through the radars. The combined positioning module can be used for collecting pose data (namely global pose) of the current position of the teaching vehicle and motion data, wherein the pose data comprises longitude and latitude coordinates, a pitch angle, a yaw angle, a roll angle and the like, and the motion data comprises acceleration, angular velocity and the like.

And step 404, when the teaching vehicle runs to the task area terminal, stopping the map building data filling function, and building a target map based on all the map building data. For example, after the teaching vehicle is driven to the target parking space (i.e., the task area end point) and the parking is completed, the teaching is clicked to end. And after the AVP system obtains the teaching ending command, stopping the graph building data filling function (namely all the graph building data are obtained), waiting for the graph building module to read all the graph building data, and building a target map based on all the graph building data, so that the graph building process is completed.

And 405, storing a target map, wherein the target map comprises a backtracking path sub-map constructed based on the backtracking data and a task path sub-map constructed based on the task data. Obviously, after the mapping module completes mapping, the range of the target map includes an observation range of the backtracking data (denoted as a backtracking path sub-map) and an observation range of the task data (denoted as a task path sub-map), as shown in fig. 4D, the path of the target map is composed of a backtracking path and a task path, and the range of the target map is the observation range along the backtracking path and the task path.

For example, the AVP system may also feed back the target map to the display interface in the form of a thumbnail, confirm and save the target map in a named manner by a human, that is, set a name for the target map, and save the target map.

For the automatic passenger-replacing parking process, the embodiment of the present application provides a passenger-replacing parking method based on a target map, as shown in fig. 5A, which is a schematic diagram of the automatic passenger-replacing parking process, a starting point of a task area of a parking lot may be configured at will, such as an elevator entrance, and a destination of the task area of the parking lot may be a target parking space, that is, a target vehicle (such as target vehicle a, target vehicle B, or target vehicle C) needs to be driven into the target parking space. Referring to fig. 5B, a schematic flow chart of a method for passenger parking based on a target map is shown, where the method includes:

and 501, in the running process of the target vehicle, acquiring the pose of the target vehicle (such as the pose of a block target vehicle provided by the combined positioning module in real time), and determining whether the target vehicle is close to the range of the target map or not based on the pose of the target vehicle. If so, executing step 502, if not, continuing to acquire the pose of the target vehicle, determining whether the target vehicle is close to the range of the target map or not based on the pose of the target vehicle, and so on.

For example, during the running process of the target vehicle, pose information (such as a position of the target vehicle, for example, longitude and latitude coordinates) of the target vehicle may be obtained in real time, and if it is determined based on the pose information that the distance between the target vehicle and the coverage area of the backtracking path sub-map in the target map is smaller than a threshold (the threshold may be configured according to experience), that is, the distance between the longitude and latitude coordinates of the target vehicle and the coverage area of the backtracking path sub-map is smaller than the threshold, it is determined that the target vehicle has approached the range of the target map, and step 502 is executed.

For example, the target map may be stored in the server after the target map is obtained, and the target map may be downloaded from the server during the driving of the target vehicle, or the target map may be downloaded from the server before the driving of the target vehicle, which is not limited thereto. Based on this target map, it is possible to determine whether the target vehicle has approached the range of the target map during travel of the target vehicle.

Step 502, when the target vehicle has approached the range of the target map, the target vehicle may be located based on the backtracking path sub-map in the target map, that is, in the process that the target vehicle travels along the backtracking path to the starting point of the task area, the target vehicle may be located based on the backtracking path sub-map in the target map.

Referring to fig. 5A, for the target vehicle a, the target vehicle a travels to the task area starting point along the backtracking path, and in the process of traveling to the task area starting point along the backtracking path, since the target map includes the backtracking path sub-map corresponding to the backtracking path, the target vehicle a can be located based on the backtracking path sub-map. Since the coverage area of the backtracking path sub-map is relatively large, the target vehicle a can be initially positioned between the starting points of the task areas in a large probability.

Step 503, if the target vehicle is successfully located before the target vehicle travels to the starting point of the task area, the automatic driving function (i.e. the automatic valet parking function) is started when the target vehicle travels to the starting point of the task area. Alternatively, if the target vehicle is successfully located when the target vehicle travels to the task area starting point, the automatic driving function may be activated when the target vehicle travels to the task area starting point. Or if the target vehicle is not positioned successfully when the target vehicle runs to the task area starting point, the target vehicle is positioned continuously based on the task path sub-map in the target map in the process that the target vehicle runs from the task area starting point to the task area end point along the task path, and the automatic driving function is started when the target vehicle is positioned successfully.

For example, referring to fig. 5A, for the target vehicle a, in the process that the target vehicle a travels to the task area starting point along the backtracking path, the target vehicle a may be located based on the backtracking path sub-map, and if the target vehicle a is successfully located before the target vehicle a travels to the task area starting point, the automatic driving function may be started for the target vehicle a when the target vehicle a travels to the task area starting point.

Or, if the target vehicle a is not successfully positioned when the target vehicle a travels to the task area starting point, after the target vehicle a travels to the task area starting point, the target vehicle a needs to travel from the task area starting point to the task area ending point along the task path, and the target vehicle a continues to be positioned based on the task path sub-map during the course of the target vehicle a traveling from the task area starting point to the task area ending point along the task path until the target vehicle a is successfully positioned, and the automatic driving function is started for the target vehicle a.

For example, the target vehicle B does not travel along the backtracking route to the task area starting point, and therefore the target vehicle B cannot be positioned based on the backtracking route sub-map, that is, the initial positioning is completed before the task area starting point with a small probability in the case of the non-inductive initial positioning. Obviously, if the target vehicle B is not successfully positioned (the non-inductive initial positioning is not yet successful) when the target vehicle B travels to the task area starting point, the target vehicle B needs to be driven to the vicinity of the task area starting point, and the positioning of the target vehicle B is completed in the following manner:

during the driving process of the target vehicle B, candidate data features (such as image data and/or point cloud data) of each position of the target vehicle B (i.e. each position during the driving process) are collected, and the target vehicle B is located based on the candidate data features. For example, if the candidate data feature of a location (i.e., the task area starting point) matches the reference data feature (e.g., the image data and/or the point cloud data) of the task area starting point, i.e., the similarity between the candidate data feature and the reference data feature is greater than the similarity threshold (configured empirically), the location may be located as the task area starting point, the location of the target vehicle B is completed, and the automatic driving function is started.

In summary, the target vehicle B may be located based on the candidate data features of each position of the target vehicle B during the driving process, that is, the target vehicle B is successfully located at the start point of the task area.

Or if the target vehicle B is not successfully positioned when the target vehicle B travels to the task area starting point, the target vehicle B needs to travel from the task area starting point to the task area ending point along the task path after the target vehicle B travels to the task area starting point, and the target vehicle B continues to be positioned based on the task path sub-map during the process that the target vehicle B travels from the task area starting point to the task area ending point along the task path until the target vehicle B is successfully positioned, and the automatic driving function is started for the target vehicle B.

For example, for the target vehicle C, the target vehicle C does not pass through the task area starting point, and the target vehicle C cannot be located based on the backtracking path sub-map, and the target vehicle C needs to travel from the task area starting point to the task area ending point along the task path.

And step 504, after the automatic driving function (namely the automatic valet parking function) is started, automatically driving the target vehicle based on the task path sub-map in the target map, and driving the target vehicle from the task area starting point to the task area end point along the task path. For example, after the target vehicle is successfully positioned, the AVP system prompts that the automatic valet parking function is ready, the driver manually confirms to start the automatic valet parking function, and after the driver manually confirms to start the automatic valet parking function, the driver can leave the target vehicle.

In a possible implementation manner, when the target vehicle runs to the starting point of the task area, if the target vehicle has no initial positioning success, the AVP system does not prompt the user to "successfully position, and the AVP function can be enabled", on this basis, the user can stop the target vehicle near the starting point of the task area, then choose to start the "AVP function", start the initial positioning function of the starting point of the task area, and the AVP system performs the initial positioning of the starting point of the task area, that is, positions the current position as the starting point of the task area, completes the positioning of the target vehicle (that is, the initial positioning of the starting point of the task area succeeds), and starts the automatic driving function. And if the initial positioning of the starting point is not successful, indicating the user to drive the target vehicle to continue driving, and performing 'task path positioning' until the positioning is successful.

According to the technical scheme, the target map comprises the backtracking path sub-map and the task path sub-map, the task path sub-map is a map of a task path between a task area starting point and a task area end point, the backtracking path sub-map is a map of a backtracking path before the task area starting point, when the vehicle is positioned based on the target map and runs to the backtracking path, the vehicle can be positioned, therefore, when the vehicle runs to the task area starting point, the vehicle is positioned successfully, and therefore the vehicle can run to the task area end point from the task area starting point based on the target map, and the automatic passenger replacing function is achieved. By the method, the self-map building and non-inductive initial positioning functions of the AVP system can be realized, and the initial positioning success rate and the use convenience of the AVP system in a self-map building scene are greatly improved. When the AVP system utilizes the backtracking data and the task data to construct the target map, the map construction range of the target map is expanded, the operation interval of the initial positioning is increased, and the success rate of the initial positioning is increased. In the mode, the backtracking data and the task data are utilized to build the map together, the map range is wider, and the map is more beneficial to success of initial positioning as soon as possible. The positioning function is triggered after the combined positioning information shows that the vehicle approaches the map range, the vehicle is supported to approach the starting point of the task area from different paths, and when the driving path in front of the starting point of the task area and the backtracking path in the process of building the map coincide, a longer initial positioning interval is possessed, and the initial positioning success rate is higher. The AVP system maintains the backtracking data with fixed storage size in real time, and uses the backtracking data and the task data to build the graph after obtaining the graph building command, thereby ensuring the sufficiency of the graph building range.

Based on the same application concept as the method, in the embodiment of the present application, a map building apparatus is provided, as shown in fig. 6, and is a schematic structural diagram of the map building apparatus, the apparatus may include:

the acquisition module 61 is used for acquiring the backtracking data of the backtracking path and storing the backtracking data in the process that the teaching vehicle travels to the starting point of the task area along the backtracking path;

a filling module 62, configured to start a map building data filling function when the teaching vehicle travels to the start point of the task area, and fill the backtracking data into map building data;

the acquisition module 61 is further configured to acquire task data of a task path in the process that the teaching vehicle travels from the task area starting point to the task area ending point along the task path;

the filling module 62 is further configured to fill the task data into the mapping data;

and the building module 63 is used for stopping the map building data filling function when the teaching vehicle runs to the task area end point, and building a target map based on all the map building data.

For example, the collecting module 61 collects the trace-back data of the trace-back path, and is specifically configured to: acquiring backtracking data of the current position of the teaching vehicle when a data acquisition condition is met each time; judging whether the quantity of backtracking data stored in the first-in first-out queue reaches n, wherein n is a positive integer; if not, storing the backtracking data of the current position as the last backtracking data of the first-in first-out queue; if so, deleting the stored first backtracking data from the first-in first-out queue, and storing the backtracking data of the current position as the last backtracking data of the first-in first-out queue.

For example, the acquisition module 61 determines that the data acquisition condition is satisfied, and specifically:

In a possible implementation, the target map includes a backtracking path sub-map constructed based on the backtracking data and a task path sub-map constructed based on the task data, and the apparatus further includes (not shown in fig. 6): the positioning module is used for positioning the target vehicle based on a backtracking path sub-map in the target map in the process that the target vehicle travels to the starting point of the task area along the backtracking path; and the automatic driving module is used for starting an automatic driving function when the target vehicle drives to the task area starting point if the target vehicle is successfully positioned before the target vehicle drives to the task area starting point, automatically driving the target vehicle based on a task path sub-map in the target map, and driving the target vehicle to the task area end point from the task area starting point along the task path.

For example, the positioning module is further configured to, if the target vehicle is not successfully positioned when the target vehicle travels to the task area starting point, position the target vehicle based on the task path sub-map in a process that the target vehicle travels from the task area starting point to the task area ending point along the task path; the automatic driving module is further used for starting an automatic driving function when the target vehicle is successfully positioned, automatically driving the target vehicle based on the task path sub-map, and driving the target vehicle to the task area end point.

For example, the positioning module is specifically configured to, when positioning the target vehicle based on a backtracking path sub-map in a target map: acquiring pose information of a target vehicle in the running process of the target vehicle; and if the distance between the target vehicle and the coverage range of the backtracking path sub-map is determined to be smaller than a threshold value based on the pose information, positioning the target vehicle based on the backtracking path sub-map.

Illustratively, the acquisition module 61 is further configured to acquire and store the reference data characteristic of the task area starting point when the teaching vehicle travels to the task area starting point;

the positioning module is further used for acquiring candidate data characteristics of each position in the running process of the target vehicle and positioning the target vehicle based on the candidate data characteristics;

the automatic driving module is further configured to, if a candidate data feature of a location matches the reference data feature, locate the location as the task area starting point, start an automatic driving function, automatically drive the target vehicle based on a task path sub-map in the target map, and drive the target vehicle along the task path from the task area starting point to the task area ending point.

Based on the same application concept as the method, the embodiment of the present application provides an intelligent driving device, as shown in fig. 7, the intelligent driving device may include: a processor 71 and a machine-readable storage medium 72, the machine-readable storage medium 72 storing machine-executable instructions executable by the processor 71; the processor 71 is configured to execute machine executable instructions to perform the following steps:

and when the teaching vehicle runs to the task area end point, stopping the map building data filling function, and building a target map based on all the map building data. Optionally, the target map includes a backtracking path sub-map constructed based on the backtracking data and a task path sub-map constructed based on the task data.

Based on the same application concept as the method, the embodiment of the application further provides a vehicle, and in the process that the vehicle runs to the starting point of the task area along the backtracking path, the vehicle acquires backtracking data of the backtracking path and stores the backtracking data;

and when the vehicle runs to the task area end point, stopping the map building data filling function of the vehicle, and building a target map based on all the map building data. Optionally, the target map includes a backtracking path sub-map constructed based on the backtracking data and a task path sub-map constructed based on the task data.

Based on the same application concept as the method, embodiments of the present application further provide a machine-readable storage medium, where a plurality of computer instructions are stored, and when the computer instructions are executed by a processor, the map construction method disclosed in the above example of the present application can be implemented.

The machine-readable storage medium may be any electronic, magnetic, optical, or other physical storage device that can contain or store information such as executable instructions, data, and the like. For example, the machine-readable storage medium may be: a RAM (random Access Memory), a volatile Memory, a non-volatile Memory, a flash Memory, a storage drive (e.g., a hard drive), a solid state drive, any type of storage disk (e.g., an optical disk, a dvd, etc.), or similar storage medium, or a combination thereof.

The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. A typical implementation device is a computer, which may take the form of a personal computer, laptop computer, cellular telephone, camera phone, smart phone, personal digital assistant, media player, navigation device, email messaging device, game console, tablet computer, wearable device, or a combination of any of these devices.

For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functionality of the units may be implemented in one or more software and/or hardware when implementing the present application.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

Furthermore, these computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above description is only an example of the present application and is not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A map construction method, characterized in that the method comprises:

2. The method of claim 1,

the collection the backtracking data of the backtracking path, and save the backtracking data, include:

acquiring backtracking data of the current position of the teaching vehicle when a data acquisition condition is met each time;

3. The method of claim 2,

the determination process of satisfying the data acquisition condition includes:

4. The method of claim 1, wherein the target map comprises a backtracking path sub-map constructed based on the backtracking data and a task path sub-map constructed based on the task data;

5. The method of claim 4, wherein after locating the target vehicle based on the traceback path sub-map in the target map, the method further comprises:

if the target vehicle is not positioned successfully when the target vehicle runs to the starting point of the task area, positioning the target vehicle based on the task path sub-map in the process that the target vehicle runs to the end point of the task area from the starting point of the task area along the task path;

and starting an automatic driving function when the target vehicle is successfully positioned, automatically driving the target vehicle based on the task path sub-map, and driving the target vehicle to the task area end point.

6. The method of claim 4 or 5, wherein the locating the target vehicle based on the backtracking path sub-map in the target map comprises:

acquiring pose information of the target vehicle in the running process of the target vehicle;

7. The method of claim 1, wherein reference data characteristics of the task area starting point are collected and saved as the teach pendant vehicle travels to the task area starting point;

in the running process of a target vehicle, acquiring candidate data characteristics of each position, and positioning the target vehicle based on the candidate data characteristics; and if the candidate data feature of one position is matched with the reference data feature, positioning the position as the starting point of the task area, starting an automatic driving function, automatically driving the target vehicle based on a task path sub-map in the target map, and driving the target vehicle from the starting point of the task area to the end point of the task area along the task path.

8. The method according to any one of claims 1-5, wherein the backtracking data comprises at least one of: image data, point cloud data, pose data and motion data; the task data includes at least one of: image data, point cloud data, pose data and motion data; the task area end point comprises a target parking space, and the target map is used for automatically parking a target vehicle to the target parking space.

9. A map building apparatus, characterized in that the apparatus comprises:

10. An intelligent driving apparatus, comprising: a processor and a machine-readable storage medium storing machine-executable instructions executable by the processor;

11. A vehicle, characterized in that,

in the process that the vehicle drives to the starting point of a task area along a backtracking path, the vehicle acquires backtracking data of the backtracking path and stores the backtracking data;