A robot companion path planning method fusing sensor and large model

Abstract

The application discloses a kind of fusion sensor and robot accompanying path planning method of big model, it is related to accompanying path planning technical field, including the following steps: based on laser radar obtains the three-dimensional point cloud graph of accompanying target, and is marked as target point cloud graph;Based on target point cloud graph obtains target space;Based on target space and target square obtains reference square;Based on reference square obtains final square;Based on direction wide side of final square;Based on reference surface;Based on reference surface, obtain starting area;Based on laser radar obtains the three-dimensional point cloud graph of environment, and is marked as environment point cloud graph;Based on starting area, reference surface and starting area to judge whether there is obstacle on path;The application is used to solve the problem that existing accompanying path planning technology cannot set appropriate obstacle recognition method based on accompanying target, resulting in large data processing amount and complex data processing.

Description

Technical Field

[0001] This invention relates to the field of accompanying path planning technology, specifically to a robot accompanying path planning method that integrates sensors and large models. Background Technology

[0002] With the rapid development of robotics technology, robots are being used more and more widely in industries, services, logistics and other fields; for example, robots that guide pedestrians and avoid obstacles; traditional robot path planning methods mainly rely on preset maps and fixed algorithms; therefore, traditional methods are difficult to handle problems such as dynamic obstacles and unknown terrain in the environment;

[0003] Because the accompanying target is autonomous, its walking path may not follow the planned route exactly. Therefore, a method is needed to determine in real time whether there are obstacles in front of the accompanying target. Existing obstacle recognition requires data analysis of the entire surrounding environment and cannot set a suitable obstacle recognition method based on the accompanying target, resulting in a large amount of data computation and complexity. For example, patent application CN109059924A discloses an incremental path planning method and system for accompanying robots based on the A* algorithm. This scheme uses the path from the previous moment to perform incremental path updates, saving path planning time, improving the reaction speed of the accompanying robot, and avoiding new obstacles in time. However, it fails to set a suitable obstacle recognition method based on the accompanying target, resulting in a large amount of data processing and complex data processing. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in the prior art. It involves acquiring a 3D point cloud map of an accompanying target based on lidar and marking it as a target point cloud map; acquiring a target space based on the target point cloud map; acquiring a reference square based on the target space and a target square acquisition method; acquiring a final square based on the reference square; acquiring the directional width based on the final square; acquiring a reference surface based on the directional width; acquiring a starting region based on the reference surface; acquiring a 3D point cloud map of the environment based on lidar and marking it as an environment point cloud map; and determining whether there are obstacles on the path based on the starting region, the reference surface, and the starting region. This addresses the problem that existing accompanying path planning technologies fail to establish suitable obstacle identification methods based on the accompanying target, resulting in large data processing volumes and complex data processing.

[0005] To achieve the above objectives, this application provides a robot accompanying path planning method that integrates sensors and large models, comprising the following steps:

[0006] A 3D point cloud map of the accompanying target is acquired using lidar and marked as the target point cloud map; the target space is then obtained based on the target point cloud map.

[0007] A reference square is obtained based on the target space and target square acquisition method;

[0008] Obtain the final square based on the reference square;

[0009] The directional width is obtained based on the final square.

[0010] The reference surface is obtained based on the width of the direction;

[0011] The starting region is obtained based on the reference surface;

[0012] The environment is acquired using a 3D point cloud map obtained by LiDAR and marked as an environmental point cloud map; the presence of obstacles on the path is determined based on the starting region, the reference surface, and the starting region.

[0013] Furthermore, obtaining the target space based on the target point cloud map includes the following sub-steps:

[0014] Make the plane formed by the X-axis and Y-axis coincide with the ground, and make the Z-axis vertically upward to establish a three-dimensional coordinate system, which is marked as the position reference coordinate system; place the target point cloud map in the position reference coordinate system;

[0015] Mark the plane formed by the X and Y axes in the coordinate system as the starting plane;

[0016] Obtain the plane that is parallel to the starting plane, is at a distance of the first distance from the starting plane, and intersects the positive Z-axis in the position comparison coordinate system; mark it as the final plane.

[0017] The space between the starting plane and the final plane is designated as the target space.

[0018] Furthermore, obtaining the reference square based on the target space and target square acquisition method includes the following sub-steps:

[0019] Obtain the point cloud coordinates of the target in the target space and mark them as the target object coordinates;

[0020] Obtain the X-axis and Y-axis data of the target object's coordinate points, and use them as the horizontal and vertical axis data of the two-dimensional coordinate points, respectively, and mark them as reference plane coordinate points;

[0021] Establish a Cartesian coordinate system and label it as the reference plane coordinate system;

[0022] Place all the coordinate points of the reference plane in the reference plane coordinate system to obtain a scatter plot, and mark it as the reference plane scatter plot;

[0023] The function is obtained by fitting the scatter plot of the reference plane and marked as the reference profile;

[0024] Using the reference contour as the target data, the target square is obtained using the target square acquisition method and marked as the reference square.

[0025] Furthermore, the method for obtaining the target square includes:

[0026] Obtain the minimum and maximum values ​​of the x-coordinate of the target data and mark them as the first x-coordinate and the second x-coordinate.

[0027] Obtain the minimum and maximum values ​​of the ordinate of the target data, and label them as the first ordinate and the second ordinate.

[0028] The first horizontal and vertical values ​​are used as the first coordinate point, and the coordinate point formed by the first horizontal and vertical values ​​is marked as the first coordinate point; the second horizontal and vertical values ​​are used as the first coordinate point, and the coordinate point formed by the second horizontal and vertical values ​​is marked as the second coordinate point; the second horizontal and vertical values ​​are used as the second coordinate point, and the coordinate point formed by the first horizontal and vertical values ​​is marked as the third coordinate point; the coordinate point formed by the first horizontal and vertical values ​​is used as the fourth coordinate point.

[0029] Connect the first coordinate point to the second coordinate point to obtain a line segment, which is marked as the first line segment; connect the second coordinate point to the third coordinate point to obtain a line segment, which is marked as the second line segment; connect the third coordinate point to the fourth coordinate point to obtain a line segment, which is marked as the third line segment; connect the fourth coordinate point to the first coordinate point to obtain a line segment, which is marked as the fourth line segment.

[0030] The rectangle formed by the first, second, third, and fourth line segments is marked as the target square.

[0031] Furthermore, obtaining the final square based on the reference square includes the following sub-steps:

[0032] Obtain the center point of the reference square and mark it as the starting midpoint;

[0033] Rotate the reference contour 360° with the starting midpoint as the rotation center; during the rotation, obtain a new reference square in real time and obtain the area of ​​the reference square in real time, and mark it as the reference area; obtain the reference square corresponding to the smallest reference area before the rotation is completed, and mark it as the final square.

[0034] Furthermore, obtaining the directional width based on the final square includes the following sub-steps:

[0035] Starting with one wide side of the final square, draw a second number of equally spaced parallel lines along the long side of the final square, and mark them as parallel lines.

[0036] For any two intersection points of a parallel line drawn with the final square, mark them as parallel intersection points; obtain the distance between the parallel intersection points and mark it as the reference interval distance.

[0037] Obtain the mean of the reference interval distances formed by all parallel drawn lines, and mark it as the mean of the reference distance;

[0038] Starting from the other wide side of the final square, obtain another reference distance mean; obtain the reference wide side corresponding to the smallest reference distance mean among the two reference distance means, and mark it as the direction wide side.

[0039] Furthermore, obtaining the reference plane based on the direction width edge includes the following sub-steps:

[0040] Using the relative position of the directional width edge and the reference profile as a reference, obtain the surface formed by the directional width edge with an arbitrary Z-axis value in the position comparison coordinate system, and mark it as the reference surface.

[0041] Furthermore, obtaining the starting region based on the reference plane includes the following sub-steps:

[0042] Project all target object coordinates onto the reference plane to obtain the projection reference points;

[0043] On the reference plane, using all the projected reference points as target data, the target square is obtained using the target square acquisition method and marked as the direction square;

[0044] Get the area enclosed by the directional square and mark it as the starting area.

[0045] Furthermore, determining whether there are obstacles on the path based on the starting region, the reference plane, and the starting region includes the following sub-steps:

[0046] Mark the point cloud coordinates in the environmental point cloud map as environmental coordinates;

[0047] Obtain the space on the side with the smallest number of coordinate points of the target object between the two sides of the reference plane, and mark it as the target space.

[0048] Furthermore, determining whether there are obstacles on the path based on the starting region, the reference plane, and the starting region also includes the following sub-steps:

[0049] Starting from the initial area, move a third distance towards the target space, and obtain in real time whether environmental coordinate points appear in the initial area. If they appear, it indicates that there is an obstacle in front of the target. Prompt the obstacle in front of the target and replan the route until no environmental coordinate points appear in the initial area.

[0050] The beneficial effects of this invention are as follows: This invention acquires a 3D point cloud map of an accompanying target based on lidar and marks it as a target point cloud map; acquires a target space based on the target point cloud map; acquires a reference square based on the target space and a target square acquisition method; acquires a final square based on the reference square; acquires the directional width based on the final square; acquires a reference surface based on the directional width; acquires a starting region based on the reference surface; acquires a 3D point cloud map of the environment based on lidar and marks it as an environment point cloud map; and determines whether there are obstacles on the path based on the starting region, the reference surface, and the starting region. The advantage lies in that a suitable obstacle recognition method is set based on the accompanying target, reducing and simplifying the data processing volume; and improving the obstacle recognition efficiency.

[0051] The present invention obtains a reference square based on the target space and target square acquisition method. Its advantage is that it can select a specific area accompanying the target and further transform it into a reference square that is easy to identify, which simplifies the identification of the movement direction of the accompanying target and improves the identification efficiency. Attached Figure Description

[0052] Figure 1 This is a flowchart of the steps of the method of the present invention;

[0053] Figure 2 This is a schematic diagram of the reference square of the present invention;

[0054] Figure 3 This is a schematic diagram of the final square shape of the present invention;

[0055] Figure 4 This is a schematic diagram of the reference interval distance of the present invention;

[0056] Figure 5 This is a schematic diagram of the directional square of the present invention. Detailed Implementation

[0057] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0058] Example 1, please refer to Figure 1 As shown, this application provides a robot accompanying path planning method that integrates sensors and a large model, including the following steps:

[0059] Step S1: Acquire a 3D point cloud map of the accompanying target based on the lidar and mark it as the target point cloud map; acquire the target space based on the target point cloud map; when acquiring the target point cloud map, due to environmental interference, it is difficult to directly acquire the target point cloud map, so the target point cloud map can be identified by combining large model training. Step S1 includes the following sub-steps:

[0060] Step S101: Align the plane formed by the X-axis and Y-axis with the ground, and make the Z-axis vertically upward to establish a three-dimensional coordinate system, which is marked as the position comparison coordinate system; place the target point cloud map in the position comparison coordinate system; establishing the position comparison coordinate system facilitates the analysis of the position of the target point cloud map; setting the position of the position comparison coordinate system facilitates subsequent processing of the target point cloud map;

[0061] Step S102: Mark the plane formed by the X-axis and Y-axis in the position reference coordinate system as the starting plane;

[0062] Step S103: Obtain a plane that is parallel to the starting plane, is at a distance of the first distance from the starting plane, and intersects the positive Z-axis in the position reference coordinate system, and mark it as the final plane; the first distance should not exceed the height of the accompanying target foot to facilitate obtaining the outer contour of the foot, for example, the first distance is 3cm;

[0063] Step S104: The space between the starting plane and the final plane is marked as the target space; here the target is to obtain the contour shape of the target foot, and the target space is the part of the foot position.

[0064] Step S2: Obtain a reference square based on the target space and the target square acquisition method; Step S2 includes the following sub-steps:

[0065] Step S201: Obtain the point cloud coordinates of the target point cloud map in the target space and mark them as the target object coordinates.

[0066] Step S202: Obtain the X-axis and Y-axis data of the target object coordinate points, and use them as the horizontal axis data and vertical axis data of the two-dimensional coordinate points, respectively, and mark them as reference plane coordinate points; here, the reference plane coordinate points are equivalent to the coordinate point positions of the target object coordinate points projected onto the starting plane in the target space, and the image is converted into a two-dimensional plane to facilitate image processing;

[0067] Step S203: Establish a Cartesian coordinate system and mark it as the reference plane coordinate system;

[0068] Step S204: Place all the reference plane coordinate points in the reference plane coordinate system to obtain a scatter plot, and mark it as the reference plane scatter plot;

[0069] Step S205: Fit the scatter plot of the reference plane to obtain a function, and mark it as the reference profile;

[0070] Step S206: Using the reference contour as the target data, obtain the target square using the target square acquisition method, and mark it as the reference square; Step S206 includes the following sub-steps:

[0071] Step S20601: Obtain the minimum and maximum values ​​of the x-coordinate of the target data and mark them as the first x-coordinate and the second x-coordinate.

[0072] Step S20602: Obtain the minimum and maximum values ​​of the ordinate of the target data, and mark them as the first ordinate value and the second ordinate value;

[0073] Step S20603: Mark the first coordinate point as the coordinate point formed by the first horizontal value and the first vertical value as the horizontal and vertical coordinates respectively; mark the second coordinate point as the coordinate point formed by the second horizontal value and the first vertical value as the horizontal and vertical coordinates respectively; mark the third coordinate point as the coordinate point formed by the second horizontal value and the second vertical value as the horizontal and vertical coordinates respectively; mark the fourth coordinate point as the coordinate point formed by the first horizontal value and the second vertical value as the horizontal and vertical coordinates respectively.

[0074] Step S20604: Connect the first coordinate point and the second coordinate point to obtain a line segment, which is marked as the first line segment; connect the second coordinate point and the third coordinate point to obtain a line segment, which is marked as the second line segment; connect the third coordinate point and the fourth coordinate point to obtain a line segment, which is marked as the third line segment; connect the fourth coordinate point and the first coordinate point to obtain a line segment, which is marked as the fourth line segment.

[0075] Step S20605: The rectangle formed by the first line segment, the second line segment, the third line segment, and the fourth line segment is marked as the target rectangle; here, the target rectangle is equivalent to a type of circumscribed rectangle.

[0076] The beneficial effects of the target square acquisition method: When the reference contour is an irregular object, it is difficult to distribute and judge it. In order to facilitate the analysis of the reference contour, a target square acquisition method is constructed to facilitate the subsequent judgment of the orientation of the accompanying target. At the same time, it is more complicated to obtain the midpoint of the reference contour. The target square is a regular object, and it is simpler to obtain the center. Therefore, the midpoint of the target square can be used as the center point of the reference contour, such as the starting midpoint obtained later. At the same time, the reference square can be quickly obtained using this method.

[0077] For practical applications, please refer to Figure 2 As shown, the diagram depicts the first coordinate point, the second coordinate point, the third coordinate point, the fourth coordinate point, the first line segment, the second line segment, the third line segment, the fourth line segment, and a reference square.

[0078] Step S3: Obtain the final square based on the reference square; Step S3 includes the following sub-steps:

[0079] Step S301: Obtain the center point of the reference square and mark it as the starting midpoint;

[0080] Step S302: Rotate the reference contour 360° with the starting midpoint as the rotation center; during the rotation, obtain a new reference square in real time and obtain the area of ​​the reference square in real time, and mark it as the reference area; obtain the reference square corresponding to the smallest reference area before the rotation is completed, and mark it as the final square; in order to determine the walking direction of the accompanying target, it can be based on the direction of the foot, which is roughly the walking direction of the accompanying target. In order to make the reference square closer to the shape of the foot, the smaller the area of ​​the reference square when it completely covers the reference contour, the closer the reference square is to the shape of the foot. Rotate the reference contour while keeping the reference plane coordinate system stationary, and obtain the closest reference contour to the final square.

[0081] For practical applications, please refer to Figure 3 As shown, this is a schematic diagram of the final square drawn.

[0082] Step S4: Obtain the directional width based on the final square; Step S4 includes the following sub-steps:

[0083] Step S401: Starting from one wide side of the final square, draw a second number of equally spaced parallel straight lines along the long side of the final square, and mark them as parallel drawing lines; Since the front and back parts of the accompanying target's shoes are different, and the front part of the shoes is more pointed than the back part, the orientation of the accompanying target is analyzed based on the shape of the reference contour, and this orientation can be identified as the walking direction of the accompanying target; Therefore, the walking direction of the accompanying target can be obtained through the shape of the reference contour; Since only the graphic data of the front and back parts of the reference contour need to be obtained, the second number is set to be small, for example, drawing two equally spaced parallel straight lines with a spacing of 2cm;

[0084] Step S402: Mark the two intersection points of any parallel line and the final square as parallel intersection points; obtain the distance between the parallel intersection points and mark it as the reference interval distance; the two wide sides of the final square can be initially identified as the two directions accompanying the front and rear of the target. Therefore, the specific directions accompanying the front and rear of the target can be determined by the shape of the close reference contour of the two wide sides of the final square.

[0085] Step S403: Obtain the mean value of the reference interval distance formed by all parallel drawn lines, and mark it as the mean reference distance;

[0086] Step S404: Starting from the other wide side of the final square, obtain another reference distance mean; obtain the reference wide side corresponding to the smallest reference distance mean between the two reference distance means, and mark it as the direction wide side; since the front part of the shoe is more pointed than the back part, the position with the smaller reference distance mean is the front end of the accompanying target, and the position with the larger reference distance mean is the rear end of the accompanying target; here, in order to obtain the movement direction of the accompanying target, which is the front end, the direction wide side is the front end of the accompanying target, which is the movement direction;

[0087] For practical applications, please refer to Figure 4 As shown, this is a schematic diagram of the reference interval distance.

[0088] Step S5: Obtain the reference plane based on the direction width edge; Step S5 includes the following sub-steps:

[0089] Step S501: Using the relative position of the direction width and the reference contour as a reference, obtain the surface formed by the direction width with an arbitrary Z-axis value in the position reference coordinate system, and mark it as the reference surface; here, in order to convert two-dimensional data into three-dimensional data, since the Z-axis can be an arbitrary value, the direction width can be drawn as a surface in the position reference coordinate system.

[0090] Step S6: Obtain the starting region based on the reference plane; Step S6 includes the following sub-steps:

[0091] Step S601: Project all target object coordinate points onto the reference surface to obtain projection reference points; projection reference points are more convenient for two-dimensional data analysis compared to three-dimensional data distribution.

[0092] Please see Figure 5 As shown, in step S602, on the reference plane, using all the projected reference points as target data, the target square is obtained using the target square acquisition method and marked as the direction square; the direction square is equivalent to the outline of the area that the target can pass through.

[0093] Step S603: Obtain the area enclosed by the directional square and mark it as the starting area; the starting area is the area that the accompanying target can pass through.

[0094] Step S7: Acquire a 3D point cloud map of the environment based on the LiDAR and mark it as the environment point cloud map; determine whether there are obstacles on the path based on the starting region, the reference surface, and the starting region; Step S7 includes the following sub-steps:

[0095] Step S701: Mark the point cloud coordinates in the environmental point cloud map as environmental coordinates;

[0096] Step S702: Obtain the space on the side with the smallest number of target object coordinate points on both sides of the reference plane, and mark it as the target space; since the reference plane is the plane in front of the target, the target space is the space in front of the target.

[0097] Step S703: Starting from the starting area, move a third distance towards the target space. In real time, obtain whether environmental coordinate points appear on the starting area. If they appear, it means there is an obstacle in front of the target. Prompt the obstacle in front of the target and replan the route until no environmental coordinate points appear on the starting area. The third distance is set as a safe distance, that is, a distance that can safely avoid obstacles, such as 1.5m. If environmental coordinate points appear, it means there is an obstacle in the walking range in front of the target. Therefore, it is necessary to prompt the obstacle in front of the target and replan the route to avoid the obstacle.

[0098] Example 2: This application also provides an electronic device, which may include: a processor, a communication interface, a memory, and a communication bus. The processor, communication interface, and memory communicate with each other via the communication bus. The memory stores computer-readable instructions, and the processor can call the instructions in the memory. When the computer-readable instructions are executed by the processor, steps such as those in a robot accompanying path planning method that integrates sensors and a large model are performed to achieve the following functions: acquiring a 3D point cloud map of the accompanying target based on a LiDAR, and marking it as a target point cloud map; acquiring a target space based on the target point cloud map; acquiring a reference square based on the target space and a target square acquisition method; acquiring a final square based on the reference square; acquiring a directional width based on the final square; acquiring a reference surface based on the directional width; acquiring a starting region based on the reference surface; acquiring a 3D point cloud map of the environment based on a LiDAR, and marking it as an environment point cloud map; and determining whether there are obstacles on the path based on the starting region, the reference surface, and the starting region.

[0099] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0100] Example 3: This application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute a robot accompanying path planning method that integrates sensors and a large model, as provided by the above methods. The method includes: acquiring a three-dimensional point cloud map of the accompanying target based on a lidar, and marking it as a target point cloud map; acquiring a target space based on the target point cloud map; acquiring a reference square based on the target space and a target square acquisition method; acquiring a final square based on the reference square; acquiring a directional width based on the final square; acquiring a reference surface based on the directional width; acquiring a starting region based on the reference surface; acquiring a three-dimensional point cloud map of the environment based on a lidar, and marking it as an environment point cloud map; and determining whether there are obstacles on the path based on the starting region, the reference surface, and the starting region.

[0101] Example 4: This application also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program performs the steps of the above-described robot accompanying path planning method integrating sensors and a large model to achieve the following functions: acquiring a 3D point cloud map of the accompanying target based on a LiDAR and marking it as a target point cloud map; acquiring a target space based on the target point cloud map; acquiring a reference square based on the target space and the target square acquisition method; acquiring a final square based on the reference square; acquiring the directional width based on the final square; acquiring a reference surface based on the directional width; acquiring a starting region based on the reference surface; acquiring a 3D point cloud map of the environment based on a LiDAR and marking it as an environment point cloud map; and determining whether there are obstacles on the path based on the starting region, the reference surface, and the starting region.

[0102] Based on the above description of the embodiments, the embodiments of the present invention can be provided as methods, systems, or computer program products. Based on this understanding, the above technical solutions, in essence or in terms of their contribution to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or certain parts of the embodiments.

[0103] In the embodiments provided in this application, it should be understood that the disclosed system or method can be implemented in other ways. The embodiments described above are merely illustrative. For example, the division of modules or units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces. The indirect coupling or communication connection between systems, modules, and units may be electrical, mechanical, or other forms.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A robot companion path planning method fusing a sensor and a large model, characterized by, It comprises the following steps: Based on the laser radar, a three-dimensional point cloud map of the target is obtained, which is marked as a target point cloud map; a target space is obtained based on the target point cloud map; wherein the target point cloud map is obtained by combining a large model training recognition; Based on the target space and the target square acquisition method, a reference square is obtained; Based on the reference square, a final square is obtained; Based on the final square, a direction wide side is obtained; the direction wide side is the front end of the target, that is, the moving direction; Based on the direction wide side, a reference surface is obtained; Based on the reference surface, a starting area is obtained; the starting area is an area where the target can pass through; Based on the laser radar, a three-dimensional point cloud map of the environment is obtained, which is marked as an environment point cloud map; based on the starting area, the reference surface and the environment point cloud map, it is judged whether there is an obstacle on the path; if there is an obstacle, the route is re-planned until no environmental coordinate point appears on the starting area; Based on the target point cloud map, a target space is obtained, comprising the following sub-steps: Make the plane formed by the X-axis and the Y-axis coincide with the ground, and make the direction of the Z-axis vertical upward to establish a three-dimensional coordinate system, which is marked as a position comparison coordinate system; make the target point cloud map in the position comparison coordinate system; The plane formed by the X-axis and the Y-axis in the position comparison coordinate system is marked as a starting plane; A plane parallel to the starting plane and having a distance of a first distance from the starting plane and having an intersection with the positive axis of the Z-axis in the position comparison coordinate system is obtained, which is marked as a final plane; The space between the starting plane and the final plane is marked as a target space; Based on the target space and the target square acquisition method, a reference square is obtained, comprising the following sub-steps: Obtain the point cloud coordinate points of the target point cloud map in the target space, which are marked as target object coordinate points; Obtain the data of the X-axis and the Y-axis in the target object coordinate points, respectively as the horizontal axis data and the vertical axis data of the two-dimensional coordinate points, which are marked as the reference plane coordinate points; A plane rectangular coordinate system is established, which is marked as a reference plane coordinate system; Place all the reference plane coordinate points in the reference plane coordinate system to obtain a scatter plot, which is marked as a reference plane scatter plot; The reference plane scatter plot is fitted to obtain a function, which is marked as a reference contour; The reference contour is taken as the target data to obtain the target square by the target square acquisition method, which is marked as the reference square; The target square acquisition method comprises: Obtain the minimum and maximum values of the horizontal coordinates of the target data, which are marked as the first horizontal value and the second horizontal value; Obtain the minimum and maximum values of the vertical coordinates of the target data, which are marked as the first vertical value and the second vertical value; The first horizontal value and the first vertical value are respectively taken as the horizontal coordinate and the vertical coordinate to form a coordinate point, which is marked as a first coordinate point; the second horizontal value and the first vertical value are respectively taken as the horizontal coordinate and the vertical coordinate to form a coordinate point, which is marked as a second coordinate point; the second horizontal value and the second vertical value are respectively taken as the horizontal coordinate and the vertical coordinate to form a coordinate point, which is marked as a third coordinate point; the first horizontal value and the second vertical value are respectively taken as the horizontal coordinate and the vertical coordinate to form a coordinate point, which is marked as a fourth coordinate point; The first coordinate point and the second coordinate point are connected to obtain a line segment, which is marked as a first line segment; the second coordinate point and the third coordinate point are connected to obtain a line segment, which is marked as a second line segment; the third coordinate point and the fourth coordinate point are connected to obtain a line segment, which is marked as a third line segment; and the fourth coordinate point and the first coordinate point are connected to obtain a line segment, which is marked as a fourth line segment; The first line segment, the second line segment, the third line segment and the fourth line segment form a rectangle, which is marked as a target square; The final square is obtained based on the reference square, including the following sub-steps: A center point of the reference square is obtained, which is marked as a starting midpoint; The reference contour is rotated 360° with the starting midpoint as the rotation center; a new reference square is obtained in real time during the rotation, and an area of the reference square is obtained in real time, which is marked as a reference area; a reference square corresponding to the minimum reference area before the rotation is completed is obtained, which is marked as a final square.

2. The method of claim 1, wherein, The direction wide side is obtained based on the final square, including the following sub-steps: A wide side of the final square is taken as a starting point, and a second number of parallel straight lines with equal intervals are drawn along the long side of the final square, which are marked as parallel drawing lines; Any two intersection points of a parallel drawing line and the final square are marked as parallel intersection points; a distance between the parallel intersection points is obtained, which is marked as a reference interval distance; A mean value of the reference interval distances formed by all the parallel drawing lines is obtained, which is marked as a reference distance mean value; Another reference distance mean value is obtained with another wide side of the final square as a starting point; a reference wide side corresponding to the minimum reference distance mean value of the two reference distance mean values is obtained, which is marked as a direction wide side.

3. The method of claim 2, wherein, The reference surface is obtained based on the direction wide side, including the following sub-steps: The reference surface is obtained based on the direction wide side, including the following sub-steps:

4. The method of claim 3, wherein, The reference surface is obtained based on the direction wide side, including the following sub-steps: All coordinate points of the target object are projected onto the reference surface to obtain projected reference points; The target square is obtained based on the target data and the target square obtaining method on the reference surface, which is marked as a direction square; A region surrounded by the direction square is obtained, which is marked as a starting region.

5. The method of claim 4, wherein, It is judged whether there is an obstacle on the path based on the starting region, the reference surface and the environment point cloud map, including the following sub-steps: The point cloud coordinate points in the environment point cloud map are marked as environment coordinate points; The space on the side with the minimum number of target object coordinate points on the two sides of the reference surface is obtained, which is marked as a front space of the target object.

6. The method of claim 5, wherein, It is judged whether there is an obstacle on the path based on the starting region, the reference surface and the environment point cloud map, including the following sub-steps: The starting region is moved to the front space of the target object by a third distance, and it is judged whether there is an environment coordinate point on the starting region in real time; if there is, it indicates that there is an obstacle in front of the target object, the obstacle in front of the target object is prompted, and the route is re-planned until there is no environment coordinate point on the starting region.

Images