Robot anti-falling method, device and equipment and storage medium

Abstract

The invention discloses a robot anti-falling method, device and equipment and a storage medium. The method comprises the steps that original point cloud is collected; segmenting a ground point cloud from the original point cloud; the ground point clouds at the same emission pitch angle are sorted according to the horizontal rotation angle when the laser radar is emitted, and the included angle characteristic value of each point of the ground point clouds is calculated through the following formula; based on the included angle characteristic value of each point of the ground point cloud, screening boundary points of the falling area from the ground point cloud; fitting boundary points of the falling area to form a boundary line; and based on the boundary line, carrying out anti-falling path planning. The laser radar is used for detecting the boundary line of the falling area, the robot is prevented from falling, and the problem of misjudgment of a traditional ramp path is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of robots, and particularly relates to a robot anti-fall method, device, equipment and storage medium. Background Art

[0002] With the progress of technology, robot products (such as floor-sweeping robots, delivery robots, sanitation robots, etc.) have gradually entered the public life. Robots usually need to have the ability to perceive the surrounding environment to ensure their safe and reliable operation. When a robot is moving, it may encounter road conditions such as steps, platforms, curbs, etc., and there is a risk of falling. Therefore, it is necessary to detect such road conditions to avoid falling. The core of realizing anti-fall kinetic energy is the detection of the fall scenario. The existing technology for detecting the fall scenario lies in judging the height difference value of the ground. For example, there are CN117031488A and CN116330291A.

[0003] CN117031488A judges that there is a height difference greater than a preset threshold between the robot and the horizontal ground in front by receiving the ranging signal sent by the laser ranging device in real time, and then triggers braking.

[0004] CN116330291A collects environmental point clouds, and determines the point clouds with a height less than or equal to the preset height in the robot coordinate system as the fall point clouds. When the number of fall point clouds reaches a certain threshold, the area corresponding to the fall point clouds is determined as the fall area, and the robot performs path planning according to the fall area.

[0005] The above detection of the fall scenario by the ground height difference value is not applicable to the downhill road condition. The robot can reach the ground below the slope through the ramp, thus avoiding falling. When the height difference between the ground on the slope and the ground below the slope reaches the threshold, it will trigger a misjudgment, thereby preventing the robot from going downhill and restricting the passable range of the robot. Summary of the Invention

[0006] In order to solve the above technical problems, the present invention proposes a robot anti-fall method, device, equipment and storage medium.

[0007] To achieve the above object, the technical solution of the present invention is as follows:

[0008] In the first aspect, the present invention discloses a robot anti-fall method, including:

[0009] Step S1: Collect original point clouds;

[0010] Step S2: Segment the ground point clouds from the original point clouds;

[0011] Step S3: Sort the ground point clouds with the same emission pitch angle according to the horizontal rotation angle when the lidar is emitted, and calculate each point P of the ground point clouds through the following formulai The included angle eigenvalue F i ;

[0012]

[0013] Where: θ i,j is the included angle between the vector formed by point P i and the j-th adjacent point P before and after; i+j and P i-j ;

[0014] K is the number of included angles;

[0015] Step S4: Based on the included angle eigenvalue of each point in the ground point cloud, screen the boundary points of the fall area from the ground point cloud;

[0016] Step S5: Fit the boundary points of the fall area to form a boundary line;

[0017] Step S6: Based on the boundary line, perform anti-fall path planning.

[0018] On the basis of the above technical solution, the following improvements can also be made:

[0019] As a preferred solution,

[0020] Step S4 is specifically: From all points of the ground point cloud, screen out the points with the included angle eigenvalue less than the threshold as the boundary points of the fall area.

[0021] As a preferred solution, Step S5 includes:

[0022] Step S5.1: Use a clustering algorithm to cluster the boundary points of the fall area, and divide the boundary points belonging to the same boundary into one category;

[0023] Step S5.2: Fit the boundary points of the same boundary to form several boundary lines.

[0024] As a preferred solution, use the quadratic curve fitting method or the B-spline curve fitting method to fit the boundary points of the fall area to form a boundary line.

[0025] In the second aspect, the present invention also discloses a robot anti-fall device, including:

[0026] An original point cloud acquisition module for acquiring an original point cloud;

[0027] A ground point cloud segmentation module for segmenting the ground point cloud from the original point cloud;

[0028] An included angle eigenvalue calculation module for sorting the ground point cloud with the same emission pitch angle according to the horizontal rotation angle during lidar emission, and calculating each point P of the ground point cloud through the following formula iThe included angle eigenvalue F i ;

[0029]

[0030] Where: θ i,j is the included angle between the vector formed by point P i and the j-th point P adjacent before and after; i+j and P i-j ;

[0031] K is the number of included angles;

[0032] A boundary point screening module, configured to screen boundary points of the fall area from the ground point cloud based on the included angle eigenvalue of each point of the ground point cloud;

[0033] A boundary line forming module, configured to fit the boundary points of the fall area to form a boundary line;

[0034] A fall prevention planning module, configured to perform fall prevention path planning based on the boundary line.

[0035] As a preferred solution, the boundary point screening module is configured to screen out points with an included angle eigenvalue less than a threshold from all points of the ground point cloud as boundary points of the fall area.

[0036] As a preferred solution, the boundary line forming module includes:

[0037] A clustering unit, configured to cluster the boundary points of the fall area by using a clustering algorithm and divide the boundary points belonging to the same boundary into one category;

[0038] A fitting unit, configured to fit the boundary points of the same boundary to form several boundary lines.

[0039] As a preferred solution, the boundary line forming module uses a quadratic curve fitting method or a B-spline curve fitting method to fit the boundary points of the fall area to form a boundary line.

[0040] In a third aspect, the present invention also discloses a computing device, including:

[0041] One or more processors;

[0042] A memory;

[0043] And one or more programs, where one or more programs are stored in the memory and configured to be executed by one or more processors, and one or more programs include instructions of any of the above robot fall prevention methods.

[0044] Fourthly, the present invention also discloses a storage medium, characterized in that the storage medium stores one or more computer-readable programs, and the one or more programs include instructions adapted to be loaded and executed by the memory to perform any of the above robot anti-falling methods.

[0045] The present invention discloses a robot anti-falling method, device, equipment and storage medium, which uses a lidar to detect the boundary line of the falling area to prevent the robot from falling and solves the problem of misjudgment of the traditional slope path. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 It is a flowchart of the robot anti-falling method provided by the embodiment of the present invention.

[0048] Figure 2 It is the semi-circular point cloud in front of the robot corresponding to the launch pitch angle α under different road conditions provided by the embodiment of the present invention;

[0049] (a) is a flat road condition;

[0050] (b) is a stepped road condition;

[0051] (c) is a downhill road condition.

[0052] Figure 3 It is a block diagram of the robot anti-falling device provided by the embodiment of the present invention.

[0053] Figure 4 It is a block diagram of the computing device provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] The following will describe in detail the preferred embodiments of the present invention with reference to the accompanying drawings.

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0056] The expression "including" an element is an "open-ended" expression, which merely means the existence of corresponding components or steps and should not be construed as excluding additional components or steps.

[0057] To achieve the object of the present invention, in some embodiments of the robot anti-fall method, such as Figure 1 shown, the robot anti-fall method is based on a lidar and includes:

[0058] Step S101: Collect the original point cloud;

[0059] Step S102: Segment the ground point cloud from the original point cloud;

[0060] Step S103: Sort the ground point clouds with the same emission pitch angle according to the horizontal rotation angle at the time of lidar emission, and calculate the included angle eigenvalue F of each point P of the ground point cloud through the following formula i of; i ;

[0061]

[0062] where: θ i,j is the included angle between the vector formed by point P i and the j-th adjacent front and rear points P i+j and P i-j ;

[0063] K is the number of included angles;

[0064] Step S104: Based on the included angle eigenvalue of each point of the ground point cloud, screen the boundary points of the fall area from the ground point cloud;

[0065] Step S105: Fit the boundary points of the fall area to form a boundary line;

[0066] Step S106: Based on the boundary line, perform anti-fall path planning.

[0067] The above steps will be elaborated in detail below.

[0068] The lidar carried by the robot adjusts the laser emission angle in real time, emits laser signals according to a period, and collects the original point cloud.

[0069] When the lidar emits laser signals, it continuously adjusts the laser emission angle to be able to capture the point clouds of all environments. Specifically, the horizontal rotation angle is adjusted to capture the point clouds at different angles around, and the pitch angle is adjusted to adjust the distance and height range that the point cloud can capture.

[0070] Step S102 segments the ground point cloud from the original point cloud to reduce the computational amount of subsequent steps and can exclude the interference of non-ground point clouds.

[0071] There are various methods for segmenting ground point clouds. For example, points with height values less than a certain threshold are classified as ground points, or a ground segmentation algorithm based on RANSAC is used.

[0072] The ground points in the current area and the ground points in the drop area both belong to the ground as described here. The ground points in these areas will be segmented together for subsequent feature calculation.

[0073] Step S103 is the key step of the present invention.

[0074] Since the lidar continuously adjusts the angle of laser emission, the inventor found the following distribution rules of the point cloud in the following ground conditions. For example Figure 2 as shown.

[0075] Figure 2 In (a), (b), and (c) respectively show the flat road condition, the step road condition, and the downhill road condition. The lower half of the subgraph is the point cloud schematic view when looking down. For the convenience of display, only the half-circle point cloud in front of the robot corresponding to the laser emission pitch angle of α is drawn.

[0076] According to the laser emission principle of the lidar, in the flat road condition, the point cloud formed by the laser emitted at a specific pitch angle is almost circular. As Figure 2 shown in (a), at this time, the angle formed by any three adjacent points is relatively large. In the step road condition, the circular point cloud is truncated in the middle. As Figure 2 shown in (b), at the truncation point, the angle between three adjacent points is relatively small. In the downhill road condition, the point cloud is non-standard circular on the slope but still smooth, without showing the truncation phenomenon as in the step road condition. As Figure 2 shown in (c), at this time, the angle formed by any three adjacent points is still relatively large.

[0077] For example: Sort the ground point cloud with the laser emission pitch angle of α according to the horizontal rotation angle when the lidar emits, and calculate the angle feature value F of each point P of the ground point cloud through the following formula i of i ;

[0078]

[0079] where: P i is the i-th point with the laser emission pitch angle of α;

[0080] θ i,j is the angle between the vector formed by point P i and the j-th adjacent front and rear points P i+j and P i-j ;

[0081] K is the number of angles.

[0082] Considering the stability of the entire system, the present invention uses the mean of multiple included angles as the included angle eigenvalue at point P i

[0083] Step S104 is specifically as follows: From all the points of the ground point cloud, select the points with included angle eigenvalues less than the threshold as the boundary points of the drop area.

[0084] Furthermore, step S105 includes:

[0085] Step S105.1: Use a clustering algorithm to cluster the boundary points of the drop area, and divide the boundary points belonging to the same boundary into one category;

[0086] Step S105.2: Fit the boundary points of the same boundary to form several boundary lines.

[0087] In the actual scenario, there may be multiple boundaries of the drop area at the same time. Therefore, before fitting, the boundary points are also clustered, and the boundary points of the same boundary are clustered into one category. The clustering algorithm can be but is not limited to using DBSCAN.

[0088] Step S106 takes that the robot does not touch or cross the boundary line as one of the constraints for path planning to avoid driving in the drop area.

[0089] The present invention can use but is not limited to using the quadratic curve fitting method or the B-spline curve fitting method to fit the boundary points of the drop area to form a boundary line.

[0090] It should be noted that in some specific usage scenarios, the lidar almost only emits laser light to the ground, and the obtained laser points are the ground points. At this time, the process of ground point segmentation, that is, step S102, can be cancelled.

[0091] In some other embodiments, the present invention also discloses a robot anti-drop device, as Figure 3 shown, including:

[0092] An original point cloud acquisition module 201 for acquiring an original point cloud;

[0093] A ground point cloud segmentation module 202 for segmenting the ground point cloud from the original point cloud;

[0094] An included angle eigenvalue calculation module 203 for sorting the ground point cloud with the same emission pitch angle according to the horizontal rotation angle at the time of lidar emission, and calculating the included angle eigenvalue F of each point P of the ground point cloud through the following formula i of i ;

[0095]

[0096] Where: θ i,j is the point P​i and the included angle of the vector formed by P i+j and P i-j and the j-th point adjacent before and after;

[0097] K is the number of included angles;

[0098] The boundary point screening module 204 is configured to screen boundary points of the falling area from the ground point cloud based on the included angle eigenvalue of each point of the ground point cloud;

[0099] The boundary line forming module 205 is configured to fit the boundary points of the falling area to form a boundary line;

[0100] The anti-fall planning module 206 is configured to perform anti-fall path planning based on the boundary line.

[0101] Further, the boundary point screening module is configured to screen out points with an included angle eigenvalue less than the threshold from all points of the ground point cloud as the boundary points of the falling area.

[0102] Further, the boundary line forming module includes:

[0103] The clustering unit is configured to cluster the boundary points of the falling area by using a clustering algorithm and divide the boundary points belonging to the same boundary into one category;

[0104] The fitting unit is configured to fit the boundary points of the same boundary to form several boundary lines.

[0105] Further, the boundary line forming module fits the boundary points of the falling area by using a quadratic curve fitting method or a B-spline curve fitting method to form a boundary line.

[0106] Further, it should be noted that: when determining the anti-fall boundary of the robot anti-fall device provided in the above embodiment, only the division of the above functional modules is used for illustration. In actual application, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the robot anti-fall device is divided into different functional modules to complete all or part of the functions described above.

[0107] In addition, the robot anti-fall device provided in the above embodiment and the embodiment of the robot anti-fall method belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be elaborated here.

[0108] In addition, in some other embodiments, as Figure 4 shown, the present invention also discloses a computing device, including:

[0109] One or more processors 301;

[0110] A memory 302;

[0111] and one or more programs, where the one or more programs are stored in the memory 302 and configured to be executed by the one or more processors 301, and the one or more programs include instructions for the robot anti-fall method disclosed in the above embodiments.

[0112] The processor 301 may include one or more processing cores, such as: a 4-core processor, an 8-core processor, etc. The processor 301 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 301 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the wake state, also known as the CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 301 may be integrated with a GPU (Graphics Processing Unit), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 301 may further include an AI (Artificial Intelligence) processor, and the AI processor is used to process computational operations related to machine learning.

[0113] The memory 302 may include one or more computer-readable storage media, and the computer-readable storage media may be non-transitory. The memory 302 may further include high-speed random access memory and non-volatile memory, such as one or more disk storage devices and flash storage devices. In some embodiments, the non-transitory computer-readable storage media in the memory 302 is used to store at least one instruction, and the at least one instruction is used to be executed by the processor 301 to implement the robot anti-fall method provided in the method embodiments of the present invention.

[0114] In addition, the computing device may optionally further include: a peripheral device interface and at least one peripheral device. The processor 301, the memory 302, and the peripheral device interface may be connected by a bus or signal lines. Each peripheral device may be connected to the peripheral device interface through a bus, signal lines, or a circuit board. Schematically, the peripheral devices include but are not limited to: a radio frequency circuit, a touch display screen, an audio circuit, and a power supply, etc.

[0115] Of course, the computing device may also include fewer or more components, and this embodiment does not limit this.

[0116] In addition, in some other embodiments, the present invention also discloses a storage medium storing one or more computer-readable programs, and the one or more programs include instructions adapted to be loaded and executed by the memory to perform the robot anti-falling method disclosed in the above embodiments.

[0117] The present invention discloses a robot anti-falling method, device, equipment and storage medium, which uses a lidar to detect the boundary line of the falling area to prevent the robot from falling and solves the problem of misjudgment of the traditional slope path.

[0118] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of the present invention claimed is defined by the appended claims and their equivalents.

Claims

1. A robot anti-fall method, characterized in that: include: Step S1: Collecting original point cloud; Step S2: Segmenting the ground point cloud from the original point cloud; Step S3: Sort the ground point clouds with the same transmission pitch angle according to the horizontal rotation angle when the laser radar is transmitted, and calculate each point P of the ground point cloud by the following formula i The angle eigenvalue F i ; Where: θ i,j Point P i P of the jth point adjacent to the front and back i+j and P i-j The angle between the vectors formed; K is the number of angles; Step S4: based on the angle feature value of each point of the ground point cloud, the boundary points of the falling area are selected from the ground point cloud; Step S5: fitting the boundary points of the falling area to form a boundary line; Step S6: Based on the boundary line, anti-fall path planning is performed.

2. The robot anti-falling method according to claim 1, characterized in that: The step S4 is specifically as follows: from all points of the ground point cloud, points whose angle characteristic values ​​are less than a threshold are selected as boundary points of the falling area.

3. The robot anti-falling method according to claim 1, characterized in that: The step S5 comprises: Step S5.1: clustering the boundary points of the falling area using a clustering algorithm, and dividing the boundary points belonging to the same boundary into one category; Step S5.2: Fit boundary points of the same boundary to form several boundary lines.

4. The robot anti-falling method according to claim 1, characterized in that: The quadratic curve fitting method or the B-spline curve fitting method is used to fit the boundary points of the falling area to form a boundary line.

5. The robot anti-fall device is characterized by: include: Original point cloud acquisition module, used to acquire original point cloud; A ground point cloud segmentation module is used to segment the ground point cloud from the original point cloud; The angle eigenvalue calculation module is used to sort the ground point clouds with the same emission pitch angle according to the horizontal rotation angle when the laser radar is emitted, and calculate each point P of the ground point cloud by the following formula i The angle eigenvalue F i ; Where: θ i,j Point P i P of the jth point adjacent to the front and back i+j and P i-j The angle between the vectors formed; K is the number of angles; The boundary point screening module is used to screen the boundary points of the falling area from the ground point cloud based on the angle feature value of each point in the ground point cloud; A boundary line forming module is used to fit the boundary points of the falling area to form a boundary line; The anti-fall planning module is used to plan the anti-fall path based on the boundary line.

6. The robot anti-fall device according to claim 5, characterized in that: The boundary point screening module is used to screen out points whose angle characteristic values ​​are less than a threshold from all points in the ground point cloud as boundary points of the falling area.

7. The robot anti-fall device according to claim 5, characterized in that: The boundary line forming module comprises: A clustering unit, used to cluster the boundary points of the falling area using a clustering algorithm, and classify the boundary points belonging to the same boundary into one category; The fitting unit is used to fit boundary points of the same boundary to form several boundary lines.

8. The robot anti-fall device according to claim 5, characterized in that: The boundary line forming module adopts a quadratic curve fitting method or a B-spline curve fitting method to fit the boundary points of the falling area to form a boundary line.

9. A computing device, characterized in that include: one or more processors; Memory; And one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, and one or more of the programs include instructions for the robot anti-fall method described in any one of claims 1-4 above.

10. A storage medium, characterized in that The storage medium stores one or more computer-readable programs, and the one or more programs include instructions, and the instructions are suitable for being loaded by the memory and executing the robot anti-fall method described in any one of claims 1-4.

Images