Obstacle detection method, obstacle avoidance method and controller for AGV (Automatic Guided Vehicle) loaded with goods shelf

By analyzing the historical scan sample set and real-time data of AGV, the shelf edge error area is divided, and the shelf leg interference points are eliminated, the problem of misjudgment of obstacles in AGV lidar scanning is solved, and the detection accuracy and robustness are improved.

CN118915757APending Publication Date: 2024-11-08HEFEI GEN SONG AUTOMATION TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411018472.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The prior art is difficult to effectively deal with the misjudgment of obstacles caused by shelf legs to AGV lidar scanning, especially when the shelf legs are suspended or the universal wheels rotate, resulting in inaccurate lidar scanning.

Method used

By obtaining the historical scan sample set of AGV, dividing the shelf edge error area, analyzing the interfering pixel points, and removing the interfering points in real-time detection, providing accurate obstacle judgment, and using an adaptive update mechanism to improve robustness.

Benefits of technology

It realizes effective handling of shelf leg interference, improves the accuracy and robustness of AGV obstacle detection, and reduces obstacle misjudgment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118915757A_ABST
    Figure CN118915757A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of AGV control, in particular to an obstacle detection method, an obstacle avoidance method and a controller for an AGV carrying a goods shelf. The invention discloses an obstacle detection method for an AGV carrying a goods shelf, and the method comprises the steps: obtaining historical laser data based on the scanning of a laser radar on the AGV, carrying out the processing of the historical laser data, obtaining a historical scanning sample set, and obtaining interference pixel points related to goods shelf legs in a divided goods shelf edge error region through pixel comparison; therefore, the obstacles can be removed in subsequent real-time detection and judgment, an accurate obstacle judgment result is obtained, and related interference of the goods shelf legs is avoided. The method solves the problem that the prior art lacks a method for effectively processing related interference of shelf legs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of AGV control technology, and more specifically to: 1. an obstacle detection method for an AGV carrying shelves; 2. an obstacle avoidance method for an AGV carrying shelves; 3. an AGV controller. Background Art

[0002] AGV (Automated Guided Vehicle) is a commonly used cargo loading tool with a rotatable loading plate on top for loading shelves and other items. AGV is also equipped with a laser radar to scan the surrounding area to determine whether there are obstacles and then avoid them.

[0003] As AGV application scenarios become more and more complex, the shelves they load also have many specifications. Some larger shelves will exceed the loading tray, causing the shelf legs to be suspended in the air. The laser radar will scan them as laser dots, which can easily cause misjudgment of obstacles. In addition, some shelf legs are equipped with universal wheels at the bottom, which further affects the scanning of the laser radar. The rotation direction of the universal wheel is uncertain and can easily change under the action of external forces, so that the active area of ​​the universal wheel will exceed the bottom of the shelf leg from a bird's-eye view, causing the laser radar to scan out laser dots outside the outer contour of the shelf, which can also easily cause misjudgment of obstacles. In addition, the above situation will be further complicated due to loading errors of the shelf or changes in the direction of the shelf.

[0004] However, there is little research on the above situation in the prior art, and there is a lack of effective methods for dealing with the interference related to shelf legs. Summary of the invention

[0005] Based on this, it is necessary to provide an obstacle detection method, obstacle avoidance method and controller for an AGV carrying shelves in order to address the problem that the prior art lacks an effective method for dealing with shelf leg-related interference.

[0006] The present invention is implemented by the following technical solutions:

[0007] In a first aspect, the present invention discloses an obstacle detection method for an AGV carrying a shelf, comprising the following steps:

[0008] Step 1: Obtain a historical scanning sample set of the AGV carrying the shelf, and determine the shelf edge error area and interference pixel points;

[0009] Among them, step one includes:

[0010] S101, controlling the laser radar on the AGV to scan its surrounding area N times to obtain N historical laser images;

[0011] S102, traversing N historical laser images for processing, obtaining N historical reduction images, and forming a historical scanning sample set;

[0012] Among them, for the nth historical laser image: first delineate the outer contour of the AGV and the outer contour of the shelf, then divide the shelf edge error area according to the center point error area of ​​the shelf, and use the divided shelf edge error area as the nth historical reduction limit image;

[0013] S103, performing pixel analysis on the pixels of the historical scanning sample set to identify interfering pixels.

[0014] Step 2: Obtain a real-time reduced image and analyze the pixels of the real-time reduced image;

[0015] If there are laser pixels other than interference pixels in the real-time reduction image, they are regarded as new obstacle pixels, indicating that a new obstacle has been detected.

[0016] The method for obtaining the real-time shrinkage image includes:

[0017] Control the laser radar on the AGV to scan the surrounding area in real time to obtain a real-time laser map;

[0018] For real-time laser images: first delineate the outer contours of the AGV and the shelf, then divide the shelf edge error zone based on the center point error zone of the shelf, and use the divided shelf edge error zone as the real-time reduction limit map.

[0019] This method for detecting obstacles in an AGV carrying shelves implements a method or process according to an embodiment of the present disclosure.

[0020] In a second aspect, the present invention discloses an obstacle avoidance method for an AGV carrying a shelf, comprising:

[0021] Perform obstacle detection according to the obstacle detection method for AGVs carrying shelves disclosed in the first aspect;

[0022] If a new obstacle is detected, obtain the position coordinates of the new obstacle pixel point;

[0023] Calculate the real-time distance to the AGV based on the position coordinates of the new obstacle pixel points;

[0024] If the real-time distance between the new obstacle pixel and the AGV is greater than the preset distance threshold ΔL, there is no need to avoid the obstacle; otherwise, obstacle avoidance is performed.

[0025] This obstacle avoidance method for an AGV carrying shelves implements the method or process according to an embodiment of the present disclosure.

[0026] In a third aspect, the present invention discloses an AGV controller, which is designed using the obstacle detection method for an AGV with shelves disclosed in the first aspect, or / and the obstacle avoidance method for an AGV with shelves disclosed in the second aspect.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. The present invention obtains a historical scanning sample set based on the laser radar scanning on the AGV, obtains the historical laser image processing, and obtains the interference pixel points related to the shelf legs through pixel analysis in the divided shelf edge error area, so as to eliminate them in the subsequent real-time detection and judgment, thereby obtaining an accurate obstacle judgment result and avoiding the interference related to the shelf legs.

[0029] 2. The present invention also adaptively updates the historical scanning sample set according to the real-time recognition situation, which can provide robustness of obstacle detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0031] Figure 1 This is a flow chart of the obstacle detection method for an AGV carrying a shelf proposed in Example 1 of the present invention;

[0032] Figure 2 This is an example diagram of an AGV loading a shelf in Embodiment 1 of the present invention;

[0033] Figure 3 for Figure 2 Schematic diagram of the corresponding area division;

[0034] Figure 4 For the general Figure 2 Example diagram of the shelf loading direction after adjusting 90°;

[0035] Figure 5 for Figure 4 Schematic diagram of the corresponding area division;

[0036] Figure 6 For use Figure 4 Schematic diagram of real-time obstacle detection based on the corresponding historical scan sample set;

[0037] Figure 7 This is a flow chart of the obstacle detection method for an AGV carrying shelves disclosed in Example 2 of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] It should be noted that when a component is referred to as being "mounted on" another component, it may be directly on the other component or there may be a central component. When a component is considered to be "set on" another component, it may be directly set on the other component or there may be a central component at the same time. When a component is considered to be "fixed to" another component, it may be directly fixed on the other component or there may be a central component at the same time.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "or / and" used herein includes any and all combinations of one or more of the related listed items.

[0041] Example 1

[0042] See also Figure 1 , Figure 1 The brief process of the obstacle detection method for the AGV carrying the shelf disclosed in this embodiment 1 is shown, which includes the following steps:

[0043] Step 1: Obtain the historical scanning sample set of the AGV carrying the shelf, and determine the shelf edge error area and interference pixel points.

[0044] Generally, step 1 is performed when the preset condition 1 is met;

[0045] Among them, the first precondition includes:

[0046] ① When the AGV has finished loading the shelf: that is, the shelf is loaded onto the AGV's loading tray (located on the top of the AGV).

[0047] ②After the shelf loaded by the AGV changes direction: that is, the loading plate of the AGV rotates with the shelf.

[0048] That is to say, after situation ① or ② occurs, step 1 is performed.

[0049] To help understand, this embodiment 1 provides an example diagram of an AGV loading a shelf, such as Figure 2As shown: Select an AGV from a certain company and load the shelf with universal wheels at the bottom onto the loading tray of the AGV.

[0050] Specifically, step one includes:

[0051] S101, control the laser radar on the AGV to scan its surrounding area N times to obtain N historical laser images.

[0052] The laser radar is installed on the AGV, which can emit lasers at 360 degrees. If there is an entity within the laser scanning range, the laser will be reflected when it touches the entity. In this way, the laser radar can convert the scanned laser data into a laser image through image conversion - the pixel points with laser in the image indicate the existence of an entity, and the pixel points without laser indicate the absence of an entity.

[0053] It should be noted that for the shelf, since it is loaded onto the loading tray of the AGV, the shelf body is higher than the AGV, and the laser radar laser can only act on the shelf legs and the universal wheels at the bottom of the shelf legs.

[0054] In order to facilitate the distinction, the pixels with laser and the pixels without laser can be marked with different colors in the historical pixel image.

[0055] Among them, the scanning frequency of the laser radar is f, and the total duration of N scans is Δt; Δt=N*f.

[0056] In order to ensure data continuity and balance processing load, the values ​​of f and N should not be too large or too small. Generally, f is 0.03-0.05s / time; N is 7-10. Then, it can be considered that during the N scanning processes, the shelf does not move relative to the AGV.

[0057] S102, traversing N historical laser images for processing, obtaining N historical reduction images, and forming a historical scanning sample set.

[0058] The processing methods of N historical laser images are the same, so let's take the nth historical laser image as an example to explain:

[0059] For the nth historical laser image: first delineate the outer contours of the AGV and the shelf, then divide the shelf edge error zone based on the center point error zone of the shelf, and use the divided shelf edge error zone as the nth historical reduction limit image.

[0060] The process of defining the outer contours of the AGV and the shelf is as follows:

[0061] The size information of the AGV and the shelves loaded thereon is obtained; specifically, the size information of the AGV and the shelves loaded thereon can be obtained by a preset measurement method.

[0062] First, the outer contour of the AGV is determined according to the installation position of the laser radar on the AGV, that is, the outer contour of the AGV from a bird's-eye view. Then, the outer contour of the shelf is determined according to the loading angle of the shelf, that is, the outer contour of the shelf from a bird's-eye view.

[0063] It should be noted that there is some overlap between the inner area of ​​the AGV outer contour and the inner area of ​​the shelf outer contour.

[0064] Then, after delineating the outer contour of the shelf, the outer contour of the shelf is divided into the shelf edge error zone according to the center point error zone of the shelf. The center point error zone of the shelf is a square area with the center of the AGV as the center, and its side length is ΔD. The shelf edge error zone is an equidistant annular area with the outer contour of the shelf as the baseline, and its width is ΔD.

[0065] Then, the divided shelf edge error area is used as the nth historical reduction limit image, which reduces the image size and reduces the pressure of subsequent processing.

[0066] For ease of understanding, Figure 2 Scan the loading direction and get Figure 3 Shelf edge error zone shown.

[0067] It should be noted that according to Figure 3 In the middle figure, the AGV moves to the right, so its main function is to detect whether there are obstacles in the moving direction. Figure 3 The scan information on the right is highlighted.

[0068] See also Figure 3 , which shows the situation in one of the historical zoom images: the outer contour of the AGV is rectangular (i.e. Figure 3 The center of the shelf is marked with the center point error zone. The outer contour of the shelf is also rectangular (i.e. Figure 3 The shelf corresponding line shown in the figure) has its edge marking the shelf edge error zone (only the Figure 3 The edge error zone on the right side of the middle shelf, the other three sides are not drawn). The laser point corresponding to the universal wheel falls within the edge error zone of the shelf.

[0069] Of course, if Figure 2 The loading direction is adjusted by 90°, and the Figure 4 The situation still has similar results - the corresponding area division is as follows Figure 5 Shown (only the Figure 5 The edge error area on the right side of the middle shelf, the other three sides are not drawn), and the corresponding historical scanning sample set is as follows Figure 6 shown.

[0070] S103, performing pixel analysis on the pixels of the historical scanning sample set to identify interfering pixels.

[0071] Since the historical scanning sample set includes N historical reduction limit images, all of which are images of the shelf edge error zone, the position information is also consistent. In other words, a certain pixel point in the historical scanning sample set appears N times in the historical scanning sample set (the nth time in the nth historical reduction limit image).

[0072] Then, if the number of times a certain pixel point has no laser exceeds a preset threshold value M, it is regarded as a historical obstacle-free point; otherwise, it is regarded as an interference pixel point.

[0073] The preset number threshold M satisfies: M∈[N / 2, N]. When the N value is large enough, the larger the M value is, the smaller the theoretical error is. N and M are selected according to actual needs.

[0074] The interfering pixels are related to the shelf legs, including the pixels corresponding to the shelf legs and the pixels corresponding to the universal wheels. These points will interfere with the subsequent obstacle detection, so they will be excluded in the subsequent obstacle detection.

[0075] For ease of understanding, use Figure 6 The historical scan sample set in the upper left is used as an example: after analyzing it, the interference pixels obtained are as follows Figure 6 Shown in upper right.

[0076] Step 2: Obtain a real-time reduced image and analyze the pixels of the real-time reduced image.

[0077] Specifically, the method for acquiring the real-time reduced image is similar to S102, including:

[0078] Control the laser radar on the AGV to scan the surrounding area in real time to obtain a real-time laser map;

[0079] Similarly, in order to facilitate distinction, pixels with laser and pixels without laser can be marked with different colors in the real-time laser image.

[0080] For real-time laser images: first delineate the outer contours of the AGV and the shelf, then divide the shelf edge error zone based on the center point error zone of the shelf, and use the divided shelf edge error zone as the real-time reduction limit map.

[0081] It should be noted that the method of defining the AGV outer contour, the shelf outer contour, and dividing the shelf edge error zone here is the same as S102 and will not be repeated.

[0082] Since the real-time shrinkage limit map is also a map of the shelf edge error zone, and its position information is consistent with the historical scanning sample set, the real-time shrinkage limit map is analyzed to see whether there are laser pixels other than interference pixels;

[0083] If there are laser pixels other than interference pixels in the real-time reduction image, they are regarded as new obstacle pixels, indicating that a new obstacle is detected. Otherwise, it means that no new obstacle is detected, and step 2 is repeated.

[0084] like Figure 6 As shown, the real-time reduction image does not show any laser pixels other than the interference pixels, that is, there are no new obstacle pixels, so no new obstacles are detected.

[0085] In addition, in order to improve the robustness of detection, a third step may be added: when the second preset condition is met, the historical scanning sample set is updated.

[0086] Specifically, the second preset condition includes: when the AGV running time reaches an integer multiple of the preset time threshold T, and no new obstacle pixels appear in the real-time reduction image obtained at this time.

[0087] Then, the method for updating the historical scan sample set includes:

[0088] The earliest historical reduction image in the historical scanning sample set is replaced with a real-time reduction image obtained when the preset condition 2 is satisfied, so as to keep the total number of images in the historical scanning sample set to N.

[0089] That is to say, step three is to perform adaptive updates based on the real-time recognition situation, thereby improving the robustness of obstacle detection.

[0090] This embodiment 1 also simultaneously discloses an AGV controller, which is designed using the above-mentioned AGV obstacle detection method with shelves, that is, the AGV is controlled according to the above-mentioned AGV obstacle detection method with shelves.

[0091] Example 2

[0092] See also Figure 7 , Figure 7 The brief process of the obstacle detection method for the AGV carrying the shelf disclosed in this embodiment 2 is shown, which includes the following steps:

[0093] Perform obstacle detection according to the obstacle detection method for AGVs carrying shelves disclosed in Example 1;

[0094] If a new obstacle is detected, obtain the position coordinates of the new obstacle pixel point;

[0095] Calculate the real-time distance to the AGV based on the position coordinates of the new obstacle pixel points;

[0096] If the real-time distance between the new obstacle pixel and the AGV is greater than the preset distance threshold ΔL, there is no need to avoid the obstacle; otherwise, obstacle avoidance is performed.

[0097] Generally, the value of ΔL is determined based on experience or the moving speed of the AGV.

[0098] This embodiment 2 also simultaneously discloses an AGV controller, which is designed using the above-mentioned AGV obstacle detection method with shelves, that is, the AGV is controlled according to the above-mentioned AGV obstacle detection method with shelves.

[0099] Example 3

[0100] This embodiment 3 discloses a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the obstacle detection method for an AGV loaded with shelves disclosed in embodiment 1, or / and the steps of the obstacle avoidance method for an AGV loaded with shelves disclosed in embodiment 2 are implemented.

[0101] This embodiment 3 also discloses a readable storage medium, in which computer program instructions are stored. When the computer program instructions are read and executed by a processor, the steps of the obstacle detection method for an AGV loaded with shelves disclosed in embodiment 1, or / and the steps of the obstacle avoidance method for an AGV loaded with shelves disclosed in embodiment 2 are executed.

[0102] This embodiment 3 also discloses a computer program product, including a computer program. When the computer program is executed by a processor, the steps of the obstacle detection method for an AGV carrying shelves disclosed in embodiment 1, or / and the steps of the obstacle avoidance method for an AGV carrying shelves disclosed in embodiment 2 are implemented.

[0103] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for detecting obstacles of an AGV carrying a shelf, characterized in that: The following steps are involved: Step 1: Obtain a historical scanning sample set of the AGV carrying the shelf, and determine the shelf edge error area and interference pixel points; Among them, step one includes: S101, controlling the laser radar on the AGV to scan its surrounding area N times to obtain N historical laser images; S102, traversing N historical laser images for processing, obtaining N historical reduction images, and forming a historical scanning sample set; Among them, for the nth historical laser image: first delineate the outer contour of the AGV and the outer contour of the shelf, then divide the shelf edge error area according to the center point error area of ​​the shelf, and use the divided shelf edge error area as the nth historical reduction limit image; S103, performing pixel analysis on the pixels of the historical scan sample set to identify interfering pixels; Step 2: Obtain a real-time reduced image and analyze the pixels of the real-time reduced image; If there are laser pixels other than interference pixels in the real-time reduction image, they are regarded as new obstacle pixels, indicating that a new obstacle has been detected; The method for obtaining the real-time shrinkage image includes: Control the laser radar on the AGV to scan the surrounding area in real time to obtain a real-time laser map; For real-time laser images: first delineate the outer contours of the AGV and the shelf, then divide the shelf edge error zone based on the center point error zone of the shelf, and use the divided shelf edge error zone as the real-time reduction limit map.

2. The obstacle detection method for an AGV carrying a shelf according to claim 1, characterized in that: When the preset condition 1 is met, proceed to step 1; Among them, the preset condition one includes: when the AGV completes loading the shelf, or when the shelf loaded by the AGV changes direction.

3. The obstacle detection method for AGVs carrying shelves according to claim 1, characterized in that: The methods for defining the outer contour of AGV and shelf include: Get the size information of the AGV and the rack it is loaded on; First, the outer contour of the AGV is determined based on the installation position of the laser radar on the AGV, and then the outer contour of the shelf is determined based on the loading angle of the shelf.

4. The obstacle detection method for AGVs carrying shelves according to claim 1, characterized in that: The center point error zone of the shelf placement is a square area with the center of the AGV as the center, and its side length is ΔD; The shelf edge error zone is an equidistant annular area with the shelf outer contour as a baseline, and its width is ΔD.

5. The obstacle detection method for AGVs carrying shelves according to claim 1, characterized in that: In the historical laser map and real-time laser map, the pixels with laser and the pixels without laser are marked with different colors.

6. The obstacle detection method for AGVs carrying shelves according to claim 1, characterized in that: For a certain pixel point in the historical scan sample set, it appears for the nth time in the nth historical reduction image, and appears N times in total in the historical scan sample set; For a certain pixel point in the historical scanning sample set, if the number of times it has no laser exceeds the preset number threshold M, it is regarded as a historical obstacle-free point; otherwise, it is regarded as an interference pixel point.

7. The obstacle detection method for an AGV carrying a shelf according to claim 6, characterized in that: The preset number threshold M satisfies: M∈[N / 2, N].

8. The obstacle detection method for an AGV carrying a shelf according to claim 1, characterized in that: The step also includes step three, when the preset condition two is met, updating the historical scanning sample set; Among them, the preset condition 2 includes: when the AGV running time reaches an integer multiple of the preset time threshold T, and no new obstacle pixels appear in the real-time reduction map obtained at this time; Methods for updating the historical scan sample set include: The earliest historical reduction image in the historical scanning sample set is replaced with a real-time reduction image obtained when the preset condition 2 is satisfied, so as to keep the total number of images in the historical scanning sample set to N.

9. An obstacle avoidance method for an AGV carrying a shelf, characterized in that: include: Perform obstacle detection according to the obstacle detection method for AGV carrying shelves as described in any one of claims 1 to 8; If a new obstacle is detected, obtain the position coordinates of the new obstacle pixel point; Calculate the real-time distance to the AGV based on the position coordinates of the new obstacle pixel points; If the real-time distance between the new obstacle pixel and the AGV is greater than the preset distance threshold ΔL, there is no need to avoid the obstacle; otherwise, obstacle avoidance is performed.

10. An AGV controller, characterized in that: The AGV controller is designed using the obstacle detection method for an AGV carrying shelves as described in any one of claims 1 to 8, or / and the obstacle avoidance method for an AGV carrying shelves as described in claim 9.