Obstacle avoidance method based on radar data and robot

Abstract

The application provides an obstacle avoidance method based on radar data and a robot, and the method comprises the following steps: acquiring radar data points collected by a laser radar in a moving process of the robot in real time; converting the collected radar data points into a two-dimensional coordinate system to obtain a radar image; performing merging processing on obstacle regions distributed on the radar image to obtain a plurality of target obstacle regions; and executing a corresponding obstacle avoidance strategy according to a relative position relationship between the robot and the plurality of target obstacle regions in the moving process. According to the application, the robot can realize good obstacle avoidance function by means of the laser radar only without depending on other sensors in the moving process.

Description

Technical Field

[0001] This application relates to the field of robotics technology, and in particular to an obstacle avoidance method and robot based on radar data. Background Technology

[0002] Robotic vacuum cleaners equipped with various sensors can use radar to collect data and construct a map of their workspace. While moving within this workspace, they rely on other sensors for obstacle avoidance. However, due to cost constraints, some robotic vacuum cleaners carry fewer and fewer sensors, relying solely on radar data for obstacle avoidance. Yet, radar data is linear data collected by a 360-degree rotating radar, consisting of isolated, discrete points, and containing both noisy and valid data. Because of these inherent limitations, relying solely on radar data cannot achieve satisfactory obstacle avoidance capabilities. Summary of the Invention

[0003] The purpose of this application is to provide an obstacle avoidance method and robot based on radar data, which can achieve good obstacle avoidance function with the help of radar data.

[0004] An obstacle avoidance method based on radar data, applied to a robot equipped with a lidar system, includes:

[0005] The radar data points collected by the lidar during the robot's movement are acquired in real time.

[0006] The collected radar data points are converted into a two-dimensional coordinate system to obtain radar images;

[0007] The obstacle regions distributed on the radar image are merged to obtain several target obstacle regions;

[0008] During the movement, the robot executes corresponding obstacle avoidance strategies based on its relative position to the target obstacle areas.

[0009] In one embodiment, the step of merging obstacle regions distributed on the radar image to obtain several target obstacle regions includes:

[0010] For any two adjacent obstacle regions in the plurality of obstacle regions, determine whether the two adjacent obstacle regions meet a preset merging condition; wherein, the merging condition includes one of the lateral distance and the longitudinal distance being less than a target body parameter, the target body parameter being the maximum value of the robot's width and length, and the other being less than a preset target value;

[0011] If so, the two adjacent obstacle regions will be merged.

[0012] After merging several pairs of adjacent obstacle regions, several target obstacle regions are obtained.

[0013] In one embodiment, before determining whether any two adjacent obstacle regions in the plurality of obstacle regions meet a preset merging condition, the method further includes:

[0014] Filter out obstacle areas in the radar image that do not meet the preset size requirements.

[0015] In one embodiment, during the movement, executing a corresponding obstacle avoidance strategy based on the relative positional relationship between the robot and the plurality of target obstacle regions includes:

[0016] In the direction of movement, if the distance between any target obstacle area and the robot is less than the preset braking distance, the robot will detour from the leftmost or rightmost side of the target obstacle area.

[0017] In one embodiment, during the movement, executing a corresponding obstacle avoidance strategy based on the relative positional relationship between the robot and the plurality of target obstacle regions includes:

[0018] If, in the direction of movement and on the left and / or right of the robot, the distance between the target obstacle area and the robot is less than a preset braking distance, check the relationship between the distance between two target obstacle areas in different directions and the robot's width; if the distance is greater than the robot's width, pass between the two target obstacle areas.

[0019] If the distance is not greater than the robot's width, determine whether it can be bypassed;

[0020] If it is possible to detour, detour from the far left of the leftmost target obstacle area, or detour from the far right of the rightmost target obstacle area;

[0021] If it is not possible to detour, move in the opposite direction of the stated direction of movement.

[0022] In one embodiment, before merging the obstacle regions distributed on the radar image to obtain a plurality of target obstacle regions, the method further includes:

[0023] Each radar data point in the radar image is magnified to obtain an adjusted radar image.

[0024] Several obstacle regions are determined from the adjusted radar image; wherein each obstacle region includes several magnified radar data points.

[0025] In one embodiment, the step of magnifying each radar data point in the radar image to obtain an adjusted radar image includes:

[0026] The magnification factor is determined based on the minimum value of the robot's size parameters and a preset adjustment factor; wherein, the robot size parameters include the robot width, robot height, and the distance between the robot's two wheels;

[0027] Using each radar data point as the center and the magnification factor as the diameter, the magnified radar data points are plotted to obtain the adjusted radar image.

[0028] In one embodiment, the step of converting the acquired radar data points into a two-dimensional coordinate system includes:

[0029] If the lidar on the robot reaches a preset frame rate threshold, for any collected lidar data point, the coordinates of the lidar data point in the two-dimensional coordinate system are determined based on the lidar rotation angle corresponding to the lidar data point, the distance between the data points, and the farthest transmission distance of the lidar.

[0030] In one embodiment, the step of converting the acquired radar data points into a two-dimensional coordinate system includes:

[0031] If the lidar on the robot does not reach the preset frame rate threshold, for any collected lidar data point, the deviation distance is determined based on the lidar rotation angle corresponding to the lidar data point and the linear velocity of the robot.

[0032] The coordinates of the radar data points in the two-dimensional coordinate system are determined based on the radar rotation angle corresponding to the radar data points, the distance between the data points, the deviation distance, and the farthest transmission distance of the lidar.

[0033] On the other hand, this application provides a robot, the robot comprising:

[0034] LiDAR is used to collect obstacle information;

[0035] processor;

[0036] Memory used to store processor-executable instructions;

[0037] The processor is configured to execute the aforementioned obstacle avoidance method based on radar data according to the obstacle information collected by the lidar.

[0038] In this application, the robot acquires radar data points collected by the LiDAR during its movement in real time, and then converts the collected radar data points into radar images. After merging and processing the obstacle areas on the radar images into target obstacle areas, the robot can execute corresponding obstacle avoidance strategies based on its relative positional relationship with the target obstacle areas. Therefore, during the robot's movement, good obstacle avoidance can be achieved solely with the help of LiDAR without relying on other sensors. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly described below.

[0040] Figure 1 This is a schematic diagram of the structure of a robot provided in one embodiment of this application;

[0041] Figure 2 A schematic flowchart illustrating an obstacle avoidance method based on radar data provided in an embodiment of this application;

[0042] Figure 3 A schematic diagram of a radar image provided in an embodiment of this application;

[0043] Figure 4 A flowchart illustrating a method for obtaining a target obstacle region according to an embodiment of this application;

[0044] Figure 5 A schematic diagram of an obstacle region provided in an embodiment of this application;

[0045] Figure 6 This is a schematic diagram of a target obstacle area provided in an embodiment of this application;

[0046] Figure 7 This is a schematic diagram of a robot obstacle avoidance motion provided in an embodiment of this application;

[0047] Figure 8 A schematic diagram of an obstacle avoidance motion provided in an embodiment of this application;

[0048] Figure 9 This is a schematic diagram of a robot obstacle avoidance motion provided in an embodiment of this application;

[0049] Figure 10 A block diagram of an obstacle avoidance device based on radar data provided in an embodiment of this application. Detailed Implementation

[0050] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0051] Similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0052] like Figure 1 As shown, this embodiment provides a robot 1, including: at least one processor 11, a memory 12, and a lidar 13. Figure 1 Taking a processor 11 as an example, the processor 11, memory 12, and lidar 13 are connected via bus 10. LiDAR 13 is used to collect obstacle information, and memory 12 stores instructions for processing obstacle information that can be executed by processor 11. The instructions are executed by processor 11 to enable robot 1 to perform all or part of the processes of the methods described in the following embodiments. In one embodiment, robot 1 may be a sweeping robot used to perform an obstacle avoidance method based on radar data.

[0053] The memory 12 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable red-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0054] This application also provides a computer-readable storage medium storing a computer program that can be executed by a processor 11 to perform the radar data-based obstacle avoidance method provided in this application.

[0055] See Figure 2 This is a flowchart illustrating an obstacle avoidance method based on radar data provided in an embodiment of this application, as shown below. Figure 2 As shown, the method may include steps 210-240.

[0056] Step 210: Acquire radar data points collected by the LiDAR during the robot's movement in real time.

[0057] The robot is equipped with a lidar system that can rotate 360 ​​degrees during robot movement. After emitting detection signals, it determines multiple radar data points based on signals reflected from obstacles. These radar data points represent the location of obstacles. The robot can acquire these radar data points in real time while moving within the workplace, thus obtaining obstacle information.

[0058] Step 220: Convert the collected radar data points into a two-dimensional coordinate system to obtain a radar image.

[0059] The robot can draw a circle with the center point of the lidar as the center and the lidar's farthest transmission distance as the radius, and generate a two-dimensional map based on the minimum bounding rectangle of the circle. A two-dimensional coordinate system is then established on this map. For example, the origin of the two-dimensional coordinate system can be either the lower left corner of the map or the upper left corner. The farthest transmission distance of the lidar is the maximum transmission distance within the required accuracy range; in other words, the radar data points detected within this maximum transmission distance meet the accuracy requirements.

[0060] The robot can calculate the coordinates of each radar data point in a two-dimensional coordinate system, thus transforming the radar data points into a two-dimensional coordinate system. After acquiring radar data points in real time, the radar data points can be transformed into a two-dimensional coordinate system. Since the robot can transform all the radar data points collected in the surrounding area into a two-dimensional coordinate system, the radar image is updated in real time during movement, and the radar image can represent the position of obstacles in the surrounding area during the robot's movement.

[0061] See Figure 3 This is a schematic diagram of a radar image provided in an embodiment of this application, as shown below. Figure 3 As shown, the white pixels in the radar image are radar data points transformed into a two-dimensional coordinate system, while the black pixels are where no radar data points exist.

[0062] Step 230: Merge the obstacle regions distributed on the radar image to obtain several target obstacle regions.

[0063] The obstacle region refers to the area where the obstacle is located. The obstacle region can be represented by radar data points on the radar image, and each obstacle region can include several radar data points. The target obstacle region is the obstacle region obtained after merging. In addition, after merging the obstacle regions on the radar image, if there is any obstacle region that has not been merged with other obstacle regions, that unmerged obstacle region is also a target obstacle region.

[0064] Robots move within a workplace by executing obstacle avoidance strategies based on obstacle regions in radar images. However, some obstacle regions in the radar image are too close together for the robot to pass through. In such cases, directly judging obstacle avoidance based on obstacle regions can lead to errors or redundant judgments. Therefore, the robot can merge closely spaced obstacle regions in the radar image. After merging mergeable obstacle regions in the radar image, the remaining obstacle regions can be used as target obstacle regions.

[0065] Step 240: During the movement, execute the corresponding obstacle avoidance strategy based on the relative positional relationship between the robot and several target obstacle areas.

[0066] During movement, the robot can check the target obstacles in the direction of movement and their relative position to itself, and then determine the movement path to avoid the obstacles based on the obstacle avoidance strategy, so as to continue moving in the workplace.

[0067] Through the above measures, the robot can generate radar images based on radar data points collected in real time during its movement, and determine its relative position to obstacles based on the obstacle areas formed by the radar data points in the radar images, thus achieving good obstacle avoidance function simply by relying on radar data.

[0068] In one embodiment, when the robot converts radar data points into a two-dimensional coordinate system, it can check whether the lidar on the robot has reached a preset frame rate threshold. Here, the frame rate of the lidar represents the number of rotations the lidar makes per second, that is, the number of times one scan is completed per second. The frame rate threshold can be an empirical value, and the frame rate threshold can increase as the robot's movement speed increases.

[0069] In one scenario, if the LiDAR reaches the frame rate threshold, the displacement caused by the robot's movement has a negligible impact on the radar data points scanned in one rotation. The robot can directly determine the coordinates of the radar data points in the two-dimensional coordinate system based on the radar rotation angle corresponding to each data point, the distance between the data points, and the maximum transmission distance of the LiDAR.

[0070] Among them, the radar rotation angle corresponding to the radar data point is the rotation angle when the lidar collects the radar data point; the data point distance is the distance between the radar data point and the center point of the lidar.

[0071] For example, the abscissa of the radar data point in the two-dimensional coordinate system can be determined by the following formula (1), and the ordinate of the radar data point in the two-dimensional coordinate system can be determined by the following formula (2):

[0072] X = Maxdis + dis * cos(angle) (1)

[0073] Y = Maxdis + dis*sin(angle) (2)

[0074] Here, X is the horizontal axis; Y is the vertical axis; Maxdis is the maximum transmission distance of the lidar; dis is the distance between data points; and angle is the radar rotation angle.

[0075] In another scenario, if the LiDAR does not reach the frame rate threshold, the displacement generated by the robot's movement will affect the radar data points scanned in one rotation. For any radar data point acquired, the robot can determine the deviation distance based on the radar rotation angle corresponding to the data point and the robot's linear velocity.

[0076] For example, the deviation distance can be determined by the following formula (3):

[0077] PT=(time(angle)-time(0))*linespeed (3)

[0078] Where PT is the deviation distance; time(angle) is the time it takes for the lidar to reach the lidar rotation angle; time(0) is the time it takes for the lidar to be at 0 degrees; and linespeed is the linear speed of the robot.

[0079] After obtaining the deviation distance, the robot can determine the coordinates of the radar data point in the two-dimensional coordinate system based on the radar rotation angle corresponding to the radar data point, the distance between the data points, the deviation distance, and the farthest transmission distance of the lidar.

[0080] For example, the abscissa of the radar data point in the two-dimensional coordinate system can be determined by the following formula (4), and the ordinate of the radar data point in the two-dimensional coordinate system can be determined by the following formula (5):

[0081] X=Maxdis+(dis+PT)*cos(angle) (4)

[0082] Y=Maxdis+(dis+PT)*sin(angle) (5)

[0083] Here, X is the horizontal axis; Y is the vertical axis; Maxdis is the farthest transmission distance of the lidar; dis is the distance between data points; PT is the deviation distance corresponding to the radar data point; and angle is the radar rotation angle.

[0084] Through the above measures, the robot can calculate the coordinates of each radar data point in the radar image, and use the pixel where the data point coordinates are located in the radar image as the radar data point, setting that pixel as a white pixel.

[0085] In one embodiment, before merging obstacle regions distributed across a radar image, the robot can determine the obstacle regions from the radar image. Radar data points in a radar image are relatively discrete and may contain noisy data. To avoid filtering out valid radar data points during radar image filtering, the robot can magnify each radar data point in the radar image to obtain an adjusted radar image. Before magnification, each radar data point occupies one pixel in the radar image. Through magnification, each radar data point can be expanded outwards (in all directions) by several pixels, thereby enlarging the radar data point; a magnified radar data point can occupy multiple pixels.

[0086] The robot can identify several obstacle regions from an adjusted radar image. Each obstacle region consists of several magnified radar data points. After magnification, the originally discrete radar data points in the radar image become somewhat connected. The robot can merge multiple connected radar data points into obstacle regions based on the principle of connected components, thus obtaining several obstacle regions in the radar image.

[0087] In one embodiment, when the robot magnifies radar data points in a radar image, the magnification factor can be determined based on the minimum value of the robot's size parameters and a preset adjustment factor. The robot size parameters include the robot's width, height, and the distance between its two wheels. The adjustment factor is a constant configured based on experience.

[0088] For example, the magnification factor can be determined by the following formula (6):

[0089] Rate=Min(W,H,Wheeldis)*a (6)

[0090] Here, Rate is the magnification factor; W is the robot width; H is the robot height; Wheeldis is the distance between the robot's two wheels; and a is the adjustment factor, which can range from 0.5 to 0.8. In this case, the magnification factor is approximately 10 times.

[0091] After obtaining the magnification factor, the robot can draw a solid circle with each radar data point as the center and the magnification factor as the diameter, representing the magnified radar data point. After magnifying each radar data point, an adjusted radar image can be obtained. Since the radar data point occupies one pixel before magnification, with a magnification factor of 10, a solid circle with a diameter of 10 pixels is drawn with that pixel as the center, resulting in the magnified radar data point.

[0092] In one embodiment, see Figure 4 This is a flowchart illustrating a method for obtaining a target obstacle region according to an embodiment of this application. Figure 4 As shown, the robot can perform the following steps 231 to 233 during the process of merging obstacle regions to obtain the target obstacle region.

[0093] Step 231: For any two adjacent obstacle regions in a number of obstacle regions, determine whether the two adjacent obstacle regions meet the preset merging conditions; wherein, the merging conditions include one of the horizontal distance and the vertical distance being less than the target body parameter, the target body parameter being the maximum value of the robot's width and length, and the other being less than the preset target value.

[0094] The robot can select the maximum value from its own width and length as the target body parameters. For any two adjacent obstacle areas, the robot can determine a first distance in the horizontal direction of the radar image based on the positions of the two adjacent obstacle areas, and then determine the lateral distance based on this first distance, which is the actual distance in the workplace. The robot can also determine a second distance in the vertical direction of the radar image based on the positions of the two adjacent obstacle areas, and then determine the longitudinal distance based on this second distance, which is the actual distance in the workplace.

[0095] The robot can determine a first distance based on the x-coordinates of the endpoints of two adjacent obstacle regions, and a second distance based on the y-coordinates of the endpoints of the two adjacent obstacle regions. Alternatively, the robot can determine the first distance based on the two distances between the x-coordinates of the endpoints of the two adjacent obstacle regions and the x-coordinate of the center point of the radar image, and the second distance based on the two distances between the y-coordinates of the endpoints of the two adjacent obstacle regions and the y-coordinate of the center point of the radar image. Here, the endpoints are the two closest points of the adjacent obstacle regions. After obtaining the first and second distances, the robot can determine the lateral and longitudinal distances between the two adjacent obstacle regions based on the conversion relationship between coordinate distances on the radar image and physical distances.

[0096] After obtaining the lateral and longitudinal distances, the robot can determine whether either of these distances is less than the target body parameter. On one hand, if both the lateral and longitudinal distances are not less than the target body parameter, it means that the two adjacent obstacle areas do not meet the merging condition. On the other hand, if one of the lateral and longitudinal distances is less than the target body parameter, the robot can continue to determine whether the other is less than a preset target value. Here, the target value can be an empirical value; for example, the target value could be 10 centimeters. In one case, if the other is less than the target value, it can be determined that the two adjacent obstacle areas meet the merging condition. In another case, if the other is not less than the target value, it can be determined that the two adjacent obstacle areas do not meet the merging condition.

[0097] In other words, two adjacent obstacle regions satisfy the merging condition under the following circumstances:

[0098] The lateral distance is less than the target fuselage parameters, and the longitudinal distance is less than the target value; or,

[0099] The longitudinal distance is less than the target fuselage parameters, and the lateral distance is less than the target value.

[0100] Step 232: If so, merge the two adjacent obstacle regions. Step 233: After merging several pairs of adjacent obstacle regions, several target obstacle regions are obtained.

[0101] If any two adjacent obstacle regions meet the merging conditions, they can be merged to obtain a merged obstacle region. For the merged obstacle region, it can be further determined whether it meets the merging conditions with its adjacent obstacle regions. If the conditions are met, merging can continue.

[0102] After merging the obstacle regions on the radar image, several obstacle regions that cannot be merged are obtained. The spacing between these unmerged obstacle regions is sufficient for the robot to pass through. At this point, the obstacle regions on the radar image are the target obstacle regions.

[0103] By taking the above measures, adjacent obstacle areas that meet the conditions can be merged to obtain several target obstacle areas, thereby reducing the number of obstacle areas and enabling subsequent obstacle avoidance movements based on the target obstacle areas, thus improving the robot's obstacle avoidance efficiency.

[0104] See Figure 5 This is a schematic diagram of an obstacle region provided in an embodiment of this application, as shown below. Figure 5As shown, in the radar image, the leftmost endpoint of obstacle region Regin1 is L1, and the rightmost endpoint is R1; in the radar image, the leftmost endpoint of obstacle region Regin2 is L2, and the rightmost endpoint is R2. The lateral distance between obstacle regions Regin1 and Regin2 is dis12, which can be the actual distance between endpoints R1 and L2 in the workplace. By comparing this lateral distance with the target fuselage parameters, it can be determined whether obstacle regions Regin1 and Regin2 should be merged.

[0105] See Figure 6 This is a schematic diagram of a target obstacle area provided in an embodiment of this application, as shown below. Figure 6 As shown, the radar image displays various target obstacle areas around the robot, indicating the obstacles the robot needs to avoid during its movement.

[0106] In one embodiment, before determining whether any two adjacent obstacle regions can be merged, the robot can filter out obstacle regions in the radar image that do not meet the preset size requirements, and then perform a merging operation on the obstacle regions that meet the size requirements. The size requirements can be preset based on experience to filter out false obstacle regions formed by noise data.

[0107] Because radar images contain noisy data, in order to avoid the impact of noisy data on obstacle avoidance, the robot can filter out the obstacle areas formed by the noisy data according to size requirements. After filtering out the noisy data, the robot can then perform the subsequent merging process with the correct obstacle areas.

[0108] In one embodiment, when the robot executes an obstacle avoidance strategy based on its relative position to several target obstacle areas during movement, the radar image can be divided into four avoidance detection areas with a center point: directly in front, directly behind, to the left, and to the right. See also Figure 7 This is a schematic diagram of the robot obstacle avoidance movement provided in an embodiment of this application. In the direction of the robot's movement, area F is the front, area L is the left side of the robot, area R is the right side of the robot, and area B is the rear of the robot.

[0109] In the direction of movement, if the distance between the robot and any target obstacle area is less than a preset braking distance, the robot can detour around the leftmost or rightmost edge of the target obstacle area. The braking distance is the distance required for the robot to stop after issuing a command. If the robot continues to move beyond the target obstacle area when the distance between the robot and the target obstacle area is less than the braking distance, it will collide with the obstacle. The robot can plan a path to avoid the target obstacle area based on the angle and distance between itself and the rightmost or leftmost edge of the target obstacle area, thus turning right or left to detour.

[0110] In one embodiment, after the robot divides the radar image into four avoidance detection zones with a center point, it may detect obstacles in at least two of the avoidance detection zones during its movement. See also Figure 8 This is a schematic diagram of an obstacle avoidance motion provided in an embodiment of this application, as shown below. Figure 8 As shown, the robot can execute the obstacle avoidance strategy through the following steps 810 to 840.

[0111] Step 810: If there are target obstacle areas and the distance between the robot and the robot is less than the preset braking distance in the direction of movement and on the left and / or right of the robot, check the relationship between the distance between the two target obstacle areas in different directions and the width of the robot; if the distance is greater than the width of the robot, pass between the two target obstacle areas.

[0112] When there are target obstacle areas in the direction of movement and on the right side of the robot, and the distance between the target obstacle area and the robot is less than the braking distance, the robot can check whether the distance between the target obstacle area in the direction of movement and the target obstacle area on the right side is greater than the robot width.

[0113] See Figure 9 This is a schematic diagram of the obstacle avoidance motion of a robot according to an embodiment of this application, as shown below. Figure 9 As shown, in the robot's running direction, area F is directly in front, area L is to the left of the robot, area R is to the right of the robot, and area B is directly behind the robot. Figure 9 In the scenario, there is a target obstacle region (Region1) in the robot's direction of movement; there is also a target obstacle region (Region2) on the robot's right side, and the distances between Region1 and Region2 and the robot are both less than the braking distance. The robot can check whether the distance between Region1 and Region2 is greater than the robot's width.

[0114] When there are target obstacle areas in both the direction of movement and on the left side of the robot, and the distance between the target obstacle areas and the robot is less than the braking distance, the robot can check whether the distance between the target obstacle area in the direction of movement and the target obstacle area on the left side is greater than the robot's width. Figure 9 For example, if there are target obstacle areas in the F region along the robot's running direction and in the L region to the left, the robot can check whether the distance between the target obstacle areas in the F region and the target obstacle areas in the L region is greater than the robot's width.

[0115] When there are target obstacle areas in the direction of movement and on the right and left sides of the robot, and the distance between each target obstacle area and the robot is less than the braking distance, the robot can check whether the distance between the target obstacle area in the direction of movement and the target obstacle area on the left is greater than the robot's width, or the robot can check whether the distance between the target obstacle area in the direction of movement and the target obstacle area on the right is greater than the robot's width.

[0116] by Figure 9 For example, if there are target obstacle areas in the F region along the robot's running direction, the L region to the left, and the R region to the right, the robot can check whether the distance between the target obstacle areas in the F region and the target obstacle areas in the L region is greater than the robot's width; or, the robot can check whether the distance between the target obstacle areas in the F region and the target obstacle areas in the R region is greater than the robot's width.

[0117] In one scenario, if the distance between two target obstacle areas is greater than the robot's width, the robot can pass between them. When the distance is greater than the robot's width, there are two scenarios: First, if the distance is no more than 1.5 times the robot's width, the center point of the line connecting the two obstacle areas can be used as the robot's detour direction. Second, if the distance is greater than 1.5 times the robot's width, the line connecting the two obstacle areas can be divided equally according to the robot's width, and the point closest to the obstacle area in the direction of travel can be chosen as the robot's detour direction. In this case, the robot can choose other points when returning, thus cleaning the area between the two obstacle areas during its operation.

[0118] In another case, the distance between the target obstacle areas in two directions is not greater than the robot's width, in which case step 820 can be continued.

[0119] Step 820: If the distance is not greater than the robot's width, determine whether it can be bypassed.

[0120] If the distance between two target obstacle areas is no greater than the robot's width, the robot cannot pass between them. The robot can determine whether it can detour from the leftmost side of the leftmost target obstacle area or from the rightmost side of the rightmost target obstacle area.

[0121] Step 830: If it is possible to detour, detour from the far left of the leftmost target obstacle area, or detour from the far right of the rightmost target obstacle area.

[0122] If there is a gap larger than the robot's width at the far left of the leftmost target obstacle area, the robot can bypass it from the far left of the leftmost target obstacle area. Figure 9 For example, there is a gap on the far left of the target obstacle area Region1, and the robot can bypass it using method 2.

[0123] If there is a gap on the far right of the rightmost target obstacle area that is larger than the robot's width, then the robot can detour from the far right of the rightmost target obstacle area.

[0124] Step 840: If you cannot detour, move in the opposite direction of the direction of movement.

[0125] If the leftmost side of the leftmost obstacle area is inaccessible, and the rightmost side of the rightmost obstacle area is inaccessible, the robot can only move in the opposite direction of its original movement, i.e., turn around and move backward. During the turnaround, the robot can avoid obstacle areas in the new direction of movement using the aforementioned obstacle avoidance strategy. Figure 9 As shown, the robot can move backward using methods 3 and 4.

[0126] Figure 10 This is a block diagram of an obstacle avoidance device based on radar data according to an embodiment of the present invention, such as... Figure 10 As shown, the device may include:

[0127] The acquisition module 1010 is used to acquire radar data points collected by the lidar during the robot's movement in real time.

[0128] The conversion module 1020 is used to convert the collected radar data points into a two-dimensional coordinate system to obtain radar images;

[0129] The merging module 1030 is used to merge obstacle regions distributed on the radar image to obtain several target obstacle regions;

[0130] The obstacle avoidance module 1040 is used to execute corresponding obstacle avoidance strategies based on the relative positional relationship between the robot and the plurality of target obstacle areas during movement.

[0131] The specific implementation process of the functions and roles of each module in the above-mentioned device can be found in the implementation process of the corresponding steps in the obstacle avoidance method based on radar data, and will not be repeated here.

[0132] The apparatuses and methods disclosed in the several embodiments provided in this application can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatuses, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0133] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0134] If a function is implemented as a software module and sold or used as an independent product, it 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 portion 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 of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

Claims

1. An obstacle avoidance method based on radar data, applied to a robot equipped with a lidar system, characterized in that... include: The radar data points collected by the lidar during the robot's movement are acquired in real time. The collected radar data points are converted into a two-dimensional coordinate system to obtain radar images; Several obstacle regions distributed on the radar image are merged to obtain several target obstacle regions; During the movement, the robot executes corresponding obstacle avoidance strategies based on the relative positional relationship between itself and the target obstacle areas. The process of converting the collected radar data points into a two-dimensional coordinate system to obtain a radar image includes: Check whether the lidar has reached the preset frame rate threshold; If the lidar reaches the preset frame rate threshold, the coordinates of the lidar data point in the two-dimensional coordinate system are determined based on the lidar rotation angle corresponding to the lidar data point, the distance between the data points, and the farthest transmission distance of the lidar. If the lidar does not reach the preset frame rate threshold, the deviation distance is determined based on the radar rotation angle corresponding to the radar data point and the linear velocity of the robot; the coordinates of the radar data point in the two-dimensional coordinate system are determined based on the radar rotation angle corresponding to the radar data point, the data point distance, the deviation distance, and the farthest emission distance of the lidar. Wherein, the radar rotation angle corresponding to the radar data point is the rotation angle when the lidar acquires the radar data point; the data point distance is the distance between the radar data point and the center point of the lidar.

2. The obstacle avoidance method based on radar data according to claim 1, characterized in that, The process of merging several obstacle regions distributed on the radar image to obtain several target obstacle regions includes: For any two adjacent obstacle regions in the plurality of obstacle regions, determine whether the two adjacent obstacle regions meet a preset merging condition; wherein, the merging condition includes one of the lateral distance and the longitudinal distance being less than a target body parameter, the target body parameter being the maximum value of the robot's width and length, and the other being less than a preset target value; If so, the two adjacent obstacle regions will be merged. After merging several pairs of adjacent obstacle regions, several target obstacle regions are obtained.

3. The obstacle avoidance method based on radar data according to claim 2, characterized in that, Before determining whether any two adjacent obstacle regions in the plurality of obstacle regions meet a preset merging condition, the method further includes: Filter out obstacle areas in the radar image that do not meet the preset size requirements.

4. The obstacle avoidance method based on radar data according to claim 1, characterized in that, During the movement, based on the relative positional relationship between the robot and the plurality of target obstacle areas, a corresponding obstacle avoidance strategy is executed, including: In the direction of movement, if the distance between any target obstacle area and the robot is less than the preset braking distance, the robot will detour from the leftmost or rightmost side of the target obstacle area.

5. The obstacle avoidance method based on radar data according to claim 1, characterized in that, During the movement, based on the relative positional relationship between the robot and the plurality of target obstacle areas, a corresponding obstacle avoidance strategy is executed, including: If, in the direction of movement and on the left and / or right of the robot, the distance between the target obstacle area and the robot is less than a preset braking distance, check the relationship between the distance between two target obstacle areas in different directions and the robot's width; if the distance is greater than the robot's width, pass between the two target obstacle areas. If the distance is not greater than the robot's width, determine whether it can be bypassed; If it is possible to detour, detour from the far left of the leftmost target obstacle area, or detour from the far right of the rightmost target obstacle area; If it is not possible to detour, move in the opposite direction of the stated direction of movement.

6. The obstacle avoidance method based on radar data according to claim 1, characterized in that, Before merging the several obstacle regions distributed on the radar image to obtain several target obstacle regions, the method further includes: Each radar data point in the radar image is magnified to obtain the adjusted radar image. Several obstacle regions are determined from the adjusted radar image; wherein each obstacle region includes several magnified radar data points.

7. The method according to claim 6, characterized in that, The step of magnifying each radar data point in the radar image to obtain the adjusted radar image includes: The magnification factor is determined based on the minimum value of the robot's size parameters and a preset adjustment factor; wherein, the robot size parameters include the robot width, robot height, and the distance between the robot's two wheels; Using each radar data point as the center and the magnification factor as the diameter, the magnified radar data points are plotted to obtain the adjusted radar image.

8. The obstacle avoidance method based on radar data according to claim 1, characterized in that, The process of converting the collected radar data points into a two-dimensional coordinate system includes: If the lidar on the robot reaches a preset frame rate threshold, for any collected lidar data point, the coordinates of the lidar data point in the two-dimensional coordinate system are determined based on the lidar rotation angle corresponding to the lidar data point, the distance between the data points, and the farthest transmission distance of the lidar.

9. The obstacle avoidance method based on radar data according to claim 1, characterized in that, The process of converting the collected radar data points into a two-dimensional coordinate system includes: If the lidar on the robot does not reach the preset frame rate threshold, for any collected lidar data point, the deviation distance is determined based on the lidar rotation angle corresponding to the lidar data point and the linear velocity of the robot. The coordinates of the radar data points in the two-dimensional coordinate system are determined based on the radar rotation angle corresponding to the radar data points, the distance between the data points, the deviation distance, and the farthest transmission distance of the lidar.

10. A robot, characterized in that, The robot includes: LiDAR is used to collect obstacle information; processor; Memory used to store processor-executable instructions; The processor is configured to execute the obstacle avoidance method based on radar data according to any one of claims 1-9 based on the obstacle information collected by the lidar.

Images