Threshold recognition method and obstacle crossing and escape method of robot

Abstract

The invention provides a threshold recognition method and an obstacle crossing and detrapping method of a robot. The threshold detrapping method comprises the steps of planning an advancing path, controlling the robot to advance according to the advancing path and go to a target point located outside a feature area, and marking feature points of obstacles to be screened according to the advancing state of the robot in the advancing process; wherein the feature region is a closed region with an exit as a threshold; when at least one to-be-screened obstacle feature point is marked, re-planning an advancing path to continue to go to the target point; the robot is controlled to advance along the edge in the feature area, and an edge path is obtained; according to the edge path and / or the to-be-screened obstacle feature points, screening out target obstacle feature points from the plurality of to-be-screened obstacle feature points; and recognizing the position of the threshold in the feature region according to the target obstacle feature point. Compared with the prior art, the method provided by the invention can quickly and accurately identify the exit position in the low-lying area, and ensures that the robot quickly gets out of the low-lying area.

Description

Technical Field

[0001] The present application relates to the field of robotics technology, and in particular to a threshold recognition method and a robot obstacle-crossing and escaping method. Background Art

[0002] With the development of science and technology, robots are increasingly used in people's daily lives. In practice, robots perform corresponding tasks by moving. Generally, when a robot moves to a low-lying area with a threshold as an exit, it is easy to be trapped in the low-lying area.

[0003] In the prior art, when a robot is trapped in a low-lying area, it will try to find an exit blindly. However, when the robot finds an exit in the above way, it may misjudge a certain position inside the room as an exit, and the accuracy of finding an exit is low. Summary of the invention

[0004] The purpose of the embodiments of the present application is to provide a threshold recognition method and a robot obstacle escaping method, which can quickly and accurately identify the exit position in a low-lying area.

[0005] In a first aspect, the present application provides a threshold recognition method, which is applied to a robot, and the method comprises:

[0006] Plan the travel path, control the robot to move along the travel path, and go to the target point outside the feature area. During the travel, mark the feature points of obstacles to be screened according to the robot's travel status; the feature area is a closed area with a threshold as the exit;

[0007] After marking at least one obstacle feature point to be screened, re-plan the path to continue to the target point;

[0008] Control the robot to move along the edge in the feature area to obtain the path along the edge;

[0009] Filtering target obstacle feature points from a plurality of obstacle feature points to be filtered according to the edge path and / or obstacle feature points to be filtered;

[0010] According to the characteristic points of the target obstacle, the position of the threshold in the characteristic area is identified.

[0011] In one embodiment, marking feature points of obstacles to be screened according to the robot's traveling state during the traveling process includes:

[0012] During the movement, determine whether the robot is slipping according to its movement status;

[0013] If the robot slips, the position where the robot slips is marked as a feature point of an obstacle to be screened.

[0014] In one embodiment, selecting a target obstacle feature point from a plurality of obstacle feature points to be selected according to the edge path and / or the obstacle feature point to be selected includes:

[0015] According to the screening conditions formed by the edge path and / or the obstacle feature points to be screened, the target obstacle feature points are screened out from the multiple obstacle feature points to be screened; wherein the screening conditions include a combination of one or more of the positional relationship between the edge path and the multiple obstacle feature points to be screened, the distance relationship between the edge path and the multiple obstacle feature points to be screened, the obstacle conditions around each obstacle feature point to be screened, and the length of a first straight line formed by sequentially connecting the multiple obstacle feature points to be screened.

[0016] In one embodiment, the screening condition includes a positional relationship between the edge path and a plurality of obstacle feature points to be screened; and screening a target obstacle feature point from the plurality of obstacle feature points to be screened according to the screening condition formed by the edge path and / or the obstacle feature points to be screened includes:

[0017] According to the positional relationship between the edge path and the plurality of obstacle feature points to be screened, a candidate obstacle feature point located outside the edge path is screened from the plurality of obstacle feature points to be screened;

[0018] According to the candidate obstacle feature points, the target obstacle feature points are determined.

[0019] In one embodiment, the screening condition includes a distance relationship between the edge path and a plurality of obstacle feature points to be screened; and screening a target obstacle feature point from the plurality of obstacle feature points to be screened according to the screening condition formed by the edge path and / or the obstacle feature points to be screened includes:

[0020] Cutting out a target path from the edge path; wherein the target path is formed when the robot is in an inclined state, and the length of the target path is greater than a first preset length;

[0021] Calculating the distances between a plurality of obstacle feature points to be screened and the target path, and screening out candidate obstacle feature points whose distances to the target path are less than a first preset distance from the plurality of obstacle feature points to be screened;

[0022] According to the candidate obstacle feature points, the target obstacle feature points are determined.

[0023] In one embodiment, a laser detection unit is provided on the robot; the screening conditions include obstacle conditions around each obstacle feature point to be screened; and according to the screening conditions formed by the edge path and / or the obstacle feature point to be screened, a target obstacle feature point is screened out from a plurality of obstacle feature points to be screened, including:

[0024] Detect whether there are laser visible obstacles around each feature point of the obstacle to be screened by a laser detection unit;

[0025] According to the detection results, candidate obstacle feature points with no laser visible obstacles around them are selected from a plurality of obstacle feature points to be selected;

[0026] According to the candidate obstacle feature points, the target obstacle feature points are determined.

[0027] In one embodiment, the screening condition includes the length of a first straight line formed by sequentially connecting a plurality of obstacle feature points to be screened; screening a target obstacle feature point from the plurality of obstacle feature points to be screened according to the screening condition formed by the edge path and / or the obstacle feature points to be screened includes:

[0028] Calculate the length of the first straight line formed by sequentially connecting multiple obstacle feature points to be screened;

[0029] Selecting candidate obstacle feature points whose first straight line length is greater than a second preset length from the plurality of obstacle feature points to be selected;

[0030] According to the candidate obstacle feature points, the target obstacle feature points are determined.

[0031] In a second aspect, the present application provides a method for a robot to overcome obstacles and escape from difficulties, the method comprising controlling the robot to execute the above-mentioned threshold recognition method to recognize the position of the threshold in a characteristic area;

[0032] Generate an escape path based on the location of the threshold in the feature area;

[0033] Control the robot to move along the escape path to escape.

[0034] In one embodiment, controlling the robot to move along an escape path to escape includes:

[0035] Controlling the robot to move along the escape path at a first moving speed, and determining whether the distance between the robot and the threshold is less than a second preset distance during the moving process;

[0036] If the distance between the robot and the threshold is less than a second preset distance, the robot is controlled to move along the escape path at a second travel speed; wherein the second travel speed is greater than the first travel speed.

[0037] In one embodiment, the robot's obstacle escaping method further includes:

[0038] When the robot moves along the escape path at the second travel speed, determining whether the robot slips;

[0039] If the judgment result indicates that the robot has slipped, an inaccessible sign is generated at the threshold.

[0040] The present application provides a threshold recognition method for identifying the exit position when the robot is trapped in a low-lying area. The specific recognition method is: first, a travel path is planned, and the robot is controlled to move along the travel path to a target point outside the low-lying area, and obstacle feature points are marked according to the robot's travel status during the movement. Each time an obstacle feature point is marked, the travel path is replanned to continue to move to the target point until the travel path cannot be planned; after the travel path cannot be planned, the robot is controlled to move along the edge in the low-lying area to form an edge path, and then the obstacle feature points are screened according to the edge path and / or obstacle feature points to obtain target obstacle feature points, and finally the exit position in the low-lying area can be identified according to the target obstacle feature points.

[0041] Compared with the prior art, the method in the present application can quickly and accurately identify the exit location in the low-lying area, ensuring that the robot can quickly escape from the low-lying area. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solution of the embodiments of the present application, the drawings required for use in the embodiments of the present application are briefly introduced below.

[0043] Figure 1 A schematic diagram of a robot provided in one embodiment of the present application;

[0044] Figure 2 A schematic diagram of the structure of a control unit provided in one embodiment of the present application;

[0045] Figure 3 A schematic diagram of a map corresponding to a characteristic area provided in the first embodiment of the present application;

[0046] Figure 4 A schematic diagram of a flow chart of a threshold recognition method provided in the first embodiment of the present application;

[0047] Figure 5 A schematic diagram of a map corresponding to a characteristic area provided in the second embodiment of the present application;

[0048] Figure 6 A schematic diagram of a map corresponding to a characteristic area provided in the third embodiment of the present application;

[0049] Figure 7 A schematic diagram of a map corresponding to a characteristic area provided in the fourth embodiment of the present application;

[0050] Figure 8 A schematic diagram of a map corresponding to a characteristic area provided in the fifth embodiment of the present application;

[0051] Fig. 9A schematic diagram of a map corresponding to a characteristic area provided in the sixth embodiment of the present application;

[0052] Fig.10 A schematic flow chart of a method for identifying a threshold provided in a second embodiment of the present application;

[0053] Fig.11 A schematic diagram of a map corresponding to a characteristic area provided in the seventh embodiment of the present application;

[0054] Fig.12 A schematic diagram of a map corresponding to a characteristic area provided in the eighth embodiment of the present application;

[0055] Fig.13 A schematic diagram of a map corresponding to a characteristic area provided in the ninth embodiment of the present application;

[0056] Fig.14 A schematic diagram of a map corresponding to a characteristic area provided in the tenth embodiment of the present application;

[0057] Fig.15 A schematic diagram of a flow chart of a method for overcoming obstacles and escaping distress for a robot provided in one embodiment of the present application;

[0058] Fig.16 A schematic diagram of a characteristic area provided for the eleventh embodiment of the present application.

[0059] Reference numerals:

[0060] 1-robot; 100-control unit; 110-bus; 120-processor; 130-memory; 200-detection unit; 210-first detection unit; 220-second detection unit; 230-laser detection unit. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0062] Similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0063] Please refer to Figure 1 , which is a schematic diagram of the structure of a robot 1 provided in an embodiment of the present application. The robot 1 in the present application may be a cleaning robot, a food delivery robot, a transport robot, etc. Figure 1As shown, the robot 1 is provided with a control unit 100 and a detection unit 200 which are interconnected; the detection unit 200 is used to detect relevant information of the robot 1, and send the detected relevant information to the control unit 100; the control unit 100 is used to receive the relevant information sent by the detection unit 200, and execute the threshold recognition method provided in the following embodiment of the present application according to the relevant information.

[0064] Please refer to Figure 2 , which is a schematic diagram of the structure of a control unit 100 provided in an embodiment of the present application. Figure 2 As shown, the control unit 100 includes: at least one processor 120 and a memory 130, Figure 2 A processor 120 is taken as an example. The processor 120 and the memory 130 are connected via a bus 110. The memory 130 stores instructions that can be executed by the processor 120, and the instructions are executed by the processor 120.

[0065] The memory 130 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, disk or optical disk.

[0066] Please refer to Figure 3 , which is a schematic diagram of a characteristic region provided in an embodiment of the present application. Figure 3As shown, the characteristic area K1 is a low-lying area, and its exit is the threshold P2; there are obstacles such as the wall P1 in the characteristic area K1, and there are also transparent obstacles such as the glass P3. In practice, the robot 1 will scan the general outline of the characteristic area K1 by laser scanning, form a map after scanning and store it, and then identify the general outline of the characteristic area K1 according to the stored map when performing related tasks. However, when the robot 1 scans the characteristic area K1, it can only scan obstacles such as the wall P1 that can be seen by the laser, but cannot scan obstacles such as the glass P3 and the threshold P2 that cannot be seen by the laser. This results in that the map formed after scanning the characteristic area K1 can only present obstacles such as the wall P1 that can be seen by the laser, but cannot present obstacles such as the glass P3 and the threshold P2 that cannot be seen by the laser.

[0067] Furthermore, because the threshold P2 cannot be presented in the map, when the robot 1 is trapped in the characteristic area K1, it cannot determine the position of the threshold P2 based on the map of the characteristic area K1, which ultimately causes the robot 1 to be unable to escape from the characteristic area K1, interrupting the robot 1 from continuing to perform related tasks. For example, when the robot 1 is a cleaning robot 1, if the cleaning robot 1 cannot escape from the characteristic area K1, the cleaning robot 1 will be unable to clean other areas outside the characteristic area K1; or, when the charging pile of the cleaning robot 1 is located in other areas outside the characteristic area K1, the cleaning robot 1 will be unable to return to the charging pile for charging.

[0068] In the prior art, in order to enable the robot 1 to successfully escape from the characteristic area K1, the robot 1 is controlled to blindly explore in the characteristic area K1 to find the threshold P2. However, this method has a low accuracy rate. To solve the above problem, the present application provides a threshold recognition method, and the working principle of the method is explained in detail below.

[0069] Please refer to Figure 4 , which is a flow chart of a threshold recognition method provided in the first embodiment of the present application, such as Figure 4 As shown, the method includes the following steps S210 to S250.

[0070] Step S210: planning a travel path, controlling the robot 1 to travel along the travel path, and heading for a target point outside the feature area. During the travel process, the feature points of obstacles to be screened are marked according to the travel state of the robot 1.

[0071] Among them, the characteristic area is the closed area where the exit is the threshold.

[0072] In this step, when the robot 1 is trapped in a characteristic area, the control unit 100 can plan a travel path on the map corresponding to the characteristic area. Figure 5As shown, when robot 1 is trapped in Figure 3 When within the characteristic area K1, the control unit 100 can plan a travel path Q1 or a travel path Q2. Both the travel path Q1 and the travel path Q2 are used to connect the current position of the robot 1 and the target point F1 outside the characteristic area K1. The travel path Q2 passes through the threshold P2 and the travel path Q2 passes through the glass P3.

[0073] After the travel path is successfully planned, the control unit 100 can control the robot 1 to travel along the travel path and go to the target point outside the feature area. When the robot 1 travels along the travel path, the control unit 100 can determine whether the robot 1 encounters an obstacle based on the travel state of the robot 1. When it is determined that the robot 1 encounters an obstacle, a feature point of an obstacle to be screened is marked at the location of the obstacle on the map corresponding to the feature area. For example, Figure 6 As shown, when the control unit 100 successfully plans the travel path Q1, the control unit 100 can control the robot 1 to move along the travel path Q1 and go to the target point F1 outside the feature area K1; when the robot 1 moves along the travel path Q1, the control unit 100 can determine whether the robot 1 encounters an obstacle according to the travel state of the robot 1. When it is determined that the robot 1 encounters an obstacle, a feature point F3 of an obstacle to be screened is marked at the location of the obstacle. For example, Figure 6 As shown, when the control unit 100 successfully plans the travel path Q2, the control unit 100 can control the robot 1 to move along the travel path Q2 and go to the target point F1 outside the feature area K1; when the robot 1 moves along the travel path Q2, the control unit 100 can determine whether the robot 1 encounters an obstacle based on the travel status of the robot 1. When it is determined that the robot 1 encounters an obstacle, a feature point F3 of an obstacle to be screened is marked at the location of the obstacle.

[0074] After the feature points of the obstacles to be screened are marked successfully, the control unit 100 may continue to execute the following step S220.

[0075] Step S220: After marking at least one feature point of an obstacle to be screened, replan the path to continue to the target point.

[0076] As described in step S210, the feature points of obstacles to be screened are marked, indicating that the robot 1 encounters an obstacle and cannot travel to the target point according to the current travel path. Therefore, in this step, after the control unit 100 marks the feature points of obstacles to be screened, it can bypass the marked feature points of obstacles to be screened, and regenerate a travel path on the map of the feature area. After the generation is completed, the control unit 100 controls the robot 1 to travel according to the regenerated travel path and go to the target point outside the feature area. During the travel process, the control unit 100 still needs to determine whether the robot 1 encounters an obstacle according to the travel state of the robot 1. If the control unit 100 determines that the robot 1 encounters an obstacle, it marks a feature point of obstacles to be screened at the location of the obstacle on the map corresponding to the feature area. After the marking is successful, the control unit 100 continues to bypass all the marked feature points of obstacles to be screened, and generates a travel path on the map of the feature area. After the generation is completed, the control unit 100 controls the robot 1 to travel according to the regenerated travel path and go to the target point outside the feature area. This process is repeated until the marked obstacle feature points to be screened and the laser visible obstacles such as walls completely close the feature area, making it impossible for the control unit 100 to plan a path to the target point. Figure 7 As shown, it is a schematic diagram of the map corresponding to the characteristic area K1 when the control unit 100 cannot plan a travel path to the target point F1; Figure 7 It can be seen that when the control unit 100 is unable to plan a path to the target point F1, the obstacle feature point F3 to be screened and the laser visible obstacles such as the wall P1 completely close the feature area K1.

[0077] When the control unit 100 is unable to plan a path to the target point, the following step S230 may be continued to be executed.

[0078] Step S230: Control the robot 1 to move along the edge in the feature area to obtain an edge path.

[0079] In this step, the control unit 100 can control the robot 1 to move along the outer contour of the feature area in the feature area according to the map of the feature area. During the movement, the control unit 100 can try to plan the path to the target point again. After the planning is successful, the robot 1 is controlled to move along the planned path. If the robot 1 successfully escapes from the feature area after the movement, the control unit 100 stops executing the threshold recognition method in this embodiment; if the robot 1 fails to escape from the feature area after the movement, the robot 1 continues to be controlled to move along the outer contour of the feature area. In the process of the robot 1 moving along the outer contour of the feature area, the above process is continuously repeated until the robot 1 moves along the outer contour of the feature area for one circle.

[0080] After the robot 1 travels along the outer contour of the feature area for one cycle, a path along the edge is generated on the map of the feature area, and the control unit 100 can then continue to execute the following step S240. Figure 8 As shown, after the robot 1 travels one circle in the feature area K1 along the outer contour of the feature area K1 , an edge path V1 is generated on the map of the feature area.

[0081] Step S240: Filtering out target obstacle feature points from a plurality of obstacle feature points to be filtered according to the edge path and / or the obstacle feature points to be filtered.

[0082] Since the position of the obstacle feature point to be screened can represent the position of the obstacle in the feature area, in this step, the control unit 100 can screen multiple obstacle feature points to be screened according to the edge path and / or the obstacle feature point to be screened, and screen out the target obstacle feature point that can represent the threshold position from the multiple obstacle feature points to be screened; after the screening is successful, the control unit 100 can continue to execute the following step S250. For example, Fig. 9 As shown, according to the edge path V1 and the obstacle feature point F3 to be screened, after the obstacle feature point F3 to be screened is screened, the target obstacle feature point J1 representing the position of the threshold P2 can be screened out.

[0083] Step S250: Identify the position of the threshold in the feature area according to the target obstacle feature points.

[0084] In this step, after successfully screening out the target obstacle feature points, the control unit 100 can directly use the location of the target obstacle feature points as the location of the threshold. At this time, the location of the threshold in the feature area is successfully identified.

[0085] Compared with the prior art, the method in the present application can quickly and accurately identify the exit position in the low-lying area, ensuring that the robot 1 can quickly escape from the low-lying area.

[0086] In one embodiment, the control unit 100 may mark the feature points of obstacles to be screened in the following manner:

[0087] (1) The control unit 100 determines whether the robot 1 has slipped according to the traveling state of the robot 1; if the judgment result shows that the robot 1 has slipped, it indicates that the robot 1 has encountered an obstacle. At this time, the control unit 100 can mark a feature point of an obstacle to be screened at the position where the robot 1 has slipped. For example, when the robot 1 has slipped, a y2cm wide A-type feature point of an obstacle to be screened can be marked at a position y1cm in front of the robot 1. Y1 can be 2 to 6, and y2 can be 11 to 30.

[0088] In one embodiment, the control unit 100 can determine whether the robot 1 slips by:

[0089] like Figure 1 As shown, the robot 1 is provided with a detection unit 200, which includes a first detection unit 210 and a second detection unit 220. The first detection unit 210 is used to detect the theoretical travel distance of the robot 1, and the second detection unit 220 is used to detect the actual travel distance of the robot 1. Exemplarily, the first detection unit 210 may be an odometer, etc., and the second detection unit 220 may be a sensor, etc.

[0090] During the movement of the robot 1, the first detection unit 210 detects the theoretical travel distance of the robot 1 in real time, and sends the detection result to the control unit 100 in real time. At the same time, during the movement of the robot 1, the second detection unit 220 detects the posture information of the robot 1 in real time, determines the actual travel distance of the robot 1 according to the posture information, and then sends the actual travel distance of the robot 1 to the control unit 100. After receiving the theoretical travel distance of the robot 1 sent by the first detection unit 210 and the actual travel distance of the robot 1 sent by the second detection unit 220, the control unit 100 can perform a difference operation on the theoretical travel distance of the robot 1 and the actual travel distance of the robot 1, and obtain a characteristic parameter after the difference operation. After that, the control unit 100 can compare the characteristic parameter with the slip detection threshold, and then determine whether the robot 1 slips according to the comparison result. Specifically, if the comparison result shows that the characteristic parameter is greater than the slip detection threshold, it is determined that the robot 1 slips; if the comparison result shows that the characteristic parameter is less than the slip detection threshold, it is determined that the robot 1 does not slip. For example, when the theoretical travel distance of robot 1 is 35 cm, the actual travel distance of robot 1 is 0.5 cm, the characteristic parameter calculated by robot 1 should be 34.5 cm. At this time, if the slip detection threshold is 20 cm, it can be determined that robot 1 has slipped.

[0091] (2) The control unit 100 determines whether the robot 1 has collided according to the traveling state of the robot 1; if the judgment result shows that the robot 1 has collided, it indicates that the robot 1 has encountered an obstacle, and the control unit 100 can mark a feature point of an obstacle to be screened at the position where the robot 1 has collided. For example, when the robot 1 has collided, a y3cm wide B-type feature point of an obstacle to be screened can be marked at a position y1cm in front of the robot 1; y1 can be 2 to 6, y3 can be 5 to 15, and y3 is not equal to y2.

[0092] (3) The control unit 100 determines whether the robot 1 looks down according to the traveling state of the robot 1; if the judgment result shows that the robot 1 looks down, it indicates that the robot 1 encounters an obstacle such as a step, and the control unit 100 can mark a feature point of an obstacle to be screened at the position where the robot 1 looks down. For example, when the robot 1 looks down, a y3cm wide B-type feature point of an obstacle to be screened can be marked at a position y1cm in front of the robot 1; y1 can be 2 to 6, y3 can be 5 to 15, and y3 is not equal to y2.

[0093] In one embodiment, the robot 1 may be provided with a downward-looking sensor, which is used to detect the distance between the bottom surface of the robot 1 and the surface of the characteristic area.

[0094] When the robot 1 is moving, the downward-looking sensor detects the first distance between the bottom surface of the robot 1 and the surface of the characteristic area in real time, and sends the detection result to the control unit 100 in real time; after receiving the first distance between the bottom surface of the robot 1 and the surface of the characteristic area detected by the downward-looking sensor, the control unit 100 compares the first distance with the set threshold value, and determines whether the robot 1 is looking downward according to the comparison result. Specifically, if the first distance is greater than the set threshold value, it is determined that the robot 1 is looking downward; if the first distance is less than the set threshold value, it is determined that the robot 1 is not looking downward.

[0095] In one embodiment, after marking the obstacle feature points to be screened according to the above method, multiple obstacle feature points to be screened can be screened according to the screening conditions formed by the edge path and / or the obstacle feature points to be screened, and the target obstacle feature points can be screened out from the multiple obstacle feature points to be screened. The screening conditions include a combination of one or more of the position relationship between the edge path and the multiple obstacle feature points to be screened, the distance relationship between the edge path and the multiple obstacle feature points to be screened, the obstacle situation around each obstacle feature point to be screened, and the length of the first straight line formed by sequentially connecting the multiple obstacle feature points to be screened.

[0096] Specifically, because the robot 1 will collide with an obstacle or look down at the feature point of the obstacle to be screened of type B, the location of the feature point of the obstacle to be screened of type B is indeed not accessible to the robot 1, and the threshold cannot be located at the location of the feature point of the obstacle to be screened of type B. Based on this, in this embodiment, when screening the feature points of the obstacle to be screened, it is not necessary to screen the feature points of the obstacle to be screened of type B.

[0097] However, in practice, although the robot 1 may slip at the threshold, it may also slip at other positions, so not all the feature points of the obstacles to be screened in category A can represent the position of the threshold. Therefore, in this embodiment, the feature points of the obstacles to be screened in category A need to be screened, and the target obstacle feature points that can represent the position of the threshold are screened from the feature points of the obstacles to be screened in category A.

[0098] The following is a detailed explanation of the principle of screening the characteristic points of Class A obstacles:

[0099] (1) The control unit 100 selects a target obstacle feature point from a plurality of Class A obstacle feature points to be screened according to the positional relationship between the edge path and the plurality of Class A obstacle feature points to be screened.

[0100] Since the thresholds such as the door sill must be located outside the edge path, in this embodiment, when screening the A-type obstacle feature points to be screened, the control unit 100 can first screen out candidate obstacle feature points located outside the edge path from multiple A-type obstacle feature points to be screened. After the screening is successful, the target obstacle feature points are determined based on the candidate obstacle feature points.

[0101] It is worth noting that when the control unit 100 determines the target obstacle feature point based on the candidate obstacle feature point, the candidate obstacle feature point can be directly used as the target obstacle feature point. Alternatively, the control unit 100 can further screen the candidate obstacle feature points using other screening methods, and the target obstacle feature point can be obtained after screening.

[0102] (2) The control unit 100 selects a target obstacle feature point from the plurality of Class A obstacle feature points to be screened according to the distance relationship between the edge path and the plurality of Class A obstacle feature points to be screened.

[0103] Because the robot 1 will definitely tilt when passing through the threshold, the tilt angle is about 2.5° to 5°. Therefore, in this embodiment, when screening the feature points of the Class A obstacles to be screened, the control unit 100 can first obtain the posture information of the robot 1 on the edge path through the second detection unit 220. After successful acquisition, the control unit 100 parses the posture information of the robot 1 on the edge path, and after parsing, it determines whether each coordinate point of the robot 1 on the edge path is in a tilted state (the edge path is composed of multiple coordinate points). Specifically, if it is found after parsing that the robot 1 has a tilt angle at a certain coordinate point on the edge path and the tilt angle is greater than the set threshold, it indicates that the robot 1 is in a tilted state at the coordinate point on the edge path; if it is found after parsing that the robot 1 has a tilt angle at a certain coordinate point on the edge path but the tilt angle is less than the set threshold, it indicates that the robot 1 is not in a tilted state at the coordinate point on the edge path.

[0104] After successfully determining the inclination state of each coordinate point of the robot 1 along the edge path, the control unit 100 can further determine whether there is a target path on the edge path that is longer than a first preset length and is formed when the robot 1 is in an inclined state along the edge. If the judgment result shows that a target path exists, the target path is intercepted from the edge path. Among them, the target path is a portion of the edge path, and the target path may be formed when the robot 1 follows the edge of a threshold, or it may be formed when the robot 1 follows the edge of other obstacles. Specifically, the above-mentioned target path will also be generated when the robot 1 follows the edge of a fan base. Exemplarily, the first preset length may be 40 to 100 cm.

[0105] After the target path is intercepted successfully, the control unit 100 calculates the distance between multiple Class A obstacle feature points to be screened and the target path. After the calculation is successful, the candidate obstacle feature points whose distance from the target path is less than the first preset distance are screened from the multiple Class A obstacle feature points to be screened; after the screening is successful, the control unit 100 can determine the target obstacle feature points based on the candidate obstacle feature points. Specifically, the first preset distance can be the distance of half the body of the robot 1; illustratively, the first preset distance can be 15 to 20 cm.

[0106] It is worth noting that there is a one-to-one correspondence between the candidate obstacle feature points and the target path. The candidate obstacle feature points can reflect the position of a target obstacle, and the target path corresponding to the candidate obstacle feature points is the edge path generated after the robot 1 follows the edge of the target obstacle. Because the target path may be the edge path formed after the robot 1 follows the edge of the threshold, the candidate obstacle feature points corresponding to the target path can reflect the position of the threshold to a certain extent.

[0107] When determining the target obstacle feature point according to the candidate obstacle feature point, the control unit 100 may directly use the candidate obstacle feature point as the target obstacle feature point. Alternatively, the control unit 100 may further screen the candidate obstacle feature points using other screening methods, and the target obstacle feature point may be obtained after screening.

[0108] Obstacles and uneven surfaces of feature areas can cause the robot 1 to tilt. In this embodiment, a tilt angle is introduced, and the robot 1 is considered to be in a tilted state only when the tilt angle of the robot 1 is greater than a set threshold. This eliminates the tilt of the robot 1 caused by uneven surfaces of feature areas, thereby improving the accuracy of identifying threshold obstacles.

[0109] (3) The control unit 100 selects a target obstacle feature point from a plurality of category A obstacle feature points to be screened according to the conditions of obstacles around each category A obstacle feature point to be screened.

[0110] Among them, Figure 1 As shown, the detection unit 200 further includes a laser detection unit 230, and the laser detection unit 230 is used to detect the obstacle situation around each feature point of the Class A obstacle to be screened.

[0111] Because there are no obstacles visible to the laser around the threshold, in this embodiment, when screening the A-type obstacle feature points to be screened, the laser detection unit 230 can be used to first detect whether there are laser visible obstacles around each A-type obstacle feature point to be screened. After the detection is successful, candidate obstacle feature points without laser visible obstacles around them are screened out from multiple A-type obstacle feature points to be screened. Exemplarily, the control unit 100 can first detect whether there are laser visible obstacles around rcm of each obstacle feature point to be screened through the laser detection unit 230. After the detection is successful, candidate obstacle feature points without laser visible obstacles around rcm are screened out from multiple A-type obstacle feature points to be screened. Exemplarily, r can be 10 to 20.

[0112] After the candidate obstacle feature points are successfully screened, the control unit 100 can determine the target obstacle feature points according to the candidate obstacle feature points.

[0113] It is worth noting that when the control unit 100 determines the target obstacle feature point based on the candidate obstacle feature point, the candidate obstacle feature point can be directly used as the target obstacle feature point. Alternatively, the control unit 100 can further screen the candidate obstacle feature points using other screening methods, and the target obstacle feature point can be obtained after screening.

[0114] (4) The control unit 100 selects a target obstacle feature point from the plurality of category A obstacle feature points to be screened according to the length of a first straight line formed by sequentially connecting the plurality of category A obstacle feature points to be screened.

[0115] Because the length of the threshold is greater than the set length, in this embodiment, when screening the obstacle feature points of Class A to be screened, multiple candidate obstacle feature points of Class A can be sequentially connected to form multiple first straight lines, and then the length of each first straight line is calculated. After the calculation is successful, the candidate obstacle feature points whose first straight line length is greater than the second preset length are screened from the candidate obstacle feature points of Class A. After the screening is successful, the target obstacle feature points are determined according to the candidate obstacle feature points. Exemplarily, the second preset length can be 40 to 100 cm.

[0116] It is worth noting that each first straight line only includes candidate obstacle feature points of type A, but does not include obstacle feature points of type B. When the control unit 100 determines the target obstacle feature points based on the candidate obstacle feature points, the candidate obstacle feature points can be directly used as the target obstacle feature points. Alternatively, the control unit 100 can further screen the candidate obstacle feature points using other screening methods, and the target obstacle feature points can be obtained after screening.

[0117] From the above content, it can be seen that the present application provides a variety of methods for screening feature points of Class A obstacles to be screened, and the screening methods are diverse; at the same time, it also ensures that the control unit 100 can accurately screen out obstacle feature points that can reflect the position of the threshold from the feature points of Class A obstacles to be screened, thereby improving the accuracy of threshold recognition.

[0118] Please refer to Fig.10 , which is a flow chart of a threshold recognition method provided in the second embodiment of the present application. Fig.10 As shown, the method includes the following steps S310 to S380.

[0119] Step S310: planning a travel path, controlling the robot 1 to travel along the travel path, and heading for a target point outside the feature area. During the travel process, the feature points of obstacles to be screened are marked according to the travel status of the robot 1.

[0120] Among them, the characteristic area is the closed area where the exit is the threshold.

[0121] In this step, the control unit 100 can plan a travel path based on the map of the stored feature area, and after the planning is successful, control the robot 1 to travel along the travel path and go to the target point outside the feature area. During the travel process, the control unit 100 determines whether the robot 1 has skidded, collided, or looked down based on the travel state of the robot 1; when the judgment result shows that the robot 1 has skidded, the position where the robot 1 has skidded is marked as a type A obstacle feature point to be screened; when the judgment result shows that the robot 1 has collided or looked down, the position where the robot 1 has collided or looked down is marked as a type B obstacle feature point to be screened.

[0122] After the feature points of the obstacles to be screened are marked successfully, the control unit 100 may continue to execute the following step S320.

[0123] Step S320: After marking at least one obstacle feature point to be screened, re-plan the travel path to continue to the target point.

[0124] In this step, after the control unit 100 marks a feature point of an obstacle to be screened, it can bypass the marked feature point of the obstacle to be screened and regenerate a travel path. After the generation is completed, the control unit 100 controls the robot 1 to travel along the regenerated travel path and go to the target point outside the feature area. During the travel process, the control unit 100 still needs to determine whether the robot 1 has slipped, collided, or looked down according to the travel state of the robot 1; if the control unit 100 determines that the robot 1 has slipped, a Class A feature point of an obstacle to be screened is marked at the position where the robot 1 has slipped on the map of the feature area; if the control unit 100 determines that the robot 1 has collided or looked down, a Class B feature point of an obstacle to be screened is marked at the position where the robot 1 has collided or looked down on the map of the feature area. After the marking is successful, the control unit 100 continues to bypass all the marked feature points of obstacles to be screened and generate a travel path. After the generation is completed, the control unit 100 controls the robot 1 to travel along the regenerated travel path and go to the target point outside the feature area. This process is repeated repeatedly until the marked feature points of the obstacles to be screened of type A, the feature points of the obstacles to be selected of type B, and the laser visible obstacles such as walls completely close the feature area, so that the control unit 100 cannot plan a path to the target point. When the control unit 100 cannot plan a path to the target point, the following step S230 can be continued.

[0125] For example, Fig.11 As shown, it is a schematic diagram of the map corresponding to the characteristic area K1 when the control unit 100 cannot plan a travel path to the target point F1; Fig.11 It can be seen that when the control unit 100 is unable to plan a path to the target point F1, the laser visible obstacles such as the feature point F4 of the type A obstacle to be screened, the feature point F5 of the type B obstacle to be screened and the wall P1 completely close the feature area K1.

[0126] Step S330: Control the robot 1 to move along the edge in the feature area to obtain an edge path.

[0127] In this step, the control unit 100 can control the robot 1 to move along the outer contour of the feature area within the feature area; after the robot 1 moves along the outer contour of the feature area for one circle, an edge path will be generated on the map of the feature area, and the control unit 100 can then continue to execute the following step S340. For example, Fig.12 As shown, after the robot 1 travels one circle in the feature area K1 along the outer contour of the feature area K1 , an edge path V2 is generated on the map of the feature area.

[0128] Step S340: screening multiple obstacle feature points to be screened according to the distance relationship between the edge path and the multiple obstacle feature points to be screened.

[0129] In this step, the control unit 100 can screen multiple Class A obstacle feature points to be screened based on the distance relationship between the edge path and the multiple Class A obstacle feature points to be screened. Specifically, the screening method is to first intercept the target path from the edge path; wherein the length of the target path is greater than a first preset length, and the target path is formed when the robot 1 is in a tilted state. After successful interception, the distance between the multiple Class A obstacle feature points to be screened and the target path is calculated. After successful calculation, the Class A candidate obstacle feature points whose distance to the target path is less than the first preset distance are screened from the multiple Class A obstacle feature points to be screened. After successful screening, the control unit 100 can proceed to execute the following step S350. Exemplary, such as Fig.13 As shown, the control unit 100 intercepts multiple target paths V3 from the edge path V2. After successful interception, the control unit 100 calculates the distance between multiple Class A obstacle feature points F4 and each target path V3. After successful calculation, candidate obstacle feature points whose distance to the target path V3 is less than the first preset distance are screened out from the multiple Class A obstacle feature points F4 to be screened.

[0130] Step S350: screening the plurality of obstacle feature points to be screened according to the positional relationship between the edge path and the plurality of obstacle feature points to be screened.

[0131] In this step, the control unit 100 can select the type A obstacle feature points located outside the edge path from the multiple type A obstacle feature points to be screened according to the positional relationship between the edge path and the multiple type A obstacle feature points to be screened; after the screening is successful, the control unit 100 can continue to execute the following step S360. Fig.13 As shown, the control unit 100 can filter out the feature points of the class A obstacles located outside the edge path V2 from the plurality of feature points F4 of the class A obstacles to be filtered.

[0132] It is worth noting that the feature points of obstacles of type A to be screened in this step refer to the feature points of obstacles of type A to be screened screened in step S340.

[0133] Step S360: screening multiple obstacle feature points to be screened according to the obstacle conditions around each obstacle feature point to be screened.

[0134] In this step, the control unit 100 can first detect whether there are laser visible obstacles around each Class A obstacle feature point to be screened through the laser detection unit 230, and after successful detection, select Class A obstacle feature points without laser visible obstacles around them from multiple Class A obstacle feature points. Fig.13As shown, the control unit 100 can detect whether there are laser visible obstacles around each Class A obstacle feature point F4 to be screened through the laser detection unit 230. After successful detection, the Class A obstacle feature points without laser visible obstacles around them are screened out from multiple Class A obstacle feature points F4.

[0135] It is worth noting that the feature points of the obstacles of type A to be screened in this step refer to the feature points of the obstacles of type A to be screened screened in step S360.

[0136] Step S370: screening the plurality of obstacle feature points to be screened according to the length of the first straight line formed by sequentially connecting the plurality of obstacle feature points to be screened, and obtaining the target obstacle feature point after screening.

[0137] In this step, the control unit 100 can calculate the length of the first straight line formed by connecting multiple Class A obstacle feature points in sequence, and after the calculation is successful, select the target obstacle feature points whose first straight line length is greater than the second preset length from the multiple Class A obstacle feature points. Fig.14 As shown, the control unit 100 executes steps S340 to S360, and can screen out the A-type obstacle feature points F6 and F7 from the multiple A-type obstacle feature points F4. At this time, the control unit 100 can calculate the length of the first straight line formed by connecting the A-type obstacle feature points F6, and calculate the length of the first straight line formed by connecting the A-type obstacle feature points F7 in sequence. After the calculation is successful, the target obstacle feature point F7 whose length of the first straight line is greater than the second preset length is screened out.

[0138] It is worth noting that the feature points of the obstacles of type A to be screened in this step refer to the feature points of the obstacles of type A to be screened screened in step S360.

[0139] Step S380: Identify the position of the threshold in the feature area according to the target obstacle feature points.

[0140] In this step, after successfully screening out the target obstacle feature points, the control unit 100 can directly use the location of the target obstacle feature points as the location of the threshold. At this time, the location of the threshold in the feature area is successfully identified.

[0141] It can be seen from the above that in this embodiment, when the robot 1 is trapped in a low-lying area, the robot 1 can quickly and accurately identify the exit position in the low-lying area, which fully ensures that the robot 1 can quickly escape from the low-lying area. Compared with the prior art that searches for an exit by blind trial, the method in the present application reduces the number of attempts of the robot 1, so that the robot 1 can more efficiently determine the exit position of the low-lying area. In addition, when searching for an exit by blind trial in the prior art, the robot 1 has a large number of attempts, which increases the number of collisions between the robot 1 and obstacles, and easily causes damage to the body of the robot 1. In the present application, the number of attempts of the robot 1 when searching for an exit is reduced, and the body of the robot 1 is effectively protected.

[0142] In one embodiment, after the control unit 100 identifies the position of the threshold in the feature area, it can control the robot 1 to escape from the obstacle. Specifically, the control unit 100 can control the robot 1 to escape from the obstacle according to the obstacle escaping method of the robot 1 provided in the following embodiment. The working principle of the obstacle escaping method of the robot 1 is explained in detail below.

[0143] Please refer to Fig.15 , which is a flow chart of the obstacle escaping method of the robot 1 provided in one embodiment of the present application. Fig.15 As shown, the method includes the following steps S410 to S430.

[0144] Step S410: Control the robot 1 to execute the threshold recognition method provided in the above embodiment to recognize the position of the threshold in the feature area.

[0145] In this step, when the robot 1 is trapped in the characteristic area, the control unit 100 may first execute the threshold recognition method provided in the above embodiment to recognize the position of the threshold in the characteristic area. After successful recognition, the control unit 100 may continue to execute the following step S420. Fig.16 As shown, the control unit 100 can use the threshold recognition method provided in the above embodiment to recognize the position of the threshold P2 in the feature area K1.

[0146] Step S420: Generate an escape path according to the position of the threshold in the feature area.

[0147] In this step, after the control unit 100 identifies the location of the threshold in the characteristic area, it can generate an escape path on the map of the characteristic area. After generating the escape path, the control unit 100 can continue to perform the following step S430. Fig.16As shown, the control unit 100 can generate an escape path V4; the planned escape path V4 passes through the threshold P2 from the current position of the robot 1 and finally points to the target point F1 outside the feature area.

[0148] Step S430: Control the robot 1 to move along the escape path to escape.

[0149] In this step, the control unit 100 can control the robot 1 to move along the escape path to escape. When moving along the escape path, if the robot 1 can cross the threshold, it indicates that the escape is successful; if the robot 1 cannot cross the threshold, it indicates that the escape is unsuccessful.

[0150] In one embodiment, the control unit 100 can control the robot 1 to move along the travel path to escape from trouble in the following manner:

[0151] The control unit 100 first controls the robot 1 to move along the escape path at the first travel speed. During the travel, the control unit 100 determines in real time whether the distance between the robot 1 and the threshold is less than the second preset distance; if the judgment result shows that the distance between the robot 1 and the threshold is not less than the second preset distance, the robot 1 is controlled to continue to move along the escape path at the first travel speed; if the judgment result shows that the distance between the robot 1 and the threshold is less than the second preset distance (when the robot 1 moves to Fig.16 When the robot 1 is at point H1 in the figure, the distance between the robot 1 and the threshold is less than the second preset distance), the control unit 100 increases the travel speed of the robot 1 to the second travel speed. After the speed increase is successful, the control unit 100 continues to control the robot 1 to travel along the escape path at the second travel speed. When the robot 1 travels along the escape path at the second travel speed, the control unit 100 can determine in real time whether the robot 1 has slipped; if it is found that the robot 1 has slipped after the determination, it indicates that the threshold is impassable, and the control unit 100 can generate impassable identification information at the threshold in the feature area map.

[0152] It is worth noting that there may be multiple thresholds in a feature area. In this embodiment, the control unit 100 can generate an escape path for each threshold. After the generation is completed, the robot 1 is controlled to move along each escape path in turn. When the robot 1 moves to the second preset distance from the threshold along each travel path, the travel speed of the robot 1 can be increased. After the travel speed is increased, the robot 1 is controlled to continue to move along the escape path. During the travel process, the control unit 100 determines in real time whether the robot 1 slips. If it is found that the robot 1 slips after judgment, the control unit 100 can generate an inaccessible identification information at the threshold in the feature area map. Finally, if inaccessible identification is generated at all thresholds in the map corresponding to the feature area, it means that the robot 1 still cannot escape from the feature area smoothly. At this time, the control unit 100 can control the robot 1 to alarm. Send a help message to the user through the alarm, so that the user can help the robot 1 escape from the feature area.

[0153] When the robot 1 follows the escape path to escape, if the robot 1 slips, the control unit 100 will control the robot 1 to stop escaping. To this end, in this embodiment, the control unit 100 can increase the slip detection threshold of the robot 1. By increasing the slip detection threshold, the control unit 100 will no longer easily detect that the robot 1 has slipped, thereby preventing the occurrence of slippage from affecting the escape of the robot 1. For example, if the initial value of the slip detection threshold is 35 cm, the slip detection threshold can be increased to 50 cm.

[0154] In this embodiment, when the threshold is impassable, the control unit 100 generates impassable identification information at the threshold in the map of the characteristic area. In this way, the next time the robot 1 is trapped in the characteristic area and looking for an exit, the area marked with impassable identification information can be directly filtered out, eliminating the need to try again to see if the area is impassable.

[0155] In several embodiments provided in the present application, the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and a part of a module, a program segment or a code contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0156] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0157] If the function is implemented in the form of a software function 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 the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

Claims

1. A threshold recognition method, applied to a robot, characterized in that: The method for identifying the threshold includes: Planning a travel path, controlling the robot to travel along the travel path, heading for a target point outside a characteristic area, and marking characteristic points of obstacles to be screened according to the travel state of the robot during the travel process; wherein the characteristic area is a closed area whose exit is a threshold; After marking at least one obstacle feature point to be screened, replan the travel path to continue to the target point; Controlling the robot to move along the edge in the feature area to obtain an edge path; Screening out a target obstacle feature point from a plurality of obstacle feature points to be screened according to the edge path and / or the obstacle feature point to be screened; According to the target obstacle feature point, the position of the threshold in the feature area is identified.

2. The method for identifying a threshold according to claim 1, characterized in that: During the traveling process, marking feature points of obstacles to be screened according to the traveling state of the robot includes: During the movement, judging whether the robot slips according to the movement state of the robot; If the robot slips, the position where the robot slips is marked as the feature point of the obstacle to be screened.

3. The method for identifying a threshold according to claim 1, characterized in that: The step of selecting a target obstacle feature point from a plurality of obstacle feature points to be screened according to the edge path and / or the obstacle feature point to be screened includes: The target obstacle feature point is screened out from a plurality of obstacle feature points to be screened according to the screening conditions formed by the edge path and / or the obstacle feature points to be screened; wherein the screening conditions include a combination of one or more of a positional relationship between the edge path and the plurality of obstacle feature points to be screened, a distance relationship between the edge path and the plurality of obstacle feature points to be screened, obstacle conditions around each obstacle feature point to be screened, and a length of a first straight line formed by sequentially connecting the plurality of obstacle feature points to be screened.

4. The method for identifying a threshold according to claim 3, characterized in that: The screening condition includes a positional relationship between the edge path and a plurality of obstacle feature points to be screened; the screening condition formed by the edge path and / or the obstacle feature points to be screened, screening out a target obstacle feature point from the plurality of obstacle feature points to be screened, includes: According to the positional relationship between the edge path and the plurality of obstacle feature points to be screened, a candidate obstacle feature point located outside the edge path is screened from the plurality of obstacle feature points to be screened; The target obstacle feature point is determined according to the candidate obstacle feature point.

5. The method for identifying a threshold according to claim 3, characterized in that: The screening condition includes a distance relationship between the edge path and a plurality of obstacle feature points to be screened; the screening condition formed by the edge path and / or the obstacle feature points to be screened, screening out a target obstacle feature point from the plurality of obstacle feature points to be screened, includes: Intercepting a target path from the edge path; wherein the target path is formed when the robot is in an inclined state, and the length of the target path is greater than a first preset length; Calculating the distances between a plurality of obstacle feature points to be screened and the target path, and screening out candidate obstacle feature points whose distances to the target path are less than a first preset distance from the plurality of obstacle feature points to be screened; The target obstacle feature point is determined according to the candidate obstacle feature point.

6. The method for identifying a threshold according to claim 3, characterized in that: The robot is provided with a laser detection unit; the screening condition includes the obstacle conditions around each obstacle feature point to be screened; the screening condition formed by the edge path and / or the obstacle feature point to be screened, screening out a target obstacle feature point from a plurality of obstacle feature points to be screened, comprises: The laser detection unit is used to detect whether there are laser visible obstacles around each feature point of the obstacle to be screened; According to the detection results, candidate obstacle feature points with no laser visible obstacles around them are selected from a plurality of obstacle feature points to be selected; The target obstacle feature point is determined according to the candidate obstacle feature point.

7. The method for identifying a threshold according to claim 3, characterized in that: The screening condition includes the length of a first straight line formed by sequentially connecting a plurality of feature points of obstacles to be screened; The step of selecting a target obstacle feature point from a plurality of obstacle feature points to be selected according to the selection conditions formed by the edge path and / or the obstacle feature points to be selected includes: Calculate the length of the first straight line formed by sequentially connecting multiple obstacle feature points to be screened; Selecting candidate obstacle feature points whose length of the first straight line is greater than a second preset length from a plurality of obstacle feature points to be selected; The target obstacle feature point is determined according to the candidate obstacle feature point.

8. A robot obstacle-crossing and escaping method, characterized in that: The robot's obstacle-crossing and escape method comprises: Controlling the robot to execute the threshold recognition method according to any one of claims 1 to 7 to recognize the position of the threshold in the feature area; Generate an escape path according to the position of the threshold in the characteristic area; The robot is controlled to move along the escape path to escape.

9. The robot obstacle escaping method according to claim 8, characterized in that: The controlling the robot to move along the escape path to escape comprises: Controlling the robot to move along the escape path at a first moving speed, and determining whether the distance between the robot and the threshold is less than a second preset distance during the moving process; If the distance between the robot and the threshold is less than the second preset distance, the robot is controlled to move along the escape path at a second travel speed; wherein the second travel speed is greater than the first travel speed.

10. The robot obstacle escaping method according to claim 9, characterized in that: The robot's obstacle-crossing and escape method also includes: When the robot moves along the escape path at the second travel speed, determining whether the robot slips; If the judgment result indicates that the robot has slipped, an inaccessible identification information is generated at the threshold.

Images