Robot movement direction planning method using obstacle points to set tangents

Abstract

The present application discloses a method for planning the movement direction of a robot using an obstacle point to set a tangent. The method includes step S1: when a laser radar scans only one obstacle point in front of the robot, a position point at a preset safety distance from the obstacle point on a line segment connecting the center of the laser radar and the obstacle point is selected as the center of the anti-collision circle, an anti-collision circle is set with the preset safety distance as the radius, and a tangent line passing through the obstacle point and tangent to the anti-collision circle is set as the positioning line scanned in front of the robot; a laser probe that supports rotation is provided in the laser radar, and the rotation center of the laser probe is the center of the laser radar; step S2: the robot sets the extension direction of the positioning line to the target planned movement direction, and then adjusts the movement direction to the target planned movement direction, and keeps from entering the anti-collision circle during the movement.

Description

Technical Field

[0001] The present application relates to the technical field of obstacle perception and navigation planning, and in particular to a method for planning the movement direction of a robot using obstacle points to set tangents. Background Art

[0002] For robots moving in two-dimensional ground, relevant technologies can only use the point cloud information of a single-line laser sensor as the real-time point cloud input for local path planning. In an environment where the distance to the target object is far or the number of point clouds collected about the target object's outline is small, the robot has insufficient perception of obstacles in space and does not simulate in advance the situation where the fuselage is close to the obstacle outline. Instead, it only performs rasterization processing on each laser point, which causes the robot to repeatedly collide with obstacles during navigation control and the robot's detour path is not smooth. Summary of the Invention

[0003] This application discloses a method for planning the movement direction of a robot using obstacle points to set tangents. The specific technical solution is as follows:

[0004] A robot movement direction planning method using an obstacle point to set a tangent, the robot movement direction planning method includes: step S1, when the laser radar scans only one obstacle point in front of the robot, on the line segment connecting the center of the laser radar and the obstacle point, select a position point at a preset safety distance from the obstacle point as the center of the anti-collision circle, set the anti-collision circle with the preset safety distance as the radius, and then set the tangent passing through the obstacle point and tangent to the anti-collision circle as the positioning line scanned in front of the robot, so that the extension line of the laser beam emitted at some scanning angles of the laser radar intersects with the positioning line; a laser probe that supports rotation is provided in the laser radar, and the rotation center of the laser probe is the center of the laser radar; step S2, the robot sets the extension direction of the positioning line as the target planning movement direction, and then adjusts the movement direction to the target planning movement direction, and keeps it from entering the anti-collision circle during the movement.

[0005] Compared to existing technologies, step S1 eliminates the obstruction of nearby obstacles in a local area by scanning only one obstacle point in front of the robot, facilitating movement direction planning using the positioning line scanned in front of the robot. In step S2, while the robot is moving in a certain direction, if it is configured not to enter the anti-collision circle and adjusts its direction, the robot may move along the arc of the anti-collision circle before and after the adjustment, making the robot's movement path smoother. Moreover, adjusting to a single direction does not further guide the robot to move back and forth between different contours of the same obstacle or return to the starting point.

[0006] Combining steps S1 and S2, it can be seen that since steps S1 and S2 are executed when the laser radar scans only one obstacle point in front of the robot, the robot movement direction planning method is suitable for scenarios where the robot is far away from the obstacle to be scanned, and the robot moves around the obstacle without touching it. When the laser radar scans only one obstacle point in front of the robot, an anti-collision circle is set to avoid collision with the obstacle, and the tangent line passing through the obstacle point and tangent to the anti-collision circle is set as the positioning line scanned in front of the robot. This can not only achieve basic positioning of the obstacle with fewer obstacle points, but also serve as a direction to walk along the obstacle contour without colliding with the obstacle, forming a smooth obstacle circumvention path, allowing the robot to smoothly avoid obstacles.

[0007] Furthermore, in step S2, before the robot adjusts its movement direction to the target planned movement direction, the robot starts from a reference position point and moves in the established movement direction to a position point that is a predetermined safety distance from the anti-collision edge position, and then adjusts its movement direction to the target planned movement direction; the anti-collision edge position is the edge position of the anti-collision circle on the side closest to the robot; when the laser radar scans only one obstacle point, the robot's position point is set as the reference position point. Therefore, before the robot adjusts its movement direction to the target planned movement direction, based on setting the anti-collision circle using only the scanned obstacle point, the robot is shortened to the scanned obstacle, but remains outside the anti-collision circle, thereby providing effective adjustment space for the robot's planned movement to bypass obstacles at close range.

[0008] Furthermore, in step S2, the method for adjusting the movement direction to the target planned movement direction includes: the robot sets a straight line passing through the anti-collision edge position and tangent to the anti-collision circle as a tangent boundary, so that the anti-collision circle and the robot are separated on both sides of the tangent boundary; then the robot is controlled to perform circular motion on one side of the tangent boundary so that the robot's movement path extends toward the target planned movement direction, and the robot is kept from crossing the tangent boundary to enter the anti-collision circle. Thus, with the tangent boundary as the boundary, the area on one side close to the robot is defined as a passable area, and the other side is defined as an impassable area. If the circular motion is performed within the passable area and the tangent boundary is regarded as the tangent of the circular motion, it is possible to walk along the tangent boundary or the outline of the obstacle, and successfully complete a detour of a limited distance without touching the obstacle.

[0009] Furthermore, the method for controlling the robot to perform circular motion on one side of the tangent boundary includes: the robot sets a position point in the opposite direction of the predetermined moving direction and at a predetermined safety distance from the anti-collision edge of the anti-collision circle as the starting point of the circular motion; the robot sets the extension direction of the tangent boundary as the tangent direction of the circular motion, and sets the radius of the circular motion to be equal to the predetermined safety distance; wherein the tangent boundary and the target planned moving direction are within a predetermined angular range; and according to the tangent direction of the circular motion and the radius of the circular motion, the robot is controlled to perform circular motion from the starting point of the circular motion until the robot's moving direction changes to the target planned moving direction. Thus, the robot achieves the purpose of moving around obstacles.

[0010] Furthermore, during the circular motion, the robot turns at a preset turning angle around the center of the circular motion and moves along an arc until the robot's movement direction changes to the target planned movement direction; the tangent boundary is tangent to the arc; the maximum value of the preset turning angle is equal to the difference between 90 degrees and the lower limit of the preset angle range, and the minimum value of the preset turning angle is equal to the difference between 90 degrees and the upper limit of the preset angle range; wherein, the center of the circular motion is located in one of the laser projection directions of the laser radar when the robot is at the reference position point. This achieves bypassing obstacles, or making the side of the robot parallel to the straight outline of the obstacle, or making the side of the robot parallel to the tangent boundary.

[0011] Furthermore, if the only obstacle point scanned by the laser radar in front of the robot is located at the edge of one side of the scanned obstacle, the only obstacle point scanned by the laser radar in front of the robot is set as the reference obstacle point. Then, when the robot moves in the predetermined direction by executing steps S1 and S2, the number of obstacle points scanned by the laser radar on one side of the reference obstacle point increases, while the laser radar does not scan any obstacle points on the other side of the reference obstacle point. The location on the side where the additional obstacle points are scanned is the location of the obstacle scanned in front of the robot. Therefore, after the robot scans the reference obstacle point, it can determine the orientation characteristics of the obstacle scanned in front by moving within a certain distance along the predetermined direction of movement, thereby prompting the robot to avoid the obstacle in advance.

[0012] Furthermore, if the robot begins circular motion on one side of the tangent boundary by executing step S2, the center of the circular motion is set to the side corresponding to the line connecting the anti-collision edge position and the reference position point, where the corresponding side is the side where the scanned obstacle point increases; the extension line of the laser beam emitted by the laser radar on the side where the scanned obstacle point increases intersects the positioning line; and, when the robot is at the reference position point, the center of the circular motion is located in the direction of the laser projection of the laser radar. Based on the setting of the center of the circular motion, this technical solution fully considers the movement trend of the robot on the side where the scanned obstacle point increases, thereby more accurately controlling the robot to move in the target planned movement direction.

[0013] Furthermore, if the laser radar only scans one obstacle point in front of the robot, which is located at a non-edge position of the scanned obstacle, the only obstacle point scanned by the laser radar in front of the robot is set as the reference obstacle point. Then, when the robot moves in a predetermined direction by executing steps S1 and S2, the number of obstacle points scanned by the laser radar on both sides of the reference obstacle point increases. Moreover, the projection points of the obstacle points scanned on both sides of the reference obstacle point on the robot's travel plane are all located on tangents passing through the reference obstacle point and tangent to the anti-collision circle. The extension lines of the laser beams emitted by the laser radar on both sides of the reference obstacle point intersect with the positioning line. The non-edge position of the obstacle is located within the area surrounded by the obstacle's outline. This improves the robot's obstacle perception capability on the scanning plane without touching the obstacle, reducing the risk of collision during navigation control.

[0014] Furthermore, starting from the robot's position, rays are set at equal angles within the laser radar's detectable field of view, forming a grid. Scanned obstacle points occupy corresponding grids, with one ray corresponding to one grid. The angle between two adjacent grids is equal to the angle between the two corresponding rays, so that: among the obstacle points scanned by the laser radar to form the positioning line, the angle between two adjacent obstacle points is obtained by equally dividing the grid division angle. The laser radar's detectable field of view is formed by the line connecting the center of the laser radar and the two ends of the boundary of the laser radar's detection area, as well as the boundary of the laser radar's detection area. By setting rays at equal angles starting from the center of the laser radar, the computational effort is reduced.

[0015] Furthermore, the upper cover of the laser radar's protective cover is parallel to the robot's roof, which is in turn parallel to the robot's travel plane. The laser probe within the laser radar scans obstacles at various angles during rotation. The robot's travel plane is coplanar with the anti-collision circle. The laser radar's protective cover is provided with an exit aperture, through which the laser beam emitted by the laser probe scans the external environment of the laser radar. The arc enclosed by the laser beams emitted from either side of the exit aperture represents the boundary of the laser radar's detection area. This allows obstacle points detected by the laser radar's rotational scanning and their positional characteristics relative to the center of the laser radar to be used as obstacle information for a given plane.

[0016] Furthermore, when the obstacle where the reference obstacle point is located is a linear obstacle, the predetermined movement direction is perpendicular to the positioning line, and the positioning line is tangent to the boundary of the laser radar's detection area, so that the reference obstacle point becomes the scanned obstacle point closest to the reference position point; wherein the positioning line is located within a contour surface of the linear obstacle scanned by the laser radar, and the contour surface of the linear obstacle scanned by the laser radar is perpendicular to the robot's travel plane. This technical solution eliminates the obstruction of obstacles on at least one side of the robot in a local area through the reference obstacle point, and uses the positioning line scanned in front of the robot to plan the movement direction. This allows the robot to subsequently move along the edge in a direction parallel to the positioning line, i.e., along the contour of the linear obstacle, while using less obstacle point information. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of a method for planning the movement direction of a robot using obstacle points to set tangents, disclosed in an embodiment of the present application.

[0018] Figure 2 Another embodiment of the present application discloses a schematic diagram in which the laser radar scans only one obstacle point C1 in front of the robot, which is located within the contour surface of the obstacle #1.

[0019] Figure 3 This is a schematic diagram of another embodiment of the present application, which discloses that the laser radar scans only one obstacle point C2 in front of the robot, which is located at the edge of one side of the obstacle #2. Implementation Method

[0020] The technical solutions in the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings in the embodiments of the present invention. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the steps as sequential processes, many of the steps therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the steps can be rearranged. The process can be terminated when its operation is completed, but it can also have additional steps not included in the accompanying drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0021] As an embodiment, in order to use less laser reflection information to plan the movement direction of the robot in an environment where the distance to the target object is far or the number of point clouds about the target object contour collected is small, and to achieve real-time guidance of the robot to avoid obstacles with less point calculation when the laser radar rotates to scan the surrounding environment, this embodiment discloses a robot movement direction planning method using obstacle points to set tangents. The executor of the robot movement direction planning method is a robot fixed with a laser radar, including but not limited to a circular robot such as a sweeping robot that moves in an indoor environment. The laser radar installed on the robot is provided with a laser probe that supports rotation, and the rotation center of the laser probe is the center of the laser radar, such as Figure 2 and Figure 3 As shown in point L, point L is located at the center of the laser radar (the cylindrical structure located at the center of the robot in the figure). One end of the laser beam emitted by the laser probe in the laser radar during the rotation will be connected to form an arc line, and is formed in front of the robot through the exit hole of the laser radar, so as to sample the outline of the obstacle in front through the circumference. Preferably, the arc sandwiched by the laser beams emitted from both sides of the exit hole is the boundary of the detection area of ​​the laser radar.

[0022] See Figure 1 It can be seen that the robot movement direction planning method includes:

[0023] Step S1: When the laser radar scans only one obstacle point in front of the robot, which can also be considered as the robot scanning only one obstacle point in front of it, a point located a preset safety distance from the obstacle point on the line segment connecting the center of the laser radar and the obstacle point (the currently scanned obstacle point) is selected as the center of the collision avoidance circle. A collision avoidance circle is set with the preset safety distance as the radius. Specifically, the collision avoidance circle is set in the direction from the obstacle point to the center of the laser radar. A tangent line passing through the obstacle point and tangent to the collision avoidance circle is then set as the positioning line scanned in front of the robot, so that the extended line of the laser beam emitted at some scanning angles of the laser radar intersects with the positioning line. Then, step S2 is executed. Thus, step S1 converts the data corresponding to the obstacles that may be encountered during the robot's movement into the collision avoidance circle based on the fact that the robot scans only one obstacle point in front of it.

[0024] In an embodiment of the present application, the scanning of obstacle points is performed by a laser radar. For the convenience of calculation, the center position of the laser radar and the current position point of the robot are set to the same point, that is, the laser beam of the laser radar scanning the obstacle point is considered to be emitted from the current position point of the robot.

[0025] Corresponding to Figure 2 In the line segment connecting the center L of the laser radar and the obstacle point C1, a point M1 at a preset safety distance from the obstacle point C1 is selected as the center of the collision avoidance circle. The obstacle point C1 is located within the contour surface of the obstacle #1. The length of the line segment C1M1 is the preset safety distance. Figure 2 The tangent line C1A1 passing through the obstacle point C1 and tangent to the anti-collision circle is set as the positioning line scanned in front of the robot, so that the extension lines of the laser beams emitted at some scanning angles of the laser radar intersect with the positioning line C1A1; Figure 2In the example, the arc enclosed by ray LA1 (representing the laser beam emitted from the right side of the exit aperture) and ray LD1 (representing the laser beam emitted from the left side of the exit aperture) is configured as the boundary of the lidar's detection area. The moment when positioning line C1A1 intersects the lidar's detection area boundary is when the lidar has scanned only one obstacle, point C1, in front of the robot. Consequently, obstacle point C1 becomes the point of tangency between the lidar's detection area boundary and a contour surface of obstacle #1. This tangency point is the farthest obstacle point currently scanned by the robot and also the closest obstacle point to the reference position. By identifying this tangency point, a collision avoidance circle is set between obstacle #1 and the robot. This simulates the area occupied by the robot when its body is close to the contour surface of obstacle #1, i.e., the high-risk area for collision with the obstacle. Fewer laser point information can be used to prompt the robot to avoid entering the collision avoidance circle. The tangent direction of the collision avoidance circle or the positioning line can then be used to circumvent the obstacle. Since the obstacle point C1 becomes the tangent point between the boundary of the laser radar detection area and a contour surface of obstacle #1 and the tangent point is the farthest obstacle point currently scanned by the robot, the obstacle point C1 is relatively close to the obstacle points on both sides (for example, Figure 2 Obstacle points A1, B1, and D1 within the contour surface of obstacle #1 are closer to the robot's current position and pose a greater risk of collision. Therefore, this embodiment uses the scanning effect of the laser radar to obtain obstacle point C1 in advance and set an anti-collision circle and its tangent, thereby eliminating the obstruction of the robot to the obstacles on both sides of obstacle point C1 in a local area, so that the positioning line scanned in front of the robot can be used to plan the movement direction.

[0026] Corresponding to Figure 3 In the line segment connecting the center L of the same laser radar and the obstacle point C2, the point M2 at a preset safety distance from the obstacle point C2 is selected as the center of the collision avoidance circle. The obstacle point C2 is located at the right edge of the obstacle #2. The length of the line segment C2M2 is the preset safety distance. Figure 3 The tangent line C2A2 passing through the obstacle point C2 and tangent to the anti-collision circle is set as the positioning line scanned in front of the robot, so that the extension lines of the laser beams emitted at some scanning angles of the laser radar intersect with the positioning line C2A2; Figure 3In the example, the arc enclosed by ray LA2 (representing the laser beam emitted from the right side of the exit aperture) and ray LD2 (representing the laser beam emitted from the left side of the exit aperture) is configured as the boundary of the lidar's detection area. The moment when positioning line C2A2 intersects the lidar's detection area boundary is when the lidar has scanned only one obstacle, C2, in front of the robot. Consequently, obstacle C2 becomes the point of tangency between the lidar's detection area boundary and the edge of obstacle #2. This point of tangency is the farthest obstacle currently scanned by the robot and also the closest obstacle to the reference position. By identifying this point of tangency, a collision avoidance circle is established between obstacle #2 and the robot. This simulates the area occupied by the robot when it approaches the edge of obstacle #2, i.e., the high-risk area for collision with the obstacle. A laser point can be used to warn the robot not to enter the collision avoidance circle. The robot can then use the tangent direction of the collision avoidance circle or the positioning line to circumvent the obstacle. Since obstacle point C2 becomes the tangent point between the boundary of the laser radar detection area and a contour surface of obstacle #2 and this tangent point is the farthest obstacle point currently scanned by the robot within obstacle #2, obstacle point C2 is located at the right side of the obstacle point (e.g. Figure 3 Obstacle points A2 and B2 within the contour surface of obstacle #2 are closer to the robot's current position and pose a greater collision risk. Therefore, this embodiment uses the scanning effect of the laser radar to obtain obstacle point C2 in advance and set an anti-collision circle and its tangent. This eliminates the obstruction of the obstacle points on one side to the robot in the local area through obstacle point C2, making it easier to use the positioning line scanned in front of the robot to plan the movement direction.

[0027] Step S2: The robot sets the extension direction of the positioning line as the target planned movement direction, then adjusts the movement direction to the target planned movement direction, and maintains the robot from entering the anti-collision circle during the movement process. In the process of adjusting the movement direction to the target planned movement direction, the robot adjusts the direction according to the planned movement mode to smoothly bypass obstacles without frequently colliding with obstacles. The generated path does not fold back and forth, but presents a smooth trajectory. In the process of the robot moving in a certain direction, if it is configured not to enter the anti-collision circle and adjusts the direction, the robot may walk along the arc line of the anti-collision circle before and after the direction adjustment, making the robot's movement path smoother and smoother. Moreover, adjusting to a single direction will not further guide the robot to reciprocate between different contour surfaces of the same obstacle or return to the walking starting point. Specifically, the robot extends along a specific direction but does not revolve around the anti-collision circle.

[0028] Preferably, the extending direction of the positioning line is as follows Figure 2As shown by the arrow ON1 pointing to the left, the extension direction of the positioning line C1A1 is parallel to the ray ON1. In order to avoid obstacle #1, the robot needs to adjust its moving direction to the target planned moving direction (pointing by the arrow ON1). You can also set the extension direction of the positioning line to Figure 2 The arrow on ON1 points in the opposite direction.

[0029] Preferably, the extending direction of the positioning line is as follows Figure 3 As shown by the arrow ON2 pointing to the left, the extension direction of the positioning line C2A2 is parallel to the ray ON2. In order to avoid obstacle #2, the robot needs to adjust its moving direction to the target planned moving direction (pointing by the arrow ON2). You can also set the extension direction of the positioning line to Figure 3 The arrow ON2 points in the opposite direction.

[0030] Combining steps S1 and S2, it can be seen that since steps S1 and S2 are executed when the laser radar scans only one obstacle point in front of the robot, the robot movement direction planning method is suitable for scenarios where the robot is far away from the obstacle to be scanned, and the robot moves around the obstacle without touching it. When the laser radar scans only one obstacle point in front of the robot, an anti-collision circle is set to avoid collision with the obstacle, and the tangent line passing through the obstacle point and tangent to the anti-collision circle is set as the positioning line scanned in front of the robot. This can not only achieve basic positioning of the obstacle with fewer obstacle points, but also serve as a direction to walk along the obstacle contour without colliding with the obstacle, forming a smooth obstacle circumvention path, allowing the robot to smoothly avoid obstacles.

[0031] The relevant existing technologies are not yet able to propose an accurate motion control solution while ensuring that no collision occurs when there is less laser point cloud data. This embodiment uses the extension direction of the positioning line as the direction for the robot to avoid obstacles, which can eliminate in advance the obstruction to the robot caused by the two sides of an obstacle point scanned by the laser radar in step S1 (the built-in laser probe rotates and scans), thereby providing more convenience for the robot's obstacle avoidance operations such as avoiding obstacles and walking along the edges.

[0032] As an embodiment, in step S2, before the robot adjusts its movement direction to the target planned movement direction, the robot starts from a reference position point and moves in the established movement direction to a position point that is the preset safety distance from the anti-collision edge position, and then adjusts its movement direction to the target planned movement direction; wherein, when the laser radar scans only one obstacle point, the robot's position point is set as the reference position point; and the anti-collision edge position is the edge position on the side of the anti-collision circle close to the robot. Before the robot adjusts its movement direction to the target planned movement direction, based on setting the anti-collision circle using only the scanned obstacle point, the robot is shortened to the scanned obstacle, but remains outside the anti-collision circle, thereby providing effective adjustment space for the robot's planned movement to bypass the obstacle at close range.

[0033] The preset safety distance is preferably the width of the robot body. Figure 2 and 3 In the case of the circular robot shown, the preset safety distance is preferably the robot's body radius. The process of moving from the reference position to a position at the preset safety distance from the collision avoidance edge in the predetermined direction does not require a large number of laser points for calculation, thus reducing the computational effort and facilitating the robot's real-time navigation process.

[0034] Schematically, in Figure 2 In the figure, the robot will start from the position where only one obstacle point is scanned (reference position point), and move in a straight line to position point G1. The anti-collision edge position is position E1, and the length of line segment G1E1 is the preset safety distance. The predetermined moving direction is Figure 2 In the figure, it is represented as the direction of ray OP1, which is parallel to line segment LG1.

[0035] Schematically, in Figure 3 In the figure, the robot will start from the position where only one obstacle point is scanned (reference position point), and move in a straight line to position G2. The anti-collision edge position is position E2, and the length of line segment G2E2 is the preset safety distance. The predetermined moving direction is Figure 3 In the figure, it is represented as the direction of ray OP2, which is parallel to line segment LG2.

[0036] Based on the above embodiment, in step S2, the method of adjusting the moving direction to the target planned moving direction includes: the robot sets a straight line passing through the anti-collision edge position and tangent to the anti-collision circle as a tangent boundary, so that the anti-collision circle and the robot are separated on both sides of the tangent boundary; the two sides of the tangent boundary are Figure 2The two sides of the line E1F1 are the side pointed by E1G1 and the side pointed by E1M1. Figure 3 The two sides corresponding to the straight line E2F2 are the side pointed by E2G2 and the side pointed by E2M2.

[0037] The robot is then controlled to perform circular motion on one side of the tangent boundary, while preventing it from crossing the boundary and entering the collision avoidance circle. The area where the robot performs circular motion is located on the side of the tangent boundary closest to the robot, while obstacles are distributed on the other side of the boundary. Thus, the area near the robot is defined as a traversable zone, while the other side is defined as an impassable zone, based on the tangent boundary. By performing circular motion within the traversable zone and treating the tangent boundary as the tangent of the circular motion, the robot can follow the tangent boundary or the outline of the obstacle, successfully completing a limited detour without colliding with the obstacle.

[0038] As an embodiment, the method for controlling the robot to make circular motion on one side of the tangent boundary includes: the robot sets a position point in the opposite direction of the predetermined moving direction and at a predetermined safety distance from the anti-collision edge position of the anti-collision circle as the starting point of the circular motion, corresponding to Figure 2 Position point G1, or Figure 3 Position point G2; forming the position point at which the robot changes the predetermined moving direction.

[0039] The robot sets the extension direction of the tangent boundary to the tangent direction of the arc motion, and sets the radius of the arc motion to be equal to the preset safety distance; the preset safety distance is preferably the body radius of the robot. Since the tangent direction of the arc motion is limited, the arc motion will cause the robot to Figure 2 The robot starts to move in a clockwise arc based on the position point G1 shown in the figure. This arc motion will make the robot move in a clockwise arc. Figure 3 Based on the position G2 shown, the robot starts to move in a clockwise arc. The dotted arc lines in each figure show the schematic diagram of the robot's circular motion. Among them, the tangent boundary and the target planned moving direction are within a preset angle range; in order to achieve the purpose of inertial motion of an arc along the tangent boundary, the tangent direction of the circular motion and the target planned moving direction to be achieved can be limited to be within the preset angle range, but will not cross the tangent boundary to enter the anti-collision circle. Preferably, the upper limit value of the preset angle range is set to 10 degrees, and the lower limit value of the preset angle range is set to 0 degrees. Accordingly, the maximum value of the angle between the target planned moving direction and the predetermined moving direction is equal to 90 degrees, and the minimum value of the angle between the target planned moving direction and the predetermined moving direction is equal to 0 degrees.

[0040] On this basis, the robot is controlled to perform circular motion from the starting point of the circular motion according to the tangent direction and radius of the circular motion, until the robot's movement direction changes to the target planned movement direction. Optionally, when the robot performs the circular motion until its movement direction changes to the target planned movement direction, the side of the robot close to the collision avoidance circle contacts the tangent boundary, thereby achieving the purpose of the robot's movement to avoid the obstacle.

[0041] Specifically, during the circular motion of the robot, it turns a preset turning angle around the center of the circular motion and walks out an arc until the robot's moving direction changes to the target planned moving direction; the tangent boundary is tangent to the arc; the maximum value of the preset turning angle is equal to the difference between 90 degrees and the lower limit of the preset angle range, and the minimum value of the preset turning angle is equal to the difference between 90 degrees and the upper limit of the preset angle range; in this embodiment, the center of the circular motion is in one of the laser projection directions of the laser radar when the robot is located at the reference position point, and the distance between the center of the circular motion and the robot performing the circular motion is maintained at the radius of the circular motion until the robot's moving direction changes to the target planned moving direction; it is achieved by bypassing obstacles, or making the side of the robot parallel to the straight contour of the obstacle, or making the side of the robot parallel to the tangent boundary.

[0042] In an embodiment in which the robot is a circular model as shown in the figure, the center of the robot's body can be the center of the robot's wheel axle, and the center of the robot's wheel axle can be set below the laser radar. A wheel can be installed on each side of the robot's chassis (not shown in the figure). After moving the starting point of the arc motion, when a speed difference is generated between the two wheels, the robot turns, and the robot can be guided not to enter the anti-collision circle during the rotation process, wherein the speed of the wheel close to the anti-collision circle or the tangent boundary side is greater than the speed of the wheel away from the anti-collision circle or the tangent boundary side, so as to drive the robot to revolve around the center of the arc motion and rotate along a preset turning angle in a certain clockwise direction to realize the change of the robot's movement direction to the target planned movement direction.

[0043] As an embodiment, when the laser radar scans only one obstacle point in front of the robot and it is located at the edge of one side of the obstacle being scanned, refer to Figure 3It can be seen that the laser probe rotates around the center L of the laser radar and scans only one obstacle point C2 at the edge of one side of obstacle #2, and no obstacle point is scanned at the edge of the other side of obstacle #2. It is possible that the extension line of the laser beam LB2 shown in the figure intersects with obstacle #2, and the extension line of the laser beam LA2 shown in the figure intersects with obstacle #2. The laser radar sets the only obstacle point scanned in front of the robot as the reference obstacle point, and the robot sets the only obstacle point scanned in front of it as the reference obstacle point C2. Then, when the robot moves in the predetermined direction by executing steps S1 and S2 disclosed in the aforementioned embodiment, the robot As the robot moves along arrow OP2, the rotational scanning laser radar (LIDR) detects more obstacles to the right of reference obstacle point C2. The LIDAR detects no obstacles to the left of reference obstacle point C2 (which may not be occupied by obstacle #2 or may be located farther away) until the robot reaches position G2. The location on the side where the additional obstacles are detected corresponds to the location of obstacle #2, which is scanned in front of the robot. As the robot moves along arrow OP2, the obstacle scanned in front of the robot is determined to be to the right of arrow OP2 (ray LG2). Therefore, after the robot scans the reference obstacle point, it can determine the orientation of the obstacle scanned in front of it by moving within a certain distance in the predetermined direction, thus prompting the robot to avoid the obstacle in advance.

[0044] On the basis of the above embodiment, if the robot starts to make an arc motion on one side of the tangent boundary by executing step S2, the center of the arc motion is set to the corresponding side of the line connecting the anti-collision edge position and the reference position point (the position point where the robot only scans one obstacle point). The corresponding side is the side where the scanned obstacle points are increased, reflecting the relative position of the obstacle scanned in front of the robot and the robot, corresponding to Figure 3 In the figure, when the robot moves to position G2, it starts to make a circular motion in the clockwise direction. The center H2 of the circular motion can be regarded as being set above the line segment E2G2. Moreover, when the robot is located at the reference position, the center H2 of the circular motion is located in the laser projection direction of the laser radar, which corresponds to Figure 2 The extended line of the laser beam emitted by the laser radar on the side where the obstacle point is scanned intersects with the positioning line, as shown in FIG. Figure 2 The line segment LA2 intersects the positioning line C2A2 at point A2. Figure 2Line segment LB2 intersects positioning line C2A2 at point B2. Line segment LB2 and line segment LA2 represent the laser beams emitted sequentially during the clockwise rotation of the laser probe. This embodiment, based on the setting of the center of the arc motion, fully considers the robot's movement tendency when scanning the side with increased obstacles, thereby more accurately controlling the robot's movement toward the planned target direction.

[0045] As an embodiment, when the laser radar scans only one obstacle point in front of the robot and the obstacle point is located at a non-edge position of the obstacle being scanned, refer to Figure 2 It can be seen that the laser probe rotates around the center L of the laser radar and scans only one obstacle point C1 in the middle of obstacle #1. The non-edge position of the obstacle is located in the area surrounded by the outline of the obstacle, for example Figure 2 The laser radar sets the only obstacle point scanned in front of the robot as the reference obstacle point, and the robot sets the only obstacle point scanned in front of it as the reference obstacle point C1; then when the robot moves in the predetermined direction by executing steps S1 and S2, the robot moves along the arrow OP1, and the number of obstacle points scanned on both sides of the reference obstacle point C1 by the laser radar supporting rotational scanning increases, and the projection points of the obstacle points scanned on both sides of the reference obstacle point C1 on the robot's travel plane are all located on the tangent line C1A1 passing through the reference obstacle point C1 and tangent to the anti-collision circle; the laser radar sets the only obstacle point scanned in front of the robot as the reference obstacle point C1; The extension lines of the laser beams emitted from both sides of the reference obstacle point C1 intersect with the positioning line. The extension line of the illustrated laser beam LB1 intersects with obstacle #1, the extension line of the illustrated laser beam LA1 intersects with obstacle #1, and the extension line of the illustrated laser beam LD1 intersects with obstacle #1. The laser beam LB1, the laser beam LA1, the laser beam LD1, the positioning line C1A1, and the laser beam LC1 are all parallel to or projected into the same robot travel plane, and may also fall into the same plane, thereby improving the robot's obstacle perception ability on the scanning plane without touching the obstacle, and reducing the robot's collision risk in navigation control.

[0046] Preferably, with the robot's position point as the starting point, rays are set at equal angles within the detectable viewing angle range of the laser radar, corresponding to the angular area divided at equal angles within the arc sandwiched by the laser beams emitted on both sides of the laser radar's exit hole. In this way, grids are divided within the corresponding scanning area, and the scanned obstacle points occupy the corresponding grids. One ray corresponds to one grid, and the angle between the two adjacent grids is equal to the angle between the corresponding two rays, forming a step angle. This realizes the equiangular acquisition of rays with the center L of the laser radar as the starting point, thereby reducing the amount of calculation.

[0047] For example, the origin of the grid coordinate system can be located at the center of the laser radar, and the horizontal axis in the grid coordinate system can be used as the predetermined moving direction of the robot.

[0048] The boundary of the detection area of ​​the laser radar is regarded as the arc boundary of the detectable viewing angle range of the laser radar. Setting rays at equal angles within the detectable viewing angle range of the laser radar is equivalent to making equal-angle divisions within the arc boundary of the detectable viewing angle range of the laser radar, wherein the center L of the laser radar and the line connecting the two ends of the boundary of the detection area of ​​the laser radar, as well as the boundary of the detection area of ​​the laser radar constitute the detectable viewing angle range of the laser radar.

[0049] Among the obstacle points scanned by the LiDAR and used to form the positioning line, the angle between two adjacent obstacle points is obtained by equally dividing the grid's angles. If there are no obstacle points on the ray, it indicates that there are no obstacles in that direction. The positioning line is the outline of the obstacle, formed by sequentially connecting the reference obstacle points scanned by the LiDAR and the extension of the laser beam at the intersection of the obstacle's contour surface. The smaller the angle between the rays, the more obstacle points there are, the more accurate the representation of the obstacle's outline, and the more precisely distinguishing between the robot's traversable and inaccessible areas. This provides greater convenience for the robot's obstacle avoidance direction.

[0050] In one embodiment, the upper surface of the lidar's protective cover (the small cylindrical structure where point L is located) is parallel to the robot's top cover (the large circular cover shown in the figure, where "large" refers to the horizontal area of ​​the large circular cover being larger than that of the small cylindrical structure). The robot's top cover is parallel to the robot's travel plane, and the robot's travel plane is coplanar with the anti-collision circle. The laser probe within the lidar scans obstacles at various angles during rotation, generally supporting 360-degree rotational scanning. However, the laser beam is emitted only through the lidar's exit aperture, parallel to the robot's travel plane. The lidar's protective cover has an exit aperture, through which the laser beam emitted by the laser probe scans the environment outside the lidar. The arc enclosed by the laser beams emitted from both sides of the exit aperture represents the boundary of the lidar's detection area. This allows obstacle points detected by the lidar's rotational scanning and their positional characteristics relative to the center of the lidar to be used as obstacle information for a given plane.

[0051] As an example, see Figure 1 and Figure 2 It can be seen that when the obstacle where the reference obstacle point is located is a straight obstacle (such as a wall or a long straight furniture), the predetermined moving direction is perpendicular to the positioning line. It can be understood that when the robot only scans the reference obstacle point in front of it, the moving direction of the robot is perpendicular to the positioning line, which corresponds to Figure 2The arrow indicating the predetermined moving direction points to OP1 and is perpendicular to the positioning line C1A1, corresponding to Figure 3 The arrow indicating the predetermined movement direction points to OP2 and is perpendicular to the positioning line C2A2. The positioning line is tangent to the boundary of the laser radar's detection area, so that the reference obstacle point becomes the closest obstacle point to the reference position point. This is equivalent to the reference obstacle point becoming the farthest obstacle point currently scanned by the laser radar.

[0052] The boundary of the detection area scanned with the center L of the laser radar as the rotation center is Figure 2 The reference obstacle point C1 is tangent to the positioning line C1A1, and becomes the farthest obstacle point scanned by the laser radar at the reference position point, and is also regarded as the obstacle point closest to the reference position point within the laser radar scanning range; Figure 2 In the figure, positioning line C1A1 is located on a contour surface scanned by the laser radar in the linear obstacle #1. The contour surface scanned by the laser radar in the linear obstacle #1 is perpendicular to the robot's travel plane. The anti-collision circle with point M1 as the center is coplanar with the robot's travel plane. Points A1, B1, C1, and D1 in the contour surface scanned by the laser radar in the linear obstacle #1 are located on the same straight line and can all be obstacle points. The boundary of the detection area scanned with the center L of the laser radar as the rotation center is Figure 3 The reference obstacle point C2 is tangent to the positioning line C2A2, and becomes the farthest obstacle point scanned by the laser radar at the reference position point, and is also regarded as the obstacle point closest to the reference position point within the laser radar scanning range; Figure 3 In the example, positioning line C2A2 lies within a contour surface of linear obstacle #2 scanned by the LiDAR. For the same LiDAR, the length of linear obstacle #2 is less than that of linear obstacle #1. The contour surface of linear obstacle #2 scanned by the LiDAR is perpendicular to the robot's travel plane, and the collision avoidance circle centered at point M2 is coplanar with the robot's travel plane. Points A2, B2, and D1 within the contour surface of linear obstacle #2 scanned by the LiDAR lie on the same straight line and may all be obstacles. However, point D2 is the endpoint of the laser beam emitted from the right side of the LiDAR's exit aperture, but it does not scan linear obstacle #2, nor any other obstacles. In summary, by eliminating the obstruction of obstacles on at least one side of the robot within a local area using reference obstacle points, and using the positioning line scanned from the front of the robot to plan its movement direction, the robot can subsequently be controlled to move along the edge in a direction parallel to the positioning line, i.e., along the contour of the linear obstacle, while using less information about obstacle points.

[0053] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the embodiments. Some portions not fully described in this invention are known to those skilled in the art. Those skilled in the art will readily appreciate that other variations or modifications based on the above description are possible. It is not necessary and impossible to provide an exhaustive list of all possible embodiments. Obvious variations or modifications arising therefrom remain within the scope of protection of this invention.

Claims

1. A method for planning the movement direction of a robot using an obstacle point to set a tangent line, characterized in that: The robot movement direction planning method includes: Step S1: When the laser radar scans only one obstacle point in front of the robot, a point at a preset safety distance from the obstacle point on the line segment connecting the center of the laser radar and the obstacle point is selected as the center of the anti-collision circle, and an anti-collision circle is set with the preset safety distance as the radius. Then, a tangent line passing through the obstacle point and tangent to the anti-collision circle is set as the positioning line scanned in front of the robot, so that the extension line of the laser beam emitted at some scanning angles of the laser radar intersects with the positioning line; a rotating laser probe is provided in the laser radar, and the rotation center of the laser probe is the center of the laser radar; Step S2: The robot sets the extension direction of the positioning line as the target planned moving direction, and then adjusts the moving direction to the target planned moving direction, and keeps from entering the anti-collision circle during the moving process.

2. The robot movement direction planning method according to claim 1, characterized in that: In step S2, before the robot adjusts its movement direction to the target planned movement direction, the robot starts from a reference position point, moves in the established movement direction to a position point that is the preset safety distance from the anti-collision edge position, and then adjusts its movement direction to the target planned movement direction; The anti-collision edge position is the edge position of the anti-collision circle on the side close to the robot; when the laser radar scans only one obstacle point, the position point of the robot is set as the reference position point.

3. The robot movement direction planning method according to claim 2, characterized in that: In step S2, the method of adjusting the moving direction to the target planned moving direction includes: The robot sets a straight line passing through the anti-collision edge position and tangent to the anti-collision circle as a tangent boundary, so that the anti-collision circle and the robot are separated on both sides of the tangent boundary; Then, the robot is controlled to move in an arc on one side of the tangent boundary so that the movement path of the robot extends toward the target planned movement direction, and the robot is kept from crossing the tangent boundary to enter the anti-collision circle.

4. The robot movement direction planning method according to claim 3, characterized in that: The method for controlling the robot to perform circular motion on one side of the tangent boundary comprises: The robot sets a position point in the opposite direction of the predetermined moving direction and at a predetermined safety distance from the anti-collision edge of the anti-collision circle as the starting point of the arc motion; The robot sets the extension direction of the tangent boundary to the tangent direction of the circular motion, and sets the radius of the circular motion to be equal to the preset safety distance; wherein the tangent boundary and the target planned movement direction are within a preset angle range; According to the tangent direction and radius of the circular motion, the robot is controlled to perform circular motion from the starting point of the circular motion until the moving direction of the robot changes to the target planned moving direction.

5. The robot movement direction planning method according to claim 4, characterized in that: During the circular motion, the robot turns at a preset turning angle around the center of the circular motion and moves along an arc until the robot's movement direction changes to the target planned movement direction; the tangent boundary is tangent to the arc; the maximum value of the preset turning angle is equal to the difference between 90 degrees and the lower limit of the preset angle range, and the minimum value of the preset turning angle is equal to the difference between 90 degrees and the upper limit of the preset angle range; The center of the circular motion is located in one of the laser projection directions of the laser radar when the robot is located at the reference position point.

6. The robot movement direction planning method according to claim 3, characterized in that: In a case where the only obstacle point scanned by the laser radar in front of the robot is located at the edge of one side of the scanned obstacle, the only obstacle point scanned by the laser radar in front of the robot is set as the reference obstacle point. Then, when the robot moves in a predetermined moving direction by executing steps S1 and S2, the obstacle points scanned by the laser radar on one side of the reference obstacle point increase, and the laser radar does not scan any obstacle points on the other side of the reference obstacle point. The position on the side where the increased obstacle points are scanned is the position of the obstacle scanned in front of the robot.

7. The robot movement direction planning method according to claim 6, characterized in that: If the robot starts to make an arc motion on one side of the tangent boundary by executing step S2, the center of the arc motion is set to be on the corresponding side of the line connecting the anti-collision edge position and the reference position point, and the corresponding side is the side where the scanned obstacle point increases; the extension line of the laser beam emitted by the laser radar on the side where the scanned obstacle point increases intersects the positioning line; Moreover, when the robot is located at the reference position point, the center of the circular motion is located in the laser projection direction of the laser radar.

8. The robot movement direction planning method according to claim 3, characterized in that: In a case where the only obstacle point scanned by the laser radar in front of the robot is located at a non-edge position of the scanned obstacle, the only obstacle point scanned by the laser radar in front of the robot is set as the reference obstacle point. Then, when the robot moves in a predetermined moving direction by executing steps S1 and S2, the number of obstacle points scanned by the laser radar on both sides of the reference obstacle point increases. Moreover, the projections of the obstacle points scanned on both sides of the reference obstacle point on the robot's travel plane are both located on a tangent line passing through the reference obstacle point and tangent to the anti-collision circle; and the extension lines of the laser beams emitted by the laser radar on both sides of the reference obstacle point intersect with the positioning line. The non-edge position of the obstacle is located within the area surrounded by the obstacle's contour line.

9. The robot movement direction planning method according to claim 6 or 8, characterized in that: Starting from the robot's position point, rays are set at equal angles within the detectable field of view of the laser radar to form a grid. The scanned obstacle points occupy corresponding grids, with one ray corresponding to one grid. The angle between two adjacent grids is equal to the angle between the two corresponding rays, so that: among the obstacle points scanned by the laser radar to form the positioning line, the angle between two adjacent obstacle points is obtained by equally dividing the grid division angle. Among them, the line connecting the center of the laser radar and the two ends of the boundary of the laser radar's detection area, as well as the boundary of the laser radar's detection area constitute the detectable viewing angle range of the laser radar.

10. The robot movement direction planning method according to claim 9, characterized in that: The upper cover of the laser radar protective cover is parallel to the top cover of the robot, and the top cover of the robot is parallel to the robot's travel plane; the laser probe provided in the laser radar scans obstacle points at various angles during rotation; the robot's travel plane is coplanar with the anti-collision circle; The protective cover of the laser radar is provided with an exit hole, and the laser beam emitted by the laser probe scans the external environment of the laser radar through the exit hole; the arc sandwiched by the laser beams emitted on both sides of the exit hole is the boundary of the laser radar's detection area.

11. The robot movement direction planning method according to claim 10, characterized in that: When the obstacle where the reference obstacle point is located is a linear obstacle, the predetermined moving direction is perpendicular to the positioning line, and the positioning line is tangent to the boundary of the detection area of ​​the laser radar, so that the reference obstacle point becomes the obstacle point scanned closest to the reference position point; wherein, the positioning line is located in a contour surface of the linear obstacle scanned by the laser radar, and the contour surface of the linear obstacle scanned by the laser radar is perpendicular to the travel plane of the robot.

Images