WO2024037262A1 - Narrow passage navigation method for robot, chip, and robot - Google Patents

Abstract

A narrow passage navigation method for a robot, a chip, and a robot. The navigation method comprises: step S1, after a robot enters a narrow passage, if the robot collides with an obstacle, fitting a positioning contour line segment for the obstacle, then turning until a current travel direction of the robot is the same as a preset passage direction of the currently fitted positioning contour line segment, or a current travel direction of the robot is deflected by a preset angle to a passable area relative to the currently fitted positioning contour line segment, so that the robot is currently not blocked by the obstacle, and then executing step S2; and step S2, the robot linearly traveling according to the current travel direction until the robot collides with an obstacle, and then executing step S1; and when the robot travels to a preset target position, stopping executing step S1, and determining that the robot has passed through the narrow passage.

Description

A navigation method, chip and robot for a robot to pass through a narrow passage Technical field

The present invention relates to the technical field of mobile robot navigation, and in particular to a navigation method, a chip and a robot for a robot to pass through a narrow passage.

Background technique

When the sweeping robot is cleaning, if it travels between two walls or a narrow path formed by two low obstacles that is slightly larger than the width of the robot's body, it will actively avoid the obstacle when it touches it, and pass through a large area. Turn to avoid obstacles and avoid obstacles or narrow passages; however, this narrow passage is a more convenient passage to the target location; generally speaking, there are two outlines of obstacles on both sides of the narrow passage; And when setting the movement path, the sweeping robot is generally set as a point. In order to facilitate the sweeping robot to move to the target location in another room area, the set movement path may need to pass through a narrow passage. Due to the requirements of calculation error and exploration efficiency, the navigation path planned by the robot is not suitable for narrow channel environments, and navigation in narrow channels is not smooth enough or repetitive.

Contents of the invention

In order to solve the problem of robots passing through narrow passages, the present invention provides a navigation method, chip and robot for robots to pass through narrow passages. The specific technical solutions are as follows:

A navigation method for a robot passing through a narrow passage. The navigation method includes: step S1. After the robot enters the narrow passage, if the robot collides with an obstacle, a positioning contour line segment is fitted to the obstacle, and then rotates until the robot The current walking direction is the same as the preset traffic direction of the currently fitted positioning contour line segment, or the current walking direction of the robot is deflected to the passable area by a preset angle relative to the currently fitted positioning contour line segment, so that the robot cannot currently If the robot is blocked by an obstacle, then execute step S2; in step S2, the robot walks in a straight line according to the current walking direction until it collides with an obstacle, and then executes step S1; when the robot walks to the preset target position, it stops executing step S1, and Make sure the robot passes through the narrow passage.

Further, in step S1, among the positioning contour segments it has fitted, the robot sets the endpoint that forms the shortest traversable path relative to the preset target position as the target endpoint of the positioning contour segment, where the shortest The passable path is located within the passable area of the narrow road that the robot enters; the robot configures the endpoints of the positioning contour line segment except the target endpoint in the direction pointing to the target endpoint as the preset passing direction of the positioning contour line segment.

Further, the robot configures the channel where the target endpoint of each positioning contour segment is located as a narrow channel; the robot sets the positioning contour segment as the boundary line of the narrow channel; whenever the robot collides with a new obstacle, the robot will The obstacle is fitted to a positioning contour segment, and then the target endpoint of the new obstacle's positioning contour segment is obtained, and the opening of the channel where the target endpoint is located is identified as a narrow crossing.

Further, in step S2, before the robot walks to the preset target position, if the robot recognizes a narrow crossing, it will recognize the narrow crossing as the exit of the narrow crossing, and then the robot will adjust its current walking direction to the current The preset travel direction of the fitted positioning contour line segment, or the deflection to the passable area by a preset angle relative to the currently fitted positioning contour line segment; then the robot walks straight in accordance with the current walking direction to leave the narrow road.

Further, the navigation method also includes: before the robot enters the narrow passage, if the robot recognizes the narrow passage, then recognize the narrow passage as the entrance of the narrow passage, and then the robot adjusts its current walking direction to the current fitted The preset travel direction of the positioning contour line segment, or the current walking direction of the robot deflects to the passable area by a preset angle relative to the currently fitted positioning contour line segment; then the robot walks in a straight line according to the current walking direction to enter the entrance of the narrow passage , and then perform step S1.

Furthermore, if the robot collides with an obstacle before the robot starts to recognize the narrow crossing, the obstacle will be marked as the latest obstacle detected before the robot starts to recognize the narrow crossing, and a positioning contour will be fitted to the obstacle. line segment, and then the robot adjusts its current walking direction to be the same as the preset direction of the currently fitted positioning contour segment, so that the robot walks along the currently fitted positioning contour segment until it collides with a new obstacle and Fit the corresponding positioning contour line segment for the new obstacle, and then start to identify the narrow crossing.

Furthermore, whenever the robot rotates to the point where its current walking direction is the same as the preset travel direction of the currently fitted positioning contour line segment, or when the robot rotates to the point where its current walking direction deflects toward the passable area relative to the currently fitted positioning contour line segment, At the preset angle, the robot walks in a straight line according to the current walking direction to a preset safe distance, and detects whether it collides with an obstacle during the straight walking process. If so, step S1 is executed. Otherwise, it is determined that the robot has passed through a narrow road, and then uses The heuristic search algorithm plans a path to the preset target position or a path to the position closest to the obstacle detected in the current walking direction, and then walks along the currently planned path until the robot Identify the entrance to another narrow passage and enter the narrow passage, and then perform step S1.

Further, the method for the robot to identify a narrow crossing includes: the robot first fits a positioning contour segment to the latest obstacle it collides with, and then the robot extracts two obstacle bypass points in the search direction of the positioning contour segment. If the robot detects the obstacle, If the distance between the two obstacle bypassing points is within the preset distance range, the channel where the two obstacle bypassing points are located is identified as a narrow channel, and the gap formed between the obstacles where the two obstacle bypassing points are located is identified. It is a narrow crossing.

Further, after the robot recognizes the narrow crossing, if it is detected that the robot's walking direction changes from the reference obstacle bypassing direction to the preset traveling direction of the currently fitted positioning contour line segment or relative to the direction of the currently fitted positioning contour line segment. The passable area is deflected by the preset angle, and the robot walks in a straight line according to the changed walking direction, then the robot recognizes that it has entered a narrow passage; wherein the changed walking direction is configured to point from the current position of the robot to the The direction of the passable area inside the channel where the two obstacle bypassing points are located; where the reference obstacle bypassing direction is the preset passing direction of the positioning contour line segment fitted last time or the direction relative to the positioning contour line segment fitted last time. The traversable area is deflected by the preset angle; wherein, the last fitted positioning contour line segment is the positioning contour line segment fitted for the latest obstacle encountered before the robot starts to recognize the narrow crossing.

Furthermore, before starting to recognize the narrow crossing, the robot collides with an obstacle and fits the corresponding positioning contour line segment; then, when the robot collides with another obstacle, it starts to recognize the narrow crossing; the robot first collides with the one it collides with. The positioning contour line segment fitted by the obstacle is the positioning contour line segment fitted last time, and the positioning contour line segment fitted by the robot for an obstacle it collides with later is the positioning contour line segment fitted currently. , making the preset travel direction of the last fitted positioning contour segment different from the preset travel direction of the currently fitted positioning contour segment.

Further, the robot sets the search direction of the positioning contour line segment in the vertical direction of the positioning contour line segment to form the width direction of the channel where the two obstacle bypassing points are located; the two obstacle bypassing points are along the positioning contour The search directions of line segments are distributed among: the latest obstacles that the robot collides with before it starts to recognize the narrow crossing, and the obstacles that the robot collides with when it starts to recognize the narrow crossing.

Further, the preset distance range is set to be greater than or equal to the body width of the robot; the preset safety distance is set to be equal to the body width of the robot; and the preset angle is set to guide the robot away from the current collision area. Obstacle, or any endpoint of the positioning contour line segment that bypasses the currently collided obstacle, in order to walk to the passable area connected to the preset target position; the shortest passable path is the positioning contour of the robot respectively. The two endpoints of the line segment are the starting point of the search, and the path with the shortest path length among the paths leading to the preset target location planned using a heuristic search algorithm is used.

A chip stores program code, and when the program code is executed, the steps of the navigation method are implemented.

A robot is equipped with a distance sensor. The robot is equipped with the chip to control the robot to use the distance sensor to detect point cloud data of obstacles and fit corresponding positioning contour segments from the point cloud data to facilitate passage. Execute the navigation method robot navigation method to pass through the narrow passage.

The beneficial technical effect of the present invention is that in a narrow channel, the robot disclosed in the present invention tentatively moves forward along the specific direction indicated by the fitted positioning contour line segment (the direction of the shortest path to the preset target position in the narrow channel), Identify narrow crossings by hitting obstacles and adjust the direction in time to avoid obstacles. Repeat the process to walk a route from the starting position to the preset target position, so as to achieve alternate twisting and advancement towards the preset target position. The said narrow passage; thereby controlling the robot to accurately avoid obstacles in the narrow passage, enabling the navigation to smoothly pass through narrow passages with a width 1cm larger than the body width and other complex narrow passage scenes, improving the smoothness and success rate of the robot passing through narrow passages.

The walking direction of the robot only needs to change from the reference obstacle bypassing direction to the preset passing direction of the currently fitted positioning contour line segment or deflect to the passable area by the preset angle relative to the currently fitted positioning contour line segment, and the robot will follow the When the changed walking direction walks in a straight line and the distance between the two obstacle bypassing points is within the preset distance range, the robot will identify the passage where the two obstacle bypassing points are located as a narrow passage and determine that it has entered the narrow passage. road.

The invention starts to identify the narrow crossing, before and after the narrow crossing, before entering the narrow crossing and in the corresponding stages of entering the narrow crossing, according to the collision situation between the robot and the obstacle and the simulation of the outline of the obstacle. According to the matching situation, adjust the walking direction of the robot, and then walk in a straight line in the narrow channel to tend to walk towards the preset target position; and based on the collision of the robot with two obstacles and the fitting of the outline of each obstacle Recognize narrow crossings to distinguish between three recognition states before starting to recognize narrow crossings, when starting to recognize narrow crossings, and recognizing narrow crossings, thereby advancing the process of the robot entering the narrow crossing and entering the narrow crossing again.

Description of drawings

Figure 1 is a flow chart of a method for robot navigation through a narrow passage disclosed in an embodiment of the present invention.

FIG. 2 is a schematic diagram of the movement trajectory of a robot passing through a narrow passage according to another embodiment of the present invention.

Implementation

The technical solutions in the embodiments of the present invention will be described in detail below with reference to the accompanying drawings in the embodiments of the present invention. To further explain various embodiments, the present invention provides drawings. These drawings are part of the disclosure of the present invention, and are mainly used to illustrate the embodiments, and can be used to explain the operating principles of the embodiments in conjunction with the relevant descriptions in the specification. With reference to these contents, those of ordinary skill in the art will be able to understand other possible implementations and advantages of the present invention. A process or method depicted as a flowchart. Although the flowcharts depict steps as a sequential process, many of the steps may be performed in parallel, concurrently, or simultaneously. Additionally, the order of steps can be rearranged. The process may be terminated when its operation is completed, but may also have additional steps not included in the figures. The processing may correspond to a method, function, procedure, subroutine, subroutine, or the like.

When the robot passes through a narrow passage, in order to overcome the excessive calculation errors produced during the process of fitting contour line segments and the sensor errors introduced by the planned walking direction, so that the robot's navigation can smoothly travel in and through the narrow passage, The invention discloses a navigation method for a robot to pass through a narrow passage; the execution subject of the navigation method is a robot equipped with a ranging sensor. In this application, the navigation method is automatically executed by the robot, which can accept human input instructions or Run pre-programmed program instructions, and can also independently formulate and execute tasks based on artificial intelligence technology. The robot may be a cylinder, integrating at least a collision sensor and a ranging sensor. Among them, the ranging sensor can be a line laser sensor, which emits a line laser to the outside to obtain the position information reflected by the obstacle. By detecting the two-dimensional point cloud data of the surrounding environment, a two-dimensional point cloud map can be constructed in a timely manner, and the robot can The number of sensors installed on the body can be one line laser sensor or multiple line laser sensors; line laser sensors include multi-line lidar and single-line lidar. Single-line lidar refers to a radar whose line beam emitted by the laser source is a single line. It is used in In the field of robots, most are service robots, which can help robots avoid obstacles. They have fast scanning speed, strong resolution, and high reliability. Compared with multi-line lidar, single-line lidar responds faster in terms of angular frequency and sensitivity. Therefore, It is more accurate in testing the distance and accuracy of surrounding obstacles. Especially for cleaning robots that walk on the ground surface, lawn mower robots that walk on lawn areas with narrow passages demarcated by boundary lines, floor washing machines, and security patrol robots, this embodiment is not applicable to the navigation method. type is limited. The robot can also be equipped with inertial sensors (including but not limited to odometers for measuring walking distance, collision sensors for detecting collisions with obstacles, and gyroscopes for measuring body rotation angles), or visual sensors ( Any type of depth information collection device can be used, including but not limited to monocular cameras, binocular cameras and other cameras).

Referring to Figure 1, it can be seen that the navigation method includes: Step S1. After the robot enters the narrow channel, if the robot collides with an obstacle, a positioning contour line segment is fitted to the obstacle as the positioning contour line segment currently fitted by the robot. , the currently fitted positioning contour line segment is characterized as the contour line segment of the latest obstacle detected by the robot; then the robot rotates, specifically rotating on the spot until the current walking direction of the robot matches the preset value of the currently fitted positioning contour line segment The traffic direction is the same, or the current walking direction of the robot is deflected to the passable area by a preset angle relative to the currently fitted positioning contour line segment, so that the robot is not blocked by obstacles; where, the current walking direction of the robot is different from the current When the preset travel directions of the fitted positioning contour segments are the same, the robot's current walking direction is parallel to the currently fitted positioning contour segment and the robot's current walking direction points to the preset target position in the narrow channel where it is located. the extension direction of the shortest path; in addition, the implementation of deflecting the preset angle to the passable area includes: there is a passable area on the left side of the currently fitted positioning contour line segment or on the left side of the obstacle where the positioning contour line segment is located (not occupied by an obstacle), the robot deflects to the left relative to the extension direction of the currently fitted positioning contour segment; to the right of the currently fitted positioning contour segment or to the right of the obstacle where the positioning contour segment is located When there is a passable area (not occupied by obstacles) on the side, the robot deflects to the right relative to the extension direction of the currently fitted positioning contour segment; the preset angle is related to the accuracy of the fitting calculation, the higher the accuracy, the The closer the currently fitted positioning contour line segment is to the obstacle, the smaller the preset angle is set, otherwise the larger it is. The preset angle is preferably within 0 to 30 degrees. The robot then executes step S2.

It should be noted that the method for the robot to fit the positioning contour line segment includes: for narrow roads, the point cloud data (radar points) collected by the line laser sensor disclosed in this embodiment are all expressed in discrete coordinates, and these discrete coordinates are specifically expressed in To represent the position of obstacles, most of them need to be filtered and grouped, and then the grouped point cloud data are fitted to obtain a fitting straight line or a fitting curve, and then based on the proportional conversion relationship between world coordinates and image coordinates, through fitting The coordinate points of the curve or the fitted straight line are used to obtain the obstacle envelope; preferably, in order to ensure the efficiency of the path planning algorithm, a simple robot particle model is used in the path search stage. Preferably, the fitted curve or the fitted straight line is virtualized Expansion processing, the expansion radius is set to the radius of the inscribed circle or the radius of the circumscribed circle of the robot outline projection, forming a fitted circle surrounding the obstacle, corresponding to the obstacle envelope; thereby enveloping the obstacle Lines are displayed in a grid map in real time (point cloud data is discretely distributed around obstacles), which not only corresponds to the laser detection distance from the obstacle to the edge of the robot, but can also be used to indicate the location of the obstacle.

Step S2: The robot walks in a straight line in the current walking direction until it collides with an obstacle, and then executes step S1; during the process of repeating step S1 and the aforementioned straight-line walking operation, if the robot walks to the preset target position, it stops executing step S1. , and determine that the robot passes through the narrow passage; preferably, the preset target position is located outside the narrow passage and is connected to the passable area inside the narrow passage to support the robot inside the narrow passage to walk straight to the preset target position.

Specifically, for step S2, whenever the robot rotates to the point where its current walking direction is the same as the preset direction of the currently fitted positioning contour line segment, or when the robot rotates to the direction where its current walking direction is relative to the direction of the currently fitted positioning contour line segment, When the passable area deflects to the preset angle, the robot walks in a straight line according to the current walking direction at a preset safe distance, and detects whether it collides with an obstacle during the straight walking process. If so, step S1 is executed to achieve accurate obstacle avoidance for the robot; otherwise, Determine that the robot has passed through a narrow passage, and determine that the robot is currently exiting the narrow passage. The preset safety distance is preferably the width of the robot's body to comply with the width setting requirements in the narrow passage environment, and then uses a heuristic search algorithm to plan the path leading to the narrow passage. The path to the preset target position, or the path leading to the closest position to the obstacle detected in the current walking direction (which can be the obstacle closest to the current position of the robot), can be the contour of the corresponding obstacle. path to the nearest local target position; then walk along the planned path. If the robot is in an area close to the preset target position, it will walk along the path leading to the preset target position. If the robot is far away from the preset target position, it will walk along the path to the preset target position. In the local area surrounded by obstacles on three sides of the preset target position, it walks along the path leading to the local target position; until the robot recognizes the entrance of another narrow lane and enters the narrow lane, it then performs the steps S1 to enable entry into a new narrow road in a new direction.

Preferably, the prerequisite for the robot to recognize the entrance of another narrow passage includes: while walking along the planned path, the robot has detected or collided with a new obstacle and started to identify the entrance of the narrow passage. Corresponding to Figure 2, the robot navigates from the start position to the target position (preset target position), encounters an obstacle at the R position, uses point cloud data to fit the line segment A0B0, and then based on the local target position R1 or the preset target position target Choose whether the current walking direction is A0 pointing to B0 or B0 pointing to A0, and then the robot chooses the direction of B0 pointing to A0 to move to the local target position R1. It may collide with the obstacle on its right side and start to identify the narrow crossing; the robot is at position R1 When it is recognized that the gap between position B1 and position A0 is a narrow crossing, the robot enters and passes through the narrow crossing and moves until it reaches the preset target position target. Among them, when the robot walks to position R5, it selects the direction in which B5 points to A5 as the current walking direction, and then walks along the current walking direction to position R6 to control the robot to rotate to a direction without obstacles and does not conflict with the positioning contour line segment. The obstacle where B5A5 is located collides with the obstacle where the positioning contour line segment B4A4 is located. The robot just passes through the corresponding narrow passage (i.e., the exit of the narrow passage), and then uses a heuristic search algorithm to plan a path to the preset target position. , and then walk straight along the planned path or navigation direction to the preset target location target.

Preferably, the point cloud data of the ranging sensor describing the environment has a limited sensing range and field of view, and cannot describe the environmental information around the robot in detail. Then, based on the global path of the navigation task, the local target location of the local plan is selected within the local map range, in a specific manner. In order to find the track point closest to the edge in the local map along the global path starting from the current position, this is used as the local target position. The global path is a path planned to the preset target location using the heuristic search algorithm disclosed in the aforementioned embodiments.

To sum up, in the narrow passage, the robot disclosed in the present invention tentatively moves forward along the specific direction indicated by the fitted positioning contour line segment (the direction of the shortest path to the preset target position in the narrow passage), and passes the collision with the obstacle. The method identifies the narrow passage and adjusts the direction in time to avoid obstacles, repeats the process, and walks a route from the starting position to the preset target position, so as to achieve alternate twisting and advancement toward the preset target position through the narrow passage; This controls the robot to accurately avoid obstacles in narrow passages, allowing the navigation to smoothly pass through narrow crossings with a width 1cm larger than the body width and other complex narrow passage scenes, improving the smoothness and success rate of the robot passing through narrow passages.

In the above embodiment, in step S1, among the positioning contour segments it has fitted, the robot sets the endpoint that forms the shortest accessible path relative to the preset target position as the target endpoint of the positioning contour segment, according to As the robot passes through the narrow passage, the distribution positions of the target endpoints in Figure 2 are position A0, position B1, position B2, position B3, position B4, and position A5, which in turn correspond to the endpoints of the positioning contour line segment A0B0 and the positioning contour. The endpoints of line segment A1B1, the endpoints of positioning contour line segment A2B2, the endpoints of positioning contour line segment A3B3, the endpoints of positioning contour line segment A4B4, and the endpoints of positioning contour line segment A5B5; wherein, the shortest accessible path is located in the narrow channel entered by the robot Within the passable area; the robot configures the endpoints of the positioning contour line segment except the target endpoint in the direction pointing to the target endpoint as the preset passing direction of the positioning contour line segment. In Figure 2, the corresponding directions are B0 pointing to A0, The arrow marked by index number ① points to the direction, the arrow marked by index number ② points, the arrow marked by index number ③ points, the arrow marked by index number ④ points, and B5 points in the direction of A5. Therefore, after the robot recognizes the narrow crossing and enters it, it calculates the precise angle at which the robot needs to turn to avoid obstacles, so as to turn the robot's walking direction to a direction pointing forward without obstacles.

As an embodiment, the robot configures the channel where the target endpoint of each positioning contour segment is located as a narrow lane; the robot sets the positioning contour segment as the boundary line of the narrow lane; whenever the robot collides with a new obstacle, Then fit a positioning contour line segment for the new obstacle, and then obtain the target endpoint of the positioning contour line segment of the new obstacle, which is equivalent to a boundary point of the channel, and identify the opening of the channel where the target endpoint is located as Narrow crossing. In this embodiment, before and after step S1 is executed, the robot uses a collision sensor to detect whether there are obstacles on its traveling path during the traveling process, and combines with the ranging sensor to detect whether there is a narrow road on its traveling path, for example , if the robot's body width (outline diameter) is D, then the channel with a width in the range of (1-ɑ)D to (1+ɑ)D can be determined as a narrow channel, where the coefficient ɑ can be combined with the robot's shape and material are set, for example, the value of ɑ is 0.03.

It should be noted that due to the errors accumulated by the robot's sensor, different obstacle materials, and different lighting environment interference, the obstacle envelope (the peripheral envelope of the obstacle) obtained by using the point cloud data collected by the laser sensor Contour, a fitting curve or the connection of multiple fitting curves) may not necessarily fit the actual obstacle location, so the robot may exclude narrow crossings from the robot's passable area when searching for a path, so the obstacle envelope Line errors can easily cause the robot to misjudge the narrow crossing; in this application, it is necessary to use robot collision to fit the positioning contour line segment of the obstacle at the right time and obtain the corresponding guidance direction and take the opportunity to enter the narrow crossing.

As an embodiment, the navigation method further includes: before the robot enters the narrow passage, if the robot recognizes the narrow passage, it identifies the narrow passage as the entrance to the narrow passage, and then the robot adjusts its current walking direction to the current The preset travel direction of the fitted positioning contour line segment, or the current walking direction of the robot, is deflected to the passable area by a preset angle relative to the currently fitted positioning contour line segment; then the robot walks in a straight line according to the current walking direction to bypass It begins to identify the target endpoint of the positioning contour line segment of the latest obstacle detected before the narrow channel, and then enters the entrance of the narrow channel to realize that it has entered the narrow channel, and then performs step S1.

Based on the above embodiment, before the robot starts to recognize the narrow crossing, if the robot collides with an obstacle, the obstacle will be marked as the latest obstacle detected before the robot starts to recognize the narrow crossing, and then the obstacle will be targeted. The positioning contour segment is fitted, and then the robot adjusts its current walking direction to be the same as the preset direction of the currently fitted positioning contour segment, so that the robot walks along the currently fitted positioning contour segment until it collides with New obstacles are found and corresponding positioning contour line segments are fitted to the new obstacles, and then narrow crossings are identified. Specifically, when the robot first collides with an obstacle at the R position (the nearest black rectangle to the right of the start position), it is the latest obstacle detected before the robot starts to recognize the narrow crossing, that is: the latest detection before it starts to recognize the narrow crossing. The target endpoint of the positioning contour segment of the obstacle corresponds to the target endpoint A0 of the positioning contour segment A0B0 in Figure 2; then the robot adjusts its walking direction, starting from the R position in Figure 2, and points in the direction of position A0 along position B0 Walk to position R1, then collide with a new obstacle at position R1 (the nearest black rectangle on the upper right side of position R1), and then adjust the current walking direction at position R1 to the direction of the arrow marked by index number ①, which can be regarded as Parallel to the positioning contour segment A1B1 fitted for the new obstacle, when the robot starts walking from position R1 along the arrow marked by index number ①, it is recognized that the robot has entered a narrow channel and also bypassed position A0.

Based on the aforementioned embodiment before entering a narrow passage, preferably, for a non-temporarily formed fixed narrow passage, after the robot faces the direction of the narrow passage, it may identify a narrow passage entrance and then deflect to the left or right for the first predetermined time. Set an angle (for example, 30 degrees) and then move in a straight line to enter the narrow crossing. If the robot collides, it means that it has reached the boundary of one side of the narrow crossing. It may exit the narrow crossing at the gap at the current position and recognize a new narrow crossing. , then deflect again in the opposite direction by a second preset angle (such as 60 degrees) and then move in a straight line to enter a new narrow crossing. If the robot collides again, it means that it has reached the boundary on the other side of the narrow crossing, and repeats , the walking positions correspond to R1 -> R2 -> R3 -> R4 -> R5 -> R6, so that through the corresponding narrow passage, the left and right twists and turns can be realized to advance through the narrow passage. Specifically, corresponding to Figure 2, the robot walks from position R1 to position R2 (executed before starting to identify a new narrow crossing), encounters an obstacle on the right side at the position, and uses point cloud data to fit the positioning contour line segment A2B2 (triggering the robot to start identifying a new narrow crossing), then identify the first narrow crossing between position B2 and position A0, and then rotate the body so that the current walking direction becomes the arrow marked by index number ② in Figure 2. , then the robot chooses the direction A2 points to B2 and walks straight to position R3 to pass through the first narrow crossing in the corresponding narrow passage (equivalent to entering the first narrow crossing); similarly, it then collides with its front at position R3 If there is an obstacle, use the point cloud data of the obstacle to fit the positioning contour line segment A3B3, then identify the second narrow intersection between position B2 and position A3, and then rotate the body so that the current walking direction becomes Figure 2 The arrow marked by index number ③ points, and then the robot walks straight along the arrow marked by index number ③ to position R4 to pass through the second narrow crossing, and then collides with the obstacle in front of it on the right at position R4, then Use the point cloud data of the obstacle to fit the positioning contour segment A4B4, then identify the third narrow intersection between position B2 and position A4, and then rotate the body so that the current walking direction changes to the index number ④ in Figure 2 The marked arrow points, and then the robot walks straight along the arrow marked by index number ④ to position R5 to pass through the third narrow crossing, and then collides with the obstacle in front of it on the right at position R5, then the obstacle's The point cloud data fits the positioning contour line segment B5A5, and then identifies the fourth narrow crossing between position B5 and position B4, and then rotates the body so that the current walking direction changes from position B5 to position A5, and then the robot follows Landing position B5 points in the direction of position A5 and walks straight to position R6 to cross the fourth narrow crossing. As shown in Figure 2, before the robot walks to the preset target position target, if the robot recognizes the narrow crossing between position B5 and position B4, it will identify the narrow crossing as the exit of the narrow crossing. Specifically, the narrow crossing is the exit of the narrow crossing. Approaching the narrow crossing of the preset target position target, the robot then adjusts its current walking direction to the preset passing direction of the currently fitted positioning contour line segment, or to the passable area relative to the currently fitted positioning contour line segment. Deflect to the preset angle, and then the robot walks in a straight line in the current walking direction to bypass the target end point B4 of the positioning contour segment of the latest obstacle it collided with before it started to identify the exit of the narrow channel, and walks in a straight line without any obstacles to the preset After a safe distance, the exit from the narrow passage is achieved at position R6; preferably, the preset target position target is set outside the narrow passage. ;Then the robot uses a heuristic search algorithm to plan a path to the preset target position target at position R6, and then starts from position R6 and walks straight along the planned path or navigation direction to the preset target position target. Therefore, whenever the robot collides with a new obstacle, a positioning contour line segment is fitted to the new obstacle, and the target endpoint of the positioning contour line segment of the new obstacle is obtained, and the channel where the target endpoint is located is The opening is identified as a narrow crossing.

Based on the above embodiments, the present invention starts to recognize the narrow crossing before starting to recognize the narrow crossing, before and after recognizing the narrow crossing, before entering the narrow crossing, and in the corresponding stages of entering the narrow crossing, according to the collision situation and the collision between the robot and the obstacle. According to the fitting of the obstacle contour, adjust the walking direction of the robot, and then walk in a straight line in the narrow channel to tend to walk towards the preset target position; and based on the collision of the robot with two obstacles and the evaluation of each obstacle The fitting situation of the contour identifies the narrow crossing to distinguish the three recognition states before starting to recognize the narrow crossing, when starting to recognize the narrow crossing, and recognizing the narrow crossing, thereby promoting the process of the robot entering the narrow crossing and entering the narrow crossing again.

In the aforementioned embodiment, the method for the robot to identify the narrow crossing includes: while triggering the robot to start identifying the narrow crossing, the robot first fits a positioning contour segment to the latest obstacle it collides with, and then the robot searches in the direction of the positioning contour segment. Two obstacle bypass points are extracted from above. If it is detected that the distance between the two obstacle bypass points is within the preset distance range, the channel where the two obstacle bypass points are located is identified as a narrow channel; and the two obstacle bypass points are The gap formed between the obstacles where each obstacle point is located is identified as a narrow crossing, preferably an opening between the two obstacle bypassing points. Further, after the robot recognizes the narrow crossing, if it is detected that the robot's walking direction changes from the reference obstacle bypassing direction to the preset traveling direction of the currently fitted positioning contour line segment or relative to the direction of the currently fitted positioning contour line segment. The passable area is deflected by the preset angle, and the robot walks in a straight line according to the changed walking direction, then the robot recognizes that it has entered a narrow passage; wherein the changed walking direction is configured to point from the current position of the robot to the The direction of the passable area inside the channel where the two obstacle bypass points are located is preferably parallel to the currently fitted positioning contour line segment. Then the reference obstacle bypassing direction is the preset passing direction of the positioning contour line segment fitted last time or the preset angle deflected to the passable area relative to the positioning contour line segment fitted last time; wherein, the positioning contour line segment fitted last time is deflected to the passable area by the preset angle; The positioning contour line segment is the positioning contour line segment fitted to the latest obstacle encountered before the robot starts to recognize the narrow crossing. Specifically, before starting to recognize the narrow crossing, the robot collides with an obstacle and fits the corresponding positioning contour line segment; then, when the robot collides with another obstacle, it starts to recognize the narrow crossing; the robot first collides with a The positioning contour line segment fitted by the obstacle is the positioning contour line segment fitted last time, and the positioning contour line segment fitted by the robot for an obstacle it collides with later is the positioning contour line segment fitted currently. , making the preset travel direction of the last fitted positioning contour segment different from the preset travel direction of the currently fitted positioning contour segment. Therefore, the robot uses the orientation and endpoints of the positioning contour segments of the obstacles it collides with twice to identify the narrow crossing and enter the narrow passage.

As an embodiment, when the robot extracts two obstacle avoidance points in the search direction of the positioning contour line segment, if it is detected that the robot's walking direction changes from the reference obstacle avoidance direction to the predetermined positioning contour line segment currently fitted, Assume that the direction of travel or the preset angle is deflected to the passable area relative to the currently fitted positioning contour line segment, and it is detected that the distance between the two obstacle avoidance points is within the preset distance range, and the robot follows the changed If the walking direction is straight, the robot will recognize its behavior as entering a narrow passage, or have a tendency to enter a narrow passage by identifying the entrance of a narrow passage. At this time, the robot enters the entrance of the narrow passage from the outside of the narrow passage, or The robot is located in front of the entrance of the passage where the two obstacle bypass points are located, and its walking direction is leading to the passable area inside the passage where the two obstacle bypass points are located, and the robot may twist left and right. Wherein, after the robot obtains the positioning contour line segment in step S1, the robot extracts two obstacle bypassing points in the search direction of the positioning contour line segment. It may be based on selecting one obstacle bypassing point and then searching along the The direction extracts another obstacle avoidance point. These obstacle avoidance points are the endpoints of the fitted contour line segments and are the two closest endpoints on the two positioning contour line segments, forming the obstacle avoidance points on the left and right sides of the robot. Specifically, the robot determines that it has started to enter a narrow crossing, which is a gap formed between the obstacles where the two obstacle bypass points are respectively located. The gap is located in the passable area. The boundary line of the channel where the gap is located can be The actual contour of the obstacle or the fitted positioning contour line segment, the distance between the boundary lines is within the preset distance range to improve the judgment accuracy. Preferably, the width of the gap is the distance between the grids where the two endpoints of the gap are located on the robot's traveling plane. The two endpoints of the gap are respectively located on the contour line of the corresponding obstacle or the fitted one. on the contour segment. No matter how frequently the robot collides with the obstacles around it before entering the channel where the two obstacle avoidance points are located, the robot's walking direction only needs to change from the reference obstacle avoidance direction to the preset passage of the currently fitted positioning contour line segment. direction or the preset angle relative to the currently fitted positioning contour line segment toward the traversable area, and the distance between the two obstacle avoidance points is within the preset distance range, the robot will The channel where each obstacle detour point is located is identified as a narrow channel, which makes the robot's navigation from the current position to the preset target position smoother and the navigation path shorter, improving the robot's navigation efficiency.

Preferably, the preset distance range is greater than or equal to the body width of the robot; preferably, the preset distance range is greater than the body width of the robot. When the shape of the robot is circular, the body width of the robot is its Body diameter. When the width of the entrance of the channel where the two obstacle bypass points are located (i.e., the gap formed between the two obstacle bypass points) is greater than or equal to the width of the robot's body, it is determined that the channel where the two obstacle bypass points are located allows the robot to enter, Then, when the robot walks to one of the obstacle bypassing points or its vicinity to collide with a new obstacle, it can adjust its walking direction to be parallel to the corresponding positioning contour line segment, or to a direction that forms a preset angle relative to the positioning contour line segment. The latter direction may be a straight line pointing to the preset target position, or a direction that meanders through the passage where the two obstacle bypass points are located to extend to the preset target position.

As an embodiment, if the robot does not collide with an obstacle while walking in a straight line in the current walking direction at a preset safe distance, the robot will maintain barrier-free walking in the narrow passage at a preset safe distance and then start using the heuristic search algorithm. Plan the navigation path. The preset target position is the end point of the navigation path. As for the specific search starting point, it can be the position where the robot first collides with the obstacle, or it can be a position selected near the obstacle, or it can be all Describe the obstacle avoidance point or nearby area. Between the aforementioned step S1 and the aforementioned step S2, it also includes: after the robot detects an obstacle (which can collide with it), and after the robot detects another obstacle (the new obstacle detected can be the same as the previously detected obstacle). The obstacles are relatively close in position and are considered to be adjacent in position. Before searching for two obstacle bypass points in the search direction of the positioning contour line segment, keep walking in the same direction, corresponding to each index number in Figure 2. The direction of the marker. To a certain extent, the robot is guaranteed to pass through narrow passage areas. Terrain traversability information is stored as the cost value of each raster in the cost map. The trafficability cost is considered in the path search stage, that is, the trafficability cost value is added to the cost function of the search algorithm, and the algorithm will tend to find a path to a low trafficability cost area.

In the foregoing embodiments, the robot generally uses the location where it collides with an obstacle for the first time as the search starting point. Then, while walking along the planned obstacle path starting from the search starting point, the robot collides with another obstacle. When an object is detected, it can also be regarded as that the robot detects another obstacle, and then obtains the positioning contour line segment through the point cloud data collected by the ranging sensor. The positioning contour line segment is a fitting of the point cloud data reflected from the other obstacle. The result (a partial line segment derived from the envelope of the other obstacle and capable of overlapping with the current position of the robot) is used to explore the gap or channel formed between the two detected obstacles. In some embodiments, the positioning contour line segment may be a fitting line segment calculated based on point cloud data pre-collected by the robot (such as the contour derived from the first collision with an obstacle) or may be parallel to the fitting line segment. The robot may not have Collision with the other obstacle, but the robot can predict the gap that may form between two different obstacles; therefore, the robot regards the positioning contour segment as the boundary line where one end of the gap is located or the boundary where one side of the channel is located Wire. Preferably, the narrow passage is a passage composed of the contour lines (fitting results) of the two obstacles as the boundary line; the minimum distance between the obstacle bypass points extracted from the contour lines of the two obstacles is greater than Or equal to the minimum value of the preset distance range; the minimum value of the preset distance range is greater than the width of the robot's body; the width of the narrow channel is within the preset distance range, and the entrance of the narrow channel is also within the preset distance range. distance range. The narrow entrance can be a small door opening in a room. The obstacles on both sides of the door opening are the four walls in the same room. The four walls are continuous and integrated. In addition, the walls on the left and right sides of the passage where the obstacle bypass point is located are It can also be approximately parallel.

It should be noted that the robot will use a heuristic search algorithm to plan a navigation path, and the preset target position is the end point of the navigation path; after planning the navigation path, the robot will walk along the navigation path. When the robot uses the ranging sensor to detect When an obstacle is detected, such as when it is detected for the first time, or even when it collides with an obstacle, the robot starts from its current position and uses a heuristic search algorithm to plan a navigation path around the currently detected obstacle and The navigation path is marked as an obstacle avoidance path, and the preset target position is the end point of the obstacle avoidance path. Specifically, when the robot starts to use its current position or the preset starting position as the search starting point, the search starting point points to the direction of the preset target position, which corresponds to the search starting point start in Figure 2 pointing to the direction of the preset target position target. , when the robot uses a heuristic search algorithm to plan its path, it searches for free rasters (grid areas that represent passable areas in a raster map) in the neighborhood of the search starting point as path nodes so that the path nodes are not located at obstacles. On the occupied area, the robot searches for multiple path nodes in the neighborhood of the search starting point. These path nodes are all free grids, and each path node corresponds to an initial search direction; then it continues to update each path node as A new search starting point, and then search free grids in the neighborhood of each search starting point in the direction of the search starting point pointing to the same preset target position. Repeat this until the preset target position is searched in the grid map, and follow the The search sequence connects the searched grids into multiple navigation paths or obstacle avoidance paths at the same time; wherein, the robot searches for one navigation path or obstacle avoidance path in each initial search direction.

Preferably, when the robot first detects the first obstacle using a line laser sensor, the robot has collided with the first obstacle; when the robot walks to the gap formed between the first obstacle and the second obstacle, the robot has collided with the third obstacle. Two obstacles; when the robot sets the preset distance range to be greater than or equal to the width of the robot's body, so that the distance between the two obstacle bypass points is greater than or equal to the width of the robot's body, the two obstacles are determined. The gap formed between the obstacles where the obstacle detour points are located is a narrow crossing that allows the robot to enter. The upper limit of the preset distance range is the sum of the robot's body width and the preset redundancy. Since the width of the narrow channel is only slightly larger than the width of the robot's body, in order to reduce the error interference of the calculated contour line in the process of identifying the narrow channel, a preset distance range is set for judgment; specifically, the minimum value of the preset distance range (Lower limit value) is larger than the robot's body width, and the maximum value (upper limit value) of the preset distance range is only slightly larger than the robot's body width, not exceeding twice the body width. The preset distance range can be determined according to the width of the robot's body and the preset fitting error. The setting of the preset redundancy can call the channel that is closer to the width of the robot's body the narrow channel. For example, if the width of the sweeping robot is 30 cm, the preset distance range may be 32 cm to 35 cm. That is, when the center of the robot is located at the center of the width direction of the entrance of the channel where the two obstacle bypass points are located, the The setting of the preset redundancy shows that the gap size between the left and right sides of the robot and the channel is between 1 cm and 2 cm. Correspondingly, the center position of the robot can use position R1, position R2, position R3, and position in sequence. R4, position R5, and position R6 are represented. To sum up, the foregoing embodiment uses the change in the walking direction before entering the gap formed between the two obstacles and the distance between the obstacle avoidance points corresponding to the two obstacles to identify whether the robot has started. Entering a narrow passage, overcoming the error caused by the contour fitting calculation of obstacles in identifying the narrow passage, improving the accuracy of narrow passage identification and the success rate of the robot entering the narrow passage.

As an embodiment, the robot sets the search direction of the positioning contour line segment in the vertical direction of the positioning contour line segment to form the width direction of the channel where the two obstacle bypassing points are located; the two obstacle bypassing points are along the The search directions of the positioning contour line segments are distributed among: the latest obstacles that the robot collides with before it starts to recognize the narrow crossing and the obstacles that the robot collides with when it starts to recognize the narrow crossing. In this embodiment, the positioning contour line segment corresponds to the obstacle where one of the obstacle bypass points is located, corresponding to the uppermost obstacle in Figure 2. The line segment AB is its positioning contour line segment, which belongs to the topmost obstacle passing through Figure 2. The fitted line segment of the lower left corner of the obstacle, but does not necessarily belong to the actual contour line of the obstacle, can form one of the boundary lines of the channel where the two obstacle bypass points are located, that is, the narrow channel identified in the previous embodiment. The boundary line, where the vertical line formed in the search direction of line segment A1B1 can pass through a corner point of the uppermost obstacle (black rectangle) shown in Figure 2 to fit the outline and distribution position of the obstacle. The two obstacle bypassing points are points on the positioning contour line segment of the obstacle closest to the left and right sides of the robot; it can also be understood as the two endpoints with the smallest distance among the endpoints of the two positioning contour line segments and the two endpoints are the smallest. There is a gap between the end points; in some embodiments, the distance between the two obstacle bypass points can also be the minimum distance between the end points of the positioning contour segments where the two obstacle bypass points are located, to determine The passability of the gap between the two obstacle bypass points.

Preferably, the preset distance range is set to be greater than or equal to the body width of the robot; the preset safety distance is set to be equal to the body width of the robot; and the preset angle is set to guide the robot away from the current collision area. Obstacle, or any endpoint of the positioning contour line segment that bypasses the currently collided obstacle, in order to walk to the passable area connected to the preset target position; the shortest passable path is the positioning contour of the robot respectively. The two endpoints of the line segment are the starting point of the search, and the path with the shortest path length among the paths leading to the preset target location planned using a heuristic search algorithm is used.

Preferably, when the robot walks inside the narrow passage, the shortest distance between the boundary lines on both sides of the entrance of the narrow passage and the corresponding side of the robot is equal to half of the preset redundancy amount; for example, the robot moves along When entering a narrow crossing along the center line of a narrow passage, the vertical distance between the robot's left side and its left side around the obstacle point is equal to half of the preset redundancy, and the vertical distance between the robot's right side and its right side around the obstacle is also equal to half of the preset redundancy amount. In this embodiment, half of the preset redundancy is preferably 1 cm to 2 cm, so as to avoid fitting calculation errors of the contour line.

In some embodiments, in the working area of the robot, there will be a working area framed by a boundary line, and a non-working area will also be formed between obstacles, or a non-working area will be formed between the working areas framed by the boundary line. The non-working area is formed between the corresponding boundary lines of two working areas, or the non-working area is formed by the gap channel between two wall obstacles; the non-working area here generally refers to a channel with a narrow width. Referred to as a narrow passage, the entrance and exit of the narrow passage are both set as narrow passages. On both sides of the narrow passage are the outlines of two obstacles; and when setting the moving path, the sweeping robot is generally set to a Viewed in a circle, in order to facilitate the sweeping robot to move to the preset target position, the set movement path may pass through a narrow passage; generally, the gap area between two or more obstacles (such as the passage between two walls) ) is displayed in the grid map in the form of a free channel; a free channel refers to connecting two different working areas. It can be a passable road for the robot on the premise that the width is greater than the width of the robot's body, and then specifically passes through the entrance of the detection channel , the exit of the passage, the length of the passage and the width of the passage to identify the narrow entrance and narrow passage that allow the robot to pass, and determine the location information covered by the narrow passage, such as identifying the grid area of a door opening under a wall, two The passage between walls and three walls form a passage with only one gap to be identified; the passage in the actual working scene of the robot has certain characteristics, including three-dimensional shape characteristics and size characteristics.

It should be added that for the aforementioned narrow crossings, narrow passages or gaps, in the area where the robot is located, there are door openings on the horizontal ground that penetrate two adjacent room areas, and a gap between two parallel walls. , the gap formed between two low obstacles, the entrance of the doorway or the entrance of the gap can be set as the opening formed between the two obstacles, which is the opening formed between the outlines of at least two obstacles, these Obstacles can exist isolated from each other; when the robot uses a line laser sensor to scan the surrounding environment, the two endpoints and width of the opening are scanned by the robot's laser sensor and converted into point cloud coordinates in the corresponding coordinate system, and then the opening is The two endpoints are scanned into corresponding point clouds and converted to the corresponding raster of the raster map. In some embodiments, when the robot senses the surrounding environment through collision, whenever the robot's collision sensor contacts the two endpoints of the opening, the contour point of the obstacle or the positioning contour line segment mark that it collides with is to the corresponding grid of the grid map; an evaluation quantity is also introduced in some implementations to represent the gap formed by the robot between obstacles or its corresponding grid (the number of free grids between two endpoints) or its proportion within the opening width range), which can represent the passability probability at the corresponding grid area. Generally speaking, when a robot uses a ranging sensor to scan a gap, it is given a higher degree of credibility than the robot. Collisions are given confidence when detecting a gap because the positioning accuracy of the ranging sensor is higher than that resulting from physical contact with the robot.

In the aforementioned embodiment, the robot collects point cloud data through the ranging sensor. The point cloud data is configured to reflect the position information of the obstacles detected by the ranging sensor. It is a collection of a series of discrete points and will carry environmental noise points (feedback The ambient light interference or the influence of the obstacle's material); the robot then fits the collected point cloud data into the outline of the obstacle to represent the local outline of the detected obstacle or the envelope of the obstacle. ; The fitting process here will specifically go through sorting, grouping (to distinguish different types of obstacles), filtering, piecewise interpolation fitting, and then fitting the curve coordinate points of each group of connections to obtain the obstacle envelope. Or a contour line, specifically a fitting line segment, a fitting curve and a combination thereof. Preferably, the robot controls the contour line obtained by the aforementioned fitting process to perform virtual expansion processing, and its expansion radius is set to the radius of the inscribed circle of the robot contour projection. Or the radius of the circumscribed circle, forming a fitted circle surrounding the obstacle, corresponding to the obstacle envelope, then the aforementioned positioning contour line segment belongs to the obstacle envelope.

Based on the foregoing embodiments, the present invention also discloses a chip that stores program code. When the program code is executed, the steps disclosed in the foregoing embodiments are implemented. When the program code corresponding to the steps of the method for navigating a robot through a narrow passage is stored in a chip, it is regarded as a computer program product. The computer program is operable to cause the computer to perform the navigation of the robot through a narrow passage. Some or all steps of any method described in the method examples. The robot equipped with the aforementioned chip inside identifies the narrow passage before and after the intersection and before entering the narrow passage. The robot uses point cloud data to obtain the positioning contour line segment and searches for relevant obstacle avoidance points. It identifies the narrow passage with a width slightly larger than the robot's body width and determines its width. Being able to enter the narrow passage and continuously enter multiple narrow passages and pass through the narrow passage and continuously pass through multiple narrow passages will also facilitate real-time planning of a navigation path with stronger environmental adaptability and smoother traffic.

The invention also discloses a robot equipped with a distance sensor. The robot is equipped with the chip, which is used to control the robot to use the distance sensor to detect point cloud data of obstacles and fit the corresponding positioning from the point cloud data. Contour line segments facilitate the passage of narrow passages by executing the robot navigation method. Before the robot starts to recognize the narrow crossing, when it starts to recognize the narrow crossing, before and after recognizing the narrow crossing, before entering the narrow crossing, and in the corresponding stages of entering the narrow crossing, the robot determines the collision between the robot and the obstacle and the fitting of the outline of the obstacle. situation, adjust the walking direction of the robot, and then walk in a straight line in the narrow channel to tend to walk towards the preset target position; and identify based on the collision of the robot with two obstacles and the fitting of the contours of each obstacle. Narrow crossings are used to distinguish three recognition states before starting to recognize narrow crossings, when starting to recognize narrow crossings, and recognizing narrow crossings, thereby promoting the process of the robot entering and re-entering narrow crossings.

It should be noted that for the sake of simple description, the foregoing method embodiments are expressed as a series of action combinations. However, those skilled in the art should know that the present application is not limited by the described action sequence. Because in accordance with this application, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily necessary for this application.

In the above embodiments, each embodiment is described with its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

In the several embodiments provided in this application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or may be Integrated into another system, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be in electrical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated units can be implemented in the form of hardware or software functional units.

If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application is essentially or contributes to the existing technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a memory, It includes several instructions to cause a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable memory. The memory can include: a flash disk. , Read-Only Memory (ROM), Random Access Memory (RAM), magnetic disk or optical disk, etc.

An embodiment of the present application has been introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, For those of ordinary skill in the art, there will be changes in the specific implementation and application scope based on the ideas of the present application. In summary, the content of this description should not be understood as a limitation of the present application.

Claims (14)

1. A navigation method for a robot through a narrow passage, characterized in that the navigation method includes:

   Step S1. After the robot enters the narrow passage, if the robot collides with an obstacle, it will fit a positioning contour segment to the obstacle, and then rotate until the current walking direction of the robot matches the preset positioning contour segment currently fitted. If the traffic direction is the same, or the current walking direction of the robot is deflected to the trafficable area by a preset angle relative to the currently fitted positioning contour line segment, so that the robot is not blocked by obstacles, then step S2 is executed;

   Step S2: The robot walks in a straight line in the current walking direction until it collides with an obstacle, and then executes step S1; when the robot walks to the preset target position, it stops executing step S1 and determines that the robot passes through the narrow passage.
2. The navigation method according to claim 1, characterized in that, in the step S1, the robot sets the end point of the shortest accessible path relative to the preset target position among the positioning contour segments it fits as the positioning The target endpoint of the contour line segment, where the shortest passable path is located within the passable area of the narrow channel entered by the robot; the robot configures the positioning contour with the endpoints other than the target endpoint of the positioning contour line segment pointing in the direction of the target endpoint. The default travel direction of the line segment.
3. The navigation method according to claim 2, characterized in that the robot configures the channel where the target endpoint of each positioning contour segment is located as a narrow channel; the robot sets the positioning contour segment as the boundary line of the narrow channel;

   Whenever the robot collides with a new obstacle, it will fit a positioning contour segment for the new obstacle, obtain the target endpoint of the positioning contour segment of the new obstacle, and identify the opening of the channel where the target endpoint is located. It is a narrow crossing.
4. The navigation method according to claim 2, characterized in that, in step S2, before the robot walks to the preset target position, if the robot recognizes a narrow crossing, it will identify the narrow crossing as the exit of the narrow crossing, Then the robot adjusts its current walking direction to the preset traffic direction of the currently fitted positioning contour line segment, or deflects it at a preset angle to the passable area relative to the currently fitted positioning contour line segment; then the robot follows the current walking direction in a straight line Walk to get out of the narrow path.
5. The navigation method according to claim 2, characterized in that the navigation method further includes: before the robot enters the narrow passage, if the robot recognizes the narrow passage entrance, then the robot identifies the narrow crossing as the entrance of the narrow passage, and then the robot Its current walking direction is adjusted to the preset traffic direction of the currently fitted positioning contour line segment, or the current walking direction of the robot is deflected to the passable area by a preset angle relative to the currently fitted positioning contour line segment; then the robot walks according to the current Walk straight in the direction to enter the entrance of the narrow passage, and then perform step S1.
6. The navigation method according to claim 5, characterized in that, before the robot starts to recognize the narrow crossing, if the robot collides with an obstacle, the obstacle is marked as the latest obstacle detected before the robot starts to recognize the narrow crossing, Then, a positioning contour segment is fitted to the obstacle, and then the robot adjusts its current walking direction to be the same as the preset direction of the currently fitted positioning contour segment, so that the robot follows the currently fitted positioning contour segment. Walk until you collide with a new obstacle and fit the corresponding positioning contour line segment for the new obstacle, and then start to identify the narrow intersection.
7. The navigation method according to claim 2, characterized in that whenever the robot rotates to its current walking direction which is the same as the preset traveling direction of the currently fitted positioning contour segment, or when it rotates to its current walking direction which is relative to the current fitted positioning contour segment, When the positioning contour line segment deflects to the passable area by the preset angle, the robot walks straight in the current walking direction at a preset safe distance, and detects whether it collides with an obstacle during the straight walking. If so, step S1 is executed, otherwise Determine that the robot has passed through a narrow passage, and then use a heuristic search algorithm to plan a path to the preset target position, or a path to the position closest to the obstacle detected in the current walking direction, and then follow The currently planned path is walked until the robot recognizes the entrance of another narrow lane and enters the narrow lane, and then executes step S1.
8. The navigation method according to any one of claims 4 to 5, characterized in that the method for the robot to identify a narrow crossing includes:

   The robot first fits the positioning contour segment to the latest obstacle it collides with, and then extracts two obstacle bypass points in the search direction of the positioning contour segment. If it is detected that the distance between the two obstacle bypass points is within the preset Assuming a distance range, the passage where the two obstacle bypassing points are located is identified as a narrow passage, and the gap formed between the obstacles where the two obstacle bypassing points are located is identified as a narrow crossing.
9. The navigation method according to claim 8, characterized in that after the robot recognizes the narrow crossing, if it is detected that the walking direction of the robot changes from the reference obstacle bypassing direction to the preset passing direction or relative direction of the currently fitted positioning contour line segment, When the currently fitted positioning contour line segment is deflected to the passable area by the preset angle, and the robot walks in a straight line according to the changed walking direction, the robot recognizes that it has entered a narrow road;

   Wherein, the changed walking direction is configured such that the current position of the robot points to the direction of the passable area inside the channel where the two obstacle bypass points are located;

   Wherein, the reference obstacle bypassing direction is the preset passing direction of the positioning contour line segment fitted last time or the preset angle deflected to the passable area relative to the positioning contour line segment fitted last time;

   Among them, the last fitted positioning contour line segment is the positioning contour line segment fitted for the latest obstacle encountered before the robot starts to recognize the narrow crossing.
10. The navigation method according to claim 9, characterized in that, before starting to identify a narrow crossing, the robot collides with an obstacle and fits the corresponding positioning contour segment; then, when the robot collides with another obstacle, it starts to identify the narrow crossing. crossing; crossing

    The positioning contour line segment fitted by the robot for an obstacle it collided with previously is the positioning contour line segment fitted last time, and the positioning contour line segment fitted by the robot for an obstacle it collided with later is the positioning contour line segment fitted by the robot. The currently fitted positioning contour segment makes the preset traveling direction of the last fitted positioning contour segment different from the preset traveling direction of the currently fitted positioning contour segment.
11. The navigation method according to claim 10, characterized in that the robot sets the search direction of the positioning contour line segment in the vertical direction of the positioning contour line segment to form the width direction of the channel where the two obstacle bypass points are located;

    The two obstacle bypass points are distributed along the search direction of the positioning contour line segment in: the latest obstacle that the robot collides with before it starts to recognize the narrow crossing, and the obstacle that the robot collides with when it starts to recognize the narrow crossing.
12. The navigation method according to claim 8, characterized in that the preset distance range is set to be greater than or equal to the body width of the robot; the preset safety distance is set to be equal to the body width of the robot;

    The preset angle is set to guide the robot away from the obstacle it currently collides with, or around any end point of the positioning contour segment of the obstacle it currently collides with, so as to walk to a possible location connected to the preset target position. access area;

    The shortest traversable path is the path with the shortest path length among the paths leading to the preset target position planned by the robot using a heuristic search algorithm using the two endpoints of the positioning contour line segment as the starting point of the search.
13. A chip stores program code, characterized in that when the program code is executed, the steps of the navigation method according to any one of claims 1 to 12 are implemented.
14. A robot equipped with a distance sensor, characterized in that the robot is equipped with the chip according to claim 13 for controlling the robot to use the distance sensor to detect point cloud data of obstacles, and to fit the corresponding point cloud data from the point cloud data The positioning contour line segment facilitates passing through the narrow passage by executing the navigation method robot navigation method according to any one of claims 1 to 12.