A method for automatically constructing SLAM maps for AGV systems

Abstract

The present invention provides an automatic SLAM map construction method for an AGV system. After the AGV is blocked, the AGV is made to retreat a set distance from the blocked point, and then the AGV is made to move around the obstacle from the initial obstacle avoidance point R1 until it stops when passing through the straight line segment P0P1 and arrives at the obstacle avoidance end point R2. The AGV is then made to move along the straight line segment R2P1. If an obstacle is encountered, the step is repeated. If the AGV moves smoothly to the transport point P1, the path along which the AGV moves smoothly is marked as a clear path, and the path along which the AGV moves around the obstacle is marked as an obstacle avoidance path, and the SLAM map is updated. By making another AGV give priority to using the clear path and obstacle avoidance path previously marked by the AGV when planning a path, and updating the SLAM map after the planning is completed, the AGV's own characteristics can be used to automatically realize the planning of the obstacle avoidance path, reduce the number of manual interventions, improve the efficiency of SLAM map construction, and continuously iterate the SLAM map to further improve the efficiency of SLAM map construction.

Description

Technical Field

[0001] The present invention relates to the technical field of map construction, and in particular to a SLAM map automatic construction method for an AGV system. Background Art

[0002] AGV (Automated Guided Vehicle), also known as automatic guided vehicle, automatic guided transporter, and automatic guided transporter, is an industrial vehicle that can automatically load goods, automatically drive along a set route or tow a cargo trolley to a designated location, and then automatically load and unload goods.

[0003] In general, an AGV system consisting of multiple AGVs needs to carry out cargo handling operations under the guidance of a SLAM map. When entering the handling site for the first time, it is necessary to first create a SLAM map by manually scanning the map, then set handling points on the map, and connect the points to form a route to guide the AGV. To ensure the implementation progress, such connections are usually not tested on actual vehicles. When the AGV cannot pass, the operator needs to enter the handling site for actual inspection, and then optimize the AGV's route. After the AGV actually passes, the route is updated in the SLAM map to build a complete SLAM map. To ensure safety, when the operator enters the handling site, the entire AGV system needs to stop running, which will greatly affect the utilization rate of the AGV. In particular, when goods are piled up at certain points, it is necessary to find another location for temporary storage. This type of temporary storage has a high probability of affecting the normal passage of the AGV, and will frequently cause abnormalities, causing the operator to frequently enter and exit the handling site, affecting the normal operation of the AGV system. Summary of the invention

[0004] The purpose of the present invention is to solve one or more problems in the prior art and to provide a method for automatically constructing a SLAM map for an AGV system.

[0005] To achieve the above object, the technical solution adopted by the present invention is that the AGV system uses a SLAM map automatic construction method, comprising the following steps:

[0006] S1. Use the contour line of the handling site as the boundary line of the SLAM map to initially establish the SLAM map;

[0007] S2. Mark all the transport points P0, P1, P2, P3, ..., P in the SLAM map i ;

[0008] S3. The AGV moves along the straight line segment P0P1 from the transport point P0 to the transport point P1. If there is any resistance, proceed to step S4. If the transport point P1 is moved smoothly, proceed to step S5.

[0009] S4. The AGV is made to retreat a set distance from the obstruction point to the initial obstacle avoidance point R1, and then the AGV is made to move around the obstacle from the initial obstacle avoidance point R1 until the AGV passes through the straight line segment P0P1 and stops, and reaches the obstacle avoidance end point R2, and then the AGV is made to move along the straight line segment R2P1. If an obstacle is encountered, step S4 is repeated. If it moves smoothly to the transfer point P1, step S5 is performed;

[0010] S5. The path along which the AGV moves smoothly is marked as a clear path, the path along which the AGV moves around an obstacle is marked as an obstacle-avoiding path, and the SLAM map is updated;

[0011] S6. Move another AGV along the straight line segment P2P3 from the transport point P2 to the transport point P3. If the straight line segment P2P3 has an intersection X with the clear path and the obstacle avoidance path marked in step S5 (the intersection X refers to the intersection point where the clear path and the obstacle avoidance path marked in step S5 meet for the first time), proceed to step S7. If the straight line segment P2P3 has no intersection with the clear path and the obstacle avoidance path marked in step S5, proceed to step S3.

[0012] S7. The AGV moves along the straight line segment P2X from the transport point P2 to the intersection X. If there is an obstacle, proceed to step S4. If it moves smoothly to the intersection X, proceed to step S8.

[0013] S8. Move the AGV along the clear path and / or obstacle-avoiding path marked in step S5 to the point Y on the straight line segment P0P1 closest to the transfer point P3. After arriving, move the AGV along the straight line segment YP3 to the transfer point P3. If there is an obstacle, proceed to step S4. If it moves smoothly to the transfer point P3, proceed to step S5.

[0014] S9. Repeat step S6 until all transport points are on the clear path and / or obstacle-avoiding path in step S5, and the construction of the SLAM map is completed.

[0015] Preferably, the obstruction in step S3 is based on the detection result of the detection system on the AGV.

[0016] Preferably, during the process of moving around the obstacle in step S4, a detection system on the AGV is used to keep a set distance between the AGV and the obstacle.

[0017] Further preferably, the detection system includes a laser rangefinder, an ultrasonic radar, and a visual camera.

[0018] Preferably, when retreating along the original route in step S4, if the set distance is not reached after retreating to the transport point P0, the AGV stops moving and issues an alarm.

[0019] Preferably, when the obstacle avoidance end point R2 is reached in step S4, if the obstacle avoidance end point R2 is the same as the obstacle avoidance initial point R1, the AGV stops moving and issues an alarm.

[0020] Preferably, before the SLAM map is updated in step S5, a step of converting the obstacle avoidance path into a clear path is also included, and the step includes:

[0021] A. Fit the obstacle avoidance path through a cubic Bezier curve to obtain a fitting path;

[0022] B. Move the AGV along the fitting path from one end point to another end point;

[0023] C. If the route is successfully reached, the obstacle avoidance path is replaced by the fitted path and the fitted path is marked as a clear path, and the SLAM map is updated. If an obstacle is encountered, step A is repeated;

[0024] D. If step A is repeated for the set number of times and there is still resistance, the AGV stops moving and sounds an alarm.

[0025] Preferably, in steps S3 to S9, when the AGV is in a clear path, its travel speed is a set speed, and when the AGV is in an obstacle avoidance path, its travel speed is one-fifth to one-third of the set speed.

[0026] Preferably, in steps S3 to S9, at any time, there is at most one AGV on the clear path and / or obstacle-avoiding path connecting two transporting points.

[0027] Preferably, the obstacle avoidance path in step S5 is detected every set running time, and if the AGV can move smoothly on the straight line segment between the two end points of the obstacle avoidance path, the obstacle avoidance path is replaced by the straight line segment to form a new smooth path.

[0028] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art:

[0029] 1. It can use the characteristics of AGV itself to automatically plan obstacle avoidance paths, reduce the number of manual interventions, and improve the efficiency of SLAM map construction.

[0030] 2. It can utilize the path successfully planned by the previous AGV to reduce the workload of the next AGV in planning the path, and further improve the efficiency of SLAM map construction.

[0031] 3. When obstacles temporarily appear on the planned path, they can also be avoided and the path can be planned reasonably. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the principle of a preferred embodiment of the present invention. At this time, no cubic Bezier curve fitting is performed, and the obstacle is a column with a rectangular cross-section.

[0033] Figure 2 It is a schematic diagram of the principle of a preferred embodiment of the present invention. At this time, no cubic Bezier curve fitting is performed, and the obstacle is a column with a circular cross-section.

[0034] Figure 3 yes Figure 1 Schematic diagram after completing cubic Bezier curve fitting.

[0035] Figure 4 yes Figure 1 Schematic diagram of path planning after the obstacles are removed.

[0036] Among them: 10. Boundary line; 11. Obstacle; 20. Clear path; 30. Obstacle avoidance path; 40. Fitting path. DETAILED DESCRIPTION

[0037] like Figures 1 to 4 As shown, the SLAM map automatic construction method for the AGV system provided by the present invention comprises the following steps:

[0038] S1. Using the contour of the handling site as the boundary line of the SLAM map 10, a preliminary SLAM map is established;

[0039] S2. Mark all the transport points P0, P1, P2, P3, ..., P in the SLAM map i ;

[0040] S3. Move the AGV along the straight line segment P0P1 from the transport point P0 to the transport point P1. If there is any obstacle, proceed to step S4. If the AGV moves smoothly to the transport point P1 (reference Figures 1 to 3 From P0' to P1'), then proceed to step S5;

[0041] S4. The AGV is made to retreat a set distance D from the obstruction point to the initial obstacle avoidance point R1, and then the AGV is made to move around the obstacle from the initial obstacle avoidance point R1 until the AGV passes through the straight line segment P0P1 and stops, and reaches the obstacle avoidance end point R2, and then the AGV is made to move along the straight line segment R2P1. If an obstacle is encountered, step S4 is repeated. If it moves smoothly to the transfer point P1, step S5 is performed;

[0042] S5. The path along which the AGV moves smoothly is marked as a clear path 20, and the path along which the AGV moves around an obstacle is marked as an obstacle-avoiding path 30, and the SLAM map is updated;

[0043] S6. Move another AGV along the straight line segment P2P3 from the transport point P2 to the transport point P3. If the straight line segment P2P3 has an intersection X with the clear path 20 and the obstacle avoidance path 30 marked in step S5 (the intersection X refers to the intersection point where the clear path and the obstacle avoidance path marked in step S5 meet for the first time), proceed to step S7. If the straight line segment P2P3 has no intersection with the clear path 20 and the obstacle avoidance path 30 marked in step S5 (refer to Figures 1 to 3 P2' to P3'), then proceed to step S3 (refer to the operation method of step S3 to move the AGV from P2' to P3');

[0044] S7. Move the AGV along the straight line segment P2X from the transport point P2 to the intersection X. If there is any resistance, proceed to step S4 (refer to the operation method of step S4 to move the AGV from P2 to X). If it moves smoothly to the intersection X, proceed to step S8;

[0045] S8. Move the AGV along the clear path 20 and / or obstacle-avoiding path 30 marked in step S5 to the point Y on the straight line segment P0P1 closest to the transfer point P3. After arriving, move the AGV along the straight line segment YP3 to the transfer point P3. If there is an obstacle, proceed to step S4. If it moves smoothly to the transfer point P3, proceed to step S5.

[0046] S9. Repeat step S6 until all transport points are on the clear path 20 and / or the obstacle-avoiding path 30 in step S5, and the construction of the SLAM map is completed.

[0047] The advantages of this setting are:

[0048] 1. It can use the characteristics of AGV itself to automatically plan obstacle avoidance paths, reduce the number of manual interventions, and improve the efficiency of SLAM map construction.

[0049] 2. It can utilize the path successfully planned by the previous AGV to reduce the workload of the next AGV in planning the path, and further improve the efficiency of SLAM map construction.

[0050] 3. When obstacles temporarily appear on the planned path, they can also be avoided and the path can be planned reasonably.

[0051] In this embodiment, the obstacle encountered in step S3 is based on the detection result of the detection system on the AGV. In step S4, during the movement around the obstacle, the detection system on the AGV is used to maintain a set distance between the AGV and the obstacle. The detection system includes a laser rangefinder, an ultrasonic radar, and a visual camera. This setting can make full use of the characteristics of the AGV itself and realize automated operation.

[0052] In this embodiment, when the AGV retreats along the original route in step S4, if it still does not reach the set distance after retreating to the transport point P0, it means that the set distance is incorrect or the transport point P0 is too close to the obstacle, and manual intervention is required. At this time, the AGV stops moving and issues an alarm; at the same time, when the AGV reaches the obstacle avoidance end point R2 in step S4, if the obstacle avoidance end point R2 is the same position as the obstacle avoidance initial point R1, it means that the transport point P1 is located inside the faulty object, and manual intervention is required. At this time, the AGV stops moving and issues an alarm.

[0053] Preferably, before the SLAM map is updated in step S5, a step of converting the obstacle avoidance path 30 into a clear path is also included, which step includes:

[0054] A. Fitting the obstacle avoidance path by a cubic Bezier curve to obtain a fitting path 40;

[0055] B. moving the AGV along the fitting path 40 from one end point to another end point of the fitting path 40;

[0056] C. If the route is successfully reached, the obstacle avoidance path 30 is replaced by the fitting path 40 and the fitting path 40 is marked as the clear path 20 (refer to Figure 2 ), and update the SLAM map. If there is an obstacle, repeat step A;

[0057] D. If step A is repeated for the set number of times and there is still resistance, the AGV stops moving and sounds an alarm.

[0058] The advantage of this setting is that it can make each path in the updated SLAM map smoother, reduce the turning angle and the number of stops of the AGV during its movement, so that the AGV can move at a set speed.

[0059] Preferably, in steps S3 to S9, when the AGV is in the clear path 20, its travel speed is a set speed, and when the AGV is in the obstacle avoidance path 30, its travel speed is one fifth to one third of the set speed.

[0060] Preferably, in steps S3 to S9, at any time, there is at most one AGV on the clear path 20 and / or the obstacle-avoiding path 30 connecting two transporting points to avoid interference.

[0061] Preferably, the obstacle avoidance path 30 in step S5 is detected every set running time. If the AGV can move smoothly on the straight segment between the two end points of the obstacle avoidance path 30, the obstacle avoidance path 30 is replaced by the straight segment to form a new unobstructed path 20. This setting can timely optimize the path in the SLAM map after the temporary obstacle is removed, making it more efficient.

[0062] It should be noted that the AGV system has multiple AGVs and a control module that can obtain real-time status information of multiple AGVs. The control module is connected to each AGV through wireless communication. The real-time status information of the AGV includes: coordinate information (X, Y, map ID, vehicle angle, etc.), control information (fork height, control status, map status), sensor information (front obstacle avoidance sensor, steering obstacle avoidance sensor, fork obstacle avoidance sensor, fork cargo sensor, temperature sensor, etc.), battery information (battery power, battery voltage, battery current, battery health, etc.), etc.; at the same time, the control module also dispatches each AGV by sending tasks and control commands to each AGV.

[0063] Of course, to facilitate manual intervention, the AGV system also includes a PC for manual intervention. The operator can input A, S, D, and W on the PC keyboard to control the left turn, forward, right turn, and backward of a single AGV. The operator can also click the logical buttons on the system interface to realize operations such as lowering and raising the fork and opening and closing the I / O interface.

[0064] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with this technology to understand the contents of the present invention and implement them accordingly. They cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatically constructing a SLAM map for an AGV system, characterized in that: The following steps are involved: S1. Use the contour line of the handling site as the boundary line of the SLAM map to initially establish the SLAM map; S2. Mark all the transport points P0, P1, P2, P3, ..., P in the SLAM map i ; S3. The AGV moves along the straight line segment P0 P1 from the transport point P0 to the transport point P1. If there is any obstruction, proceed to step S4. If the AGV moves smoothly to the transport point P1, proceed to step S5. S4. The AGV is made to retreat a set distance from the obstruction point to the initial obstacle avoidance point R1, and then the AGV is made to move around the obstacle from the initial obstacle avoidance point R1 until the AGV passes through the straight line segment P0P1 and stops, and reaches the obstacle avoidance end point R2, and then the AGV is made to move along the straight line segment R2P1. If an obstacle is encountered, step S4 is repeated. If it moves smoothly to the transfer point P1, step S5 is performed; S5. The path along which the AGV moves smoothly is marked as a clear path, the path along which the AGV moves around an obstacle is marked as an obstacle-avoiding path, and the SLAM map is updated; S6. Move another AGV along the straight line segment P2 P3 from the transport point P2 to the transport point P3. If the straight line segment P2 P3 has an intersection X with the clear path and the obstacle avoidance path marked in step S5, proceed to step S7. If the straight line segment P2P3 has no intersection with the clear path and the obstacle avoidance path marked in step S5, proceed to step S3. S7. The AGV moves along the straight line segment P2X from the transport point P2 to the intersection X. If there is an obstacle, proceed to step S4. If it moves smoothly to the intersection X, proceed to step S8. S8. Move the AGV along the clear path and / or obstacle-avoiding path marked in step S5 to the point Y on the straight line segment P0P1 closest to the transfer point P3. After arriving, move the AGV along the straight line segment YP3 to the transfer point P3. If there is an obstacle, proceed to step S4. If it moves smoothly to the transfer point P3, proceed to step S5. S9. Repeat step S6 until all transport points are on the clear path and / or obstacle-avoiding path in step S5, and the construction of the SLAM map is completed.

2. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: The obstruction in step S3 is based on the detection result of the detection system on the AGV.

3. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: In step S4, during the movement around the obstacle, the detection system on the AGV is used to keep a set distance between the AGV and the obstacle.

4. The method for automatically constructing a SLAM map for an AGV system according to claim 2 or 3, characterized in that: The detection system includes a laser rangefinder, an ultrasonic radar, and a visual camera.

5. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: When retreating along the original route in step S4, if the set distance is not reached after retreating to the transport point P0, the AGV stops moving and issues an alarm.

6. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: When arriving at the obstacle avoidance end point R2 in step S4, if the obstacle avoidance end point R2 is the same as the obstacle avoidance initial point R1, the AGV stops moving and issues an alarm.

7. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: Before the SLAM map is updated in step S5, a step of converting the obstacle-avoiding path into a clear path is also included, which step includes: A. Fit the obstacle avoidance path through a cubic Bezier curve to obtain a fitting path; B. Move the AGV along the fitting path from one end point to another end point; C. If the route is successfully reached, the obstacle avoidance path is replaced by the fitted path and the fitted path is marked as a clear path, and the SLAM map is updated. If an obstacle is encountered, step A is repeated; D. If step A is repeated for the set number of times and there is still resistance, the AGV stops moving and sounds an alarm.

8. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: In steps S3 to S9, when the AGV is in a clear path, its travel speed is a set speed, and when the AGV is in an obstacle avoidance path, its travel speed is one fifth to one third of the set speed.

9. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: In steps S3 to S9, at any time, there is at most one AGV on the clear path and / or obstacle-avoiding path connecting two transport points.

10. The method for automatically constructing a SLAM map for an AGV system according to claim 1, characterized in that: Every set operation time, the obstacle avoidance path in step S5 is detected. If the AGV can move smoothly on the straight line segment between the two end points of the obstacle avoidance path, the obstacle avoidance path is replaced by the straight line segment to form a new smooth path.

Images