A double-deck drum AGV and method

Abstract

The application relates to the technical field of drum AGV, and discloses a double-layer drum AGV and a method, which comprises the following steps: an outer shell is provided, a bottom plate is fixed to the upper surface of the outer shell, a blocker is installed in the inner part of the outer shell, a double-layer drum module is installed in the inner part of the outer shell, a lever floating type chassis is installed on the lower surface of the bottom plate, a driving wheel is connected to the lever floating type chassis, a laser radar is fixed to the outer side of the bottom plate, and a downward-looking camera is embedded in the lower surface of the bottom plate; the driving wheel constitutes a two-drive differential walking steering system, and the laser radar is linked with a safety detection system. Through the differential steering structure of the lever floating type chassis and the symmetrical driving wheel, the steering angles of the driving wheels on the two sides are independently controlled through a servo motor, self-adaptive path tracking under complex ground conditions is realized, and the technical defects of travel trajectory deviation and steering response lag caused by uneven ground are solved.

Description

Technical Field

[0001] This invention relates to the field of roller AGV technology, specifically to a double-layer roller AGV and its method. Background Technology

[0002] In industrial automation scenarios, traditional AGVs mostly adopt rigid chassis and single-drive steering structures, which have inherent limitations in their mechanical design. When the ground is tilted or uneven, the rigid connection causes uneven stress distribution between the drive wheels and the ground, which can easily cause the driving trajectory to deviate from the preset path. Frequent correction operations not only increase energy consumption but also accelerate the wear of the steering mechanism, seriously restricting the stability of long-cycle operations.

[0003] Existing AGVs based on QR code navigation generally use a fixed scanning frequency positioning mode, which exposes significant defects under high-speed operation. Due to the mismatch between vehicle displacement speed and image acquisition rate, QR code recognition is prone to lag or missed detection at high speeds, forcing the vehicle to repeatedly perform deceleration and repositioning operations, resulting in interruption of the material handling task and failing to meet the efficiency requirements of modern high-cycle production lines.

[0004] Conventional obstacle avoidance systems rely on horizontally mounted lidar, whose monitoring plane is parallel to the ground, creating a vertical blind spot. For obstacles below the radar scanning plane (such as debris on the ground or protruding parts of equipment), the system cannot generate an effective warning signal. Such missed detections can easily lead to equipment collisions or material spills, forcing companies to add manual inspections, significantly increasing the complexity of production line maintenance and safety risks. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a double-layer roller AGV and method, which solves the problems of trajectory deviation and sluggish steering response caused by uneven ground in roller AGVs.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a double-layer roller AGV, comprising: a shell, a base fixed to the upper surface of the shell, a stopper installed inside the shell, a double-layer roller module installed inside the shell, a lever-floating chassis installed on the lower surface of the base, a drive wheel connected to the lever-floating chassis, a laser radar fixed to the outer side of the base, and a downward-looking camera embedded in the lower surface of the base; the drive wheel constitutes a two-drive differential walking and steering system, the laser radar is linked with a safety detection system, the roller module is equipped with a material detection sensor, and the stopper can be raised and lowered to control the docking of materials with production line equipment.

[0007] Preferably, the drive wheels are symmetrically arranged on both sides of the lever floating chassis, and the steering angle is controlled by an independent servo motor. The servo motor is connected to the two-wheel drive differential travel and steering system to realize differential steering and path correction.

[0008] Preferably, the lidar is installed on the outside of the base, with its scanning direction tilted downwards, covering the area in front of and to the side of the AGV, and is linked with the safety detection system to generate dynamic obstacle avoidance commands.

[0009] Preferably, the downward-facing camera is embedded in the center of the bottom of the base, with its lens facing the QR code on the ground, and the scanning frequency is dynamically adjusted according to the AGV's driving speed.

[0010] Preferably, the blocker is a liftable structure, and its lifting end is equipped with a contact sensing module for detecting the docking status with the production line equipment. The lifting action is linked with the start and stop signal of the roller module.

[0011] Preferably, the material detection sensors are symmetrically arranged along the conveying direction of the double-layer roller module to detect the edge offset of the material tray and feed the signal back to the control system to trigger roller position adjustment.

[0012] Preferably, the base and the lever-floating chassis are connected by a floating connection mechanism to adapt to the undulations of the ground, and the base is provided with a cable channel for integrating the connection harnesses of the control system and various sensors.

[0013] Preferably, the top of the outer casing has a reserved wireless communication module for real-time interaction with the host computer to exchange task instructions and status information.

[0014] A method for operating a double-layer roller AGV includes the following steps: S1. Initialization phase: After the AGV is powered on, the control system performs a self-test, initializes the parameters of the two-drive differential travel and steering system, QR code positioning and navigation system, safety detection system, material detection system and roller conveyor system, and activates the downward-looking camera to scan the ground QR code and match and locate it with the pre-stored data. S2. Task Reception and Path Planning: After receiving the handling task from the host computer, the optimal path is generated by combining the current QR code coordinates and the target location and then sent to the drive wheel through the path planning algorithm. S3, Driving Control: The drive wheels travel at a differential speed along the planned path, and the downward-looking camera provides real-time feedback on the position deviation. The control system dynamically adjusts the speed and steering angle of the drive wheels. S4. Picking operation: After recognizing the QR code of the target storage location, the lowering barrier connects to the production line, the roller module receives the material and detects the pallet position, and automatically adjusts if it is not in place; S5. Material handling monitoring: The material detection system continuously monitors the handling status and triggers the drive wheels to slow down or stop when an abnormality occurs. S6. Unloading operation: After identifying the target equipment's QR code, the stopper is lowered and the roller module is driven in reverse to push the material to the production line. After detecting that there is no residue, the stopper is raised. S7. Return and Standby: After completing the mission, the vehicle autonomously plans its return route and drives the wheel-mounted lidar to continuously avoid obstacles and return to the standby position.

[0015] Preferably, the initialization in S1 includes: During self-testing, the matching accuracy between the downward-looking camera and the preset QR code is simultaneously calibrated, and the security detection system verifies the lidar point cloud data acquisition function. The path planning algorithm in S2 includes: Based on the distance between adjacent QR code coordinates and the maximum steering angle of the drive wheels, calculate the minimum turning radius path and generate the drive wheel differential parameters; The dynamic adjustment in S3 includes: Based on the tilt feedback data of the lever floating chassis, the speed difference between the two drive wheels is adjusted to compensate for the path deviation caused by uneven ground. The automatic adjustment in S4 specifically refers to: When the material detection system detects pallet misalignment, it controls the roller module to intermittently rotate forward and backward to move the pallet to the center position. The detection of no residue in S6 includes: If the material detection system does not detect the pallet after the roller module has been idling for 3 seconds, it determines that unloading is complete. After the stoppers in S4 and S6 descend, the roller module starts rotating after a 0.5-second delay to ensure that the production line and the AGV docking mechanism are completely unlocked. The security detection systems in S3 and S7 continue to execute: When the lidar detects an obstacle within 1 meter ahead, it prioritizes triggering differential steering of the drive wheels to avoid it. If the obstacle exceeds the steering capability, it will brake urgently. After completing the task, the AGV autonomously plans the return route, keeping the lidar and the downward-facing camera running until it returns to the standby position.

[0016] This invention provides a double-layer roller AGV and a method thereof. It has the following beneficial effects: 1. This invention employs a differential steering structure with a lever-floating chassis and symmetrical drive wheels. A servo motor independently controls the steering angle of both drive wheels, enabling adaptive path tracking under complex terrain conditions. Compared to existing technologies using rigid chassis or single-drive steering, this invention overcomes the technical shortcomings of uneven ground causing trajectory deviation and sluggish steering response, allowing the AGV to maintain millimeter-level tracking accuracy even on bumpy or sloping surfaces.

[0017] This invention combines dynamic scanning frequency adjustment technology of a downward-looking camera with a QR code positioning and navigation system. By matching the driving speed and QR code recognition rate in real time, it achieves continuous and accurate positioning at high speeds. Existing fixed-frequency scanning schemes are prone to signal loss when the AGV accelerates, leading to path planning interruptions. This solution completely eliminates positioning blind spots at high speeds through a dynamic adaptation mechanism, ensuring seamless connection of transport tasks.

[0018] This invention constructs a three-dimensional obstacle avoidance perception system through the collaborative design of an inclined-mounted lidar and a floating connecting base. Compared with the traditional horizontally mounted lidar solution, its downward-tilted scanning angle expands the obstacle detection range, enabling it to identify obstacles ahead and also detect low-lying risk sources such as ground protrusions and scattered parts, thus solving the obstacle avoidance failure problem caused by monitoring blind spots in existing technologies. Attached Figure Description

[0019] Figure 1 This is a perspective view of the present invention; Figure 2 This is a schematic diagram of the base portion of the present invention; Figure 3 This is a schematic diagram of the contact sensing module structure of the present invention; Figure 4 This is a partial structural schematic of the lever-floating chassis of the present invention; Figure 5 This is a schematic diagram of the drive wheel section of the present invention; Figure 6 This is a flowchart of the method of the present invention.

[0020] The components include: 1. Outer shell; 2. Blocker; 3. Roller module; 4. Base; 5. Drive wheel; 6. LiDAR; 7. Camera; 8. Lever floating chassis; 9. Contact sensing module; 10. Floating connection mechanism; and 11. Material detection sensor. Detailed Implementation

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see the appendix Figure 1 -Appendix Figure 6This invention provides a double-layer roller AGV, comprising: a shell 1, a base 4 fixed to the upper surface of the shell 1, a stopper 2 installed inside the shell 1, a double-layer roller module 3 installed inside the shell 1, a lever-floating chassis 8 installed on the lower surface of the base 4, drive wheels 5 connected to the lever-floating chassis 8, a laser radar 6 fixed to the outer side of the base 4, and a downward-looking camera 7 embedded in the lower surface of the base 4; the drive wheels 5 constitute a two-drive differential walking and steering system, the laser radar 6 is linked with a safety detection system, the roller module 3 is equipped with a material detection sensor 11, the stopper 2 can lift and control the docking of materials with production line equipment, the drive wheels 5 are symmetrically arranged on both sides of the lever-floating chassis 8, and the steering angle is controlled by an independent servo motor, the servo motor is signal-connected to the two-drive differential walking and steering system to realize differential steering and path correction, and the laser radar 6 is installed on the outer side of the base 4 with its scanning direction tilted downward. The system covers the area in front of and to the side of the AGV and generates dynamic obstacle avoidance commands in conjunction with the safety detection system. The downward-looking camera 7 is embedded in the center of the bottom of the base 4, with its lens facing the QR code on the ground. The scanning frequency is dynamically adjusted according to the AGV's speed. The blocker 2 is a liftable structure, and its lifting end is equipped with a contact sensing module 9 for detecting the docking status with the production line equipment. The lifting action is linked with the start and stop signals of the roller module 3. The material detection sensor 11 is symmetrically arranged along the conveying direction of the double-layer roller module 3 to detect the edge offset of the material tray and feed the signal back to the control system to trigger the roller position adjustment. The base 4 and the lever floating chassis 8 are connected by a floating connection mechanism 10 to adapt to the ground undulations. The base 4 has a cable channel inside for integrating the connection harnesses of the control system and various sensors. The top of the outer shell 1 has a reserved wireless communication module for real-time interaction with the host computer for task commands and status information.

[0023] Specifically, the double-layer roller AGV achieves differential steering control of both wheels through independent servo motors of drive wheels 5, and, in conjunction with the adaptive terrain compensation mechanism of the lever floating chassis 8, corrects path deviation in real time during movement; the downward-facing camera 7 dynamically adjusts the scanning frame rate according to the real-time vehicle speed to ensure continuous capture of ground QR codes and rapid matching with pre-stored coordinate data during high-speed driving; the lidar 6 adopts a tilted downward multi-dimensional scanning mode to simultaneously monitor obstacles ahead and low-lying risk sources on the ground, generating a graded obstacle avoidance strategy; the stopper 2 verifies the docking status through the contact sensing module 9 during lifting and lowering, and controls the delayed start and stop logic of the double-layer roller module 3 in conjunction; the material detection sensor 11 detects the pallet deviation based on a symmetrical layout, triggering intermittent forward and reverse rotation of the rollers to achieve material self-correction; the wireless communication module uploads the equipment operating status in real time and receives dynamic scheduling instructions, and the floating connection mechanism 10 and the cable channel work together to ensure the stability of signal transmission under complex working conditions.

[0024] Please see the appendix Figure 1 -Appendix Figure 5This invention provides a method for operating a double-layer roller AGV, comprising the following steps: S1. Initialization phase: After the AGV is powered on, the control system performs a self-test, initializes the parameters of the two-drive differential travel and steering system, QR code positioning and navigation system, safety detection system, material detection system and roller conveyor system, and activates the downward-looking camera 7 to scan the ground QR code and match and position it with the pre-stored data. S2. Task reception and path planning: After receiving the handling task from the host computer, the optimal path is generated by combining the current QR code coordinates and the target position and then sent to the drive wheel 5 through the path planning algorithm. S3, Driving control: Drive wheels 5 travel at differential speed along the planned path, and the downward-looking camera 7 provides real-time feedback on position deviation. The control system dynamically adjusts the speed and steering angle of drive wheels 5. S4. Retrieval Operation: After recognizing the QR code of the target storage location, the lowering stopper 2 connects to the production line, the roller module 3 receives the material and detects the pallet position, and automatically adjusts if it is not in place; S5. Material handling monitoring: The material detection system continuously monitors the handling status and triggers drive wheel 5 to slow down or stop when an abnormality occurs. S6. Unloading operation: After identifying the QR code of the target equipment, the stopper 2 is lowered and the roller module 3 is driven in reverse to push the material to the production line. After detecting that there is no residue, the stopper 2 is raised. S7, Return and Standby: After completing the mission, the vehicle autonomously plans its return route, and the drive wheel 5 carrying the lidar 6 continuously avoids obstacles to return to the standby position.

[0025] Initialization in S1 includes: During self-testing, the matching accuracy between the downward-looking camera 7 and the preset QR code is simultaneously calibrated, and the security detection system verifies the point cloud data acquisition function of the lidar 6. The path planning algorithm in S2 includes: Based on the distance between adjacent QR code coordinates and the maximum steering angle of drive wheel 5, calculate the minimum turning radius path and generate drive wheel differential parameters; Dynamic adjustments in S3 include: Based on the tilt feedback data of the lever floating chassis 8, the speed difference of the two drive wheels 5 is adjusted to compensate for the path deviation caused by uneven ground. The automatic adjustment in S4 is as follows: When the material detection system detects pallet offset, it controls the roller module 3 to intermittently rotate forward and backward to move the pallet to the center position. The absence of residue detected in S6 includes: If the material detection system does not detect the pallet after the roller module 3 has been idling for 3 seconds, it determines that unloading is complete. After the stopper 2 in S4 and S6 descends, the roller module 3 starts rotating after a delay of 0.5 seconds to ensure that the production line and the AGV docking mechanism are completely unlocked. The security detection system in S3 and S7 continues to run: When the lidar 6 detects an obstacle within 1 meter ahead, it will first trigger the differential steering of the drive wheel 5 to avoid it. If the obstacle exceeds the steering capability, it will brake urgently. After the mission is completed, the AGV will autonomously plan the return route and keep the lidar 6 and the downward-facing camera 7 running until it returns to the standby position.

[0026] Specifically, the drive wheel 5 performs dynamic steering based on differential parameters generated by the path planning algorithm, and the real-time terrain compensation function of the lever floating chassis 8 eliminates trajectory deviations caused by ground undulations; the downward-looking camera 7 maintains the positioning accuracy of the QR code with a speed-adaptive scanning mechanism, seamlessly connecting navigation signals during high-speed movement; the lidar 6 constructs a three-dimensional obstacle avoidance model through tilt scanning, triggering steering avoidance or emergency braking strategies in stages; the stopper 2 simultaneously verifies the production line docking status during lifting and lowering actions, and controls the delayed start and stop sequence of the roller module 3; the material detection system captures the pallet offset through a symmetrical sensor array, and the drive roller intermittently rotates forward and backward to complete self-correction; the wireless communication module realizes bidirectional real-time transmission of task instructions and running data. Its technical value is reflected in: the integrated control of differential steering and floating compensation breaks through the path tracking bottleneck of traditional AGVs, the dynamic positioning mechanism ensures continuous navigation capability under high-speed conditions; the three-dimensional perception system achieves full-coverage obstacle avoidance in three-dimensional space, significantly reducing the risk of sudden collisions; the intelligent docking and material self-adjustment functions form a closed-loop loading and unloading process, greatly improving the reliability of production line docking.

[0027] Working principle: Drive wheel 5 achieves differential steering through an independent servo motor, which, together with the floating mechanism of lever floating chassis 8, compensates for ground undulations in real time. During movement, it dynamically adjusts the speed difference between the two sides according to the path deviation. The downward-looking camera 7 adaptively adjusts the scanning frequency based on the real-time speed of the AGV, accurately captures the ground QR code and matches it with the pre-stored map data to achieve continuous positioning. LiDAR 6 performs three-dimensional scanning of the AGV's movement direction at a downward-tilted detection angle, simultaneously integrating ground protrusion detection and obstacle recognition functions to generate multi-level safety warning signals. The stopper 2 triggers the contact sensing module 9 through lifting action, and after descending to the correct position, it controls the start and stop sequence of the double-layer roller module 3 to ensure that the physical lock of the production line docking mechanism is completely released before material transfer is started. The double-layer roller module 3 is based on the pallet offset feedback from the material detection sensor 11. According to reports, the system performs intermittent forward and reverse operations to autonomously correct the material position. Material detection sensors 11 are symmetrically distributed along the roller conveyor axis, capturing the pallet edge offset in real time and generating position compensation commands. The wireless communication module continuously receives handling tasks issued by the host computer and synchronously uploads the AGV's real-time position, obstacle avoidance status, and material loading data. During operation, the control system completes the calibration of each sensor in the initialization phase, generates the optimal trajectory by combining the QR code coordinates and the steering limit of the drive wheel 5 in the path planning phase, and dynamically integrates the positioning data of the downward-looking camera 7 and the obstacle avoidance information of the lidar 6 to achieve closed-loop control in the driving phase. When picking up goods, the stopper 2 descends to trigger the roller delay start mechanism. After unloading, the material detection sensors 11 confirm that there is no residue and the actuator resets. When the task ends, the AGV autonomously activates the return navigation mode, and each sensing module continues to work until it safely returns to the standby point.

[0028] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A double-layer roller AGV, characterized in that, include: The outer shell (1) has a base (4) fixed on its upper surface, a stopper (2) installed inside the outer shell (1), a double-layer roller module (3) installed inside the outer shell (1), a lever floating chassis (8) installed on the lower surface of the base (4), a drive wheel (5) connected to the lever floating chassis (8), a laser radar (6) fixed on the outer side of the base (4), and a downward-looking camera (7) embedded on the lower surface of the base (4); the drive wheel (5) constitutes a two-wheel drive differential walking and steering system, the laser radar (6) is linked with the safety detection system, the roller module (3) is equipped with a material detection sensor (11), and the stopper (2) can lift and control the docking of materials with production line equipment.

2. The double-layer roller AGV according to claim 1, characterized in that, The drive wheels (5) are symmetrically arranged on both sides of the lever floating chassis (8), and the steering angle is controlled by an independent servo motor. The servo motor is connected to the two-wheel drive differential walking and steering system to realize differential steering and path correction.

3. The double-layer roller AGV according to claim 1, characterized in that, The lidar (6) is installed on the outside of the base (4), with its scanning direction tilted downwards, covering the area in front of and to the side of the AGV, and is linked with the safety detection system to generate dynamic obstacle avoidance commands.

4. The double-layer roller AGV according to claim 1, characterized in that, The downward-facing camera (7) is embedded in the center of the bottom of the base (4), with its lens facing the QR code on the ground. The scanning frequency is dynamically adjusted according to the AGV's driving speed.

5. A double-layer roller AGV according to claim 1, characterized in that, The blocker (2) is a liftable structure, and its lifting end is equipped with a contact sensing module (9) for detecting the docking status with the production line equipment. The lifting action is linked with the start and stop signal of the roller module (3).

6. A double-layer roller AGV according to claim 1, characterized in that, The material detection sensors (11) are symmetrically arranged along the conveying direction of the double-layer roller module (3) to detect the edge offset of the material tray and feed the signal back to the control system to trigger roller position adjustment.

7. A double-layer roller AGV according to claim 1, characterized in that, The base (4) and the lever floating chassis (8) are connected by a floating connection mechanism (10) to adapt to the undulation of the ground. The base (4) is provided with a cable channel inside for integrating the connection harnesses of the control system and each sensor.

8. A double-layer roller AGV according to claim 1, characterized in that, The top of the outer shell (1) has a reserved wireless communication module for real-time interaction with the host computer to exchange task instructions and status information.

9. A method for operating a double-layer roller AGV, wherein the double-layer roller AGV according to any one of claims 1-8 is characterized in that, Includes the following steps: S1. Initialization stage: After the AGV is powered on, the control system performs a self-test, initializes the parameters of the two-drive differential walking and steering system, QR code positioning and navigation system, safety detection system, material detection system and roller conveyor system, and activates the downward camera (7) to scan the ground QR code and match and position it with the pre-stored data. S2, Task reception and path planning: After receiving the transport task from the host computer, the optimal path is generated by combining the current QR code coordinates and the target position and sent to the drive wheel (5) through the path planning algorithm. S3, Driving control: The drive wheels (5) travel at a differential speed along the planned path, and the downward-looking camera (7) provides real-time feedback on the position deviation. The control system dynamically adjusts the speed and steering angle of the drive wheels (5). S4. Picking operation: After recognizing the QR code of the target storage location, the lowering blocker (2) docks with the production line, and the roller module (3) receives the material and detects the pallet position. If it is not in place, it will be automatically adjusted. S5. Handling monitoring: The material detection system continuously monitors the handling status and triggers the drive wheel (5) to slow down or stop when an abnormality occurs. S6. Unloading operation: After identifying the target equipment QR code, the stopper (2) is lowered and the roller module (3) is driven in reverse to push the material to the production line. After detecting that there is no residue, the stopper (2) is raised. S7, Return and Standby: After completing the mission, the vehicle autonomously plans the return route and the drive wheel (5) carrying the laser radar (6) continuously avoids obstacles and returns to the standby position.

10. The method for operating a double-layer roller AGV according to claim 9, characterized in that, The initialization in S1 includes: During self-testing, the matching accuracy between the downward-looking camera (7) and the preset QR code is simultaneously calibrated, and the security detection system verifies the point cloud data acquisition function of the lidar (6). The path planning algorithm in S2 includes: Based on the distance between adjacent QR code coordinates and the maximum steering angle of the drive wheel (5), calculate the minimum turning radius path and generate the drive wheel differential parameters; The dynamic adjustment in S3 includes: Based on the tilt feedback data of the lever floating chassis (8), the speed difference of the two drive wheels (5) is adjusted to compensate for the path deviation caused by uneven ground. The automatic adjustment in S4 specifically refers to: When the material detection system detects pallet offset, it controls the roller module (3) to intermittently rotate forward and backward to move the pallet to the center position; The detection of no residue in S6 includes: If the material detection system does not detect the pallet after the roller module (3) has been idling for 3 seconds, it is determined that the unloading is complete. After the stopper (2) in S4 and S6 descends, the roller module (3) starts to rotate after a delay of 0.5 seconds to ensure that the production line and the AGV docking mechanism are completely unlocked; The security detection systems in S3 and S7 continue to execute: When the lidar (6) detects an obstacle within 1 meter in front, it will trigger the differential steering of the drive wheel (5) to avoid it. If the steering capability is exceeded, it will brake in an emergency. After the task is completed, the AGV will autonomously plan the return route and keep the lidar (6) and the downward camera (7) running until it returns to the standby position.

Images