Vision-based automatic goods shelf cross beam deformation detection mechanism of stacker and detection method

Abstract

The invention discloses a mechanism and method for automatically detecting the deformation of a goods shelf cross beam of a stacking machine based on vision, which improve the operation stability and safety of a whole three-dimensional warehouse system and enable the stacking machine to be safer when goods are stored and taken out. Comprising a stacking machine rack (1) and a lifting cargo carrying table (2) embedded in the stacking machine rack (1). Binocular cameras (3) are installed on the two sides of a frame body of the lifting cargo carrying table (2), a switch supporting POE power supply is arranged in a control box (5) on the lifting cargo carrying table (2), the switch is connected with the binocular cameras (3) through a first network cable, and the switch in the control box (5) is connected with an edge computer in an electric control cabinet (6) through a second network cable. The edge computer is connected with a programmable controller for controlling the stacking machine through a third network cable; the operation stability and safety of the whole three-dimensional warehouse system are improved, and the stacking machine is safer and more intelligent.

Description

Technical Field

[0001] This invention relates to a stacker crane, and more particularly to a vision-based automatic detection mechanism and method for detecting deformation of rack beams in a stacker crane. Background Technology

[0002] With the continuous development of the automated storage and retrieval system (AS / RS) industry, the requirements for the safety performance of equipment used in AS / RS are becoming increasingly stringent. Previously, after a period of operation, maintenance personnel needed to enter the aisles between the storage racks to inspect and measure the deformation of each rack beam. However, due to the large number of storage locations in AS / RS, many of which are in the air, manual observation and measurement were difficult, leading to a high rate of missed inspections. If a rack beam is significantly deformed, the stacker crane's forks often push pallets and their contents off the rack when it extends or retracts the forks. Alternatively, the deformed beam may scrape against the pallet when the stacker crane's forks retract, causing the pallet to fall. Both of these situations pose safety hazards and economic losses to the AS / RS. Therefore, developing a mechanism that can automatically detect beam deformation before the stacker crane's forks extend or retract to solve these problems has become an urgent issue. Summary of the Invention

[0003] This invention provides a vision-based automatic detection mechanism and method for stacker cranes to detect the deformation of rack beams, which improves the stability and safety of the entire automated warehouse system and makes stacker cranes safer when storing and retrieving goods.

[0004] The present invention solves the above technical problems through the following technical solutions: A vision-based automatic detection mechanism for rack beam deformation of a stacker crane includes a stacker crane frame and a lifting platform nested within the stacker crane frame. Binocular cameras are installed on both sides of the lifting platform frame. A PoE-enabled switch is installed in the control box on the lifting platform. The switch is connected to the binocular cameras via a first network cable. The switch in the control box is connected to an edge computer in the electrical control cabinet via a second network cable. The edge computer is connected to the programmable controller that controls the stacker crane via a third network cable. PoE is also known as a local area network-based power supply system.

[0005] The edge computer sets a deviation threshold for the distance between the lifting platform frame and the rack beam; the edge computer controls a binocular camera to capture 3D point cloud data of the rack beam and compares the 3D point cloud data with the reference 3D point cloud data of the rack beam. If the deviation threshold is exceeded, a command is issued to prohibit the stacker crane's forks from extending or retracting.

[0006] A vision-based automatic detection method for the deformation mechanism of a stacker crane's rack beams, characterized by the following steps: Step 1: Based on the warehouse racking, calibrate the horizontal and vertical positions of the stacker crane in the racking aisle; after calibration, select a certain storage location in the aisle as the reference storage location, move the stacker crane to the reference storage location, and the programmable controller controlling the stacker crane sends instructions to the edge computer. The edge computer controls the binocular camera to capture 3D point cloud data of the basic racking beams, and uses the captured 3D point cloud data as the standard value A of the racking beams; The second step involves comparing the 3D point cloud data of the shelf beams collected by the binocular camera controlled by the edge computer with the standard value A of the shelf beams. The absolute value of the maximum difference obtained after the comparison is defined as the beam position deviation value B. If the beam position deviation value B is less than 5 mm, it is defined as the beam in normal position. If the beam position deviation value B is greater than or equal to 5 mm and less than 8 mm, it is defined as the beam in warning position. If the beam position deviation value B is greater than or equal to 8 mm, it is defined as the beam in fault position. Third step: When the stacker crane is working in the warehouse aisle, the rack beam self-detection function in the edge computer of the stacker crane is activated by the programmable controller. Step 4: Once the lifting platform in the stacker crane frame is aligned with a storage location on the rack, the edge computer controls the binocular camera to collect the 3D point cloud data of the rack beam in real time and compare it with the standard value A of the rack beam. If the crossbeam is in the normal position, the crossbeam normal position signal is transmitted to the programmable controller. After receiving the crossbeam normal position signal, the programmable controller issues a fork extension and retraction operation command. If the crossbeam is in the crossbeam warning position, the crossbeam warning position signal is transmitted to the programmable controller. After receiving the crossbeam warning position signal, the programmable controller records the position and then issues a fork extension and retraction operation command. If the crossbeam is in a fault position, the crossbeam fault position signal is transmitted to the programmable controller. After receiving the crossbeam fault position signal, the programmable controller records the fault position, alarms, and issues a command to prohibit the extension and retraction of the forks.

[0007] This invention utilizes two binocular vision cameras installed on both sides of the loading platform to collect shelving data, learn from the shelving, and automatically measure storage location data. This significantly shortens the initial debugging time. Furthermore, this technology addresses issues such as shelving subsidence due to ground settlement, shelving deformation due to shelving quality problems, and excessive deviations in storage location data due to insufficient installation accuracy. It ensures the stability and safety of the stacker crane during each loading and unloading operation, preventing forks from colliding with shelving or goods, and avoiding situations where forks fail to detach from goods, thus avoiding misalignment. This enhances the stability and safety of the entire automated warehouse system, making the stacker crane safer and more intelligent. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the structure of the present invention in the main viewing direction; Figure 2 This is a schematic diagram of the structure of the present invention in a side view. Detailed Implementation

[0009] The present invention will now be described in detail with reference to the accompanying drawings: A vision-based automatic detection mechanism for rack beam deformation of a stacker crane includes a stacker crane frame 1 and a lifting platform 2 nested in the stacker crane frame 1. Binocular cameras 3 are installed on both sides of the lifting platform 2. A PoE-enabled switch is installed in a control box 5 on the lifting platform 2. The switch is connected to the binocular cameras 3 via a first network cable. The switch in the control box 5 is connected to an edge computer in an electrical control cabinet 6 via a second network cable. The edge computer is connected to a programmable controller that controls the stacker crane via a third network cable.

[0010] The edge computer sets a deviation threshold for the distance between the frame of the lifting platform 2 and the shelf beam; the edge computer controls the binocular camera 3 to capture 3D point cloud data of the shelf beam 4, and compares the 3D point cloud data with the reference 3D point cloud data of the shelf beam 4. If the deviation threshold is exceeded, a command is issued to prohibit the stacker crane's forks from extending or retracting.

[0011] A vision-based automatic detection method for the deformation mechanism of a stacker crane's rack beams, characterized by the following steps: Step 1: Based on the warehouse racking, calibrate the horizontal and vertical positions of the stacker crane in the racking aisle; after calibration, select a certain storage location in the aisle as the reference storage location, move the stacker crane to the reference storage location, and the programmable controller controlling the stacker crane sends instructions to the edge computer. The edge computer controls the binocular camera 3 to capture 3D point cloud data of the basic racking beams, and uses the captured 3D point cloud data as the standard value A of the racking beams; The second step involves comparing the 3D point cloud data of the shelf beam 4 collected by the binocular camera 3 under the control of the edge computer with the standard value A of the shelf beam. The absolute value of the maximum difference obtained after the comparison is defined as the beam position deviation value B. If the beam position deviation value B is less than 5 mm, it is defined as the beam in normal position. If the beam position deviation value B is greater than or equal to 5 mm and less than 8 mm, it is defined as the beam in warning position. If the beam position deviation value B is greater than or equal to 8 mm, it is defined as the beam in fault position. Third step: When the stacker crane is working in the warehouse aisle, the rack beam self-detection function in the edge computer of the stacker crane is activated by the programmable controller. Step 4: When the lifting platform 2 in the stacker crane frame 1 is aligned with a storage location on the shelf, the edge computer controls the binocular camera 3 to collect the 3D point cloud data of the shelf beam 4 of that storage location in real time and compare it with the standard value A of the shelf beam. If the crossbeam is in the normal position, the crossbeam normal position signal is transmitted to the programmable controller. After receiving the crossbeam normal position signal, the programmable controller issues a fork extension and retraction operation command. If the crossbeam is in the crossbeam warning position, the crossbeam warning position signal is transmitted to the programmable controller. After receiving the crossbeam warning position signal, the programmable controller records the position and then issues a fork extension and retraction operation command. If the crossbeam is in a fault position, the crossbeam fault position signal is transmitted to the programmable controller. After receiving the crossbeam fault position signal, the programmable controller records the fault position, alarms, and issues a command to prohibit the extension and retraction of the forks.

Claims

1. A visual-based automatic detection mechanism for deformation of a rack beam of a stacker, comprising a stacker frame (1) and a lifting loading platform (2) nested in the stacker frame (1); characterized in that, Both sides of the frame of the lifting loading platform (2) are provided with binocular cameras (3), and a switch supporting POE power supply is arranged in a control box (5) on the lifting loading platform (2), the switch is connected with the binocular cameras (3) through a first network cable, and the switch in the control box (5) is connected with an edge computer in an electric control cabinet (6) through a second network cable, and the edge computer is connected with a programmable controller for controlling the stacker through a third network cable.

2. The visual-based automatic detection mechanism for beam deformation of a racking system according to claim 1, wherein, A deviation threshold of the distance between the frame of the lifting loading platform (2) and the shelf cross beam is arranged in the edge computer; the edge computer controls the binocular cameras (3) to shoot 3D point cloud data of the shelf cross beam (4), and compares the 3D point cloud data with reference 3D point cloud data of the shelf cross beam (4), if the deviation threshold is exceeded, an instruction of prohibiting the telescoping of the forks of the stacker is sent out.

3. The detection method of the visual-based stacker automatic detection shelf cross beam deformation mechanism according to claim 1, characterized by the following steps: First step, according to the warehouse shelf, the horizontal and vertical positions of the stacker in the shelf aisle are calibrated; after the calibration is completed, a certain storage location in the aisle is selected as a reference storage location, the stacker is run to the reference storage location, the programmable controller of the stacker sends an instruction to the edge computer, the edge computer controls the binocular cameras (3) to shoot 3D point cloud data of the basic shelf cross beam, and the shot 3D point cloud data is defined as the shelf cross beam standard value A; Second step, the 3D point cloud data of the shelf cross beam (4) collected by the edge computer controls the binocular cameras (3) is compared with the shelf cross beam standard value A, the absolute value of the maximum difference obtained after the comparison is defined as the cross beam position deviation value B; the cross beam position deviation value B is less than 5mm, which is defined as the normal position of the cross beam; the cross beam position deviation value B is greater than or equal to 5mm and less than 8mm, which is defined as the pre-warning position of the cross beam; the cross beam position deviation value B is greater than or equal to 8mm, which is defined as the fault position of the cross beam; Third step, when the stacker works in the warehouse aisle, the shelf cross beam self-detection function in the edge computer of the stacker is started through the programmable controller; Fourth step, when the lifting loading platform (2) in the stacker frame (1) is aligned with a storage location on the shelf, the edge computer controls the binocular cameras (3) to collect the 3D point cloud data of the shelf cross beam (4) of the storage location in real time, and compares it with the shelf cross beam standard value A; If the cross beam is in the normal position of the cross beam, the cross beam normal position signal is transmitted to the programmable controller, and the programmable controller receives the cross beam normal position signal and sends out the fork telescoping operation instruction; If the cross beam is in the pre-warning position of the cross beam, the cross beam pre-warning position signal is transmitted to the programmable controller, the programmable controller receives the cross beam pre-warning position signal, records the storage location, and then sends out the fork telescoping operation instruction; If the cross beam is in the fault position of the cross beam, the cross beam fault position signal is transmitted to the programmable controller, the programmable controller receives the cross beam fault position signal, records the storage location and alarms, and sends out the fork telescoping operation instruction.

Images