Rack stacking method, device and equipment based on unmanned forklift and medium

Abstract

The invention relates to the technical field of unmanned forklifts, and discloses a material frame stacking method, device and equipment based on an unmanned forklift and a medium. The method comprises the steps that a pallet fork forks a material frame, and a fork arm is controlled to be lifted to a first preset height; the radar scans the feeding and discharging frame, a first target pose is determined, the pallet fork is adjusted to the first target pose, and the first target pose enables the feeding frame to be prevented from making contact with the discharging frame limiting block; the fork arm is controlled to descend to a second preset height, stress data of the pallet fork is continuously read based on the sensor, the fork arm is controlled to be lifted to a third preset height, and the third preset height makes the feeding and discharging frame disengaged from contact; the radar scans the feeding and discharging racks, determines a second target pose, adjusts the pallet fork to the second target pose, controls the fork arm to descend until the stress data reaches the no-load range, stops descending, and enables the feeding and discharging racks to be aligned through the second target pose; and visual verification is carried out based on a camera of the unmanned forklift, and material frame stacking is completed when the visual verification passes. According to the invention, the problem of butt-joint card limiting during material rack stacking is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of unmanned forklift technology, and specifically to a rack stacking method, apparatus, equipment and medium based on unmanned forklifts. Background Technology

[0002] In existing rack stacking technologies, there are significant differences in compatibility between two different rack structures: The first type of rack (four-legged stacking rack) connects to the uprights through protruding foot cups, and the sensor can directly scan the connection structure, allowing for an error greater than 20mm in any direction, resulting in a high stacking success rate; while the second type of rack (clamping rack) uses a connection structure with triangular groove left and right guides and front / end hard limits. Although it can eliminate lateral errors greater than 20mm through the triangular grooves, the longitudinal limit blocks have no guiding function. When the longitudinal positioning error is greater than 10mm, the upper rack is prone to getting stuck on the limit block of the lower rack, leading to connection failure.

[0003] Therefore, how to solve the limiting problem of the docking card when stacking the second type of material rack and achieve stable stacking of the second type of material rack is a problem that needs to be solved. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, and medium for stacking racks based on unmanned forklifts to solve the problem of docking card limiting during the stacking of second-type racks.

[0005] In a first aspect, the present invention provides a rack stacking method based on an unmanned forklift, which is applied to an unmanned forklift equipped with radar, camera and sensors; The method includes: The unmanned forklift's forks pick up the loading rack and the fork arms of the unmanned forklift are raised to the first preset height, so that the unmanned forklift moves to the target position of the unloading rack. The radar scanning of the loading and unloading racks of the unmanned forklift determines the first target pose and adjusts the forks to the first target pose. The first target pose prevents the loading rack from contacting the limit block of the unloading rack. The fork arm is controlled to descend to the second preset height. Based on the sensors of the unmanned forklift, the force data of the forks is continuously read. According to the changes in the force data, the fork arm is controlled to rise to the third preset height. The third preset height causes the loading rack and unloading rack to disengage. Based on the radar rescanning of the loading rack and unloading rack, the second target pose is determined and the forks are adjusted to the second target pose. The fork arms are controlled to descend until the force data reaches the no-load range, and the descent of the fork arms is stopped. The second target pose aligns the loading rack and unloading rack. Visual verification is performed using cameras on unmanned forklifts, and the stacking of loading and unloading racks is completed when the visual verification is passed.

[0006] This invention utilizes forks to pick up the loading rack, raising the fork arms to a first preset height to prevent collisions between the loading and unloading racks. The forks are then moved to the target position, providing an unobstructed view for subsequent radar scanning of the loading and unloading rack docking structure, ensuring the integrity and accuracy of subsequent pose detection. By scanning the loading and unloading racks with radar, the first target pose is determined, and the forks are adjusted to this pose to ensure no contact between the loading and unloading rack limit blocks, thus avoiding jamming issues at the source. Then, the fork arms are controlled to descend from the first preset height to a second preset height. Based on changes in force data collected by sensors, the fork arms are controlled to rise to a third preset height, ensuring complete disengagement of the loading and unloading racks. This eliminates contact friction and interference risks for subsequent translational alignment, further avoiding jamming. The loading and unloading racks are re-scanned by radar to determine a second target pose for complete alignment, and the forks are adjusted. The fork arms are controlled to descend until the force data falls within the no-load range, eliminating minor deviations that may occur after rough docking and achieving high-precision alignment of the loading and unloading racks. Finally, visual verification using a camera ensures the reliability of the stacking operation and effectively solves the problem of limiting the docking card when stacking the second type of material rack.

[0007] In one optional implementation, determining the pose of the first target based on radar scanning of the loading and unloading racks of the unmanned forklift includes: The first point cloud data is obtained by scanning the bottom of the loading rack with radar, and the second point cloud data is obtained by scanning the top of the unloading rack. The first point cloud data and the second point cloud data are respectively model registered to obtain the first pose data and the second pose data. The first pose data and the second pose data are transformed into coordinate systems respectively, and the first relative deviation between the first pose data and the second pose data after the coordinate system transformation is calculated. Based on the first relative deviation, the first target position of the forks is calculated when the limit blocks of the loading rack and unloading rack are at a preset safe distance.

[0008] This embodiment uses point cloud data scanned by radar and avoids the constraint of the loading rack being stuck to the limit to calculate the first target pose that the forks need to reach, so that the loading rack can avoid being stuck to the limit and achieve a rough docking with the unloading rack without collision interference.

[0009] In one optional implementation, the forklift is controlled to descend to a second preset height, and the sensors of the unmanned forklift continuously read the force data of the forks. Based on the changes in the force data, the forklift is controlled to rise to a third preset height, including: During the process of controlling the fork arm to descend from the first preset height to the second preset height, the mean and standard deviation of the force data of the fork are calculated based on the continuous reading of the sensor, and used as the force benchmark of the fork. The control arm is continuously lowered from the second preset height until the force data and force reference meet the preset conditions, and the height of the control arm is obtained. The third preset height is determined based on the height of the fork arm, and the fork arm is controlled to be raised to the third preset height.

[0010] This embodiment establishes a force reference by analyzing the changes in force data during the dynamic descent of the fork arm. Based on the force data and the force reference, the height of the fork arm when the upper and lower racks just come into contact is determined. A third preset height is determined based on the fork arm height, and the fork arm is controlled to be raised to this third preset height. This eliminates the risk of contact interference for the subsequent translation and alignment of the upper rack with the lower rack.

[0011] In one alternative implementation, visual verification is performed using a camera on the unmanned forklift, including: The first image and the second image are obtained by taking pictures of the triangular grooves of the loading rack and the unloading rack with a camera. Identify the first and second images to determine whether the triangular grooves of the loading rack and the unloading rack are aligned. When the triangular grooves of the loading rack and the unloading rack are closed, visual verification is confirmed. If the triangular grooves of the loading rack and the unloading rack are not closed, the visual verification is deemed unsuccessful.

[0012] In this embodiment, the camera is used to photograph the triangular grooves of the loading and unloading racks to determine whether the two triangular grooves are aligned, thereby verifying whether the docking accuracy of the loading and unloading racks meets the stacking requirements.

[0013] In one optional implementation, the second preset height is the difference between the first preset height and the first preset value, where the first preset value is less than a preset safety threshold. The third preset height is the sum of the fork arm height and the second preset value, where the second preset value is less than the height of the limit block of the unloading rack.

[0014] This embodiment determines the second preset height by setting the difference between the first preset height and the first preset value, ensuring that the loading and unloading racks remain in a non-contact state during the process of the fork arm descending from the first preset height to the second preset height. By setting the sum of the fork arm height and the second preset value as the third preset height, the loading and unloading racks are completely de-contacted, eliminating the risk of contact interference for subsequent alignment of the loading rack with the unloading rack.

[0015] In one optional implementation, determining the pose of the second target based on a radar rescan of the loading and unloading racks includes: The third point cloud data is obtained by rescanning the bottom of the loading rack with radar, and the fourth point cloud data is obtained by rescanning the top of the unloading rack. The third point cloud data and the fourth point cloud data are registered with the model to obtain the third pose data and the fourth pose data. The third pose data and the fourth pose data are transformed into coordinate systems respectively, and the second relative deviation between the third pose data and the fourth pose data after the coordinate system transformation is calculated. Based on the second relative deviation, the second target pose of the forks is calculated when the loading rack and unloading rack are aligned.

[0016] This embodiment uses point cloud data scanned by radar and the constraint of perfect alignment between the loading rack and the unloading rack to calculate the second target pose that the forks need to reach, so as to achieve precise docking between the loading rack and the unloading rack.

[0017] In one alternative implementation, after visual verification based on the camera of the unmanned forklift, the method further includes: If visual verification fails, control the fork arm to rise to the fifth preset height, return to the radar-based rescan of the loading and unloading racks, determine the second target pose and adjust the forks to the second target pose, control the fork arm to descend until the force data reaches the no-load range, stop the fork arm descent, and determine whether visual verification has been passed. An alarm signal is issued if the visual verification fails for a preset number of consecutive times.

[0018] This embodiment re-performs the fine docking when the visual verification fails, and then performs visual verification again. If the visual verification fails for a preset number of consecutive times, an alarm signal is issued to ensure operational safety and stacking quality.

[0019] Secondly, the present invention provides a rack stacking device based on an unmanned forklift, which is applied to an unmanned forklift equipped with radar, camera and sensors; The device includes: The first control module is used to enable the forks of the unmanned forklift to pick up the loading rack and control the fork arms of the unmanned forklift to rise to the first preset height, so that the unmanned forklift can move to the target position of the unloading rack. The second control module is used to scan the loading rack and unloading rack based on the radar of the unmanned forklift, determine the first target pose and adjust the forks to the first target pose, so that the loading rack avoids contacting the limit block of the unloading rack. The third control module is used to control the fork arm to descend to the second preset height. Based on the sensors of the unmanned forklift, it continuously reads the force data of the fork and controls the fork arm to rise to the third preset height according to the changes in the force data. The third preset height causes the loading rack and unloading rack to disengage. The fourth control module is used to rescan the loading rack and unloading rack based on radar, determine the second target pose and adjust the forks to the second target pose, control the fork arms to descend until the force data reaches the no-load range, stop the fork arms to descend, and align the loading rack and unloading rack with the second target pose. The verification module is used for visual verification based on the camera of the unmanned forklift, and the stacking of the loading rack and unloading rack is completed when the visual verification is passed.

[0020] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the rack stacking method based on the first aspect or any corresponding embodiment described above.

[0021] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the rack stacking method based on an unmanned forklift as described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a schematic diagram of a camera according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the material rack stacking according to an embodiment of the present invention; Figure 3 This is a schematic diagram of an unmanned forklift according to an embodiment of the present invention; Figure 4 This is a flowchart of a rack stacking method based on an unmanned forklift according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the fork arm being raised to a first preset height according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the forks being positioned in a first target pose according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the contact between the loading rack and the unloading rack according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the fork arm being raised to a third preset height according to an embodiment of the present invention; Figure 9 This is a schematic diagram of the forks being positioned in a second target pose according to an embodiment of the present invention; Figure 10 This is a schematic diagram of the fork arm descending to the point where the force data reaches the no-load range according to an embodiment of the present invention; Figure 11This is a structural block diagram of a rack stacking device based on an unmanned forklift according to an embodiment of the present invention; Figure 12 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of 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, 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.

[0025] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0026] Figure 1 This is a schematic diagram of a camera according to an embodiment of the present invention, such as... Figure 1 As shown, the traditional rack stacking method uses a 3D camera placed below the fork arm and in the center of the two forks. Figure 1 (Green circle in the image) detects the position and orientation of the loading and unloading racks, calculates the relative deviation, and controls the chassis and upper structure to eliminate the relative deviation. However, there are obvious adaptation differences for the two types of racks with different structures. Figure 2 This is a schematic diagram of the rack stacking according to an embodiment of the present invention, such as... Figure 2 As shown, the left side represents the first type of rack, whose carrier structure is a four-legged stacking rack. It connects to the uprights via protruding foot cups, and the sensor can directly scan the connection structure, resulting in a high stacking success rate. The right side represents the second type of rack, whose carrier structure is a stackable rack. It uses a docking structure with triangular groove left and right guides and front / end hard limits. Although the triangular grooves can eliminate lateral errors greater than 20mm, the longitudinal limit blocks have no guiding function. When the longitudinal positioning error is greater than 10mm, the upper rack is prone to getting stuck on the lower rack's limit blocks, leading to docking failure.

[0027] This invention utilizes forks to pick up the loading rack, raising the fork arms to a first preset height to prevent collisions between the loading and unloading racks. The forks are then moved to the target position, providing an unobstructed view for subsequent radar scanning of the loading and unloading rack docking structure, ensuring the integrity and accuracy of subsequent pose detection. By scanning the loading and unloading racks with radar, the first target pose is determined, and the forks are adjusted to this pose to ensure no contact between the loading and unloading rack limit blocks, thus avoiding jamming issues at the source. Then, the fork arms are controlled to descend from the first preset height to a second preset height. Based on changes in force data collected by sensors, the fork arms are controlled to rise to a third preset height, ensuring complete disengagement of the loading and unloading racks. This eliminates contact friction and interference risks for subsequent translational alignment, further avoiding jamming. The loading and unloading racks are re-scanned by radar to determine a second target pose for complete alignment, and the forks are adjusted. The fork arms are controlled to descend until the force data falls within the no-load range, eliminating minor deviations that may occur after rough docking and achieving high-precision alignment of the loading and unloading racks. Finally, visual verification using a camera ensures the reliability of the stacking operation and effectively solves the problem of limiting the docking card when stacking the second type of material rack.

[0028] According to an embodiment of the present invention, a rack stacking method based on an unmanned forklift is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0029] This embodiment provides a rack stacking method based on an unmanned forklift, which can be used with an unmanned forklift equipped with radar, camera and sensors. Figure 3 This is a schematic diagram of an unmanned forklift according to an embodiment of the present invention, as shown below. Figure 3 As shown, the radar uses 3D navigation radar, also known as 3D LiDAR. At least one is installed on the top of the unmanned forklift (scan0), used for global SLAM (Simultaneous Localization and Mapping) to ensure accurate navigation to the target area in the work environment. Another is installed in the center below the fork arm (scan2), with a field of view of no less than 180×50°, used to detect the pose of the loading and unloading racks. 2D cameras are used, installed on the left side of the left fork (cam11) and the right side of the right fork (cam12), respectively, to visually identify whether the left and right triangular grooves at the docking point of the loading and unloading racks are fully engaged, thus verifying the docking accuracy. Weighing sensors, also known as pressure sensors, are installed at the ends of the left and right fork zippers (…). Figure 3 The gray portion (in the image) is used to sense the force on the upper and lower surfaces of the forks.

[0030] This embodiment provides a rack stacking method based on an unmanned forklift, which can be used with unmanned forklifts. Figure 4 This is a flowchart of a rack stacking method based on an unmanned forklift according to an embodiment of the present invention, such as... Figure 4 As shown, the process includes the following steps: Step S401: The forks of the unmanned forklift pick up the loading rack and the fork arms of the unmanned forklift are raised to the first preset height, so that the unmanned forklift moves to the target position of the unloading rack.

[0031] Specifically, the rack to be stacked is defined as the unloading rack, and the rack to be stacked on top of the unloading rack is defined as the loading rack. The unmanned forklift travels to the loading rack storage area via a top-mounted 3D navigation radar, extends its forks, and inserts them into the bottom fork holes of the loading rack to retrieve it. Then, the unmanned forklift raises its forks to a first preset height to prevent collisions with the unloading rack. Maintaining a constant fork height, the loading rack is smoothly transported to the target position, ≥0.2m in front of the unloading rack stacking station. This distance ensures that the 3D LiDAR in the center under the forks can completely scan the docking structure of the unloading rack, providing a positional basis for subsequent rack stacking.

[0032] Step S402: Based on the radar scanning of the loading rack and unloading rack of the unmanned forklift, determine the first target pose and adjust the forks to the first target pose. The first target pose prevents the loading rack from contacting the limit block of the unloading rack.

[0033] Specifically, a 3D LiDAR located beneath the fork arm scans the loading and unloading racks separately. Based on the scanned point cloud data and constraints to avoid the loading rack getting stuck, the first target pose the forks need to reach is calculated. Large-scale positional deviations are eliminated by forklift chassis translation, and residual attitude deviations are corrected by the mast tilt adjustment, fork carriage lateral movement, and pitch functions of the superstructure, collaboratively adjusting the forks to the first target pose. Since the forks are mounted on the loading rack, changes in the fork's pose will cause changes in the loading rack's pose. Therefore, after the loading rack's pose is adjusted, it avoids getting stuck and achieves a coarse docking with the unloading rack without collision or interference.

[0034] In step S403, the fork arm is lowered to the second preset height. Based on the sensor of the unmanned forklift, the force data of the fork is continuously read. According to the change of the force data, the fork arm is raised to the third preset height. The third preset height causes the loading rack and unloading rack to disengage.

[0035] Specifically, to control the fork arm to descend from a first preset height to a second preset height, load cells installed at the ends of the left and right fork carriage chains continuously collect force data on the upper and lower surfaces of the left and right sides of the fork. The average of the data collected by the two sensors is used as the force data for the fork. Based on the real-time force data, a third preset height is determined corresponding to the point where the loading rack and unloading rack disengage. The fork arm is then raised to this third preset height to eliminate the risk of contact interference for subsequent alignment of the loading rack with the unloading rack.

[0036] Step S404: Based on the radar rescanning of the loading rack and unloading rack, determine the second target pose and adjust the forks to the second target pose, control the fork arms to descend until the force data reaches the no-load range, stop the fork arms to descend, and align the loading rack and unloading rack with the second target pose.

[0037] Specifically, referring to step S402, the loading rack and unloading rack are rescanned. Based on the scanned point cloud data and the constraint that the loading rack and unloading rack are perfectly aligned, the second target pose that the forks need to reach is calculated. Through the coordinated actions of the forklift chassis translation and the superstructure (mast tilt adjustment, fork carriage lateral / tilt), residual deviations are dynamically fine-tuned to eliminate them, and the forks are adjusted to the second target pose. At this point, the loading rack and unloading rack are precisely aligned. After the pose adjustment is completed, the fork arms are controlled to descend, and the force data is continuously read and compared with the pre-calibrated no-load range. This no-load range is the range of force data when the forks are not supporting the loading rack. Therefore, when the force data falls into this no-load range, it is determined that the loading rack has been stably placed on the unloading rack, and the descent of the fork arms is stopped.

[0038] Step S405: Visual verification is performed based on the camera of the unmanned forklift. Upon passing the visual verification, the stacking of the loading rack and unloading rack is completed.

[0039] Specifically, 2D cameras are installed on the left side of the left fork and the right side of the right fork respectively to photograph the left and right triangular grooves at the joint between the loading rack and the unloading rack, and visual verification is performed to determine whether the two triangular grooves are completely closed (no misalignment or gaps), thereby verifying whether the docking accuracy of the loading and unloading racks meets the stacking requirements. When the visual verification is passed, it means that the loading rack has been accurately stacked on the unloading rack. At this time, the forks are maintained to support the loading rack, completing the high-precision stacking operation of the loading and unloading racks.

[0040] This invention utilizes forks to pick up the loading rack, raising the fork arms to a first preset height to prevent collisions between the loading and unloading racks. The forks are then moved to the target position, providing an unobstructed view for subsequent radar scanning of the loading and unloading rack docking structure, ensuring the integrity and accuracy of subsequent pose detection. By scanning the loading and unloading racks with radar, the first target pose is determined, and the forks are adjusted to this pose to ensure no contact between the loading and unloading rack limit blocks, thus avoiding jamming issues at the source. Then, the fork arms are controlled to descend from the first preset height to a second preset height. Based on changes in force data collected by sensors, the fork arms are controlled to rise to a third preset height, ensuring complete disengagement of the loading and unloading racks. This eliminates contact friction and interference risks for subsequent translational alignment, further avoiding jamming. The loading and unloading racks are re-scanned by radar to determine a second target pose for complete alignment, and the forks are adjusted. The fork arms are controlled to descend until the force data falls within the no-load range, eliminating minor deviations that may occur after rough docking and achieving high-precision alignment of the loading and unloading racks. Finally, visual verification using a camera ensures the reliability of the stacking operation and effectively solves the problem of limiting the docking card when stacking the second type of material rack.

[0041] This embodiment provides a rack stacking method based on an unmanned forklift, which can be used with the aforementioned unmanned forklift. The method specifically includes the following steps: Step S501: The forks of the unmanned forklift pick up the loading rack and the fork arms of the unmanned forklift are raised to a first preset height, causing the unmanned forklift to move to the target position of the unloading rack. For details, please refer to [link to details]. Figure 4 Step S401 of the illustrated embodiment will not be described again here.

[0042] In some alternative implementations, Figure 5 This is a schematic diagram of the fork arm being raised to a first preset height according to an embodiment of the present invention, as shown below. Figure 5 As shown, after the forks of the unmanned forklift pick up the loading rack, the control arm is raised to the first preset height.

[0043] Step S502: Based on the radar scanning of the loading rack and unloading rack of the unmanned forklift, determine the first target pose and adjust the forks to the first target pose. The first target pose prevents the loading rack from contacting the limit block of the unloading rack.

[0044] Specifically, step S502 above, based on the radar scanning of the loading and unloading racks of the unmanned forklift, determines the pose of the first target, including: Step S5021: Obtain first point cloud data by scanning the bottom of the loading rack with radar, and obtain second point cloud data by scanning the top of the unloading rack.

[0045] Specifically, a 3D LiDAR installed in the center below the fork arm is used to scan the key docking structure at the bottom of the loading rack (including the lower edge of the triangular groove and the support surface) to obtain the first point cloud data; at the same time, the key docking structure at the top of the unloading rack (including the upper edge of the triangular groove and the longitudinal limit block) is scanned to obtain the second point cloud data, ensuring that the two sets of point cloud data can fully reflect the core docking features of the loading and unloading racks.

[0046] Step S5022: Perform model registration on the first point cloud data and the second point cloud data respectively to obtain the first pose data and the second pose data.

[0047] Specifically, an improved NDT (Normal Distributions Transform) point cloud matching algorithm optimized for small-sized material rack structures is used to register the first point cloud data with a pre-stored standard 3D model of the loading rack to obtain the first pose data of the loading rack in the radar coordinate system; similarly, the second point cloud data is registered with a pre-stored standard 3D model of the unloading rack to obtain the second pose data of the unloading rack in the radar coordinate system.

[0048] Step S5023: Perform coordinate system transformation on the first pose data and the second pose data respectively, and calculate the first relative deviation between the first pose data and the second pose data after coordinate system transformation.

[0049] Specifically, based on the pre-calibrated translation and rotation matrices between the LiDAR coordinate system and the forklift body coordinate system, the first pose data and the second pose data are uniformly transformed to the forklift body coordinate system. The first relative deviation between the transformed first pose data and the second pose data is calculated, including longitudinal distance deviation, lateral alignment deviation, attitude tilt angle deviation, and vertical height deviation.

[0050] Step S5024: Based on the first relative deviation, calculate the first target pose of the forks when the limit blocks of the loading rack and unloading rack are at a preset safe distance.

[0051] Specifically, based on the first relative deviation, and constrained by maintaining a preset safe distance (a distance that ensures no jamming, generally greater than or equal to 50mm) between the limit blocks of the upper and lower racks, and combined with the preset relative positional relationship between the forks and the upper rack, the aforementioned NDT point cloud matching algorithm is used to reverse-calculate the first target pose that the forks need to reach, providing a precise basis for subsequent pose adjustments during coarse docking. Optionally, the pose calculation process is existing technology and will not be elaborated here.

[0052] In some alternative implementations, Figure 6 This is a schematic diagram of the forks being positioned in a first target pose according to an embodiment of the present invention, as shown below. Figure 6As shown, adjust the position of the forks to the first target position. At this time, the limit blocks of the loading rack and unloading rack are at a preset safe distance.

[0053] Step S503: Control the fork arm to descend to the second preset height. Based on the sensor of the unmanned forklift, continuously read the force data of the fork. According to the change of the force data, control the fork arm to rise to the third preset height. The third preset height causes the loading rack and unloading rack to disengage. The second preset height is the difference between the first preset height and the first preset value. The first preset value is less than the preset safety threshold.

[0054] Specifically, step S503 includes: In step S5031, during the process of controlling the fork arm to descend from the first preset height to the second preset height, the mean and standard deviation of the force data of the fork continuously read by the sensor are calculated as the force benchmark of the fork.

[0055] Specifically, the first preset value is an empirical value less than a preset safety threshold, generally less than 50mm. The difference between the first preset height and this first preset value is determined as the second preset height, ensuring that the loading and unloading racks remain in a non-contact state during the descent of the fork arm from the first preset height to the second preset height. During this dynamic descent, the readings of the weighing sensor, i.e., the force data of the forks, are continuously acquired. The mean and standard deviation of all force data during this descent are calculated through rolling statistics, and this is used as the force benchmark for the forks. This benchmark can accurately reflect the force noise distribution characteristics under interference such as the inertia of the fork arm descent and the vibration of the fork carriage, providing a reliable reference for subsequent contact determination.

[0056] Step S5032: Control the fork arm to continuously descend from the second preset height until the force data and force reference meet the preset conditions, and obtain the fork arm height.

[0057] Specifically, the control arm continues to descend from a second preset height, continuously acquiring force data from the load cell. This force data is compared in real time with a force reference until the force data and the force reference at a certain moment meet a preset condition. This preset condition is: force data < mean - 3 × standard deviation. At this point, it is determined that the loading and unloading racks have just made contact, and the height of the fork arm at this moment is obtained. Optionally, this step, based on the premise that the noise of the load cell's force data follows a normal distribution, adopts the 3-sigma principle to ensure the accuracy of contact determination and effectively avoid misjudgments.

[0058] Step S5033: Determine a third preset height based on the fork arm height, and control the fork arm to be raised to the third preset height. The third preset height is the sum of the fork arm height and the second preset value. The second preset value is less than the height of the limit block of the unloading rack.

[0059] Specifically, the second preset value is an empirical parameter, which must be strictly less than the height of the limit block of the unloading rack. It is typically chosen to ensure that the loading rack is completely detached from the unloading rack without exceeding the limit block height. The sum of the forklift height and the second preset value is determined as the third preset height. The unmanned forklift controls the forklift to rise to this third preset height, completely disengaging the loading rack from the unloading rack, thus eliminating the risk of contact interference for subsequent alignment of the loading rack with the unloading rack.

[0060] In some alternative implementations, Figure 7 This is a contact diagram of the loading rack and unloading rack according to an embodiment of the present invention, as shown below. Figure 7 As shown, the gray line segment is represented by L1, which is a preset safety distance generally greater than or equal to 50mm. With the unmanned forklift stationary, the fork arm is lowered until the force data and force reference meet the preset conditions. At this point, the fork arm height allows the loading rack and unloading rack to just touch. Figure 8 This is a schematic diagram of the fork arm being raised to a third preset height according to an embodiment of the present invention, as shown below. Figure 8 As shown, the gray line segment is represented by L1, which is a preset safety distance generally greater than or equal to 50mm. The sum of the fork arm height and the second preset value is determined as the third preset height. Without moving the unmanned forklift, the fork arm is raised to this third preset height, at which point the loading rack and unloading rack are disengaged.

[0061] Step S504: Based on the radar rescanning of the loading rack and unloading rack, determine the second target pose and adjust the forks to the second target pose, control the fork arms to descend until the force data reaches the no-load range, stop the fork arms to descend, and align the loading rack and unloading rack with the second target pose.

[0062] Specifically, step S504 above, based on radar rescanning of the loading rack and unloading rack, determines the pose of the second target, including: In step S5041, the bottom of the loading rack is rescanned by radar to obtain the third point cloud data, and the top of the unloading rack is rescanned to obtain the fourth point cloud data. For details, please refer to step S5021, which will not be repeated here.

[0063] Step S5042 involves performing model registration on the third and fourth point cloud data respectively to obtain the third pose data and the fourth pose data. For details, please refer to step S5022, which will not be repeated here.

[0064] Step S5043 involves performing coordinate system transformations on the third and fourth pose data respectively, and calculating the second relative deviation between the third and fourth pose data after the coordinate system transformation. For details, please refer to step S5023, which will not be repeated here.

[0065] Step S5044: Based on the second relative deviation, calculate the second target pose of the forks when the loading rack and unloading rack are aligned. See step S5024 for details, which will not be repeated here.

[0066] In some alternative implementations, Figure 9 This is a schematic diagram of the forks being positioned in a second target pose according to an embodiment of the present invention, as shown below. Figure 9 As shown, adjust the position of the forks to the second target position, at which point the loading rack and unloading rack are aligned. Figure 10 This is a schematic diagram illustrating the lowering of the fork arm according to an embodiment of the present invention until the force data reaches the no-load range, as shown below. Figure 10 As shown, Figure 9 Although the upper and lower racks are aligned, there is still a distance between them before they can make contact. Therefore, the fork arms are controlled to descend. When the fork arms descend to the point where the force data reaches the no-load range, there is no distance between the upper and lower racks, thus achieving precise stacking.

[0067] Step S505: Visual verification is performed based on the camera of the unmanned forklift. When the visual verification is passed, the stacking of the loading rack and unloading rack is completed.

[0068] Specifically, step S505 above performs visual verification based on the camera of the unmanned forklift, including: Step S5051: Based on the camera, the triangular grooves of the loading rack and the unloading rack are photographed to obtain the first image and the second image.

[0069] Specifically, 2D cameras installed on the left side of the left fork and the right side of the right fork are used to take pictures of the left and right triangular grooves at the docking point of the loading rack and unloading rack, respectively, to obtain a first image containing the docking state of the left triangular groove and a second image containing the docking state of the right triangular groove.

[0070] Step S5052: Identify the first image and the second image, and determine whether the triangular grooves of the loading rack and the unloading rack are aligned.

[0071] Specifically, a large number of image samples of closed and open triangular slots are used to train a deep learning network, enabling it to accurately classify closed and open triangular slots. Optionally, the structure of the deep learning network is not restricted. In practical use, the first and second images are input into the trained deep learning network, and the network determines whether the two triangular slots are closed by judging whether there is no misalignment and no obvious gap between them.

[0072] Step S5053: When the triangular grooves of the loading rack and the unloading rack are closed, visual verification is confirmed.

[0073] Specifically, if the image recognition result shows that the two triangular slots are not misaligned and have no gaps, it is determined that the loading and unloading racks have achieved high-precision alignment and pass the visual verification.

[0074] Step S5054: If the triangular grooves of the loading rack and the unloading rack are not closed, it is determined that the visual verification has not been passed.

[0075] Specifically, if the image recognition results show that there is obvious misalignment or gap between the two triangular grooves, it is determined that the docking accuracy of the loading and unloading racks does not meet the standard, and the visual verification is not passed.

[0076] Step S506: If the visual verification fails, control the fork arm to rise to the fifth preset height, return to the radar-based rescanning of the loading and unloading racks, determine the second target pose and adjust the forks to the second target pose, control the fork arm to descend until the force data reaches the no-load range, stop the fork arm descent, and determine whether the visual verification has been passed.

[0077] Specifically, if the visual verification fails, the fork arm is raised to a fifth preset height. This height must ensure that the loading rack is completely detached from the unloading rack and does not exceed the limit block to avoid contact interference during subsequent adjustments. Return to step S504, re-execute the fine docking, and perform visual verification again to determine whether it passes.

[0078] Step S507: If the visual verification fails for a preset number of consecutive times, an alarm signal is issued.

[0079] Specifically, if the visual verification fails for a preset number of consecutive times (e.g., 3 times), it is determined that there is an uncorrectable deviation in the precision docking of the upper and lower material racks, and an alarm signal is issued to remind the staff to check for problems such as abnormal material rack structure, sensor failure, or interference from the working environment, so as to ensure operational safety and stacking quality.

[0080] This invention utilizes forks to pick up the loading rack, raising the fork arms to a first preset height to prevent collisions between the loading and unloading racks. The forks are then moved to the target position, providing an unobstructed view for subsequent radar scanning of the loading and unloading rack docking structure, ensuring the integrity and accuracy of subsequent pose detection. By scanning the loading and unloading racks with radar, the first target pose is determined, and the forks are adjusted to this pose to ensure no contact between the loading and unloading rack limit blocks, thus avoiding jamming issues at the source. Then, the fork arms are controlled to descend from the first preset height to a second preset height. Based on changes in force data collected by sensors, the fork arms are controlled to rise to a third preset height, ensuring complete disengagement of the loading and unloading racks. This eliminates contact friction and interference risks for subsequent translational alignment, further avoiding jamming. The loading and unloading racks are re-scanned by radar to determine a second target pose for complete alignment, and the forks are adjusted. The fork arms are controlled to descend until the force data falls within the no-load range, eliminating minor deviations that may occur after rough docking and achieving high-precision alignment of the loading and unloading racks. Finally, visual verification using a camera ensures the reliability of the stacking operation and effectively solves the problem of limiting the docking card when stacking the second type of material rack.

[0081] This embodiment also provides a rack stacking device based on an unmanned forklift, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0082] This embodiment provides a rack stacking device based on an unmanned forklift, such as Figure 11 As shown, it includes: The first control module 1101 is used to enable the forks of the unmanned forklift to pick up the loading rack and control the fork arms of the unmanned forklift to be raised to a first preset height, so that the unmanned forklift moves to the target position of the unloading rack.

[0083] The second control module 1102 is used to scan the loading rack and unloading rack based on the radar of the unmanned forklift, determine the first target pose and adjust the forks to the first target pose, so that the loading rack avoids contacting the limit block of the unloading rack.

[0084] The third control module 1103 is used to control the fork arm to descend to the second preset height. Based on the sensors of the unmanned forklift, it continuously reads the force data of the fork and controls the fork arm to rise to the third preset height according to the changes in the force data. The third preset height causes the loading rack and unloading rack to disengage.

[0085] The fourth control module 1104 is used to rescan the loading rack and unloading rack based on radar, determine the second target pose and adjust the forks to the second target pose, control the fork arms to descend until the force data reaches the no-load range, stop the fork arms to descend, and align the loading rack and unloading rack with the second target pose.

[0086] The verification module 1105 is used for visual verification based on the camera of the unmanned forklift, and the stacking of the loading rack and unloading rack is completed through visual verification.

[0087] In some alternative implementations, the second control module 1102 includes: The first scanning unit is used to obtain first point cloud data by scanning the bottom of the loading rack with radar and to obtain second point cloud data by scanning the top of the unloading rack.

[0088] The first registration unit is used to perform model registration on the first point cloud data and the second point cloud data respectively to obtain the first pose data and the second pose data.

[0089] The first transformation unit is used to perform coordinate system transformation on the first pose data and the second pose data respectively, and to calculate the first relative deviation between the first pose data and the second pose data after the coordinate system transformation.

[0090] The first calculation unit is used to calculate the first target pose of the forks when the limit blocks of the loading rack and unloading rack are at a preset safe distance, based on the first relative deviation.

[0091] In some alternative implementations, the third control module 1103 includes: The second calculation unit is used to calculate the mean and standard deviation of the fork force data continuously read by the sensor during the process of controlling the fork arm to descend from the first preset height to the second preset height, and to use the result as the force benchmark for the fork.

[0092] The first control unit is used to control the fork arm to continuously descend from the second preset height until the force data and force reference meet the preset conditions, and to obtain the fork arm height.

[0093] The second control unit is used to determine a third preset height based on the height of the fork arm and control the fork arm to be raised to the third preset height.

[0094] In some alternative implementations, the verification module 1105 includes: The imaging unit is used to capture images of the triangular grooves of the loading rack and the unloading rack using a camera, thereby obtaining a first image and a second image.

[0095] The recognition unit is used to recognize the first image and the second image, and to determine whether the triangular grooves of the loading rack and the unloading rack are aligned.

[0096] The first determining unit is used to determine whether visual verification has been passed when the triangular grooves of the loading rack and the unloading rack are closed.

[0097] The second determining unit is used to determine that visual verification has not been passed when the triangular grooves of the loading rack and the unloading rack are not closed.

[0098] In some optional implementations, the second preset height is the difference between the first preset height and the first preset value, where the first preset value is less than a preset safety threshold. The third preset height is the sum of the fork arm height and the second preset value, where the second preset value is less than the height of the limit block of the unloading rack.

[0099] In some alternative implementations, the fourth control module 1104 includes: The second scanning unit is used to rescan the bottom of the loading rack to obtain the third point cloud data based on the radar, and to rescan the top of the unloading rack to obtain the fourth point cloud data.

[0100] The second registration unit is used to register the third point cloud data and the fourth point cloud data into models to obtain the third pose data and the fourth pose data.

[0101] The second transformation unit is used to perform coordinate system transformation on the third pose data and the fourth pose data respectively, and to calculate the second relative deviation between the third pose data and the fourth pose data after the coordinate system transformation.

[0102] The second calculation unit is used to calculate the second target pose of the forks when the loading rack and unloading rack are aligned, based on the second relative deviation.

[0103] In some alternative implementations, after the verification module 1105, the device further includes: The judgment module is used to control the fork arm to rise to the fifth preset height when the visual verification fails, return to the radar-based rescanning of the loading and unloading racks, determine the second target pose and adjust the forks to the second target pose, control the fork arm to descend until the force data reaches the no-load range, stop the fork arm descent, and determine whether the visual verification has been passed.

[0104] The alarm module is used to issue an alarm signal when the visual verification fails for a preset number of consecutive times.

[0105] The rack stacking device based on an unmanned forklift provided in this embodiment of the invention can execute the rack stacking method based on an unmanned forklift provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0106] Figure 12This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0107] The following is a detailed reference. Figure 12 The diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 1201, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 1202 or a program loaded from memory 1208 into random access memory (RAM) 1203. The RAM 1203 also stores various programs and data required for the operation of the electronic device. The processor 1201, ROM 1202, and RAM 1203 are interconnected via a bus 1204. An input / output (I / O) interface 1205 is also connected to the bus 1204.

[0108] Typically, the following devices can be connected to I / O interface 1205: input devices 1206 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 1207 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 1208 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1209. Communication device 1209 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 12 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0109] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 1209, or installed from a memory 1208, or installed from a ROM 1202. When the computer program is executed by the processor 1201, it performs the functions defined in the unmanned forklift-based rack stacking method of the embodiments of the present invention.

[0110] Figure 12 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0111] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the rack stacking method based on an unmanned forklift shown in the above embodiments is implemented.

[0112] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0113] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for stacking racks based on unmanned forklifts, characterized by, The application is applied to unmanned forklifts which are equipped with radars, cameras and sensors; The method comprises: The forks of the unmanned forklift are used to pick up a loading rack, and the fork arms of the unmanned forklift are controlled to be lifted to a first preset height, and the unmanned forklift is moved to a target position of a discharging rack; The loading rack and the discharging rack are scanned by the radar of the unmanned forklift, a first target pose is determined, and the forks are adjusted to the first target pose, so that the loading rack avoids contacting the limiting block of the discharging rack; The fork arms are controlled to be lowered to a second preset height, the force data of the forks are continuously read by the sensor of the unmanned forklift, the fork arms are controlled to be lifted to a third preset height according to the change of the force data, and the third preset height makes the loading rack and the discharging rack disengage from each other; The loading rack and the discharging rack are rescanned by the radar, a second target pose is determined, the forks are adjusted to the second target pose, the fork arms are controlled to be lowered to the empty load range of the force data, and the lowering of the fork arms is stopped, and the second target pose makes the loading rack and the discharging rack aligned; Visual verification is performed by the camera of the unmanned forklift, and the stacking of the loading rack and the discharging rack is completed when the visual verification is passed.

2. The method of claim 1, wherein, The scanning of the loading rack and the discharging rack by the radar of the unmanned forklift and the determination of the first target pose comprise: First point cloud data is obtained by scanning the bottom of the loading rack, and second point cloud data is obtained by scanning the top of the discharging rack; The first point cloud data and the second point cloud data are respectively subjected to model registration to obtain first pose data and second pose data; The first pose data and the second pose data are respectively subjected to coordinate system conversion, and a first relative deviation between the first pose data and the second pose data after the coordinate system conversion is calculated; Based on the first relative deviation, the first target pose of the forks when the distance between the loading rack and the limiting block of the discharging rack is a preset safety distance is calculated.

3. The method of claim 1, wherein, The lowering of the fork arms to the second preset height, the continuous reading of the force data of the forks by the sensor of the unmanned forklift, and the lifting of the fork arms to the third preset height according to the change of the force data comprise: During the lowering of the fork arms from the first preset height to the second preset height, the force data of the forks continuously read by the sensor are used to calculate the mean value and the standard deviation as the force reference of the forks; The fork arms are continuously lowered from the second preset height until the force data and the force reference meet a preset condition, and the height of the fork arms is obtained; Based on the height of the fork arms, the third preset height is determined, and the fork arms are controlled to be lifted to the third preset height.

4. The method of claim 1, wherein, The visual verification by the camera of the unmanned forklift comprises: First and second pictures are obtained by shooting the triangular grooves of the loading rack and the discharging rack by the camera; The first and second pictures are identified to determine whether the triangular grooves of the loading rack and the discharging rack are closed. determining that the visual verification is passed when the triangular groove of the upper loading rack and the triangular groove of the lower loading rack are closed; determining that the visual verification is not passed when the triangular groove of the upper loading rack and the triangular groove of the lower loading rack are not closed.

5. The method of claim 3, wherein, The second preset height is a difference between the first preset height and a first preset value, and the first preset value is less than a preset safety threshold; The third preset height is a sum of the height of the fork arm and a second preset value, and the second preset value is less than the height of the limiting block of the lower loading rack.

6. The method of claim 1, wherein, The method further comprises the following steps after the visual verification based on the camera of the unmanned forklift: When the visual verification is not passed, controlling the fork arm to lift by a fifth preset height, returning to the step of rescanning the upper loading rack and the lower loading rack based on the radar, determining the second target pose of the fork and adjusting the fork to the second target pose, controlling the fork arm to descend to the range of the load data reaching the empty load, stopping the step of descending the fork arm, and determining whether the visual verification is passed; When the visual verification is not passed for a continuous preset number of times, an alarm signal is sent. The unmanned forklift is provided with a radar, a camera, and a sensor; The device comprises:

7. The method of claim 1, wherein, A first control module is configured to make the fork of the unmanned forklift pick up the upper loading rack, control the fork arm of the unmanned forklift to lift to a first preset height, and make the unmanned forklift move to a target position of the lower loading rack. A second control module is configured to scan the upper loading rack and the lower loading rack based on the radar of the unmanned forklift, determine a first target pose and adjust the fork to the first target pose, and make the upper loading rack avoid contacting the limiting block of the lower loading rack. A third control module is configured to control the fork arm to descend to a second preset height, continuously read the load data of the fork based on the sensor of the unmanned forklift, and control the fork arm to lift to a third preset height according to the change of the load data, and make the upper loading rack and the lower loading rack be separated.

8. A pallet stacker based on unmanned fork truck, characterized by, A fourth control module is configured to rescan the upper loading rack and the lower loading rack based on the radar, determine a second target pose and adjust the fork to the second target pose, control the fork arm to descend to the range of the load data reaching the empty load, stop the fork arm from descending, and make the upper loading rack and the lower loading rack be aligned. ​ ​ ​ ​ ​ A verification module is configured to perform visual verification based on a camera of the unmanned forklift, and the stacking of the upper loading rack and the lower loading rack is completed when the visual verification is passed.

9. An electronic device, comprising: The application further discloses a computer readable storage medium. A memory and a processor are in communication connection with each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the unmanned forklift-based rack stacking method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for enabling a computer to perform the unmanned forklift-based rack stacking method according to any one of claims 1 to 7.

Images