Warehouse management method and system based on AGV robot

Abstract

The invention discloses a warehouse management method and system based on an AGV robot, and relates to the technical field of robot warehouse management, and the method comprises the steps: deploying a sensor in a warehouse to collect warehouse information, and collecting positioning data, speed data and image data through the AGV robot; a warehouse article replenishment task is generated based on the storage information, an initial path is planned according to AGV robot positioning data, and an artificial potential field and nonlinear programming are adopted to optimize the initial path to obtain an optimized path; the AGV robot arrives at the replenishment position according to the optimized path, and based on the image data, the visual perception technology is used for positioning the accurate position information of the replenishment goods to carry out replenishment operation; and displaying the storage information in real time, generating a replenishment record according to the replenishment operation, and storing the replenishment record in a database. The AGV robot has higher flexibility and safety, the adaptability of the AGV robot to the complex warehouse environment is improved, and efficient warehouse inventory management is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot warehouse management, and particularly to a warehouse management method and system based on an AGV robot. Background Art

[0002] With the increasing complexity of modern warehouse management systems, especially in large-scale and efficient logistics warehouses, traditional manual operations can no longer meet the requirements of fast, accurate, and efficient management. To solve these problems, automated warehouse management systems have emerged, and the application of AGV (Automated Guided Vehicle) robots has become the key to improving warehouse operation efficiency. AGV robots, equipped with a variety of sensors and visual perception systems, can automatically complete operations such as cargo storage, transportation, and replenishment, greatly improving the automation level and efficiency of warehouse operations.

[0003] However, there are still deficiencies in existing AGV robots. When facing complex and changing warehouse environments, the automatic path planning and cargo recognition of existing AGV robots cannot meet the requirements of efficient and flexible warehouse management, and the accurate positioning of items is often affected by the warehouse environment, resulting in a significant reduction in the accuracy and stability of recognition, which affects the efficiency of warehouse inventory replenishment. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a warehouse management method and system based on an AGV robot, which solves the problems that the existing technology cannot meet the requirements of efficient and flexible warehouse management and has low accuracy and stability in cargo recognition.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a warehouse management method based on an AGV robot, which includes,

[0008] Deploy sensors in the warehouse to collect warehousing information, and collect positioning data, speed data, and image data through the AGV robot;

[0009] Generate a warehouse item replenishment task based on the warehousing information, plan an initial path according to the AGV robot positioning data, and optimize the initial path using artificial potential field and nonlinear programming to obtain an optimized path;

[0010] The AGV robot reaches the replenishment position according to the optimized path, and uses visual perception technology based on the image data to locate the accurate position information of the replenishment item for replenishment operations;

[0011] The warehousing information is displayed in real time, and replenishment records are generated based on replenishment operations and stored in the database.

[0012] As a preferred solution of the warehouse management method based on AGV robots according to the present invention, wherein: generating a warehouse item replenishment task based on the warehousing information, planning an initial path according to the AGV robot positioning data, and optimizing the initial path by using artificial potential field and nonlinear programming to obtain an optimized path means comparing the quantity of warehouse items in the warehousing information with a set item threshold. If the quantity of warehouse items is less than the set item threshold, a corresponding warehouse item replenishment task is generated. The warehouse item replenishment task includes the pick-up point location, the replenishment point location, the replenishment items, and the replenishment quantity.

[0013] Construct a three-dimensional model of the warehouse based on the warehousing information, and mark the obstacle information and the AGV robot position in the three-dimensional model.

[0014] After receiving the warehouse item replenishment task, the AGV robot uses the 3D A * algorithm to obtain the pick-up initial path of the AGV robot according to the positioning data and the pick-up point location.

[0015] Calculate the attractive force U a according to the AGV robot positioning data and the pick-up point location, and synchronously calculate the repulsive force U r ;

[0016] Combine the attractive force U a and the repulsive force U r to form a resultant force F. Based on the AGV robot, a coordinate system is constructed, and the resultant force F is decomposed according to the coordinate system direction to obtain F x and F y respectively. Calculate the ratio of the decomposed force F x and F y to the mass of the AGV robot to obtain the acceleration a x and a y in the coordinate system direction. Synchronously decompose the AGV robot speed v according to the coordinate system direction to obtain v x and v y , and update the initial path nodes according to the decomposed speed and acceleration.

[0017] After updating the path node coordinates, construct a path optimization objective function J based on the nonlinear programming technique:

[0018]

[0019] where w(t) is the angular velocity of the AGV robot, α1, α2, and α3 are weight coefficients, and t f is the moving time of the AGV robot;

[0020] Set physical constraint conditions for the path optimization objective function J through the kinematic model;

[0021] Use the interior point method to solve the objective function J to obtain the optimized AGV robot speed and acceleration, and further update the updated path node coordinates through the optimized speed and acceleration to form a picking optimization path;

[0022] The AGV robot reaches the picking point through the picking optimization path to obtain the replenishment items, then generates a replenishment optimization path according to the picking point position and the replenishment point position, and reaches the replenishment point position according to the replenishment optimization path.

[0023] As a preferred solution of the warehouse management method based on AGV robots described in the present invention, wherein: the replenishment operation of positioning the accurate position information of the replenishment items using visual perception technology based on image data means that after the AGV robot reaches the replenishment point position, it acquires the replenishment point image I, performs image equalization operation on the replenishment point image I to obtain the average environmental illumination image as the flat field image R, and sets a short exposure time to re-take the replenishment point image as the dark field image D, and corrects the replenishment point image I according to the flat field image and the dark field image;

[0024] Convert the corrected replenishment point image into a grayscale image, apply Gaussian blur for denoising, and calculate the Hessian matrix of each pixel point in the replenishment point image through the SURF algorithm;

[0025] Perform eigenvalue decomposition based on the Hessian matrix to obtain eigenvalues λ1 and λ2;

[0026] Take the denoised replenishment point image as the target image and the replenishment point image I as the source image, extract the key points of the target image and the source image respectively to form key point sets, calculate the Euclidean distance of the key points in the two key point sets for minimum matching to obtain matching point pairs;

[0027] Randomly select four point pairs from the matching point pairs, and estimate the homography matrix G by the least squares method;

[0028] Train the YOLOv3 algorithm and use the YOLOv3 algorithm to extract the position coordinates (i1, j1) of the goods from the source image and calculate the position coordinate depth Z1, and calculate the X1 and Y1 coordinates of the three-dimensional coordinates of the goods through the position coordinates of the goods;

[0029] Combine the coordinate depth Z1 with the X1 and Y1 coordinates of the three-dimensional coordinates of the goods to form the complete three-dimensional coordinates of the goods (X1, Y1, Z1), and the AGV robot performs replenishment operations according to the three-dimensional coordinates of the goods.

[0030] As a preferred solution of the warehouse management method based on AGV robots according to the present invention, the following steps are included: After marking the obstacle information and the positions of AGV robots in the three-dimensional model, the positions of idle AGV robots can be browsed in real time in the three-dimensional warehouse model. Whenever a replenishment task for warehouse items occurs, the distances between the replenishment task and the positions of all idle AGV robots are calculated based on the replenishment point position, and the idle AGV robot closest to the replenishment point position is selected to receive the replenishment task and the AGV robot is marked as being in a working state.

[0031] As a preferred solution of the warehouse management method based on AGV robots according to the present invention, the following steps are included: Sensors are deployed in the warehouse to collect warehousing information, and positioning data, speed data, and image data are collected through AGV robots, which means that lidar and RFID sensors are deployed in the warehouse to collect warehouse environment data and warehouse goods data, and an IMU sensor, a GPS sensor, and a camera are installed on the AGV robot.

[0032] As a preferred solution of the warehouse management method based on AGV robots according to the present invention, the following steps are included: The real-time display of the warehousing information means that the warehousing information collected by the sensors is displayed in real time through a visualization interface, and the three-dimensional warehouse model is synchronously displayed in the visualization interface.

[0033] As a preferred solution of the warehouse management method based on AGV robots according to the present invention, the following steps are included: The generation of replenishment records according to replenishment operations and the storage of them in the database means that replenishment records are generated for each replenishment operation, and the replenishment records and warehousing information are stored in the database. The database regularly detects the integrity and security of the stored data, and synchronously uploads the stored data to the cloud for backup.

[0034] In a second aspect, the present invention provides a warehouse management system based on AGV robots, including

[0035] An information acquisition module, which is used to deploy sensors to collect warehousing information, and collect positioning data, speed data, and image data through AGV robots;

[0036] A warehouse management module, which is used to generate replenishment tasks according to the warehousing information, allocate AGV robots to plan replenishment paths, and perform replenishment operations based on the accurate positions of replenishment items identified from the image data collected by AGV robots;

[0037] A display and storage module, which is used to display the warehousing information in real time, and generate replenishment records according to replenishment operations and store them in the database.

[0038] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the warehouse management method based on an AGV robot as described in the first aspect of the present invention is implemented.

[0039] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the warehouse management method based on an AGV robot as described in the first aspect of the present invention is implemented.

[0040] The beneficial effects of the present invention are as follows: by collecting warehousing information and AGV robot information to generate replenishment tasks, planning and optimizing the replenishment path based on the AGV robot, it has higher flexibility and security, and accurately identifies the three-dimensional coordinates of goods for replenishment through image recognition and coordinate point matching, improving the adaptability of the AGV robot to complex warehouse environments and realizing efficient warehouse inventory management. Description of the Drawings

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0042] Figure 1 It is a flowchart of the warehouse management method based on an AGV robot in Embodiment 1.

[0043] Figure 2 It is a structural diagram of the warehouse management system based on an AGV robot in Embodiment 1. Detailed Embodiments

[0044] To make the above objects, features, and advantages of the present invention more obvious and understandable, the detailed embodiments of the present invention will be described in detail below with reference to the drawings in the specification.

[0045] Many specific details are set forth in the following description to facilitate a thorough understanding of the present invention, but the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0046] Second, the "one embodiment" or "embodiment" referred to herein means a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The appearances of "in one embodiment" in different places in this specification do not all refer to the same embodiment, nor are they separate or alternative embodiments that exclude each other with other embodiments.

[0047] Embodiment 1, referring to Figure 1 and Figure 2 , is the first embodiment of the present invention. This embodiment provides a warehouse management method based on an AGV robot, including the following steps:

[0048] S1. Deploy sensors in the warehouse to collect warehousing information, and collect positioning data, speed data, and image data through the AGV robot;

[0049] Specifically, deploying sensors in the warehouse to collect warehousing information and collecting positioning data, speed data, and image data through the AGV robot means deploying lidar and RFID sensors in the warehouse to collect warehouse environment data and warehouse goods data, and installing an IMU sensor, a GPS sensor, and a camera on the AGV robot.

[0050] The step of deploying lidar and RFID sensors in the warehouse to collect environmental data and goods data first enhances the data acquisition ability of the entire warehouse management system. The lidar can provide an accurate environmental map for the AGV robot by precisely measuring obstacles and spatial layouts in the warehouse environment, helping it identify and avoid obstacles, and then perform precise path planning. Especially in a complex warehousing environment, the high precision and real-time performance of the lidar effectively make up for the limitations of other traditional sensors (such as ultrasonic or infrared sensors) in environmental recognition. Through RFID technology, the robot can obtain the accurate location, quantity, and relevant information of goods in real time, thereby generating corresponding replenishment tasks. This not only greatly improves the transparency and accuracy of warehouse management, but also reduces the error rate caused by manual operations and improves the overall warehousing efficiency. The combined use of the IMU sensor and the GPS sensor provides more stable and accurate positioning data for the AGV robot, greatly enhancing the adaptability of the AGV robot. Especially in a complex and dynamic warehouse environment, it can ensure that the robot can maintain a stable position under any circumstances. The application of the camera provides additional visual information for the AGV robot, enabling it to accurately identify the position and status of goods during the item replenishment process. Especially in a complex warehouse environment, the camera can not only assist the robot in avoiding obstacles, but also achieve goods recognition and precise three-dimensional positioning. The collaborative work of multiple sensors enhances the robustness and flexibility of the system.

[0051] S2. Generate a replenishment task for warehouse items based on the warehousing information, plan an initial path according to the positioning data of the AGV robot, and optimize the initial path using artificial potential field and nonlinear programming to obtain an optimized path;

[0052] Specifically, generating a replenishment task for warehouse items based on the warehousing information, planning an initial path according to the positioning data of the AGV robot, and optimizing the initial path using artificial potential field and nonlinear programming to obtain an optimized path means comparing the quantity of warehouse items in the warehousing information with the set item threshold. If the quantity of warehouse items is less than the set item threshold, a corresponding replenishment task for warehouse items is generated. The replenishment task for warehouse items includes the pick-up point location, replenishment point location, replenishment items, and replenishment quantity;

[0053] Construct a 3D model of the warehouse based on the warehousing information, and mark the obstacle information and the location of the AGV robot in the 3D model;

[0054] After receiving the replenishment task for warehouse items, the AGV robot uses the 3D A * algorithm to obtain the pick-up initial path of the AGV robot according to the positioning data and the pick-up point location;

[0055] Calculate the attractive force U a :

[0056]

[0057] where q is the positioning coordinate of the AGV robot, q g is the pick-up point location, and ∈ is the attractive force intensity factor;

[0058] Synchronously calculate the repulsive force U r :

[0059]

[0060] where q o is the obstacle location, κ is the repulsive force gain factor, and ρ is the repulsive force action range threshold;

[0061] Combine the attractive force U a and the repulsive force U r to form a combined force F. Based on the AGV robot, construct a coordinate system, and decompose the combined force F according to the coordinate system direction to obtain F x and F y , respectively calculate the ratio of the decomposed force F x and F y and the mass of the AGV robot to obtain the accelerations a x and a y, decompose the AGV robot speed v in the direction of the coordinate system synchronously to obtain v x and v y , update the initial path nodes according to the decomposed speed and acceleration:

[0062] x(t + 1) = x(t) + (v x (t) + a x (t) * Δt) * Δt;

[0063] y(t + 1) = y(t) + (v y (t) + a y (t) * Δt) * Δt;

[0064] where x(t) and y(t) are the coordinates of the initial path nodes, x(t + 1) and y(t + 1) are the coordinates of the updated path nodes, and Δt is the time step;

[0065] After updating the path node coordinates, construct the path optimization objective function J based on the nonlinear programming technique:

[0066]

[0067] where w(t) is the angular velocity of the AGV robot, obtained from the speed data, α1, α2, and α3 are weight coefficients, and t f is the movement time of the AGV robot;

[0068] Set physical constraint conditions for the path optimization objective function J through the kinematic model:

[0069] a(t) ≤ a max , v(t) ≤ v max , w(t) ≤ w max ;

[0070] where a max is the maximum acceleration of the AGV robot, v max is the maximum speed of the AGV robot, and w max is the maximum angular velocity of the AGV robot;

[0071] Use the interior point method to solve the objective function J to obtain the optimized AGV robot speed and acceleration, and further update the coordinates of the updated path nodes through the optimized speed and acceleration to form a picking optimization path;

[0072] The AGV robot reaches the picking point through the picking optimization path to obtain the replenishment items, then generates a replenishment optimization path according to the picking point position and the replenishment point position, and reaches the replenishment point position according to the replenishment optimization path.

[0073] Generate replenishment tasks based on warehousing information and AGV robot positioning data, and optimize based on preliminary path planning, enabling replenishment tasks to quickly respond to inventory changes, avoiding problems such as untimely inventory monitoring and lagging replenishment decisions in traditional methods. Conduct path planning based on AGV robot positioning data, which can ensure the precise execution of tasks, improve the adaptability and safety of path planning. Construct a three-dimensional model of the warehouse through warehousing information, and mark obstacle information and AGV robot positions in the model, greatly enhancing the accuracy of path planning and providing more comprehensive information for path optimization. By comprehensively considering the influence of picking points and obstacles, a flexible dynamic adjustment mechanism is provided in path planning. The design of the attraction strength factor and repulsion gain factor enables the AGV to precisely control its movement trajectory, not only avoiding the limitations of fixed path planning but also dynamically adjusting the path in real time according to environmental changes. The introduction of nonlinear programming makes path optimization more precise, and it can find the optimal path under various physical constraints. Path optimization in the prior art usually relies on simple heuristic algorithms and is difficult to handle complex constraint conditions. However, the present invention combines nonlinear programming with the interior point method, making the path optimization process smoother, more accurate, and effectively avoiding unstable factors in the path selection process. Add physical constraints (such as maximum acceleration, maximum speed, maximum angular velocity, etc.) to the path optimization objective function, effectively avoiding performance bottlenecks of AGV robots due to exceeding motion limits during path planning. Through the setting of constraint conditions, the stability and safety of the robot during actual operation can be ensured. Through the dynamic adjustment and optimization of the path, the AGV robot has stronger adaptability in a complex warehouse environment, can cope with real-time changing environmental conditions, and significantly improves the execution efficiency and accuracy of replenishment tasks.

[0074] Furthermore, after marking obstacle information and AGV robot positions in the three-dimensional model, view the positions of idle AGV robots in real time in the three-dimensional warehouse model. Whenever a replenishment task for warehouse items appears, calculate the distances between the replenishment task and the positions of all idle AGV robots according to the replenishment point position, and select the idle AGV robot closest to the replenishment point position to receive the replenishment task and mark the AGV robot as being in a working state.

[0075] By marking obstacle information and the position of AGV robots in a 3D model, the AGV robots can identify and avoid obstacles in real time in a complex warehouse environment. The 3D model can dynamically update and display the position of obstacles and their changes, enabling the AGV robots to adjust their paths in a changing environment in a timely manner, avoiding collisions and invalid paths, and significantly improving the flexibility and safety of the robots in actual operations. By browsing the positions of idle AGV robots in real time, the present invention can ensure the reasonable allocation and scheduling of tasks. When a replenishment task appears, the system can accurately understand which AGV robots are idle, avoiding repeated allocation or waiting situations. Dynamic and real-time task scheduling can maximize the efficiency of warehouse operations, reduce the waiting time of idle AGV robots, and improve the utilization rate of AGV robots, thereby optimizing the resource allocation of the entire warehouse management system. By calculating the distance between the replenishment point position and the positions of idle AGV robots, the present invention can accurately select the idle robot closest to the replenishment point to execute the task. This design not only improves the response speed of the replenishment task but also reduces the driving distance of the AGV robot during task execution, thereby shortening the replenishment time and path and saving the overall time of warehouse operations.

[0076] S3. The AGV robot reaches the replenishment position according to the optimized path and uses visual perception technology based on image data to locate the accurate position information of the replenishment item for replenishment operation;

[0077] Specifically, using visual perception technology based on image data to locate the accurate position information of the replenishment item for replenishment operation means that after the AGV robot reaches the replenishment point position, it acquires the replenishment point image I, performs image equalization operation on the replenishment point image I to obtain the average environmental illumination image as the flat-field image R. For each acquired image, perform image equalization operation (such as histogram equalization or gamma correction), which can extract the main structural information from the acquired image and remove the influence of uneven illumination in the image, and set a short exposure time to re-take the replenishment point image as the dark-field image D. Set the camera to an extremely short exposure time (such as 1 millisecond or shorter) to reduce the input of ambient light. Through this short exposure time, the captured image is mainly composed of the noise and dark current signal of the sensor, and the information in the image is almost completely determined by the inherent characteristics of the sensor, simulating the dark-field image. Correct the replenishment point image I according to the flat-field image and the dark-field image:

[0078]

[0079] where C is the corrected replenishment point image and m is the normalization scale factor;

[0080] Convert the corrected replenishment point image to a grayscale image, apply Gaussian blur for denoising, and calculate the Hessian matrix of each pixel point in the replenishment point image through the SURF algorithm:

[0081]

[0082] where H(i, j) is the value of the pixel point (i, j) in the Hessian matrix, and C * is the replenishment point image after denoising;

[0083] Perform eigenvalue decomposition based on the Hessian matrix to obtain eigenvalues λ1 and λ2:

[0084] λ1, λ2 = eigenvaluesH(i, j);

[0085] Compare the eigenvalues with the set threshold. If both eigenvalues are greater than the threshold, record the pixel point (i, j) as a key point;

[0086] Take the denoised replenishment point image as the target image and the replenishment point image I as the source image. Extract the key points of the target image and the source image respectively to form key point sets, and calculate the Euclidean distance between the key points in the two key point sets for minimum matching to obtain matching point pairs;

[0087] Randomly select four point pairs from the matching point pairs and estimate the homography matrix G by the least squares method;

[0088] Train the YOLOv3 algorithm and use the YOLOv3 algorithm to extract the position coordinates (i1, j1) of the goods from the source image and calculate the position coordinate depth Z1. Calculate the X1 and Y1 coordinates of the three-dimensional coordinates of the goods through the position coordinates of the goods:

[0089]

[0090] where c X and c Y are the coordinates of the camera principal point, and f X and f Y are the camera focal lengths;

[0091] Combine the coordinate depth Z1 with the X1 and Y1 coordinates of the three-dimensional coordinates of the goods to form the complete three-dimensional coordinates of the goods (X1, Y1, Z1), and the AGV robot performs replenishment operations according to the three-dimensional coordinates of the goods.

[0092] By performing image equalization on the replenishment point images, the present invention can effectively remove the influence of uneven illumination in the images, enhance the quality and details of the images, which not only improves the recognizability of the item features in the images, but also makes the subsequent image processing more accurate. Especially when performing key point extraction and object matching, clearer feature information can be obtained. By setting an extremely short exposure time to capture dark field images, the noise and dark current signals caused by the inherent characteristics of the camera sensor can be captured. This strategy reduces the influence of ambient light, thereby improving the accuracy of the images. By combining flat field images and dark field images for image correction, the AGV robot can obtain more accurate image data, which is crucial for the subsequent item detection and replenishment tasks. By using the Hessian matrix to process the replenishment point images, the present invention can effectively extract the key feature points in the images. These feature points play a decisive role in subsequent item matching and position localization. Especially when matching between different images or perspectives, the Hessian matrix can provide accurate local geometric information to help the AGV robot accurately locate the positions of goods in the warehouse. By using YOLOv3 to extract the position coordinates of the goods and calculate their depth information, and combining with the internal parameter data of the camera, the three-dimensional coordinates of the items can be further deduced to ensure the accuracy of the replenishment operation. By organically combining multiple steps such as image equalization, correction, feature extraction, and YOLOv3 object detection, an efficient and accurate warehouse item positioning and replenishment system is formed. Each step not only makes up for the deficiencies of the previous step, but also improves the accuracy and efficiency of the overall system through multi-level information processing. Especially in complex environments, the collaborative effect of these steps enables the AGV robot to complete the replenishment task with high accuracy and efficiency, thus exerting greater automation potential in warehouse management.

[0093] S4. Real-time display the warehousing information and generate a replenishment record according to the replenishment operation and store it in the database;

[0094] Specifically, real-time display of the warehousing information means real-time display of the warehousing information collected by the sensors through the visualization interface, and synchronously display the three-dimensional model of the warehouse in the visualization interface.

[0095] Further, generating a replenishment record according to the replenishment operation and storing it in the database means generating a replenishment record for each replenishment operation and storing the replenishment record and the warehousing information in the database. The database regularly detects the integrity and security of the stored data, and synchronously uploads the stored data to the cloud for backup.

[0096] This embodiment also provides a warehouse management system based on an AGV robot, including:

[0097] An information acquisition module, which is used to deploy sensors to collect warehousing information, and collect positioning data, speed data, and image data through an AGV robot;

[0098] A warehouse management module, which is used to generate replenishment tasks according to the warehousing information, allocate AGV robots to plan replenishment paths, and identify the precise positions of replenishment items based on the image data collected by the AGV robots for replenishment operations;

[0099] A display and storage module, which is used to display the warehousing information in real time, and generate replenishment records according to the replenishment operations and store them in a database.

[0100] This embodiment also provides a computer device applicable to the situation of the warehouse management method based on an AGV robot, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the warehouse management method based on an AGV robot proposed in the above embodiment.

[0101] This computer device can be a terminal. This computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of this computer device is used to provide computing and control capabilities. The memory of this computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of this computer device is used to communicate with an external terminal in a wired or wireless manner. The wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of this computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of this computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad set on the outer shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0102] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the warehouse management method based on an AGV robot as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, abbreviated as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, abbreviated as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, abbreviated as EPROM), programmable read-only memory (Programmable Red-Only Memory, abbreviated as PROM), read-only memory (Read-Only Memory, abbreviated as ROM), magnetic memory, flash memory, magnetic disk or optical disc.

[0103] In summary, the present invention generates replenishment tasks by collecting warehousing information and AGV robot information, plans and optimizes the replenishment path based on the AGV robot, has higher flexibility and security, and accurately identifies the three-dimensional coordinates of goods for replenishment through image recognition and coordinate point matching, improving the adaptability of the AGV robot to complex warehouse environments and realizing efficient warehouse inventory management.

[0104] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A warehouse management method based on AGV robots, characterized in that: include, Deploy sensors in the warehouse to collect storage information, and use AGV robots to collect positioning data, speed data, and image data; Generate warehouse item replenishment tasks based on storage information, plan the initial path according to the AGV robot positioning data, and use artificial potential field and nonlinear programming to optimize the initial path to obtain the optimized path; The AGV robot reaches the replenishment location according to the optimized path, and uses visual perception technology based on image data to locate the precise location information of the replenishment items for replenishment operations; The warehouse information is displayed in real time and replenishment records are generated based on replenishment operations and stored in the database.

2. The warehouse management method based on the AGV robot according to claim 1, characterized in that: The generating of the warehouse item replenishment task based on the storage information, planning the initial path according to the positioning data of the AGV robot, and optimizing the initial path by using the artificial potential field and nonlinear programming to obtain the optimized path refers to comparing the number of warehouse items in the storage information with the set item threshold, and if the number of warehouse items is less than the set item threshold, generating the corresponding warehouse item replenishment task, wherein the warehouse item replenishment task includes the pickup point location, the replenishment point location, the replenishment items, and the replenishment quantity; Build a 3D warehouse model based on storage information, and mark obstacle information and AGV robot positions in the 3D model; After receiving the warehouse item replenishment task, the AGV robot uses 3D A * The algorithm obtains the initial path for the AGV robot to pick up goods; Calculate the attraction U based on the AGV robot positioning data and the pickup point location a , and calculate the repulsive force U according to the obstacle information r ; The attraction U a and the repulsive force U r The force is synthesized to form a comprehensive force F. A coordinate system is constructed based on the AGV robot, and the comprehensive force F is decomposed according to the direction of the coordinate system to obtain F x and F y , calculate the decomposition force F x and F y The ratio of the mass of the AGV robot to obtain the acceleration a in the coordinate system direction x and a y , synchronously decompose the AGV robot speed v according to the coordinate system direction to obtain v x and v y , the initial path nodes are updated according to the decomposed speed and acceleration; After updating the path node coordinates, the path optimization objective function J is constructed based on nonlinear programming technology: Where w(t) is the angular velocity of the AGV robot, α1, α2 and α3 are weight coefficients, t f is the moving time of the AGV robot; Set physical constraints for the path optimization objective function J through the kinematic model; The interior point method is used to solve the objective function J to obtain the optimized AGV robot speed and acceleration. The updated path node coordinates are further updated by the optimized speed and acceleration to form an optimized pickup path. The AGV robot reaches the pickup point through the optimized pickup path to obtain the replenishment items. After that, it generates a replenishment optimized path based on the pickup point location and the replenishment point location and reaches the replenishment point location according to the replenishment optimized path.

3. The warehouse management method based on the AGV robot according to claim 2, characterized in that: The replenishment operation based on the image data using visual perception technology to locate the precise position information of the replenishment item refers to obtaining a replenishment point image I after the AGV robot arrives at the replenishment point, performing an image equalization operation on the replenishment point image I to obtain an average image of ambient light as a flat field image R, setting a short exposure time to re-shoot the replenishment point image as a dark field image D, and correcting the replenishment point image I according to the flat field image and the dark field image; The corrected replenishment point image is converted into a grayscale image and denoised by applying Gaussian blur. The Hessian matrix of each pixel in the replenishment point image is calculated by the SURF algorithm. Perform eigenvalue decomposition based on the Hessian matrix to obtain eigenvalues ​​λ1 and λ2; The denoised replenishment point image is used as the target image, and the replenishment point image I is used as the source image. The key points of the target image and the source image are extracted to form a key point set respectively. The Euclidean distance of the key points in the two key point sets is calculated to perform minimum matching to obtain matching point pairs. Randomly select four point pairs from the matching point pairs and estimate the homography matrix G by the least squares method; Train the YOLOv3 algorithm and use the YOLOv3 algorithm to extract the location coordinates (i1, j1) of the goods from the source image and calculate the location coordinate depth Z1, and calculate the X1 and Y1 coordinates of the three-dimensional coordinates of the goods through the location coordinates of the goods; The coordinate depth Z1 is combined with the X1 and Y1 coordinates of the cargo's three-dimensional coordinates to form a complete cargo three-dimensional coordinate (X1, Y1, Z1), and the AGV robot performs replenishment operations based on the cargo's three-dimensional coordinates.

4. The warehouse management method based on the AGV robot according to claim 3, characterized in that: After marking the obstacle information and the AGV robot position in the three-dimensional model, the idle AGV robot positions are browsed in real time in the warehouse three-dimensional model. Whenever a warehouse item replenishment task occurs, the distance between the replenishment task and the positions of all idle AGV robots is calculated according to the replenishment point position, and the idle AGV robot closest to the replenishment point position is selected to receive the replenishment task and the AGV robot is marked as working.

5. The warehouse management method based on the AGV robot according to claim 4, characterized in that: The deployment of sensors in the warehouse to collect storage information and the collection of positioning data, speed data and image data through the AGV robot refers to the deployment of lidar and RFID sensors in the warehouse to collect warehouse environment data and warehouse cargo data, and the installation of IMU sensors, GPS sensors and cameras on the AGV robot.

6. The warehouse management method based on the AGV robot according to claim 5, characterized in that: The real-time display of the warehouse information refers to displaying the warehouse information collected by sensors in real time through a visual interface, and synchronously displaying the three-dimensional model of the warehouse in the visual interface.

7. The warehouse management method based on the AGV robot according to claim 6, characterized in that: Generating replenishment records according to replenishment operations and storing them in a database means generating a replenishment record for each replenishment operation and storing the replenishment record and warehousing information in the database, the database regularly performs integrity and security checks on the stored data, and synchronously uploads the stored data to the cloud for backup.

8. A warehouse management system based on an AGV robot, based on the warehouse management method based on an AGV robot according to any one of claims 1 to 7, characterized in that: include, The information acquisition module is used to deploy sensors to collect warehouse information and collect positioning data, speed data, and image data through AGV robots; The warehouse management module is used to generate replenishment tasks based on storage information, assign AGV robots to plan replenishment routes, and perform replenishment operations based on the image data collected by the AGV robots to identify the precise location of the replenishment items; The display and storage module is used to display warehouse information in real time and generate replenishment records based on replenishment operations and store them in the database.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the warehouse management method based on the AGV robot according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the warehouse management method based on the AGV robot according to any one of claims 1 to 7 are implemented.

Images