Automatic logistics warehousing system and management method

Abstract

The invention discloses an automatic logistics warehousing system and a management method, and relates to the technical field of logistics warehousing management. The system comprises a path module, wherein each optimal path at least passes through a storage rack where to-be-checked goods are stored; equally dividing the same batch of warehoused goods into two groups according to the checking task, wherein the sum of the storage shelves of the to-be-checked goods on all the optimal paths of each group of warehoused goods is consistent with the number of the to-be-checked goods; wherein the first group is an inventory group, and the second group is a control group; the checking task comprises goods quantity checking and goods state checking, and after all visual images shot by the checking group are spliced by the image analysis module, the checking module performs primary checking on the to-be-checked goods according to the visual images of the checking group. According to the method, the optimal warehousing path is obtained by calculating the transportation path of the batch of warehoused goods each time, and the time consumed during warehousing is shortened.

Description

Technical Field

[0001] The present invention belongs to the technical field of logistics warehousing management, and in particular relates to an automated logistics warehousing system and a management method. Background Art

[0002] In recent years, with the widespread use of stereoscopic warehouses and the continuous improvement of the level of automated equipment management, the operating efficiency of warehouses can be greatly improved. Compared with traditional warehouses, the advantage of automated stereoscopic warehouses is that they avoid manual on-site operations and are directly dominated by computer control technology, realizing the mechanization of goods storage and handling and the visualization of goods data. Warehouse inventory, as a basic requirement of automated stereoscopic warehouses, also needs to continuously improve its level.

[0003] However, the current automated logistics warehouse inventory is difficult and is mostly done manually, with the help of stackers to take goods out to the entry and exit counter for manual inspection, or handheld radio frequency identification terminals are used to scan the shelf labels where the goods are stored to record these materials, and finally output a chart showing the storage status of the materials. This operation method is somewhat feasible for traditional small warehouses. For automated logistics warehouses, if they want to achieve the goal of uninterrupted unmanned management around the clock, manual inventory will greatly reduce work efficiency, and the accuracy of the inventory of goods cannot be guaranteed.

[0004] The existing automated logistics warehousing does not achieve accurate and fast inventory of goods inside the warehouse, and requires a lot of manpower and material resources to complete the inventory of warehouse goods, resulting in low efficiency of warehousing management. For this reason, we propose an automated logistics warehousing system and management method. Summary of the invention

[0005] The purpose of the present invention is to provide an automated logistics warehousing system and management method to solve the problem that the existing automated logistics warehousing cannot accurately and quickly count the goods inside the warehouse, and a large amount of manpower and material resources are required to complete the inventory of warehouse goods, resulting in low warehousing management efficiency.

[0006] In order to solve the above technical problems, the present invention is achieved through the following technical solutions: The present invention is an automated logistics warehousing system, comprising a processing module, a space construction module, a path module, an operation module, an image analysis module and an inventory module; Processing module: used to receive data information from each module and perform relevant processing, and also used to issue command control to each module; Space construction module: used to construct a three-dimensional coordinate system according to the shelf arrangement of the stereoscopic warehouse, and calibrate the pre-storage coordinates of the pre-storage points ; Path module: used to calculate the optimal path for the transport device to travel to the pre-storage coordinates according to the pre-storage coordinates of the pre-stored goods; Operation module: used to perform visual image scanning on each cargo position on the path through a visual scanning device arranged on the stacker, wherein the scanning radiation range of the visual image scanning is a circular image with a diameter of R; Image analysis module: used to obtain the visual images taken by the visual scanning device, process and analyze them, detect the status information of goods on each shelf through the deep learning network model, detect and locate the barcodes attached to the surface of each shelf, and decode the barcodes through the preset barcode image processing library; Inventory module: used to compare the goods status information on each shelf with the decoded barcode information to realize warehouse goods inventory.

[0007] Preferably, it also includes a storage module: used to receive and store the inventory results, including the quantity information and the status information of the goods; and the storage module is read by the system administrator; The space construction module pre-stores the coordinates Indicates the location information of the cargo space allocated in the shelf, where x is the number of the shelf group where the logistics storage parts are located, y is the number of columns corresponding to the group of shelves, and z is the number of layers corresponding to the group of shelves. The final result of the cargo space allocation is determined by the coordinates.

[0008] Preferably, when calculating the optimal path for each pre-storage coordinate in the path module, different optimal paths are planned and calculated according to the number of goods entering the warehouse in the same batch. The location information of the cargo space allocated in the shelf when each cargo enters the warehouse is imported into the three-dimensional coordinate system and converted into pre-storage coordinates, and intuitively displayed in the three-dimensional coordinate system; the end point of each of the optimal paths is the pre-storage coordinate position of a cargo entering the warehouse.

[0009] Preferably, the transport device carrying the goods has a coordinate of The pre-stored goods are transported from the starting point of the stereoscopic warehouse to the target shelf group position by the conveyor belt, and then the pre-stored goods are transferred to the surface of the stacker by the in-and-out platform near the conveyor belt position of each shelf group. Finally, the stacker stores the pre-stored goods at the cargo position corresponding to the pre-storage coordinates of the shelf group.

[0010] Preferably, the path module calculates the optimal path for each incoming goods to travel to the pre-storage coordinates, including satisfying the following conditions: each of the optimal paths passes through at least one shelf where the goods to be counted are stored, and the shelf where the goods to be counted are stored is completely exposed to the scanning radiation range of the visual scanning device; the same batch of incoming goods is divided into two groups according to the inventory task, and the sum of the shelves where the goods to be counted are stored through all the optimal paths of each group of the incoming goods is consistent with the number of goods to be counted; wherein the first group is the inventory group and the second group is the control group; the inventory task includes the inventory of the number of goods and the inventory of the status of goods, all the visual images taken by the inventory group are spliced ​​and processed by the image analysis module, and the inventory module performs the initial inventory of the goods to be counted according to the visual images of the inventory group, and re-outputs the goods profit and loss information and the abnormal goods status information obtained after the initial inventory to the path module, and plans the optimal path of all pre-stored goods in the control group through the path module, wherein the intersection of the optimal paths for each goods to enter the warehouse is the cargo location coordinates of the goods in the abnormal state after the initial inventory.

[0011] Preferably, the path module determines the coordinates of the goods with abnormal status based on the goods profit and loss information and the abnormal goods status information obtained by the inventory group during the initial inventory, and calculates the optimal path for the pre-stored goods in the control group to travel to the pre-storage coordinates, wherein the number of intersections of the optimal path for all goods in the control group to enter the warehouse is consistent with the number of abnormal goods.

[0012] A management method for an automated logistics warehousing system, which constructs a three-dimensional coordinate system in advance according to the arrangement of shelves in a three-dimensional warehouse, comprises the following steps: S1. The warehouse system administrator issues an inventory command to the warehouse system, waiting for a new batch of goods to enter the warehouse and counting the goods stored on the shelves, dividing the batch of goods into two groups according to the quantity and entering the warehouse separately; S2, the path module plans the storage locations of all goods in the inventory group on the shelves, and calculates the optimal path for each to-be-stored goods to be transported to the pre-storage coordinates; S3, transporting the goods to be stored to the command coordinate position according to the optimal path by the stacker, and at the same time performing visual image scanning on each cargo position on the path by a visual scanning device arranged on the stacker; wherein the cargo positions scanned on the path include the cargo positions to be counted; S4, the image analysis module obtains the visual images taken by the visual scanning device during the transportation of all goods, and performs splicing processing and analysis on them, and detects the quantity of goods on each shelf and the status of goods through the deep learning network model; at the same time, the barcode attached to the surface of each shelf to be counted is detected and located, and the barcode is decoded through barcode image processing; S5. Match and compare the decoded barcode content with the cargo status information on each shelf in the visual image to perform a preliminary warehouse cargo inventory. Re-output the cargo profit and loss information and cargo status abnormality information obtained after the initial inventory to the path module; S6. The path module determines the coordinates of the goods with abnormal status based on the goods profit and loss information and the abnormal status information of the goods obtained by the initial inventory count, and calculates the optimal path for the pre-stored goods in the control group to travel to the pre-stored coordinates, wherein the intersection of the optimal paths for the goods in the control group with intersections is the cargo location coordinates of the goods with abnormal status after the initial inventory count; S7, repeating steps S3-S5 to complete the acquisition of visual images and image processing and analysis of the control group, wherein the control group has two visual images of abnormal goods from different perspectives, and the visual images obtained from the initial inventory are compared with the visual images obtained from the secondary inventory. The comparison content is divided into two parts: one is to compare whether the status of goods without abnormal status is consistent to ensure the accuracy of the detection results of the initial inventory, and the other is to focus on determining the specific status of the abnormal goods based on the multi-perspective visual images, including the two states of poor quality and damaged packaging; S8. The final inventory data of the two inventory counts are sorted out and output to the storage module for the system administrator to review.

[0013] Preferably, the visual image stitching analysis process comprises the following steps: a. Preprocess the received visual image, including grayscale conversion and noise reduction; b. Identify the pre-processed visual image to determine whether there are goods on the cargo space. If there are goods, continue with subsequent barcode detection and positioning. If there are no goods, end the process; c. Roughly locate the barcode, and make adjustments based on the rough positioning to obtain its accurate position; d. Extract and crop the barcode position, input it into the decoding algorithm for decoding, and obtain the cargo information reserved by the barcode in the warehousing system.

[0014] Preferably, in the visual image splicing analysis process, a probabilistic SVM is used to perform a binary classification operation on the image for the presence or absence of goods for cargo detection, and its formula is shown in formula (1): (1) Among them, A and B are fitting parameters, and f is the output of the sample standard SVM; The image analysis module collects cargo images and performs binary classification operations to classify the images into two categories: cargo and no cargo, thus constructing a cargo binary classification dataset. The data set is divided into a training set and a test set. Some images are selected to construct a basic training set. At the same time, the cross-platform computer vision library is used to expand the basic training set images, including a series of operations such as image denoising, image rotation, and motion blur, to generate a cargo classification training set, expand the data scale and data richness of the training set, use the training set to train the SVM classification model, and test it on the test set; After obtaining the classification result through the SVM classification model, the classification result is sent to the image analysis module. When the classification result is "1" and the goods exist, the image analysis module starts the subsequent barcode detection and positioning. When the classification result is "0" and the goods do not exist, the image analysis module is in standby mode.

[0015] The present invention has the following beneficial effects: The present invention synchronizes the inventory operation with the warehousing operation by setting a path module in the warehousing system. When it is necessary to inventory the goods on the shelves, wait for a batch of goods to enter the warehouse, and set a visual scanning device on the stacker that transports the goods. When the stacker carries the goods to the designated cargo position of the designated shelf, the visual scanning device on the surface of the stacker during transportation performs a visual scan on the cargo position on the shelf during transportation to obtain a visual scanning image, and all visual images taken when a batch is entered into the warehouse are spliced ​​and sorted to obtain a complete visual image of a whole group of cargo positions on the shelves to be inventoried. In this way, the goods can be easily inventoried, and there is no need for staff to manually go to the shelf position with a handheld identification device to inventory the goods. The warehousing operation is completed, and the goods are also inventoried once. In addition, an inventory can be performed each time a batch of goods needs to be entered into the warehouse.

[0016] The present invention calculates each transportation path of a batch of incoming goods in the path module, thereby obtaining the optimal path for warehousing, which not only reduces the time spent on warehousing, but also each time the stacker passes through the cargo location to be counted at least once during the driving process, and the cargo location is completely exposed to the scanning radiation range of the visual scanning device. In this way, after all the goods are put into the warehouse, all the cargo locations to be counted can be photographed to complete the collection of visual images, which greatly reduces the time required for inventory. At the same time, the cargo location to be counted can be clearly photographed, so that the inventory effect is better.

[0017] The present invention divides the same batch of goods to be stored into two groups, an inventory group and a control group. First, the goods in the inventory group are stored. The inventory group performs an initial inventory of the goods to be stored, thereby obtaining goods profit and loss information and goods status abnormality information, and transmitting this information to a path module. The path module recalculates the path of the control group according to the coordinate points of the goods in the abnormal state to obtain the optimal path of the control group, and the intersection of the optimal paths for storing the goods in the control group is the cargo position coordinate point of the goods in the abnormal state after the initial inventory, which means that after the control group completes the storage and visual image acquisition, the goods in the abnormal state of the initial inventory can be photographed twice at different angles, and the two photographed pictures are compared with the initial inventory picture at the same time to determine the specific state of the goods in the abnormal state, thereby completing a more detailed inventory of the goods; at the same time, it is compared whether the state of the goods without abnormal state is consistent with the initial inventory result, so as to ensure the accuracy of the detection result of the initial inventory. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0019] Figure 1 A system block diagram of the automated logistics warehousing system provided by the present invention; Figure 2 A flow chart of the management method of the automated logistics warehousing system provided by the present invention. DETAILED DESCRIPTION

[0020] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0021] In the description of the present invention, it should be noted that the terms "upper", "lower", "inner", "outer", "front end", "rear end", "two ends", "one end", "the other end" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0022] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "provided with", "connected", etc. should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0023] See also Figure 1 , an automated logistics warehousing system, comprising a processing module, a space construction module, a path module, an operation module, an image analysis module and an inventory module; Processing module: used to receive data information from each module and perform relevant processing, and also used to issue command control to each module; Space construction module: used to construct a three-dimensional coordinate system according to the shelf arrangement of the stereoscopic warehouse, and calibrate the pre-storage coordinates of the pre-storage points ; Path module: used to calculate the optimal path for the transport device to travel to the pre-storage coordinates according to the pre-storage coordinates of the pre-stored goods; Operation module: used to perform visual image scanning on each cargo position on the path through a visual scanning device arranged on the stacker, wherein the scanning radiation range of the visual image scanning is a circular image with a diameter of R; When calculating the optimal path, a greedy algorithm is used, and a calculation model is first established before the calculation. The calculation model satisfies the following two constraints. The storage constraint is shown in formula (2): (2) Among them, k is the total number of goods entering the warehouse, t is the specified time for goods to enter the warehouse, z is the weight of the goods, and X is the order of entering the warehouse; The priority constraint condition is as shown in formula (3): (3) in, represents the i-th incoming goods, represents the delivery time of the Bth cargo, and P is the priority constraint; Before calculating the optimal path in the path module, a storage location optimization data model is established based on the shelf's bearing capacity and stability and the impact of storage location allocation on the efficiency of warehousing operations; Based on the assumption that each group of shelves in the automated warehouse has m rows and n layers based on shelf stability, the weight of each shelf position and the sum of its layer are minimized by using formula (4): (4) in, represents the cargo space in the i-th column and j-th layer, and L represents the cargo space load parameter; Based on the assumption that the automatic storage and warehousing platforms use the same conveyor belt, Image analysis module: used to obtain the visual images taken by the visual scanning device, process and analyze them, detect the status information of goods on each shelf through the deep learning network model, detect and locate the barcodes attached to the surface of each shelf, and decode the barcodes through the preset barcode image processing library; Inventory module: used to compare the goods status information on each shelf with the decoded barcode information to realize warehouse goods inventory.

[0024] The storage module is also included: for receiving and storing the inventory results, including the quantity information and the status information of the goods; and for the system administrator to review; The spatial construction module stores coordinates in advance Indicates the location information of the cargo space allocated in the shelf, where x is the number of the shelf group where the logistics storage parts are located, y is the number of columns corresponding to the group of shelves, and z is the number of layers corresponding to the group of shelves. The final result of the cargo space allocation is determined by the coordinates.

[0025] Among them, when calculating the optimal path of each pre-storage coordinate in the path module, different optimal paths are planned and calculated according to the number of goods entering the warehouse in the same batch. The location information of the cargo space allocated in the shelf when each cargo enters the warehouse is imported into the three-dimensional coordinate system and converted into pre-storage coordinates, and intuitively displayed in the three-dimensional coordinate system; the end point of each optimal path is the pre-storage coordinate position of a cargo entering the warehouse.

[0026] Among them, the coordinates of the transport device carrying the goods from the starting point, that is, in the three-dimensional coordinate system, are The pre-stored goods are transported from the starting point of the stereoscopic warehouse to the target shelf group position by the conveyor belt, and then the pre-stored goods are transferred to the surface of the stacker by the in-and-out platform near the conveyor belt position of each shelf group. Finally, the stacker stores the pre-stored goods at the cargo position corresponding to the pre-storage coordinates of the shelf group.

[0027] Among them, the path module calculates the optimal path for each incoming goods to travel to the pre-storage coordinates, including satisfying the following conditions: each optimal path passes through at least one shelf where the goods to be counted are stored, and the shelf where the goods to be counted are stored is completely exposed to the scanning radiation range of the visual scanning device; the same batch of incoming goods are divided into two groups according to the inventory task, and the sum of the shelves where the goods to be counted are stored through all the optimal paths of each group of incoming goods is consistent with the number of goods to be counted; the first group is the inventory group, and the second group is the control group; the inventory task includes the inventory of the number of goods and the inventory of the status of goods. After all the visual images taken by the inventory group are spliced ​​and processed by the image analysis module, the inventory module performs the initial inventory of the goods to be counted based on the visual images of the inventory group, and re-outputs the goods profit and loss information and the abnormal goods status information obtained after the initial inventory to the path module, and plans the optimal path of all pre-stored goods in the control group through the path module, wherein the intersection of the optimal paths for each goods to enter the warehouse is the cargo location coordinates of the goods in the abnormal state after the initial inventory.

[0028] Among them, the path module determines the coordinates of the goods with abnormal status based on the goods profit and loss information and abnormal goods status information obtained by the inventory group during the initial inventory, and calculates the optimal path for the pre-stored goods in the control group to travel to the pre-storage coordinates. Among them, the number of intersections of the optimal path for all goods in the control group to enter the warehouse is consistent with the number of abnormal goods.

[0029] See also Figure 2 A management method for an automated logistics warehousing system is provided, wherein a three-dimensional coordinate system is constructed in advance according to the arrangement of shelves in a three-dimensional warehouse, and the three-dimensional coordinate system comprises the following steps: S1. The warehouse system administrator issues an inventory command to the warehouse system, waiting for a new batch of goods to enter the warehouse and counting the goods stored on the shelves, dividing the batch of goods into two groups according to the quantity and entering the warehouse separately; S2, the path module plans the storage locations of all goods in the inventory group on the shelves, and calculates the optimal path for each to-be-stored goods to be transported to the pre-storage coordinates; S3, transporting the goods to be stored to the command coordinate position according to the optimal path by the stacker, and at the same time performing visual image scanning on each cargo position on the path by a visual scanning device arranged on the stacker; wherein the cargo positions scanned on the path include the cargo positions to be counted; S4, the image analysis module obtains the visual images taken by the visual scanning device during the transportation of all goods, and performs splicing processing and analysis on them, and detects the quantity of goods on each shelf and the status of goods through the deep learning network model; at the same time, the barcode attached to the surface of each shelf to be counted is detected and located, and the barcode is decoded through barcode image processing; S5. Match and compare the decoded barcode content with the cargo status information on each shelf in the visual image to perform a preliminary warehouse cargo inventory. Re-output the cargo profit and loss information and cargo status abnormality information obtained after the initial inventory to the path module; S6. The path module determines the coordinates of the goods with abnormal status based on the goods profit and loss information and the abnormal status information of the goods obtained by the initial inventory count, and calculates the optimal path for the pre-stored goods in the control group to travel to the pre-stored coordinates, wherein the intersection of the optimal paths for the goods in the control group with intersections is the cargo location coordinates of the goods with abnormal status after the initial inventory count; S7, repeating steps S3-S5 to complete the acquisition of visual images and image processing and analysis of the control group, wherein the control group has two visual images of abnormal goods from different perspectives, and the visual images obtained from the initial inventory are compared with the visual images obtained from the secondary inventory. The comparison content is divided into two parts: one is to compare whether the status of goods without abnormal status is consistent to ensure the accuracy of the detection results of the initial inventory, and the other is to focus on determining the specific status of the abnormal goods based on the multi-perspective visual images, including the two states of poor quality and damaged packaging; S8. The final inventory data of the two inventory counts are sorted out and output to the storage module for the system administrator to review.

[0030] Among them, the visual image splicing analysis processing includes the following steps: a. Preprocess the received visual image, including grayscale conversion and noise reduction; b. Identify the pre-processed visual image to determine whether there are goods on the cargo space. If there are goods, continue with subsequent barcode detection and positioning. If there are no goods, end the process; c. Roughly locate the barcode, and make adjustments based on the rough positioning to obtain its accurate position; d. Extract and crop the barcode position, input it into the decoding algorithm for decoding, and obtain the cargo information reserved by the barcode in the warehousing system.

[0031] Among them, in the visual image splicing analysis process, for cargo detection, the probabilistic SVM is used to perform a binary classification operation on the image to determine whether there is cargo or not. The formula is shown in formula (1): (1) Among them, A and B are fitting parameters, and f is the output of the sample standard SVM; The image analysis module collects cargo images and performs binary classification operations to classify the images into two categories: cargo and no cargo, thus constructing a cargo binary classification dataset. The data set is divided into a training set and a test set. Some images are selected to construct a basic training set. At the same time, the cross-platform computer vision library is used to expand the basic training set images, including a series of operations such as image denoising, image rotation, and motion blur, to generate a cargo classification training set, expand the data scale and data richness of the training set, use the training set to train the SVM classification model, and test it on the test set; After obtaining the classification result through the SVM classification model, the classification result is sent to the image analysis module. When the classification result is "1" and the goods exist, the image analysis module starts the subsequent barcode detection and positioning. When the classification result is "0" and the goods do not exist, the image analysis module is in standby mode.

[0032] Among them, in the process of detecting and locating the barcode, the image is processed by the image processing algorithm, the barcode is accurately located after the visual image processing operation, and it is cropped and passed into the decoding algorithm to improve the decoding efficiency and accuracy; the steps of barcode detection and positioning are as follows: First, resize the image according to the barcode specification information saved in the system to adjust it to a suitable size, then convert the image to a grayscale image for subsequent operations, and perform a Gaussian smoothing filter, adjust the corresponding parameters and thresholds to achieve good results. Next, calculate the gradient difference of the grayscale image in the horizontal and vertical directions, subtract the gradient value in the x direction from the gradient value in the y direction, and take the absolute value of the result to obtain the initial graphic position of the barcode; then perform mean filtering on the image to eliminate high-frequency noise; then perform binarization to basically obtain the area where the barcode is located; then remove the gaps between the barcodes through closing operations; perform corrosion operations on the image to remove some isolated points; then perform multiple expansion operations to connect the barcode area blocks to make the barcode area a whole; finally, detect the largest connected area in the image, and find the minimum base circumscribed rectangle, which is the location of the barcode. At this point, the detection and positioning of the barcode is completed.

[0033] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. An automated logistics warehousing system, characterized in that: It includes processing module, space construction module, path module, operation module, image analysis module and inventory module; Processing module: used to receive data information from each module and perform relevant processing, and also used to issue command control to each module; Space construction module: used to construct a three-dimensional coordinate system according to the shelf arrangement of the stereoscopic warehouse, and calibrate the pre-storage coordinates of the pre-storage points ; Path module: used to calculate the optimal path for the transport device to travel to the pre-storage coordinates according to the pre-storage coordinates of the pre-stored goods; The path module calculates the optimal path for each incoming goods to travel to the pre-storage coordinates, including satisfying the following conditions: each of the optimal paths passes through at least one shelf where the goods to be counted are stored, and the shelf where the goods to be counted are stored is completely exposed to the scanning radiation range of the visual scanning device; the same batch of incoming goods are divided into two groups according to the inventory task, and the sum of the shelves where the goods to be counted are stored in all the optimal paths of each group of the incoming goods is consistent with the number of goods to be counted; the first group is the inventory group, and the second group is the control group; the inventory task includes the inventory of the number of goods and the inventory of the status of goods. After all the visual images taken by the inventory group are spliced ​​and processed by the image analysis module, the inventory module performs the initial inventory of the goods to be counted according to the visual images of the inventory group, and re-outputs the goods profit and loss information and the abnormal goods status information obtained after the initial inventory to the path module, and plans the optimal path of all pre-stored goods in the control group through the path module, wherein the intersection of the optimal paths for each goods to enter the warehouse is the cargo location coordinates of the goods in the abnormal state after the initial inventory; The path module determines the coordinates of the goods with abnormal status based on the goods profit and loss information and the abnormal status information of the goods obtained by the initial inventory check by the inventory check group, and calculates the optimal path for the pre-stored goods of the control group to travel to the pre-stored coordinates, wherein the number of intersections of the optimal path for all goods entering the control group is consistent with the number of abnormal goods; Operation module: used to perform visual image scanning on each cargo position on the path through a visual scanning device arranged on the stacker, wherein the scanning radiation range of the visual image scanning is a circular image with a diameter of R; Image analysis module: used to obtain the visual images taken by the visual scanning device, process and analyze them, detect the status information of goods on each shelf through the deep learning network model, detect and locate the barcodes attached to the surface of each shelf, and decode the barcodes through the preset barcode image processing library; Inventory module: used to compare the goods status information on each shelf with the decoded barcode information to realize warehouse goods inventory.

2. The automated logistics warehousing system according to claim 1, characterized in that: It also includes a storage module: used to receive and store inventory results, including goods quantity information and goods status information; and the system administrator can access it; The space construction module pre-stores the coordinates Indicates the location information of the cargo space allocated in the shelf, where x is the number of the shelf group where the logistics storage parts are located, y is the number of columns corresponding to the group of shelves, and z is the number of layers corresponding to the group of shelves. The final result of the cargo space allocation is determined by the coordinates.

3. The automated logistics warehousing system according to claim 2, characterized in that: When calculating the optimal path for each pre-storage coordinate in the path module, different optimal paths are planned and calculated according to the number of goods entering the warehouse in the same batch. The location information of the cargo space allocated in the shelf when each cargo enters the warehouse is imported into the three-dimensional coordinate system and converted into pre-storage coordinates, and intuitively displayed in the three-dimensional coordinate system; the end point of each optimal path is the pre-storage coordinate position of a cargo entering the warehouse.

4. The automated logistics warehousing system according to claim 3, characterized in that: The transport device carrying the cargo has a coordinate of The pre-stored goods are transported from the starting point of the stereoscopic warehouse to the target shelf group position by the conveyor belt, and then the pre-stored goods are transferred to the surface of the stacker by the in-and-out platform near the conveyor belt position of each shelf group. Finally, the stacker stores the pre-stored goods at the cargo position corresponding to the pre-storage coordinates of the shelf group.

5. A management method for an automated logistics warehousing system, applied to the automated logistics warehousing system of claim 4, characterized in that: Constructing a three-dimensional coordinate system in advance according to the arrangement of shelves in the stereoscopic warehouse includes the following steps: S1. The warehouse system administrator issues an inventory command to the warehouse system, waiting for a new batch of goods to enter the warehouse and counting the goods stored on the shelves, dividing the batch of goods into two groups according to the quantity and entering the warehouse separately; S2, the path module plans the storage locations of all goods in the inventory group on the shelves, and calculates the optimal path for each to-be-stored goods to be transported to the pre-storage coordinates; S3, transporting the goods to be stored to the command coordinate position according to the optimal path by the stacker, and at the same time performing visual image scanning on each cargo position on the path by a visual scanning device arranged on the stacker; wherein the cargo positions scanned on the path include the cargo positions to be counted; S4, the image analysis module obtains the visual images taken by the visual scanning device during the transportation of all goods, and performs splicing processing and analysis on them, and detects the quantity of goods on each shelf and the status of goods through the deep learning network model; at the same time, the barcode attached to the surface of each shelf to be counted is detected and located, and the barcode is decoded through barcode image processing; S5. Match and compare the decoded barcode content with the cargo status information on each shelf in the visual image to perform a preliminary warehouse cargo inventory. Re-output the cargo profit and loss information and cargo status abnormality information obtained after the initial inventory to the path module; S6. The path module determines the coordinates of the goods with abnormal status based on the goods profit and loss information and the abnormal status information of the goods obtained by the initial inventory count, and calculates the optimal path for the pre-stored goods in the control group to travel to the pre-stored coordinates, wherein the intersection of the optimal paths for the goods in the control group with intersections is the cargo location coordinates of the goods with abnormal status after the initial inventory count; S7, repeating steps S3-S5 to complete the acquisition of visual images and image processing and analysis of the control group, wherein the control group has two visual images of abnormal goods from different perspectives, and the visual images obtained from the initial inventory are compared with the visual images obtained from the secondary inventory. The comparison content is divided into two parts: one is to compare whether the status of goods without abnormal status is consistent to ensure the accuracy of the detection results of the initial inventory, and the other is to focus on determining the specific status of the abnormal goods based on the multi-perspective visual images, including the two states of poor quality and damaged packaging; S8. The final inventory data of the two inventory counts are sorted out and output to the storage module for the system administrator to review.

6. A management method for an automated logistics warehousing system according to claim 5, characterized in that: The visual image splicing analysis process comprises the following steps: Preprocessing the received visual image, including grayscale conversion and noise reduction; Identify the pre-processed visual image to determine whether there are goods on the cargo space. If there are goods, continue with subsequent barcode detection and positioning. If there are no goods, end the process. The barcode is roughly positioned, and then adjusted based on the rough positioning to obtain its accurate position; The barcode position is extracted and cropped, and input into the decoding algorithm for decoding to obtain the cargo information reserved by the barcode in the warehousing system.

7. A management method for an automated logistics warehousing system according to claim 6, characterized in that: In the visual image splicing analysis process, the probability SVM is used to perform a binary classification operation on the image to determine whether there is cargo or not. The formula is shown in formula (1): （1） Among them, A and B are fitting parameters, and f is the output of the sample standard SVM; The image analysis module collects cargo images and performs binary classification operations to classify the images into two categories: cargo and no cargo, thus constructing a cargo binary classification dataset. The data set is divided into a training set and a test set. Some images are selected to construct a basic training set. At the same time, the cross-platform computer vision library is used to expand the basic training set images, including a series of operations such as image denoising, image rotation, and motion blur, to generate a cargo classification training set, expand the data scale and data richness of the training set, use the training set to train the SVM classification model, and test it on the test set; After obtaining the classification result through the SVM classification model, the classification result is sent to the image analysis module. When the classification result is "1" and the goods exist, the image analysis module starts the subsequent barcode detection and positioning. When the classification result is "0" and the goods do not exist, the image analysis module is in standby mode.

Images