A low-cost and fast automatic inventory counting method for a three-dimensional warehouse

Abstract

The present invention provides a method for low-cost and rapid automatic inventory checking of a stereoscopic warehouse. By pasting a piece of reflective paper on each storage location in the stereoscopic warehouse, installing a set of illuminating optoelectronics on both the left and right sides of the load platform of the stacker / mechanical grasping device in each aisle of the stereoscopic warehouse, and installing a camera above the load platform, the inventory checking control system schedules the stacker to quickly run a single trip in the aisle. At this time, the optoelectronics on the load platform will continuously irradiate the reflective paper on the storage location and feed back the irradiation result to the inventory checking control system. The inventory checking control system checks whether there is a storage location based on the irradiation feedback result, where 0 indicates no goods and 1 indicates there are goods. After running through one aisle, the inventory checking control system calculates the storage locations with anomalies and then schedules the stacker to take out the goods at the abnormal storage locations, take pictures and upload them to the inventory checking control system. The system AI analyzes the types and quantities of the goods at the storage locations and generates data for manual confirmation.

Description

Technical Field

[0001] The invention relates to the field of stereoscopic warehouses, and in particular to a low-cost, fast and automatic inventory method for stereoscopic warehouses. Background Art

[0002] In some hospital drug material warehouses with a high degree of automation or finished product warehouses, semi-finished product warehouses, and raw material warehouses of production factories, three-dimensional high-bay warehouses with automatic stacking and palletizing machines are deployed to realize the automated and dense storage of materials. This method can realize the automatic entry and exit of the warehouse, improve the utilization rate of the warehouse capacity, realize data management, and improve business processing capabilities. However, inventory counting in the warehouse has always been a difficult problem to solve, because mechanical failures, network failures, and system failures are inevitable during the entry and exit operations. When failures occur, human intervention is generally required, which will cause the actual inventory quantity to be inconsistent with the warehouse management software. At this time, inventory counting is required to match the physical objects with the system data. Generally, the internal depth of the three-dimensional warehouse is relatively deep, the height is high, there are many cargo spaces, and the internal space is small, so manual inventory counting is very difficult.

[0003] Existing methods:

[0004] Manually conduct inventory in the warehouse, record the actual inventory of each location, and compare it with the system to make adjustments;

[0005] Install photoelectric sensors at each cargo location, and transmit the illumination results to the inventory system to detect the actual inventory status of the cargo location;

[0006] Install RFID readers at each cargo location, install RFID chips on cargo containers, and report the reading results to the inventory system;

[0007] Disadvantages of existing methods:

[0008] Manual inventory in the warehouse requires riding a hoist to each cargo location to check the actual inventory situation and compare it with the system. This operation is dangerous and inefficient. The installation and testing method of installing a photoelectric camera at each cargo location has high installation and testing costs, and the system development and design is complex. The installation and debugging cost of installing an RFID reader at each cargo location is high, and there are usually hundreds to tens of thousands of cargo locations, so the development cost is even higher. Summary of the invention

[0009] In order to solve the above technical problems, the present invention proposes a low-cost, fast and automatic inventory counting method for a three-dimensional warehouse. The inventory of each cargo location inside the three-dimensional warehouse is counted in a simple and low-cost manner, and the counting results are reported to the inventory management system to generate an inventory difference report. The system inventory can be adjusted by comparing the inventory results on the system.

[0010] To achieve the above object, the technical solution adopted by the present invention is:

[0011] A method for low-cost and rapid automatic inventory of a three-dimensional warehouse, characterized by the following steps:

[0012] S1: Stick reflective paper on the inner side of each storage location in the three-dimensional warehouse, install optoelectronic components on the load platform of the stacker crane, and ensure that the optoelectronics can just irradiate the reflective paper when the load platform stops at each storage location;

[0013] S2: The inventory control system creates an inventory list and specifies the inventory location area;

[0014] S3: The inventory list is sent to the inventory control system;

[0015] S4: The inventory control system mobilizes the load platform of the stacker crane to run rapidly within the specified storage location range for optoelectronic scanning;

[0016] S5: The optoelectronic components on the load platform irradiate the reflective paper and receive the reflection signal of the reflective paper on the storage location;

[0017] S6: When there is an obstruction at the storage location, the optoelectronics returns false, and when there is no obstruction at the storage location, the optoelectronics returns true;

[0018] S7: The inventory control system checks whether there is goods at the storage location according to the returned true / false and records it in the system;

[0019] S8: After scanning the empty storage locations, exclude the empty storage locations from the inventory list;

[0020] S9: The inventory control system then mobilizes the stacker crane to each storage location with goods;

[0021] S10: The stacker crane picks up the goods at the storage location onto the load platform of the stacker crane;

[0022] S11: The photographing system on the load platform takes a photograph of the goods on the load platform;

[0023] S12: The photo is uploaded to the AI server to identify the type and quantity;

[0024] S13: The inventory control system records the recognition result;

[0025] S14: The recognition result is uploaded to the inventory control system, and the inventory control system updates and adjusts the inventory data according to the actual result;

[0026] As a preferred technical solution of the present invention: The reflective paper is optoelectronic reflective paper, and the optoelectronic reflective paper is signal-connected to the optoelectronic components.

[0027] As a preferred technical solution of the present invention: In step S8, after scanning the empty storage locations, the inventory control system records the storage location data where the inventory data in the three-dimensional warehouse is different from the actual empty storage locations and marks them as abnormal storage locations.

[0028] As the preferred technical solution of the present invention: in step S8, after the empty storage location scan is completed, the empty storage location data is excluded from the inventory list to provide a preparation list for the second round of AI identification inventory. The exclusion method adopts the binary half search method, which is as follows: first, the two Arrays are sorted according to ID, the lowest ID is marked as minID, and the highest ID is marked as maxID. The middle element is searched by the formula (minID+maxID) / 2), and the element to be compared is judged on the left or right side of the Array, and it is halved in turn, and the search area is updated to finally obtain the comparison search result. If minID>maxID appears in the end, it means that there is no target element, the comparison fails, and it is marked as abnormal inventory. The entire operation time complexity is reduced to O(log(sourceArray.length+compareArray.length)).

[0029] As a preferred technical solution of the present invention: in step S8, after the empty cargo locations are scanned, a new inventory list is generated according to the comparison results, the empty cargo locations are screened out, abnormal cargo locations are recorded, and a second round of AI inventory list is provided.

[0030] As a preferred technical solution of the present invention: in step S8, after the scanning of the empty cargo locations is completed, according to the second round of inventory list, the inventory control system drives the stacker to the cargo locations with goods to perform AI inventory, identify the cargo locations, and obtain the final inventory difference data report.

[0031] In the above structure: The purpose of the present invention is to count the inventory of each cargo location inside the three-dimensional warehouse in a simple and low-cost manner, and report the counting results to the inventory management system to generate an inventory difference report, and the system inventory can be adjusted by comparing the inventory results on the system.

[0032] The method of the present invention sticks a piece of reflective paper on each cargo location in the stereoscopic warehouse, installs a group of illuminating photoelectrics on the left and right sides of the loading platform of the stacker / mechanical grabbing device in each lane of the stereoscopic warehouse, and installs a camera above the loading platform. The inventory control system dispatches the stacker to run a single trip quickly in the lane. At this time, the photoelectrics on the loading platform will continuously illuminate the reflective paper on the cargo location, and feed back the irradiation results to the inventory control system. The inventory control system counts whether there is a cargo location based on the irradiation feedback results, 0 for no goods, and 1 for goods. After running through an lane, the inventory control system calculates the abnormal cargo location and then dispatches the stacker to the abnormal cargo location to take out the cargo location, takes a photo and uploads it to the inventory control system. The system AI analyzes the type and quantity of the cargo location and generates data for manual confirmation.

[0033] The system framework of the present invention is divided into a three-dimensional storage repository, a stacker goods access mechanism, irradiation optoelectronics, reflective paper, a camera, and an inventory control system. The stacker is scheduled and controlled through the inventory control system. During the operation of the stacker, inventory data is generated and fed back to the inventory control system. The inventory control system generates inventory difference data. The AI module in the inventory control system analyzes the abnormal data to obtain an inventory difference report, and the operator quickly adjusts the system inventory to the data corresponding to the physical objects.

[0034] The overall inventory of the automated stereoscopic warehouse is divided into three steps:

[0035] In the first step, the reflective paper on the shelves is irradiated by the optoelectronics installed on both sides of the stacker's loading platform, and it is judged whether there is a box in the cargo location according to the feedback signal.

[0036] In the second step, the empty cargo locations are excluded from the inventory list.

[0037] In the third step, the stacker is mobilized to pick up goods and take pictures at each cargo location with goods. The AI identifies the actual inventory data and compares it with the inventory data in the system to obtain an inventory difference report.

[0038] The specific steps are as follows:

[0039] 1. The inventory control system creates an inventory business document and specifies the starting cargo location for inventory.

[0040] 2. The inventory control system schedules the stacker and the loading platform to quickly go to the specified column and layer for a scan.

[0041] 3. During the cargo location scanning process, the optoelectronic components on both sides of the loading platform irradiate the reflective paper on both sides of the cargo locations.

[0042] 4. After the optoelectronic light irradiates the reflective paper, the light will be returned to the optoelectronic components, and the inventory control system judges that there is no box at this position.

[0043] 5. When the optoelectronic light irradiates the box at the cargo location, the light will not be reflected back, and the inventory control system judges that there is a box at this position.

[0044] 6. Until the scanning of the last cargo location is completed.

[0045] 7. The inventory control system compares the recorded inventory data with the actual inventory data and lists the abnormal cargo locations.

[0046] 8. The inventory control system excludes the empty cargo locations from the inventory list.

[0047] 9. The inventory control system then schedules the stacker loading platform to go to the cargo locations with goods, take out the boxes onto the loading platform, and take pictures.

[0048] 10. The photo data is transmitted to the inventory control system, and the AI module identifies the materials and quantities in the photos.

[0049] 11. The inventory control system records the materials and quantities in this location in the world, and then takes pictures of the goods picked from the next occupied location.

[0050] 12. After taking pictures of all abnormal locations, the inventory control system generates an inventory variance report.

[0051] 13. Manually confirm the inventory adjustment of the inventory control system according to the variance report to complete the inventory count.

[0052] During the running process of the stacker crane, it irradiates through the irradiation photoelectric on the load platform of the stacker crane and the reflective paper on the location, forming a series of irradiation result signals. The result signals are fed back to the inventory control system and compared with the inventory data recorded by the inventory control system. At this time, the actual storage situation of the goods on the location can be detected. The inventory control system marks the locations where the physical goods do not match the system inventory as abnormal locations. Generally, there are very few goods in abnormal locations.

[0053] After the inventory control system obtains the abnormal location data, it then schedules the stacker crane to the abnormal location for picking operations. After the goods are picked onto the load platform of the stacker crane, the inventory control system then controls the camera above the load platform to take pictures and uploads them to the inventory control system. The AI module of the inventory control system analyzes and obtains the categories and quantities of the materials in the abnormal location and records them in the inventory variance report of the inventory control system.

[0054] Operators only need to view the inventory variance report to confirm the inventory adjustment, which is fast, simple, and low-cost.

[0055] Compared with the prior art, the beneficial effects of the present invention are:

[0056] The inventory counting cost of the present invention is very low. Only a set of photoelectrics need to be installed on both sides of the stacker crane in each lane, reflective paper is pasted on the location, and a camera is installed on the load platform.

[0057] The inventory counting efficiency of the present invention is high. The inventory counting work is divided into two steps. First, all locations are quickly scanned continuously using the photoelectrics on the load platform to filter out empty locations.

[0058] After the present invention filters out the empty locations, the number of locations to be inventoried will be greatly reduced. Then, the goods in the occupied locations are taken out through the load platform, pictures are taken, and the AI identifies the material categories and quantities, which are recorded by the system. In this way, there is no need to use the load platform to take pictures and identify each location, which will greatly improve the work efficiency.

[0059] The present invention is an automated inventory count, without the need for manual inspection in the lane, reducing the risk.

[0060] The system of the present invention can automatically generate a variance report, and all abnormal location inventory adjustments can be completed by manual confirmation.

[0061] Through the low-cost automatic stereoscopic warehouse inventory method of the present invention, there is no need for manual entry into the aisle for inventory, no need to install optoelectronic devices at all storage locations, no need to install RFID readers at all storage locations, and no need to attach RFID tags to boxes and containers; only by installing optoelectronic devices and cameras on the loading platform and attaching reflective paper to the storage locations, first excluding empty storage locations, and then performing AI recognition inventory, low-cost automation of inventory is achieved through software means. Description of the Drawings

[0062] Figure 1 It is a flowchart of a low-cost and fast automatic inventory method for a stereoscopic warehouse;

[0063] Figure 2 It is a physical operation diagram of the present invention. Detailed Embodiments

[0064] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments:

[0065] As Figure 1 shown, the present invention proposes a low-cost and fast automatic inventory method for a stereoscopic warehouse, including the following steps:

[0066] S1: Attach reflective paper to the inner side of each storage location in the stereoscopic warehouse, install optoelectronic components on the loading platform of the stacker, and ensure that the optoelectronic device can just irradiate the reflective paper when the loading platform stops at each storage location;

[0067] S2: The inventory control system creates an inventory list and specifies the inventory location area;

[0068] S3: The inventory list is sent to the inventory control system;

[0069] S4: The inventory control system mobilizes the loading platform of the stacker to run quickly within the specified storage location range for optoelectronic scanning;

[0070] S5: The optoelectronic component on the loading platform irradiates the reflective paper and receives the reflection signal of the reflective paper on the storage location;

[0071] S6: When there is a storage location blocking on the storage location, the optoelectronic device returns false, and when there is no storage location blocking, the optoelectronic device returns true;

[0072] S7: The inventory control system checks whether there is goods in the storage location according to the returned true / false and records it in the system;

[0073] S8: After scanning the empty storage locations, exclude the empty storage locations from the inventory list;

[0074] S9: The inventory control system then mobilizes the stacker to go to each storage location with goods;

[0075] S10: The stacker takes the goods on the storage location to the loading platform of the stacker;

[0076] S11: The photographing system on the cargo platform takes a photograph of the cargo on the cargo platform;

[0077] S12: Photos are uploaded to the AI ​​server to identify the type and quantity;

[0078] S13: The inventory control system records the recognition result;

[0079] S14: The recognition results are uploaded to the inventory control system, and the inventory control system updates and adjusts the inventory data according to the actual results;

[0080] The reflective paper is photoelectric reflective paper, and the photoelectric reflective paper is signal-connected to the photoelectric component.

[0081] In step S8, when the empty storage locations are scanned, the inventory control system records the storage location data in the three-dimensional warehouse that is different from the actual empty storage locations, and marks it as an abnormal storage location.

[0082] In step S8, after the empty storage location scan is completed, the empty storage location data is excluded from the inventory list to provide a preparation list for the second round of AI identification inventory. The exclusion method uses the binary half search method, which is as follows: first sort the two arrays according to ID, mark the lowest ID as minID, and the highest ID as maxID, search the middle element by the formula (minID+maxID) / 2), judge whether the element to be compared is on the left or right side of the Array, halve it in turn, update the search area and finally get the comparison search result. If minID>maxID appears in the end, it means that there is no target element, the comparison fails, and it is marked as abnormal inventory. The entire operation time complexity is reduced to O(log(sourceArray.length+compareArray.length)).

[0083] In step S8, when the empty storage locations are scanned, a new inventory list is generated according to the comparison results, the empty storage locations are screened out, abnormal storage locations are recorded, and a list for the second round of AI inventory is provided.

[0084] In step S8, after the scanning of the empty cargo locations is completed, according to the second round of inventory list, the inventory control system drives the stacker to the cargo locations with goods to perform AI inventory, identify the cargo locations, and obtain the final inventory difference data report.

[0085] The purpose of the present invention is to count the inventory of each cargo location in the three-dimensional warehouse in a simple and low-cost manner, and report the counting results to the inventory management system to generate an inventory difference report, so that the system inventory can be adjusted by comparing the inventory results on the system.

[0086] In the method of the present invention, a reflective paper is pasted on each storage location. On both sides of the load platform of the stacker / mechanical grasping device in each aisle of the automated storage and retrieval system (AS / RS), a set of irradiation optoelectronics is installed, and a camera is installed above the load platform. The inventory control system schedules the stacker to quickly run a single trip in the aisle. At this time, the optoelectronics on the load platform will continuously irradiate the reflective paper on the storage location, and feed back the irradiation result to the inventory control system. The inventory control system checks whether there is an item at the storage location based on the irradiation feedback result. 0 means no item, and 1 means there is an item. After finishing scanning one aisle, the inventory control system calculates the storage locations with abnormalities and then schedules the stacker to pick up the items at the abnormal storage locations, take pictures and upload them to the inventory control system. The system's AI analyzes the types and quantities of the items at the storage locations and generates data for manual confirmation.

[0087] As Figure 2 shown, the system framework of the present invention is divided into an automated storage and retrieval system (AS / RS), a stacker goods storage and retrieval mechanism, irradiation optoelectronics, reflective paper, a camera, and an inventory control system. The inventory control system schedules and controls the stacker. During the operation of the stacker, inventory data is generated and fed back to the inventory control system. The inventory control system generates inventory difference data. The AI module in the inventory control system analyzes the abnormal data to obtain an inventory difference report, and the operator quickly adjusts the system inventory to the data corresponding to the physical inventory.

[0088] The overall inventory of the automated storage and retrieval system (AS / RS) is divided into three steps:

[0089] In the first step, the optoelectronics installed on both sides of the stacker load platform irradiate the reflective paper on the shelf, and based on the feedback signal, it is judged whether there is a box at the storage location;

[0090] In the second step, the empty storage locations are excluded from the inventory list;

[0091] In the third step, the stacker is mobilized to pick up and take pictures of the items at each storage location with items. The AI identifies the actual inventory data and compares it with the inventory data in the system to obtain an inventory difference report.

[0092] The specific steps are as follows:

[0093] 1. The inventory control system creates an inventory business document and specifies the starting storage location for the inventory;

[0094] 2. The inventory control system schedules the stacker and the load platform to quickly go to the specified column and layer for a scan;

[0095] 3. During the storage location scanning process, the optoelectronic components on both sides of the load platform irradiate the reflective paper on both sides of the storage location;

[0096] 4. After the optoelectronic light irradiates the reflective paper, the light will be returned to the optoelectronic components, and the inventory control system then judges that there is no box at this position;

[0097] 5. When the photoelectric light irradiates the boxes in the storage location, no light will be reflected back. The inventory control system determines that there are boxes in this location.

[0098] 6. Until the scanning of the last storage location is completed;

[0099] 7. The inventory control system compares the recorded inventory data with the actual inventory count data and lists the abnormal storage locations.

[0100] 8. The inventory control system excludes the empty storage locations from the inventory count list.

[0101] 9. The inventory control system then schedules the stacker crane's load platform to go to the storage location with goods, pick up the box onto the load platform, and take a photo.

[0102] 10. The photo data is transmitted to the inventory control system, and the AI module identifies the materials and quantities in the photo.

[0103] 11. The inventory control system records the materials and quantities in the real world at this storage location, and then goes to the next storage location with goods to pick up the goods and take a photo.

[0104] 12. After taking photos of all the goods at the abnormal storage locations, the inventory control system generates an inventory count difference report.

[0105] 13. Manually confirm the inventory adjustment of the inventory control system according to the difference report to complete the inventory count.

[0106] During the running process of the stacker crane, the irradiation photoelectric on the stacker crane's load platform irradiates the reflective paper on the storage location to form a series of irradiation result signals. The result signals are fed back to the inventory control system and compared with the inventory data recorded by the inventory control system. At this time, the actual goods storage situation at the storage location can be detected. The inventory control system marks the storage locations where the physical goods do not match the system inventory as abnormal storage locations. Generally, there are very few abnormal storage locations.

[0107] After the inventory control system obtains the abnormal storage location data, it schedules the stacker crane to go to the abnormal storage location to pick up the goods. After the goods are picked up onto the stacker crane's load platform, the inventory control system then controls the camera above the load platform to take a photo and upload it to the inventory control system. The AI module of the inventory control system analyzes it to obtain the category and quantity of the materials at the abnormal storage location and records them in the inventory count difference report of the inventory control system.

[0108] Operators only need to view the inventory count difference report to confirm the inventory adjustment, which is fast, simple, and low-cost.

[0109] In this embodiment, through the comparison of the time efficiency of the three-dimensional storage location inventory count with 3,500 storage locations, 3 aisles, and 3 stacker cranes:

[0110] Fully manual: Approximately 30 days for 1 person, 720 hours;

[0111] AI Full-location Recognition and Inventory Count: Based on the camera exposure time t exp is 0.1 second, the average time t type of the product specification and type detection algorithm is 5 seconds, and the average time t num of the quantity detection algorithm is 0.5 second. Then the detection time for the entire location is 5.6 seconds. The time required to detect a single aisle of 3500 / 3 is approximately 1.8 hours. Assuming the time t1 for the stacker's load platform to move to the next location is about 5 seconds, and the time t2 for goods to enter and exit the load platform is about 13 seconds, the estimated time required to inventory a single aisle is about 7.7 hours. Since 3 stackers can operate in parallel, the total time is 7.7 hours. The calculation method for the total required time T is as follows:

[0112] T = 3500·(t exp +t type +t num +t1+t2) / 3

[0113] After substituting the values of each variable, the calculated result is 27533 seconds, approximately 7.7 hours.

[0114] Photoelectric Pre-scanning Method: By installing a reflective photoelectric device on the load platform, all empty locations can be pre-excluded. Assuming that half of the locations are empty and the scanning time for all empty locations is 0.2 hour, then the total time for all inventory counts is 7.7 / 2 + 0.2 = 4.05 hours.

[0115] The inventory count cost of the present invention is very low. For each aisle, only a set of photoelectrics needs to be installed on both sides of the stacker, a reflective paper is pasted on the location, and a camera is installed on the load platform;

[0116] The inventory count efficiency of the present invention is high. The inventory count work is divided into two steps. First, all locations are quickly scanned continuously using the photoelectrics on the load platform to filter out the empty locations;

[0117] After the present invention filters out the empty locations, the number of locations to be inventoried will be greatly reduced. Then, the goods in the locations with goods are taken out through the load platform, the goods category and quantity are recognized by photographing with AI, and the system records them. In this way, there is no need for the load platform to take pictures and recognize at each location, which will greatly improve the work efficiency;

[0118] The present invention is an automated inventory count, without the need for manual inspection inside the aisle, reducing the risk;

[0119] The system of the present invention can automatically generate a difference report, and all inventory adjustments for abnormal locations can be completed after manual confirmation.

[0120] Through the low-cost automatic stereoscopic warehouse inventory method of the present invention, there is no need for manual entry into the roadway for inventory, no need to install optoelectronic devices at all storage locations, no need to install RFID readers at all storage locations, and no need to attach RFID tags to boxes and containers; only by installing optoelectronic devices and cameras on the loading platform and attaching reflective paper to the storage locations, first excluding empty storage locations, and then conducting AI recognition and inventory, a low-cost automated inventory is achieved through software means.

[0121] The above are only the preferred embodiments of the present invention, and are not any other form of limitation to the present invention. Any modification or equivalent change made according to the technical essence of the present invention still belongs to the scope protected by the present invention.

Claims

1. A method for low-cost and rapid automatic inventory of a three-dimensional warehouse, characterized in that, It includes the following steps: S1: Stick reflective paper on the inner side of each storage location in the automated storage and retrieval system (AS / RS). Install optoelectronic components on the load platform of the stacker crane. Ensure that the optoelectronic device can just irradiate the reflective paper when the load platform stops at each storage location; S2: The inventory control system creates an inventory list and designates the inventory location area; S3: The inventory list is sent down to the inventory control system; S4: The inventory control system mobilizes the load platform of the stacker crane to run quickly within the designated storage location range for optoelectronic scanning; S5: The optoelectronic components on the load platform irradiate the reflective paper and receive the reflection signal of the reflective paper on the storage location; S6: When there is an obstruction on the storage location, the optoelectronic device returns false, and when there is no obstruction on the storage location, the optoelectronic device returns true; S7: The inventory control system checks whether there is goods in the storage location according to the returned true / false and records it in the system; S8: After the scanning of empty storage locations is completed, the empty storage locations are excluded from the inventory list, providing a preparation list for the second-round AI identification and inventory. The exclusion method uses the binary half search method, specifically as follows: First, sort the two columns according to the ID, mark the lowest ID as minID, and the highest ID as maxID. Search for the middle element through the formula (minID + maxID) / 2, judge whether the compared element is on the left or right side of the column, and halve it in turn to update the search area and finally obtain the comparison search result. If minID > maxID finally appears, it means that there is no target element and the comparison fails, which is marked as abnormal inventory. The reflective paper is optoelectronic reflective paper, and the optoelectronic reflective paper is signal-connected to the optoelectronic components. In step S8, after the scanning of empty storage locations is completed, the inventory control system records the storage location data where the inventory data in the AS / RS is different from the actual empty storage locations and marks them as abnormal storage locations. In step S8, after the scanning of empty storage locations is completed, a new inventory list is generated according to the comparison result, the empty storage locations are screened out, the abnormal storage locations are recorded, and a list for the second-round AI inventory is provided. In step S8, after the scanning of empty storage locations is completed, according to the second-round inventory list, the inventory control system drives the stacker crane to go to the storage locations with goods for AI inventory, identify the storage locations, and obtain the final inventory difference data report. ​ ​ ​ 2. The method for low-cost and rapid automatic inventory checking of a three-dimensional warehouse according to claim 1, characterized in that ​ 3. A low-cost and fast automatic inventory method for a three-dimensional warehouse according to claim 1, characterized in that, ​ 4. A method for low-cost and rapid automatic inventory of a three-dimensional warehouse according to claim 1, characterized in that, ​ 5. A method for low-cost and rapid automatic inventory of a three-dimensional warehouse according to claim 1, characterized in that, ​

Images