CN-121998546-A - Warehouse goods accurate inventory method based on cooperation of RFID and unmanned aerial vehicle

CN121998546ACN 121998546 ACN121998546 ACN 121998546ACN-121998546-A

Abstract

The invention discloses a warehouse goods accurate inventory method based on cooperation of RFID and unmanned aerial vehicles, which comprises the steps of S1, binding and putting a warehouse goods on a rack after RFID labels are attached to the warehouse goods, storing goods inventory information in a warehouse management system, S2, establishing a three-dimensional model and a space coordinate system of the warehouse, wherein each warehouse location corresponds to a unique three-dimensional space coordinate, S3, executing inventory tasks according to a preset route by the unmanned aerial vehicle, S4, binding the RFID label with the strongest signal when each warehouse location hovers before the unmanned aerial vehicle so that the warehouse location corresponds to the goods one by one, S5, comparing RFID signal values acquired when the unmanned aerial vehicle flies to the next warehouse location with the last warehouse location, binding the label with the strongest signal to the current warehouse location, S6, automatically generating inventory reports by the system after the inventory tasks are finished, comparing RFID real-time data with WMS inventory records and marking difference items. The invention can realize efficient and accurate unmanned inventory and ensure the double accuracy of inventory quantity and inventory position of the inventory.

Inventors

WANG YIJIE
ZHAO XIAOFENG

Assignees

华瑭大昌商业(上海)有限公司

Dates

Publication Date: 20260508
Application Date: 20251222

Claims (11)

1. The accurate warehouse goods checking method based on the cooperation of the RFID and the unmanned aerial vehicle is characterized by comprising the following steps of: S1, binding and putting the warehoused goods on shelves after RFID labels are attached to the warehoused goods, and storing the goods inventory information in a warehouse management system; s2, establishing a three-dimensional model and a space coordinate system of the warehouse, wherein each warehouse position corresponds to a unique three-dimensional space coordinate; S3, the unmanned aerial vehicle executes inventory tasks according to a preset route; S4, binding RFID tags with strongest signals when the unmanned aerial vehicle hovers in front of each bin position, so that the bin positions correspond to cargoes one by one; S5, comparing the RFID signal value acquired when the unmanned aerial vehicle flies to the next position with the RFID signal value of the last position, and binding the tag with the strongest signal to the current position; And S6, after the checking task finishes the return of the unmanned aerial vehicle, automatically generating a checking report, comparing the RFID real-time data with the WMS inventory record, and marking a difference item for manual review.
2. The method for accurately checking the warehouse goods based on the cooperation of the RFID and the unmanned aerial vehicle according to claim 1, wherein the step S1 comprises the steps of randomly selecting 10% sample size after the goods arrive, randomly checking the goods in a sampling way, attaching an RFID label to complete the binding relationship between the label and the goods, and scanning the goods and the warehouse position number to complete the binding relationship between the goods and the warehouse position after the goods are put on a shelf.
3. The accurate inventory method of the warehouse goods based on the cooperation of the RFID and the unmanned aerial vehicle is characterized in that step S2 is used for completing the construction of a point cloud map in an off-line training stage, wherein the unmanned aerial vehicle carrying the laser radar is used for carrying out point cloud acquisition on the three-dimensional scanning of the whole scene of the warehouse, preprocessing the acquired point cloud data, extracting key geometric features, establishing a mapping relation between a point cloud feature descriptor and a spatial position, and storing all feature-position pairs into a point cloud feature database to form a point cloud feature library for positioning.
4. The precise inventory method for warehouse goods based on cooperation of RFID and unmanned aerial vehicle as claimed in claim 3, wherein the detailed chart construction flow of step S2 is as follows: i. Three-dimensional space scanning, namely planning a flight path of the unmanned aerial vehicle, ensuring to cover all areas of a warehouse, and performing multi-view scanning by using a high-precision laser radar; arranging a plurality of reflecting column calibration points with known coordinates in a warehouse to obtain a group of reference sets { P1 (x 1, y1, z 1), P2 (x 2, y2, z 2) }, pn (xn, yn, zn) }, wherein the projection center of the unmanned aerial vehicle nest on the ground is taken as the origin (0, 0) of the world coordinate system, and all the scanning point clouds are ensured to be unified to the same global coordinate system; And thirdly, performing point cloud data processing, namely performing downsampling, denoising and ground segmentation on the original point cloud, extracting key characteristic points, initializing a local coordinate system, converting the local point cloud into a global coordinate system and fusing the global coordinate system with the existing map.
5. The precise inventory method of warehouse goods based on the cooperation of RFID and unmanned aerial vehicle as claimed in claim 4, wherein the step S2 of extracting the characteristics of the point cloud and constructing the database is carried out by the following steps of extracting the point cloud blocks and the key frames, distributing a unique identifier and precise pose information for each key frame, extracting the characteristic descriptors for the point cloud P1 of the key frame to obtain a first characteristic vector: V1= [ FPFH1, curvature mean, curvature variance, plane normal vector, line segment direction, dot density, altitude statistics ]; the vector V1 and the corresponding geographic position P1 (x 1, y1, z 1) are bound and used as a record to be stored in a database, a second key frame point cloud P2 is recorded, the whole process is repeated until all the point clouds are collected, and a priori map feature database supporting rapid retrieval based on feature similarity is obtained.
6. The precise inventory method of the warehouse goods based on the cooperation of the RFID and the unmanned aerial vehicle as claimed in claim 1, wherein the size ratio of the warehouse three-dimensional model established in the step S2 to the actual warehouse is 1:1, the warehouse three-dimensional model comprises a receiving and transmitting area, high-level shelves in a storage area, the heights of each layer of the shelves are in one-to-one correspondence with real objects, and the center position of each shelf corresponds to a corresponding three-dimensional space coordinate point (x, y, z).
7. The precise inventory method of warehouse goods based on the cooperation of RFID and unmanned aerial vehicles according to claim 1 is characterized in that two points of space coordinates in the step S3 are connected to form a straight line, each space point of the warehouse location is connected to form a preset route of the unmanned aerial vehicle, the unmanned aerial vehicle executes different flight inventory tasks according to different routes, the unmanned aerial vehicle hovers briefly at each warehouse location to collect all RFID tag signals of the current position, the tag with the strongest signal is bound to the current warehouse location based on the matching of RFID signal strength, and if the strongest signal strength of a certain warehouse location is lower than a preset threshold value, the tag is rejected, so that the empty warehouse location cannot be bound with goods.
8. The method for accurately checking the warehouse goods based on the cooperation of the RFID and the unmanned aerial vehicle according to any one of claims 3 to 5, wherein the online positioning stage of the step S3 comprises i. Real-time point cloud acquisition, wherein the unmanned aerial vehicle acquires local point cloud in real time in the checking process, and extracts a real-time feature vector V_current. And ii, feature matching and position estimation, namely searching and comparing feature vectors which are most similar to the V_current in the prior map, so as to obtain the real-time position of the P_current.
9. The precise inventory method of warehouse goods based on the cooperation of RFID and unmanned aerial vehicle as claimed in claim 8, wherein the steps S4 and S5 find the best matching position in the feature database according to the tag signal collected in real time by combining the following two algorithms: finding K reference points most similar to the real-time vector in the feature database, and taking the position coordinate average value of the K points as a final positioning result; And (3) weighting the K nearest neighbor positioning algorithm, namely carrying out weighted average on the K nearest neighbor points according to the similarity, wherein the higher the similarity is, the larger the point weight is.
10. The precise inventory method of warehouse goods based on the cooperation of RFID and unmanned aerial vehicle as claimed in claim 9, wherein the step S4 or S5 adopts Euclidean distance or Manhattan distance to measure similarity, and the specific matching and binding process is as follows: i. the method comprises the steps of obtaining a current signal vector, namely converting a tag signal read currently into a vector V_current, wherein each dimension of the vector represents a known tag ID, and the value of the tag ID is a RSSI value read currently; acquiring a reference feature vector, namely, taking out a reference feature vector V_ref of the current library L_current from a feature database; Calculating the similarity by calculating the similarity of the current vector V_current and the reference feature vector V_ref by using the inverse of the Euclidean distance or the Manhattan distance, and introducing a weight in the calculation to give a higher weight W_i to the tags with strong signals in V_ref (representing that it is the inherent tag of this bin), the weight W_i being a function of the RSSI value of this tag in V_ref; if the dimension corresponding to the label of the current stock position has larger positive change in DeltaV and the signal change of other labels is smaller or negative, judging that the new goods are put in the current stock position; and V, deciding and binding, namely traversing the delta V vector to find the label T_best with the most increased signal strength, setting a minimum gain threshold, and binding the library L_current with the label T_best when the maximum gain exceeds the minimum gain threshold.
11. The method for accurately checking the warehouse goods based on the cooperation of the RFID and the unmanned aerial vehicle according to claim 1, wherein the checking report mark difference item automatically output in the step S6 comprises two differences of an inventory position error and an inventory position lack due inventory.

Description

Warehouse goods accurate inventory method based on cooperation of RFID and unmanned aerial vehicle Technical Field The invention relates to a warehouse goods checking method, in particular to a warehouse goods accurate checking method based on cooperation of RFID (Radio Frequency Identification radio frequency identification) and an unmanned aerial vehicle, which is suitable for automatic inventory checking in scenes such as a large warehouse, a logistics center and the like. Background Under the background of industry 4.0, the high-precision and real-time perception of the spatial position of a physical entity in a complex environment has become a core premise for realizing intelligent decision. Radio Frequency Identification (RFID) technology has great potential in the fields of asset tracking, warehouse logistics, intelligent manufacturing and the like by virtue of the inherent characteristics of non-line-of-sight, non-contact and multi-target concurrent identification. However, conventional RFID systems mainly serve identification, and their positioning accuracy is limited to the order of meters or even tens of meters, which is fundamentally derived from their single-dimensional, low-robustness ranging model based on Received Signal Strength Indication (RSSI). The model simplifies complex space electromagnetic wave propagation into an ideal path loss model, and the multipath effect, shadow fading, polarization mismatch of reader and tag antennas and space non-isotropic radiation of radio frequency signals which are commonly existing in an actual environment cause high nonlinearity and uncertainty between an RSSI value and a real distance, thereby severely restricting the application efficiency of the model in a precise positioning scene. In warehouse management, inventory checking is an important link for ensuring account agreement. The checking refers to a work flow for checking the actual quantity of the inventory periodically or temporarily by an enterprise, and aims to check the consistency of accounts and ensure the accuracy of inventory management. In recent years, with the development of low-altitude economy, unmanned aerial vehicle technology is gradually applied to warehouse management, and particularly to the field of inventory. The RFID technology has the advantages of flexible maneuvering, wide coverage range and the like due to the characteristics of non-contact and batch identification, and the cooperation of the unmanned aerial vehicle and the RFID becomes an ideal choice for automatic inventory in the industry. However, the following disadvantages still exist in the prior art: 1. The library position positioning is inaccurate, signals emitted by the RFID are continuously obtained in batches in the flight process of the unmanned aerial vehicle, but specific library positions corresponding to each tag cannot be accurately judged, only the number counting can be realized, and the library position level management requirement cannot be met. 2. The data processing is complex, how to efficiently integrate RFID data collected by an unmanned aerial vehicle and automatically compare the RFID data with a WMS inventory system still needs to be optimized. Therefore, a precise inventory method for warehouse goods based on cooperation of RFID and unmanned aerial vehicle is needed to improve the level of intelligent inventory. Disclosure of Invention The invention aims to solve the technical problem of providing a warehouse goods accurate checking method based on cooperation of RFID and unmanned aerial vehicle, which not only can realize high-precision RFID fingerprint positioning, but also can realize efficient and low-risk unmanned checking, and can ensure the double accuracy of the stock quantity and the stock position of checking. The invention provides a warehouse goods accurate inventory method based on cooperation of RFID and unmanned aerial vehicles, which comprises the following steps of S1, binding and putting a warehouse goods on a shelf after the RFID labels are attached to the warehouse goods, storing goods inventory information in a warehouse management system, S2, establishing a three-dimensional model and a space coordinate system of the warehouse, wherein each warehouse location corresponds to a unique three-dimensional space coordinate, S3, executing inventory tasks according to a preset route by the unmanned aerial vehicle, S4, binding the RFID label with the strongest signal when the unmanned aerial vehicle hovers in front of each warehouse location, enabling the warehouse location to correspond to goods one by one, S5, comparing RFID signal values acquired when the unmanned aerial vehicle flies to the next warehouse location with the RFID signal values acquired by the last warehouse location, binding the label with the strongest signal values to the current warehouse location, S6, automatically generating inventory reports after the inventory tasks are finished, comparing RFID real