Unmanned warehouse management method and system with real-time inventory function

Abstract

The present invention relates to the field of warehouse management technology, specifically to an unmanned warehouse management method and system with a real-time inventory function. A signal strength-probability statistical model is constructed through the return signal data generated during the historical inventory process. In the current inventory cycle, the stability of the return signal of each RFID tag is checked. When stability is confirmed, preliminary positioning based on the arrival time difference is performed and the probability of the storage unit to which it belongs is determined. Finally, the preliminary positioning is verified by combining the probability of the storage unit and the inventory information of the previous inventory cycle to avoid the situation where the positioning jumps and the verification cannot be made, thereby making it impossible to confirm whether the RFID tag has actually changed position. The present application can utilize the existing RFID system to achieve low-cost, high-reliability material positioning, thereby achieving real-time and accurate inventory, and providing an information basis for the management of unmanned warehouses.

Description

Technical Field

[0001] The present invention relates to the technical field of warehouse management, and in particular to an unmanned warehouse management method and system with a real-time inventory function. Background Art

[0002] Unmanned warehouse management is a management model that utilizes modern technologies such as automated equipment, robotics, the Internet of Things (IoT), artificial intelligence (AI), and big data analytics to perform a series of operations within a warehouse, including storage, handling, picking, and packaging. This model aims to reduce reliance on manual operations and improve the efficiency and accuracy of warehouse management.

[0003] Real-time inventory counting is the core of unmanned warehouse management. Data from real-time inventory counting can guide operations such as storage, handling, picking, and packaging. Existing inventory counting technologies primarily rely on RFID tags. However, RFID-based positioning technology typically suffers from poor positioning accuracy and is susceptible to interference. In particular, changes in the position of RFID tags over short distances can easily cause jumps in positioning information. Therefore, accurate real-time inventory counting cannot be achieved using RFID positioning technology alone. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an unmanned warehouse management method and system with real-time inventory function to solve the above technical problems.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] The unmanned warehouse management method with real-time inventory function of the present invention comprises the following steps:

[0007] Obtaining historical inventory return signal data for a plurality of storage units on shelves within a warehouse, and obtaining the strength of return signals returned multiple times by the RFID tag within a current inventory cycle and the arrival times of the return signals returned multiple times, wherein the historical inventory return signal data includes historical signal strengths of multiple historical return signals returned to the RFID tag by the RFID tag within multiple historical inventory time periods, and the multiple returned return signals are generated based on radio signals transmitted multiple times by an RFID base station;

[0008] Building a signal strength-probability statistical model between multiple storage units and the RFID base station based on the historical inventory return signal data, wherein the signal strength-probability statistical model characterizes the corresponding relationship between different signal strengths of the return signal and the probability of belonging to the storage unit;

[0009] Calculating a signal steady state and an average signal strength of the RFID tag based on the strength of the return signals returned multiple times;

[0010] Determining the probability that the RFID tag belongs to one or more storage units based on the average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model; and performing positioning based on the arrival time of the return signal of multiple returns of the RFID tag in the stable state to obtain a preliminary positioning;

[0011] Performing positioning verification based on the probability, the preliminary positioning, and inventory information from a previous inventory cycle to obtain a positioning result of the RFID tag; and performing warehouse management based on the positioning result of the RFID tag.

[0012] In one embodiment of the present application, it further includes:

[0013] Extracting identification information of an RFID tag in an unstable signal state, and extracting a positioning result of the identification information from a previous inventory cycle containing a positioning result of the identification information;

[0014] A pre-set camera is called to capture an image containing the positioning result, and the image containing the positioning result is sent to a target object, wherein the target object is a remote manager or a server containing a visual recognition model.

[0015] In one embodiment of the present application, a signal strength-probability statistical model between multiple storage units and RFID base stations is constructed based on the historical inventory return signal data, including:

[0016] Get the reference signal strength of multiple storage units , wherein the reference signal strength is obtained based on experiments, is the storage unit number;

[0017] Calculate the standard deviation of the historical signal strength of multiple return signals returned to the RFID tag within each historical inventory time period in the historical inventory return signal data and the average , Indicates the time period sequence number;

[0018] The data that meets the target conditions in the historical inventory return signal data is retained to obtain target data, wherein the target conditions include: the standard deviation of the historical signal strength of multiple return signals The average standard deviation of the historical signal strength of multiple return signals is less than the preset standard deviation threshold The corresponding reference signal strength Deviation rate is less than the preset deviation rate threshold, where ;

[0019] Calculate the average strength of the historical return signal corresponding to each storage unit in the target data and intensity standard deviation , and based on the average intensity and the intensity standard deviation Constructing a probability density distribution function based on Gaussian distribution and a signal strength reference range that conforms to the three standard deviations, wherein there are multiple RFID base stations, and the number of probability density distribution functions and signal strength reference ranges of each storage unit is consistent with the number of RFID base stations;

[0020] A signal strength-probability statistical model is constructed based on the probability density distribution functions of multiple storage units and a signal strength reference range.

[0021] In one embodiment of the present application, a signal strength-probability statistical model is constructed based on the probability density distribution function of multiple storage units and the signal strength reference range, including:

[0022] constructing a signal strength data axis, and mapping the signal strength reference ranges of the plurality of storage units to the signal strength data axis;

[0023] The signal strength data axis is divided into multiple signal strength value intervals. For the signal strength reference range of one or more storage units included in any signal strength value interval, the initial probability is obtained by integrating the probability density distribution function of one or more storage units. ,in, is the sequence number of the signal strength value interval;

[0024] Based on the initial probability of one or more storage units in each signal strength value interval Calculate the probability of one or more memory cells ,in, The number of storage units covered by the signal strength value range;

[0025] Based on the probability of storage cells corresponding to multiple signal strength value intervals Construct a signal strength-probability statistical model.

[0026] In one embodiment of the present application, calculating the signal stability state and average signal strength of the RFID tag based on the strength of the return signals returned multiple times includes:

[0027] Calculating the standard deviation of the strengths of the return signals of the multiple returns;

[0028] When the standard deviation of the strength of the return signals of the multiple returns is less than a preset standard deviation threshold, the signal state is determined to be stable; otherwise, the signal state is determined to be unstable;

[0029] When the signal state is stable, an average value of the strengths of the return signals of the multiple returns is calculated.

[0030] In one embodiment of the present application, determining the probability that an RFID tag belongs to one or more storage units based on the average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model includes:

[0031] Determining a signal strength value interval to which multiple average signal strengths of the RFID tags in a stable state belong in the signal strength-probability statistical model, and obtaining multiple probabilities of one or more storage units within the signal strength value interval;

[0032] Multiple probabilities of the same storage unit are summed and the summed probabilities of the multiple storage units are normalized to obtain the probability that the RFID tag belongs to one or more storage units.

[0033] In one embodiment of the present application, positioning is performed based on the arrival time of return signals from multiple returns of an RFID tag in a stable state to obtain a preliminary positioning, including:

[0034] Taking the arrival time of the reference RFID base station as the reference time, calculate the time difference between the arrival time of other RFID base stations and the reference time;

[0035] A distance difference equation is constructed based on the base station coordinates of multiple RFID tags and the time difference, wherein the distance difference equation represents the corresponding relationship between the difference between the reference distance and the station tag distance and the arrival time difference, the reference distance is the distance between the reference RFID base station and the RFID tag, and the station tag distance is the distance between the other RFID base stations and the RFID tag;

[0036] Substitute multiple time differences into the distance difference equation, and fit the distance difference equation in combination with the least squares method to obtain the preliminary positioning of the RFID tag.

[0037] In one embodiment of the present application, positioning verification is performed in combination with the probability, the preliminary positioning, and the inventory information of the previous inventory cycle to obtain a positioning result of the RFID tag, including:

[0038] Acquire a pre-built three-dimensional model of a shelf, wherein the three-dimensional model of the shelf includes coordinate ranges of a plurality of storage units;

[0039] Determining the storage unit to which the RFID tag belongs in the current inventory cycle based on the three-dimensional model of the shelf and the preliminary positioning;

[0040] Comparing the storage unit to which the RFID tag belongs in the current inventory cycle with the storage unit to which it belongs in the previous inventory cycle, and if the storage unit to which the RFID tag belongs in the current inventory cycle is consistent with the storage unit to which it belongs in the previous inventory cycle, using the storage unit corresponding to the preliminary positioning as the storage unit to which it belongs in the current inventory cycle;

[0041] When the storage unit to which the RFID tag belongs in the current inventory cycle is inconsistent with the storage unit to which it belongs in the previous inventory cycle, the probability of the storage unit to which it belongs in the current inventory cycle is determined, and when the probability of the storage unit to which it belongs in the current inventory cycle is greater than a preset probability threshold, the storage unit to which it belongs corresponding to the preliminary positioning is used as the storage unit to which it belongs in the current inventory cycle; when the probability of the storage unit to which it belongs in the current inventory cycle is less than or equal to the preset probability threshold, the preliminary positioning is marked as uncertain, and the storage unit to which the RFID tag belongs in the previous inventory cycle is used as the storage unit to which it belongs in the current inventory cycle.

[0042] In one embodiment of the present application, warehouse management is performed based on the positioning results of RFID tags, including:

[0043] The inventory information of the current inventory cycle is constructed based on all RFID tags, and the inventory information of the current inventory cycle is compared with the inventory information of the previous inventory cycle to obtain the difference information; and the material warehousing, material delivery and material location change management are carried out based on the difference information.

[0044] This application also provides an unmanned warehouse management system with real-time inventory function, including:

[0045] an acquisition module, configured to acquire historical inventory return signal data of a plurality of storage units on shelves in a warehouse, and acquire the strength of return signals returned multiple times by the RFID tag within a current inventory cycle and the arrival time of the return signals returned multiple times, wherein the historical inventory return signal data includes historical signal strengths of multiple historical return signals returned to the RFID tag by the RFID tag within multiple historical inventory time periods, and the multiple returned return signals are generated based on radio signals transmitted multiple times by the RFID base station;

[0046] a model building module for building a signal strength-probability statistical model between multiple storage units and the RFID base station based on the historical inventory return signal data, wherein the signal strength-probability statistical model represents the corresponding relationship between different signal strengths of the return signal and the probability of belonging to the storage unit;

[0047] A feature extraction module is used to calculate the signal stability state and average signal strength of the RFID tag based on the strength of the return signal of the multiple returns;

[0048] a preliminary positioning module, configured to determine a probability that an RFID tag belongs to one or more storage units based on an average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model; and to perform positioning based on the arrival times of return signals from multiple returns of the RFID tag in the stable state to obtain a preliminary positioning;

[0049] The positioning verification module is used to perform positioning correction based on the probability that the RFID tag belongs to one or more storage units and the preliminary positioning to obtain the positioning result of the RFID tag; and to perform warehouse management based on the positioning result of the RFID tag.

[0050] The beneficial effects of the present invention are as follows: the unmanned warehouse management method and system with real-time inventory function of the present invention constructs a signal strength-probability statistical model through the return signal data generated during the historical inventory process, and in the current inventory cycle, performs stability verification on the return signal of each RFID tag, and performs preliminary positioning based on the arrival time difference and determines the probability of the storage unit to which it belongs when stability is confirmed. Finally, the preliminary positioning is verified in combination with the probability of the storage unit and the inventory information of the previous inventory cycle to avoid the situation where the positioning jumps and the verification cannot be performed, thereby failing to confirm whether the RFID tag has actually changed position. The present application can utilize the existing RFID system to achieve low-cost, high-reliability material positioning, thereby achieving real-time and accurate inventory information, and providing an information basis for the management of unmanned warehouses. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0052] Figure 1 This is an application scenario diagram of a digital intelligent archive management method based on machine vision shown in an embodiment of the present application;

[0053] Figure 2 This is a flow chart of a digital intelligent file management method based on machine vision shown in one embodiment of the present application;

[0054] Figure 3 This is a schematic diagram of a rough inspection process in an embodiment of the present application;

[0055] Figure 4 This is a structural diagram of an unmanned warehouse management system with real-time inventory function shown in one embodiment of the present application. DETAILED DESCRIPTION

[0056] The following describes the embodiments of the present invention through specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0057] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention. Therefore, the drawings only show the layers related to the present invention and are not drawn according to the number, shape and size ratio of the layers in actual implementation. In actual implementation, the type and number of each layer can be changed at will, and the layer layout may also be more complicated.

[0058] In the following description, numerous details are set forth to provide a more thorough explanation of the embodiments of the present invention; however, it is apparent to one skilled in the art that the embodiments of the present invention may be practiced without these specific details.

[0059] Figure 1 This is an application scenario diagram of an unmanned warehouse management method with real-time inventory function shown in an embodiment of the present application, such as Figure 1 As shown, the present application includes at least three ceiling-mounted RFID base stations (readers) for reading information of all RFID tags within the coverage area. The RFID tags in the present application are ultra-high frequency (UHF) passive RFID tags. The RFID base station can provide energy for the RFID tags by transmitting wirelessly, and then the RFID tags return the pre-written identity information to the RFID base station.

[0060] The metal frame storage shelves used in this application will interfere with the RFID echo signal, which can easily cause tag positioning to jump during the asset inventory process. Therefore, the following method is used for reliable positioning.

[0061] Figure 2 This is a flow chart of an unmanned warehouse management method with real-time inventory function shown in one embodiment of the present application. Figure 2 As shown: The unmanned warehouse management method with real-time inventory function of this embodiment may include steps S210 to S250:

[0062] S210, obtaining historical inventory return signal data for multiple storage units on shelves in the warehouse, and obtaining the strength of return signals returned by the RFID tag multiple times during a current inventory cycle and the arrival time of the multiple returned return signals, wherein the historical inventory return signal data includes historical signal strengths of multiple historical return signals returned to the RFID tag by the RFID tag within multiple historical inventory time periods, and the multiple returned return signals are generated based on radio signals transmitted multiple times by the RFID base station;

[0063] In this application, each time an automatic inventory is performed, a period of time is required to send radio frequency signals and analyze echo signals, and this period of time is the inventory cycle.

[0064] During each inventory count, the RFID base station transmits 10-20 radio frequency signals, each with an interval of 50ms. In addition, in order to suppress the multipath effect, this application adopts a coding and modulation scheme with strong anti-multipath capabilities, such as direct sequence spread spectrum (DSSS) or frequency hopping spread spectrum (FHSS). The return signal and the radio frequency signal are mapped based on the timestamp. The following table is a schematic table of RFID data in this embodiment, which includes various interference and instability situations. During the actual inventory process, if the warehouse is static, the signal quality will be better.

[0065] Table 1. RFID data sheet

[0066]

[0067] S220, constructing a signal strength-probability statistical model between the plurality of storage units and the RFID base station based on the historical inventory return signal data, wherein the signal strength-probability statistical model represents a corresponding relationship between different signal strengths of the return signal and the probability of belonging to the storage unit;

[0068] Under stable conditions, the RFID return signal is related to distance. The distance between the storage unit in the container and the RFID base station is fixed. Therefore, under stable signal conditions, the RSSI of the echo signal should be within a stable range. Therefore, based on the above logic, this application constructs a signal strength-probability statistical model based on historical data, including:

[0069] S221, obtaining reference signal strengths of multiple storage units , wherein the reference signal strength is obtained based on experiments, is the storage unit number;

[0070] Reference signal strength Through experimental testing, using the reference signal strength The historical echo data can be numerically screened to avoid the situation where the echo data appears stable when the RFID tag is blocked, but the value is significantly different from the value at other times, which leads to an excessively large reference range for subsequent analysis.

[0071] S222, calculating the standard deviation of the historical signal strengths of multiple return signals returned to the RFID tag within each historical inventory time period in the historical inventory return signal data and the average , Indicates the time period sequence number;

[0072] Standard deviation The formula for calculating the stability of the reaction signal and the standard deviation is:

[0073]

[0074] Where, Indicates the number of returned signals, Indicates the The strength of the return signal.

[0075] S223, retaining the data that meets the target condition in the historical inventory return signal data to obtain target data;

[0076] The target conditions include: (1) the standard deviation of the historical signal strength of multiple return signals Less than the preset standard deviation threshold; (2) the average standard deviation of the historical signal strength of multiple return signals The corresponding reference signal strength Deviation rate is less than the preset deviation rate threshold, where ;

[0077] When target condition (1) is met, it means that the return signal strength of the RFID tag within the historical inventory period is relatively stable. When target condition (2) is met, it means that the return signal strength of the RFID tag within the historical inventory period is also close to the reference value. Therefore, it can be used as the target data for subsequent analysis.

[0078] S224, calculating the average strength of the historical return signal corresponding to each storage unit in the target data and intensity standard deviation , and based on the average intensity and the intensity standard deviation Constructing a probability density distribution function based on Gaussian distribution and a signal strength reference range that conforms to the three standard deviations, wherein there are multiple RFID base stations, and the number of probability density distribution functions and signal strength reference ranges of each storage unit is consistent with the number of RFID base stations;

[0079] Step S223 calculates the average and standard deviation of the return signal strength within a count cycle. In this step, the average and standard deviation of the return signal strength corresponding to a single RFID tag are calculated. A probability density distribution function and a signal strength reference range that meets the three standard deviations are then constructed. The mathematical expression of the probability density function is:

[0080]

[0081] Where, represents the probability density, Indicates the value of the returned signal strength.

[0082] In addition, in order to make the total probability of the probability density function within three times the standard deviation be 1, it needs to be normalized.

[0083] The range of three times the standard deviation is the reference range of the return signal strength of the RFID tag in the storage unit under normal circumstances, which can be expressed mathematically as .

[0084] S225, constructing a signal strength data axis, and mapping the signal strength reference ranges of the plurality of storage units to the signal strength data axis;

[0085] Figure 3 is a schematic diagram of the intensity data axis in one embodiment of the present application, such as Figure 3 As shown, the probability density distribution function is obtained And the signal strength reference range, the signal strength reference range and the storage unit serial number are mapped to a two-dimensional coordinate system, the horizontal axis of the two-dimensional coordinate system is the intensity data axis, and the vertical axis is the serial number axis.

[0086] In addition, since this application includes multiple RFID base stations, their corresponding reference signal strength reference ranges are different, so this application only shows the signal strength data axis corresponding to one RFID base station. For signal strength data axes of other RFID base stations, please refer to Figure 3 Understand.

[0087] S226, dividing the signal strength data axis into a plurality of signal strength value intervals, integrating the signal strength reference range of one or more storage units contained in any signal strength value interval based on the probability density distribution function of the one or more storage units to obtain an initial probability ,in, is the sequence number of the signal strength value interval; Figure 3 As shown, there are overlapping parts in the multiple signal strength reference ranges in the two-dimensional coordinate system. The probability that the return signal strength of the RFID tag in the storage unit falls into the corresponding overlapping part can be obtained by integrating the probability density function. However, since there may be multiple reference ranges overlapping with each other in reality, it is difficult to define the overlapping sections. Therefore, the present application directly divides the signal strength data axis to obtain multiple signal strength intervals, and then integrates the probability density function of the signal strength reference range within each signal strength interval to obtain the probabilities of multiple signal strength intervals.

[0088] S227, based on the initial probability of one or more storage units in each signal strength value interval Calculate the probability of one or more memory cells, ,in, The number of storage units covered by the signal strength value range;

[0089] Finally, the initial probability of one or more storage units in each signal strength value interval is By redistributing the signal strength so that the total probability of each signal strength interval is 100%, we can obtain the probability of the storage unit within each signal strength interval. This probability reflects the probability that the return signal received by the RFID base station will belong to one or more storage units when the strength falls within one of the intervals.

[0090] S228, based on the probability of the storage unit corresponding to the multiple signal strength value intervals Construct a signal strength-probability statistical model.

[0091] S230, calculating the signal stability state and average signal strength of the RFID tag based on the strength of the return signals returned multiple times; after the signal strength-probability statistical model is constructed, before the signal strength-probability statistical model is combined, stability judgment and average signal strength calculation need to be performed, specifically including:

[0092] S231, calculating the standard deviation of the strength of the return signals returned multiple times; the calculation formula of the standard deviation is referred to above and will not be repeated here.

[0093] S232, when the standard deviation of the strengths of the return signals of the multiple returns is less than a preset standard deviation threshold, determining that the signal state is stable; otherwise, determining that the signal state is unstable;

[0094] S233: When the signal state is stable, calculate the average value of the strengths of the return signals returned multiple times.

[0095] The stability of the signal is evaluated by calculating the fluctuation (standard deviation) of the signal strength. ), indicating that the environmental interference is small and the path is single. The signal fluctuates violently (such as ), there may be multipath effects, metal interference or dynamic environment influences.

[0096] When the signal is stable, extract the representative value (average value) of the signal intensity for use in subsequent analysis.

[0097] In addition, in this application, for RFID tags in an unstable signal state, this application collects image information of the corresponding position to perform remote manual / artificial intelligence inventory, including:

[0098] (1) extracting identification information of an RFID tag in an unstable signal state, and extracting the positioning result of the identification information from the last inventory cycle containing the positioning result of the identification information;

[0099] (2) calling a pre-set camera to capture an image containing the positioning result, and sending the image containing the positioning result to a target object, wherein the target object is a remote manager or a server containing a visual recognition model.

[0100] S240, determining a probability that the RFID tag belongs to one or more storage units based on an average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model; and performing positioning based on arrival times of return signals from multiple returns of the RFID tag in the stable state to obtain a preliminary positioning;

[0101] To determine the probability, one simply needs to determine the signal strength value range within which the average signal strength of the RFID tag in a stable state falls within the signal strength-probability statistical model, and then determine the probability of one or more storage cells within that signal strength value range. For example, if interval A has a probability of containing two storage cells of 70% and 30%, respectively, then if the average signal strength of the RFID tag in a stable state falls within interval A, then the corresponding probabilities of the two storage cells are 70% and 30%, respectively.

[0102] However, since there are multiple RFID base stations, the average signal strength of the same RFID tag may also be multiple. For example, if there are three RFID base stations, the probability of falling into the storage unit is:

[0103] Table 2. Storage unit belonging probability table

[0104]

[0105] Multiple probabilities of the same storage unit are summed and the summed probabilities of the multiple storage units are normalized to obtain the probability that the RFID tag belongs to one or more storage units.

[0106] For example, if the total probability of storage unit 1 is 170%, the total probability of storage unit 2 is 100%, and the total probability of storage unit 3 is 30%, after normalization, the probability of storage unit 1 is 57%, the probability of storage unit 2 is 33%, and the probability of storage unit 3 is 10%.

[0107] The above process can be considered a positioning process based on statistics and big data. In this application, each RFID tag is also positioned based on the Time Difference of Arrival (TDOA) method, which specifically includes:

[0108] S241, using the arrival time of the reference RFID base station as a reference time, calculating the time difference between the arrival time of other RFID base stations and the reference time;

[0109] For example, this application includes three RFID base stations, namely BS-001, BS-002, and BS-003; BS-001 is used as the base RFID base station. Therefore, it is necessary to calculate the arrival time of each return signal received by BS-002 and BS-003. The arrival time of each return signal received by BS-001 The difference, that is and ;

[0110] S242: constructing a distance difference equation based on the base station coordinates of the multiple RFID tags and the time difference, wherein the distance difference equation represents a corresponding relationship between a difference between a reference distance and a station-tag distance and an arrival time difference, where the reference distance is the distance between the reference RFID base station and the RFID tag, and the station-tag distance is the distance between the other RFID base stations and the RFID tag;

[0111] Time Difference and The corresponding distance difference is and , is the electromagnetic wave speed; the corresponding distance difference equation is:

[0112]

[0113] Where, is the coordinate of the RFID tag, is the positioning coordinate of the benchmark RFID base station BS-001, are the coordinates of base station BS-002, The coordinates of base station BS-003.

[0114] S243: Substitute the multiple time differences into the distance difference equation, and fit the distance difference equation using the least square method to obtain a preliminary positioning of the RFID tag.

[0115] Finally, the multiple time differences calculated based on the multiple return signals obtained during the current inventory cycle and the known coordinates of the three RFID base stations are substituted into the above equation and fitted with the least squares method to obtain a preliminary positioning result. The least squares fitting process includes:

[0116] Construct the objective function. The goal in this embodiment is to minimize the residual sum of squares of all equations, that is:

[0117]

[0118] Where, represents the total number of equations (determined by the number of base station pairs), for example, there are two pairs in the above equation group. ik represents base station i in the k-th base station pair, jk represents base station j in the k-th base station pair;

[0119] Then, choose the initial estimate , the target position can be roughly estimated by geometric methods (such as hyperbola intersection).

[0120] In the initial estimate Perform a first-order Taylor expansion at , and get the residual function :

[0121]

[0122] in,

[0123]

[0124] The residual Substituting in, we can get the linear equation system:

[0125]

[0126] Where, is the Jacobian matrix, ;

[0127] Update the estimate by solving a system of linear equations, ;

[0128] Repeat the above process until the residual converges to less than the threshold, and the preliminary positioning can be obtained.

[0129] S250, performing positioning verification based on the probability, the preliminary positioning, and the inventory information of the previous inventory cycle to obtain a positioning result of the RFID tag; and performing warehouse management based on the positioning result of the RFID tag.

[0130] After obtaining the preliminary positioning calculated by the time difference of arrival method and the probabilistic positioning obtained by the data statistics method, the positioning verification can be carried out, which includes:

[0131] S251, obtaining a pre-built three-dimensional model of a shelf, wherein the three-dimensional model of the shelf includes coordinate ranges of a plurality of storage units;

[0132] The shelf's 3D model is pre-built, and the coordinates of the shelf's corner points are collected using the UWB tags of these corner points. This is then combined with the shelf's geometric information to complete the modeling. The storage unit is then deduced based on the corresponding dimensions and relative position information to the shelf's corner points.

[0133] S252, determining the storage unit to which the RFID tag belongs in the current inventory cycle based on the three-dimensional model of the shelf and the preliminary positioning;

[0134] By comparing the initial positioning with the shelf corner position, the serial number of the storage unit to which it belongs can be directly obtained. The positioning target in this application is the serial number of the storage unit to which it belongs. However, due to the unreliability of RFID positioning, this application introduces the inventory information of the previous inventory cycle and the results of probability statistics for verification, as follows:

[0135] S253, comparing the storage unit to which the RFID tag belongs in the current inventory count cycle with the storage unit to which it belongs in the previous inventory count cycle, and if the storage unit to which the RFID tag belongs in the current inventory count cycle is consistent with the storage unit to which it belongs in the previous inventory count cycle, using the storage unit corresponding to the preliminary positioning as the storage unit to which it belongs in the current inventory count cycle;

[0136] If the current RFID tag's location information matches the location information from the previous inventory cycle, it can be inferred that the item associated with the RFID tag has not been shipped out of the warehouse or its location has changed. Since the location of items rarely changes during storage, this situation is consistent with management practices and statistical probability is not used in this case.

[0137] S254: When the storage unit to which the RFID tag belongs in the current inventory cycle is inconsistent with the storage unit to which it belongs in the previous inventory cycle, the probability of the storage unit to which it belongs in the current inventory cycle is determined, and when the probability of the storage unit to which it belongs in the current inventory cycle is greater than a preset probability threshold, the storage unit to which it belongs corresponding to the preliminary positioning is used as the storage unit to which it belongs in the current inventory cycle; when the probability of the storage unit to which it belongs in the current inventory cycle is less than or equal to the preset probability threshold, the preliminary positioning is marked as uncertain, and the storage unit to which the RFID tag belongs in the previous inventory cycle is used as the storage unit to which it belongs in the current inventory cycle.

[0138] If the positioning information of the current RFID tag is inconsistent with the positioning information in the inventory information of the previous inventory cycle, then it is possible that the position of the item corresponding to the current RFID tag has changed. In order to exclude the position jump caused by the poor accuracy of the RFID positioning technology, this application uses the statistical probability extracted in the previous article for verification. That is, if the initial positioning of the RFID tag jumps from storage unit A to storage unit B, and at the same time, the probability of the signal strength belonging to the RFID tag is greater than 50%, then it can be verified that the position of the item corresponding to the current RFID tag has indeed changed. On the contrary, if the probability of the signal strength belonging to the RFID tag is small, then verification cannot be performed at this time. At this time, the inventory results of the previous inventory cycle are used and further verification can be waited for the next inventory cycle.

[0139] The above process can effectively avoid the situation where the item location information is incorrectly changed due to poor RFID positioning accuracy, and has strong reliability for location updates.

[0140] After completing the automatic inventory count for the current inventory cycle, the location and status information of all materials (whether they can be read and whether they have been shipped out) can be obtained. The inventory information for the current inventory cycle is then constructed based on all RFID tags. The inventory information for the current inventory cycle is compared with the inventory information for the previous inventory cycle to obtain the difference information. Based on this difference information, the inventory entry, shipment, and location change management of materials are carried out.

[0141] The unmanned warehouse management method with real-time inventory function of the present invention constructs a signal strength-probability statistical model through the return signal data generated during the historical inventory process. In the current inventory cycle, the stability of the return signal of each RFID tag is checked. When stability is confirmed, preliminary positioning based on the arrival time difference is performed and the probability of the storage unit to which it belongs is determined. Finally, the preliminary positioning is verified in combination with the probability of the storage unit and the inventory information of the previous inventory cycle to avoid the situation where the positioning jumps and the verification cannot be made, thereby making it impossible to confirm whether the RFID tag has actually changed position. The present application can utilize the existing RFID system to achieve low-cost, high-reliability material positioning, thereby realizing real-time and accurate inventory information, and providing an information basis for the management of unmanned warehouses.

[0142] This application also provides an unmanned warehouse management system with real-time inventory function, including:

[0143] an acquisition module, configured to acquire historical inventory return signal data of a plurality of storage units on shelves in a warehouse, and acquire the strength of return signals returned multiple times by the RFID tag within a current inventory cycle and the arrival time of the return signals returned multiple times, wherein the historical inventory return signal data includes historical signal strengths of multiple historical return signals returned to the RFID tag by the RFID tag within multiple historical inventory time periods, and the multiple returned return signals are generated based on radio signals transmitted multiple times by the RFID base station;

[0144] a model building module for building a signal strength-probability statistical model between multiple storage units and the RFID base station based on the historical inventory return signal data, wherein the signal strength-probability statistical model represents the corresponding relationship between different signal strengths of the return signal and the probability of belonging to the storage unit;

[0145] A feature extraction module is used to calculate the signal stability state and average signal strength of the RFID tag based on the strength of the return signal of the multiple returns;

[0146] a preliminary positioning module, configured to determine a probability that an RFID tag belongs to one or more storage units based on an average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model; and to perform positioning based on the arrival times of return signals from multiple returns of the RFID tag in the stable state to obtain a preliminary positioning;

[0147] The positioning verification module is used to perform positioning correction based on the probability that the RFID tag belongs to one or more storage units and the preliminary positioning to obtain the positioning result of the RFID tag; and to perform warehouse management based on the positioning result of the RFID tag.

[0148] The unmanned warehouse management system with real-time inventory function of the present invention constructs a signal strength-probability statistical model through the return signal data generated during the historical inventory process. In the current inventory cycle, the stability of the return signal of each RFID tag is checked. When stability is confirmed, preliminary positioning based on the arrival time difference is performed and the probability of the storage unit to which it belongs is determined. Finally, the preliminary positioning is verified in combination with the probability of the storage unit and the inventory information of the previous inventory cycle to avoid the situation where the positioning jumps and the verification cannot be made, thereby making it impossible to confirm whether the RFID tag has actually changed position. The present application can utilize the existing RFID system to achieve low-cost, high-reliability material positioning, thereby realizing real-time and accurate inventory information, and providing an information basis for the management of unmanned warehouses.

[0149] This embodiment further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, any one of the methods in this embodiment is implemented, wherein the method is the execution logic of this system.

[0150] This embodiment also provides an electronic terminal, including: a processor and a memory;

[0151] The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory, so that the terminal executes any one of the methods in this embodiment.

[0152] Regarding the computer-readable storage medium in this embodiment, those skilled in the art will appreciate that all or part of the steps in the aforementioned method embodiments can be implemented using hardware associated with the computer program. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps in the aforementioned method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0153] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used for communication, and the processor and the transceiver are used to run computer programs so that the electronic terminal executes the various steps of the above method.

[0154] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage.

[0155] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0156] In the above embodiments, although the present invention has been described in conjunction with specific embodiments of the present invention, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art based on the foregoing description. The embodiments of the present invention are intended to cover all such replacements, modifications and variations that fall within the broad scope of the appended claims.

[0157] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical principles disclosed herein are intended to be covered by the claims of the present invention.

Claims

1. An unmanned warehouse management method with real-time inventory function, characterized in that: Including steps: Obtaining historical inventory return signal data for a plurality of storage units on shelves within the warehouse, and obtaining the strength of return signals returned multiple times by the RFID tag within a current inventory cycle and the arrival times of the return signals returned multiple times, wherein the historical inventory return signal data includes historical signal strengths of multiple historical return signals returned to the RFID tag by the RFID tag within multiple historical inventory time periods, and the multiple returned return signals are generated based on radio signals transmitted multiple times by the RFID base station; Building a signal strength-probability statistical model between multiple storage units and the RFID base station based on the historical inventory return signal data, wherein the signal strength-probability statistical model characterizes the corresponding relationship between different signal strengths of the return signal and the probability of belonging to the storage unit; Calculating a signal steady state and an average signal strength of the RFID tag based on the strength of the return signals returned multiple times; Determining the probability that the RFID tag belongs to one or more storage units based on the average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model; and performing positioning based on the arrival time of the return signal of multiple returns of the RFID tag in the stable state to obtain a preliminary positioning; Performing positioning verification based on the probability, the preliminary positioning, and inventory information from a previous inventory cycle to obtain a positioning result of the RFID tag; and performing warehouse management based on the positioning result of the RFID tag.

2. The unmanned warehouse management method with real-time inventory function according to claim 1 is characterized in that: Also includes: Extracting identification information of an RFID tag in an unstable signal state, and extracting a positioning result of the identification information from a previous inventory cycle containing a positioning result of the identification information; A pre-set camera is called to capture an image containing the positioning result, and the image containing the positioning result is sent to a target object, wherein the target object is a remote manager or a server containing a visual recognition model.

3. The unmanned warehouse management method with real-time inventory function according to claim 1 is characterized in that: Building a signal strength-probability statistical model between a plurality of storage units and an RFID base station based on the historical inventory return signal data includes: Get the reference signal strength of multiple storage units , wherein the reference signal strength is obtained based on experiments, is the storage unit number; Calculate the standard deviation of the historical signal strength of multiple return signals returned to the RFID tag within each historical inventory time period in the historical inventory return signal data and the average , Indicates the time period sequence number; The data that meets the target conditions in the historical inventory return signal data is retained to obtain target data, wherein the target conditions include: the standard deviation of the historical signal strength of multiple return signals The average standard deviation of the historical signal strength of multiple return signals is less than the preset standard deviation threshold The corresponding reference signal strength Deviation rate is less than the preset deviation rate threshold, where ; Calculate the average strength of the historical return signal corresponding to each storage unit in the target data and intensity standard deviation , and based on the average intensity and the intensity standard deviation Constructing a probability density distribution function based on Gaussian distribution and a signal strength reference range that conforms to the three standard deviations, wherein there are multiple RFID base stations, and the number of probability density distribution functions and signal strength reference ranges of each storage unit is consistent with the number of RFID base stations; A signal strength-probability statistical model is constructed based on the probability density distribution functions of multiple storage units and a signal strength reference range.

4. The unmanned warehouse management method with real-time inventory function according to claim 3 is characterized in that: A signal strength-probability statistical model is constructed based on the probability density distribution functions of multiple storage units and the signal strength reference range, including: constructing a signal strength data axis, and mapping the signal strength reference ranges of the plurality of storage units to the signal strength data axis; The signal strength data axis is divided into multiple signal strength value intervals. For the signal strength reference range of one or more storage units included in any signal strength value interval, the initial probability is obtained by integrating the probability density distribution function of one or more storage units. ,in, is the sequence number of the signal strength value interval; Based on the initial probability of one or more storage units in each signal strength value interval Calculate the probability of one or more memory cells , ,in, The number of storage units covered by the signal strength value range; Based on the probability of storage cells corresponding to multiple signal strength value intervals Construct a signal strength-probability statistical model.

5. The unmanned warehouse management method with real-time inventory function according to claim 1 is characterized in that: Calculating the signal stability state and average signal strength of the RFID tag based on the strength of the return signals returned multiple times includes: Calculating the standard deviation of the strengths of the return signals of the multiple returns; When the standard deviation of the strength of the return signals of the multiple returns is less than a preset standard deviation threshold, the signal state is determined to be stable; otherwise, the signal state is determined to be unstable; When the signal state is stable, an average value of the strengths of the return signals of the multiple returns is calculated.

6. The unmanned warehouse management method with real-time inventory function according to claim 4 is characterized in that: Determining the probability that the RFID tag belongs to one or more storage units based on the average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model includes: Determining a signal strength value interval to which multiple average signal strengths of the RFID tags in a stable state belong in the signal strength-probability statistical model, and obtaining multiple probabilities of one or more storage units within the signal strength value interval; Multiple probabilities of the same storage unit are summed and the summed probabilities of the multiple storage units are normalized to obtain the probability that the RFID tag belongs to one or more storage units.

7. The unmanned warehouse management method with real-time inventory function according to claim 1 is characterized in that: Positioning is performed based on the arrival time of multiple return signals from the RFID tag in a stable state to obtain a preliminary positioning, including: Taking the arrival time of the reference RFID base station as the reference time, calculate the time difference between the arrival time of other RFID base stations and the reference time; A distance difference equation is constructed based on the base station coordinates of multiple RFID tags and the time difference, wherein the distance difference equation represents the corresponding relationship between the difference between the reference distance and the station tag distance and the arrival time difference, the reference distance is the distance between the reference RFID base station and the RFID tag, and the station tag distance is the distance between the other RFID base stations and the RFID tag; Substitute multiple time differences into the distance difference equation, and fit the distance difference equation in combination with the least squares method to obtain the preliminary positioning of the RFID tag.

8. The unmanned warehouse management method with real-time inventory function according to claim 1 is characterized in that: The positioning result of the RFID tag is obtained by combining the probability, the preliminary positioning, and the inventory information of the previous inventory cycle to perform positioning verification, including: Acquire a pre-built three-dimensional model of a shelf, wherein the three-dimensional model of the shelf includes coordinate ranges of a plurality of storage units; Determining the storage unit to which the RFID tag belongs in the current inventory cycle based on the three-dimensional model of the shelf and the preliminary positioning; Comparing the storage unit to which the RFID tag belongs in the current inventory cycle with the storage unit to which it belongs in the previous inventory cycle, and if the storage unit to which the RFID tag belongs in the current inventory cycle is consistent with the storage unit to which it belongs in the previous inventory cycle, using the storage unit corresponding to the preliminary positioning as the storage unit to which it belongs in the current inventory cycle; When the storage unit to which the RFID tag belongs in the current inventory cycle is inconsistent with the storage unit to which it belongs in the previous inventory cycle, the probability of the storage unit to which it belongs in the current inventory cycle is determined, and when the probability of the storage unit to which it belongs in the current inventory cycle is greater than a preset probability threshold, the storage unit to which it belongs corresponding to the preliminary positioning is used as the storage unit to which it belongs in the current inventory cycle; when the probability of the storage unit to which it belongs in the current inventory cycle is less than or equal to the preset probability threshold, the preliminary positioning is marked as uncertain, and the storage unit to which the RFID tag belongs in the previous inventory cycle is used as the storage unit to which it belongs in the current inventory cycle.

9. The unmanned warehouse management method with real-time inventory function according to claim 1, characterized in that: Warehouse management based on the positioning results of RFID tags, including: The inventory information of the current inventory cycle is constructed based on all RFID tags, and the inventory information of the current inventory cycle is compared with the inventory information of the previous inventory cycle to obtain the difference information; and the material warehousing, material delivery and material location change management are carried out based on the difference information.

10. An unmanned warehouse management system with real-time inventory function is characterized by: include: an acquisition module, configured to acquire historical inventory return signal data of a plurality of storage units on shelves in a warehouse, and acquire the strength of return signals returned multiple times by the RFID tag within a current inventory cycle and the arrival time of the return signals returned multiple times, wherein the historical inventory return signal data includes historical signal strengths of multiple historical return signals returned to the RFID tag by the RFID tag within multiple historical inventory time periods, and the multiple returned return signals are generated based on radio signals transmitted multiple times by the RFID base station; a model building module for building a signal strength-probability statistical model between multiple storage units and the RFID base station based on the historical inventory return signal data, wherein the signal strength-probability statistical model represents the corresponding relationship between different signal strengths of the return signal and the probability of belonging to the storage unit; A feature extraction module is used to calculate the signal stability state and average signal strength of the RFID tag based on the strength of the return signal of the multiple returns; a preliminary positioning module, configured to determine a probability that an RFID tag belongs to one or more storage units based on an average signal strength of the RFID tag in a stable state and the signal strength-probability statistical model; and to perform positioning based on the arrival times of return signals from multiple returns of the RFID tag in a stable state to obtain a preliminary positioning; The positioning verification module is used to perform positioning correction based on the probability that the RFID tag belongs to one or more storage units and the preliminary positioning to obtain the positioning result of the RFID tag; and to perform warehouse management based on the positioning result of the RFID tag.

Images