Warehouse material real-time positioning deviation correction method, device, system and storage medium

Abstract

This invention relates to a method, apparatus, system, and storage medium for real-time positioning and correction of stored materials. The method includes: acquiring positioning tag signals and material images of stored materials; performing trajectory calculations on the positioning tag signals and material images to generate a fused positioning trajectory; calculating regional activity on the fused positioning trajectory based on preset warehouse zoning information to obtain real-time regional activity; performing material access analysis on the real-time regional activity and historical access information of the stored materials to obtain dynamic access information; performing trajectory offset analysis on the fused positioning trajectory and the real-time regional activity based on a preset desired storage coordinate set to obtain in-warehouse correction information; and performing handling and storage planning based on the dynamic access information, the real-time regional activity, and the in-warehouse correction information to obtain in-warehouse handling tasks. This invention can generate a high-precision fused positioning trajectory by fusing positioning tag signals and material images.

Description

Technical Field

[0001] This invention relates to the technical field of warehousing, and in particular to a method, apparatus, system, and storage medium for real-time positioning and correction of stored materials. Background Technology

[0002] With the deepening of intelligent transformation in the manufacturing industry, smart warehousing integrating the Internet of Things and visual perception has become key to improving logistics efficiency and accuracy. Current warehouse material positioning methods rely solely on single wireless positioning, which is susceptible to interference in complex warehouse environments, leading to unstable or lost positioning data and difficulty in providing consistently high-precision real-time location feedback. Furthermore, existing methods only identify the location of materials, failing to determine if they have deviated from their intended position. Their task scheduling is often relatively static, failing to be closely integrated with the real-time workload of each warehouse zone and the dynamic frequency of material access, thus preventing the generation of corresponding corrective actions and limiting further improvements in overall warehouse operational efficiency. Summary of the Invention

[0003] The main objective of this invention is to provide a method, apparatus, system, and storage medium for real-time positioning and correction of warehouse materials. This invention can generate a high-precision fused positioning trajectory by fusing positioning tag signals and material images. It can effectively overcome the interference problem of a single wireless signal in a complex environment and provide continuous and stable real-time position feedback, thereby laying a reliable data foundation for subsequent accurate correction.

[0004] To achieve the above objectives, the present invention provides a method for real-time positioning and correction of stored materials, comprising: The system collects location tag signals and material images of stored materials, performs trajectory calculations on the location tag signals and material images, and generates a fused location trajectory. Based on the preset warehouse partition information, the fused positioning trajectory is used to calculate the regional activity to obtain the real-time regional activity. The real-time regional activity is then combined with the historical storage and retrieval information of the stored materials to perform material storage and retrieval analysis to obtain dynamic storage and retrieval information. Based on a preset expected storage coordinate set, trajectory offset analysis is performed on the fused positioning trajectory and the real-time regional activity to obtain in-database correction information; Based on the dynamic access information, the real-time regional activity, and the in-database correction information, a transport and storage plan is developed to obtain the in-database transport task.

[0005] Further, the process of collecting the location tag signals and material images of the stored materials, and performing trajectory calculations on the location tag signals and material images to generate a fused location trajectory includes: Extract the first timestamp and the first coordinate sequence from the positioning tag signal, and identify the corresponding second coordinate sequence from the material image; Based on the first timestamp, the first coordinate sequence and the second coordinate sequence are aligned and paired to form a set of coordinate pairs; A fusion decision is made for each coordinate pair in the coordinate pair set based on a preset coordinate distance threshold. When the spatial distance between two coordinate points in the coordinate pair is less than the preset coordinate distance threshold, the coordinate pair is fused and calculated to obtain the positioning trajectory point; otherwise, the corresponding coordinate point in the first coordinate sequence is used as the interpolation trajectory point. By integrating all the positioning trajectory points and the interpolated trajectory points, the fused positioning trajectory is obtained.

[0006] Further, the step of calculating the regional activity of the fused positioning trajectory based on preset warehouse partition information to obtain real-time regional activity includes: Based on the location tag signal, obtain the warehouse partition information of the warehouse where the stored materials are located; Based on the warehouse partition information, identify the warehouse partition to which each positioning coordinate in the fused positioning trajectory belongs; The number of coordinates of the location coordinates existing in each of the warehouse partitions within a preset time window is counted, and the number of coordinates is used to calculate the partition activity to obtain the initial activity level. The initial activity level is compared with the activity threshold in the warehouse partition information. If the initial activity level is higher than the activity threshold, a corresponding active status identifier is generated based on the warehouse partition; otherwise, a corresponding inactive status identifier is generated based on the warehouse partition. The real-time regional activity level is obtained by integrating all the active status identifiers, the inactive status identifiers, and the initial activity level.

[0007] Further, the step of performing material access analysis by combining the real-time regional activity level with the historical access information of the stored materials to obtain dynamic access information includes: Based on the real-time regional activity and the historical access information, the target warehouse partition that is currently active and has historical access records is identified. Based on the target warehouse partition, extract the corresponding number of historical access events from the historical access information; The number of access events occurring in each of the target warehouse partitions in the fused positioning trajectory is counted according to the preset time window. Calculate the ratio of the actual number of access events to the historical number of access events for each target warehouse partition to obtain the access change rate; The dynamic access information is generated by integrating all the access change rates and the number of access events.

[0008] Furthermore, the trajectory offset analysis based on the fused positioning trajectory and the real-time regional activity, using a preset expected storage coordinate set, yields in-database correction information, including: Extract the current positioning coordinate sequence from the fused positioning trajectory, and extract the target expected coordinates corresponding to the stored materials from the expected storage coordinate set; Calculate the coordinate deviation between the current positioning coordinate sequence and the target desired coordinates; The coordinate deviation is compared with a preset deviation threshold. If the coordinate deviation is higher than the deviation threshold, it is determined that the storage material has been misaligned, and an offset identifier is generated. If the coordinate deviation is not higher than the deviation threshold, the positioning is determined to be normal and the process ends. Based on the real-time regional activity, traverse the free storage coordinates around the stored materials and calculate the storage distance between each free storage coordinate and the target desired coordinate. The free storage coordinate with the smallest storage distance is selected as the recommended correction coordinate. The offset identifier, the coordinate deviation, and the recommended correction coordinate are combined to generate the in-library correction information.

[0009] Further, the step of performing transport and storage planning based on the dynamic access information, the real-time regional activity, and the in-database correction information to obtain the in-database transport task includes: The dynamic access information is matched with a preset handling equipment status table to obtain available handling equipment information; Traverse each material identifier to be corrected in the correction information in the library, extract the current position coordinates corresponding to the material identifier to be corrected from the fused positioning trajectory, and search for the target storage coordinates corresponding to the material identifier to be corrected from the expected storage coordinate set; Based on the current location coordinates and the target storage coordinates, a spatial offset vector is calculated. The spatial offset vector, the material identifier to be corrected, and the target storage coordinates are combined to obtain a basic handling instruction. Based on the real-time regional activity, all the basic transport instructions are prioritized to obtain a priority label sequence; Based on the priority tag sequence, the available handling equipment information and the basic handling instructions are used to plan the handling tasks, thereby obtaining the in-warehouse handling tasks.

[0010] Further, the step of planning the handling task based on the available handling equipment information and the basic handling instructions according to the priority tag sequence to obtain the in-warehouse handling task includes: Based on the priority tag sequence, the basic handling instruction is bound to the available handling equipment information to obtain the equipment to be assigned; Based on the basic handling instructions, a path is planned for the equipment to be assigned, and the initial handling path obtained is compared with the preset planned task path for path conflict detection. If conflicting paths exist, route planning and conflict detection are performed again until the conflict is resolved. When there are no conflicting paths, all the basic transport instructions, the devices to be assigned, and the initial transport path are bound together to obtain the in-warehouse transport task.

[0011] The present invention also provides a real-time positioning and correction device for warehouse materials, applied to any of the above-described real-time positioning and correction methods for warehouse materials, comprising: The acquisition module is used to acquire the positioning tag signals and material images of the stored materials, perform trajectory calculation on the positioning tag signals and the material images, and generate a fused positioning trajectory. The analysis module is used to perform regional activity calculation on the fused positioning trajectory based on preset warehouse partition information to obtain real-time regional activity, and to perform material access analysis on the real-time regional activity and the historical access information of the stored materials to obtain dynamic access information. The association module is used to perform trajectory offset analysis on the fused positioning trajectory and the real-time regional activity based on a preset expected storage coordinate set, so as to obtain in-database correction information. The processing module is used to perform transport and storage planning based on the dynamic access information, the real-time regional activity, and the in-database correction information to obtain in-database transport tasks.

[0012] The present invention also provides a real-time positioning and correction system for warehouse materials, comprising: Memory, used to store programs; A processor is used to execute the program to implement each step of the real-time positioning and correction method for warehouse materials as described in any of the above-mentioned methods.

[0013] The present invention also provides a storage medium storing computer instructions for causing a computer to perform any of the methods described above.

[0014] The present invention provides a method, apparatus, system, and storage medium for real-time positioning and correction of stored materials, which has the following beneficial effects: By fusing positioning tag signals with material images to generate high-precision fused positioning trajectories, interference problems caused by single wireless signals in complex environments can be effectively overcome, providing continuous and stable real-time position feedback, thus laying a reliable data foundation for subsequent accurate deviation correction. By automatically comparing and analyzing real-time positioning data with preset desired coordinates, it is possible to accurately and automatically identify whether there is a deviation in the actual position of the material, achieving a leap from simple positioning to intelligent diagnosis of material status. By comprehensively considering real-time area activity, dynamic access information, and deviation analysis results, the generation of deviation correction and handling tasks can be closely aligned with the dynamic workload of the warehouse and the real-time access needs of materials, realizing a transformation from static, fixed task scheduling to dynamic, adaptive resource planning. Attached Figure Description

[0015] Figure 1 This is a flowchart of a real-time positioning and correction method for warehouse materials provided by the present invention; Figure 2 This is a structural diagram of a real-time positioning and correction device for warehouse materials provided by the present invention; Figure 3 This is a structural diagram of a real-time positioning and correction system for warehouse materials provided by the present invention.

[0016] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0019] Reference Figure 1 As shown, the present invention provides a method for real-time positioning and correction of stored materials, comprising: Step S1: Collect the positioning tag signal and material image of the stored materials, perform trajectory calculation on the positioning tag signal and the material image, and generate a fused positioning trajectory; Specifically, the positioning tag signals originate from UWB (Ultra-Wideband) or RFID active tags attached to material carriers (such as pallets or turnover boxes). These signals are received by a fixed base station network deployed within the warehouse, which calculates a series of three-dimensional coordinate points with timestamps, forming the first coordinate sequence. Parallel-deployed visual acquisition units (such as industrial cameras fixed in aisles or on shelves) simultaneously capture image streams containing materials and their identifiers (such as QR codes or specific patterns). By processing the image stream in real time, while identifying the material, the system uses camera calibration parameters and image geometric relationships to calculate another set of coordinate points for the material in the global coordinate system, forming the second coordinate sequence.

[0020] Based on the hardware clock synchronization mechanism or application layer timestamp, the first coordinate sequence and the second coordinate sequence are aligned on the time axis. Using the timestamp as an index, the first coordinate sequence and the second coordinate sequence at similar times are paired to form a set of coordinate pairs. For each coordinate pair, the Euclidean distance between the two points is calculated. If this distance is less than a fusion threshold preset based on sensor characteristics and installation geometry, the two data streams are considered mutually corroborating and valid. In this case, a weighted average is calculated using preset weights to output the positioning trajectory points. If visual coordinates are missing due to occlusion or other reasons, or if the pairing distance exceeds the threshold indicating significant data inconsistencies, the system automatically downgrades, outputting the first coordinate sequence at the current time as the interpolated trajectory points to ensure trajectory continuity. By traversing all time slices and integrating all positioning trajectory points and interpolated trajectory points, a fused positioning trajectory is finally generated that can utilize high-precision visual calibration and compensate for visual blind spots by leveraging the continuity of wireless signals.

[0021] Step S2: Calculate the regional activity of the fused positioning trajectory based on the preset warehouse partition information to obtain the real-time regional activity. Analyze the real-time regional activity with the historical storage and retrieval information of the stored materials to obtain dynamic storage and retrieval information. Specifically, the warehouse is pre-divided into multiple zones, such as by aisles or functional areas (picking area, storage area, replenishment area). This predefined warehouse zone information, containing the spatial boundary coordinates of each zone, is loaded. The output fused positioning trajectory is continuously scanned, mapping each coordinate point in the trajectory to its specific warehouse zone based on its spatial location. Within a sliding time window (e.g., the last 5 minutes), the number of all trajectory points appearing in each zone is accumulated. This number directly reflects the frequency of material movement within that zone, constituting the initial activity level. For decision-making purposes, this initial activity level is typically compared to a preset static or dynamic baseline threshold for that zone. If it exceeds the threshold, the zone is marked as high-activity; otherwise, it is marked as normal or "low-activity." Finally, a real-time area activity mapping table is output.

[0022] This stage introduces a time dimension, linking the aforementioned real-time snapshots with historical storage and retrieval information of warehouse materials. Historical storage and retrieval information comes from the Warehouse Management System (WMS) database, recording details such as the entry and exit times, quantities, and involved zones for each material in past cycles. First, based on the real-time area activity mapping table, zones currently marked as highly active are selected as key target warehouse zones. Next, for these target zones, the number of historical storage and retrieval events within a comparable time period (e.g., a 5-minute window of equal length) is extracted from the historical storage and retrieval information as a baseline. Simultaneously, backtracking analysis is performed within the current sliding time window, fusing the number of events that actually occurred in these target zones and conform to storage and retrieval behavior characteristics (e.g., the trajectory start or end point is in front of the shelf) within the current target zone. By calculating the ratio of the actual number of storage and retrieval events to the number of historical storage and retrieval events, a dynamic storage and retrieval change rate is obtained. If this ratio is significantly greater than 1, it indicates an abnormally high level of storage and retrieval activity in that zone, potentially associated with urgent orders or concentrated picking operations. Integrating the storage and retrieval change rates and real-time event counts for all target zones generates dynamic storage and retrieval information.

[0023] Step S3: Based on the preset expected storage coordinate set, perform trajectory offset analysis on the fused positioning trajectory and the real-time regional activity to obtain in-database correction information; Specifically, it relies on a preset expected storage coordinate set, which is a digital mapping of the theoretical storage location of materials (such as the precise center coordinates of a specific shelf or storage location) recorded in the warehouse management system (WMS). The coordinate sequence of the specified material at the current moment and over a recent period is extracted from the fused positioning trajectory as observation values. Based on the material's unique identifier, its corresponding target expected coordinates are retrieved from the expected storage coordinate set as theoretical benchmark values. The deviation between the observed coordinate sequence and the target expected coordinates is calculated. Offset determination is not a simple static threshold comparison but incorporates dynamic adaptability. The calculated coordinate deviation is compared with a preset base deviation threshold. This base deviation threshold can be set based on the overall accuracy of the warehouse positioning system. If the real-time area activity indicator of the current material's zone is high, it indicates that the area is busy and dynamic, and slight positioning errors or temporary placements are normal. Therefore, a more lenient dynamic threshold is used for comparison to avoid generating too many unnecessary correction alarms in busy areas. Only when the coordinate deviation continuously exceeds this dynamic threshold is it finally determined that the material has experienced a positioning offset requiring processing, and an offset identifier is generated.

[0024] After confirming the offset, actionable correction suggestions are generated. At this point, the warehouse digital map is accessed, and a search is conducted within a certain physical radius of the target desired coordinates for currently available storage coordinates. During the search, coordinates within zones marked as highly active in real-time are actively avoided to prevent the correction task from exacerbating regional congestion. The distances between all eligible available coordinates and the target desired coordinates are calculated, and the coordinate with the smallest distance is selected as the recommended correction coordinate. The offset identifier, coordinate deviation, and recommended correction coordinates are encapsulated to generate a complete in-warehouse correction message.

[0025] Step S4: Based on the dynamic access information, the real-time regional activity, and the in-database correction information, perform a transport and storage plan to obtain an in-database transport task.

[0026] Specifically, each record in the correction information within the warehouse is parsed into a correction and handling request to be executed, including material identification, current location (derived from the fused positioning trajectory), and recommended correction coordinates. Simultaneously, the access change rate of each material is obtained from dynamic access information, and the activity status of each partition is obtained from real-time area activity. The primary task of planning is task priority ranking, establishing a comprehensive evaluation model whose input variables include: coordinate deviation in the correction information (larger deviation generally indicates higher priority), access change rate in the dynamic access information (high change rate indicates urgent material demand), and the activity status of the partition where the target coordinates are located. This model assigns a dynamic priority label to each correction request through preset rules or weights. For example, a material with a large deviation, low activity in its partition, but a rapidly increasing access change rate may be assigned the highest priority to ensure rapid restoration of its accessibility. Subsequently, the task-resource matching and conflict resolution phase begins, reading information on available handling equipment (such as AGV status and location). Based on dynamic priority, the system attempts to assign high-priority tasks to the nearest available device with the lightest current load. During assignment, a proactive conflict simulation is performed: based on the warehouse map, device speed, and already assigned tasks, the system simulates whether the task's path, if executed, will overlap with or deadlock planned tasks in time and space. If a conflict is detected, a conflict resolution mechanism is triggered. Possible strategies include reselecting a suboptimal device for the current task, adjusting its path, or slightly delaying its execution sequence. This process iterates until a conflict-free device allocation and path scheme is found for all tasks. All decision results are instantiated into specific control instructions. Each in-warehouse handling task instruction explicitly specifies the device number, the coordinates of the start and end points of the handling action, the preset safe path point sequence, and the task priority.

[0027] This invention provides a real-time positioning and correction method for warehouse materials. By fusing positioning tag signals and material images to generate a high-precision fused positioning trajectory, it effectively overcomes the interference problem of single wireless signals in complex environments, providing continuous and stable real-time position feedback, thus laying a reliable data foundation for subsequent accurate correction. By automatically comparing and analyzing real-time positioning data with preset desired coordinates, it can accurately and automatically identify whether there is a deviation in the actual position of the material, achieving a leap from simple positioning to intelligent diagnosis of material status. By comprehensively considering real-time area activity, dynamic access information, and deviation analysis results, the generation of correction and handling tasks can closely align with the dynamic workload of the warehouse and the real-time access needs of materials, realizing a transformation from static, fixed task scheduling to dynamic, adaptive resource planning.

[0028] In one embodiment, the step of collecting the location tag signal and material image of the stored materials, and performing trajectory calculation on the location tag signal and the material image to generate a fused location trajectory includes: Extract the first timestamp and the first coordinate sequence from the positioning tag signal, and identify the corresponding second coordinate sequence from the material image; Based on the first timestamp, the first coordinate sequence and the second coordinate sequence are aligned and paired to form a set of coordinate pairs; Specifically, the algorithm receives a first coordinate sequence and a second coordinate sequence. The alignment process is based solely on the timestamp. For each coordinate point Pi (with a timestamp Tiu) in the first coordinate sequence, the algorithm searches for a coordinate point Vj in the second coordinate sequence whose timestamp is closest to Tiu. The search range is typically limited to a maximum time tolerance window preset according to sensor characteristics, such as ±50 milliseconds. If such a coordinate point Vj is found within the window, the algorithm successfully creates a coordinate pair (Pi, Vj) and adds it to the coordinate pair set. If there are no data points in the second coordinate sequence within the corresponding time window (e.g., visual recognition failure), Pi is considered an unpaired point and will be handled specially in subsequent steps. Conversely, if a coordinate point Vj in the second coordinate sequence cannot find a match within the corresponding time window of the first sequence, that coordinate point Vj is temporarily set aside. One pairing strategy employs interpolation: when a point Tiu in the first sequence lies between two points Vj (time Tjv) and Vj+1 (time Tj+1v) in the second sequence, the estimated visual coordinates of Tiu at time Tjv and Tj+1v can be calculated linearly using interpolation, thus pairing it with Pi. After completion, the coordinate pair set contains a large number of temporally aligned (first coordinate, second coordinate) pairs, as well as a list of unpaired points.

[0029] A fusion decision is made for each coordinate pair in the coordinate pair set based on a preset coordinate distance threshold. When the spatial distance between two coordinate points in the coordinate pair is less than the preset coordinate distance threshold, the coordinate pair is fused and calculated to obtain the positioning trajectory point; otherwise, the corresponding coordinate point in the first coordinate sequence is used as the interpolation trajectory point. Specifically, for coordinate pairs marked as fusionable, a fusion calculation is performed. Based on a prior assessment of the characteristics of the two types of sensors: the second coordinate sequence has higher instantaneous absolute accuracy under conditions of no obstruction and successful recognition, and is therefore assigned a higher fusion weight; while the first coordinate sequence has better temporal continuity and penetration, but contains more high-frequency noise, and is assigned a relatively lower weight. Coordinates Pi and Vj are weighted and averaged according to their respective weights to calculate a new coordinate point Fi. This point is the positioning trajectory point. For coordinate pairs marked as non-fusionable, the above fusion calculation is not performed, and a degradation processing strategy is adopted instead. The coordinate point Pi belonging to the first coordinate sequence is extracted from the coordinate pair, adopted as is, and used as the trajectory point at the current moment. It is also given a special identifier and marked as the interpolation trajectory point.

[0030] By integrating all the positioning trajectory points and the interpolated trajectory points, the fused positioning trajectory is obtained.

[0031] The method provided in this embodiment generates a fused positioning trajectory by calculating the trajectory of the positioning tag signal and the material image. It can comprehensively utilize the continuity of wireless signals and the high precision of visual recognition, effectively overcoming the limitations of single signals being susceptible to interference or obstruction in complex warehousing environments. This provides a stable, continuous, and more accurate material movement path, laying a reliable data foundation for all subsequent analyses. By aligning and pairing the two coordinate sequences based on timestamps and performing fusion judgment and differentiation processing according to preset thresholds, it can intelligently identify and integrate valid data. At the same time, it uses backup data sources to fill gaps, thereby ensuring the integrity and continuity of the location information output while maintaining trajectory accuracy.

[0032] In one embodiment, the step of calculating the regional activity of the fused positioning trajectory based on preset warehouse partition information to obtain real-time regional activity includes: Based on the location tag signal, obtain the warehouse partition information of the warehouse where the stored materials are located; Based on the warehouse partition information, identify the warehouse partition to which each positioning coordinate in the fused positioning trajectory belongs; The number of coordinates of the location coordinates existing in each of the warehouse partitions within a preset time window is counted, and the number of coordinates is used to calculate the partition activity to obtain the initial activity level. Specifically, a time window is maintained, for example, a window of length T, whose end is always aligned with the current time, and whose beginning is the current time minus T. As time progresses, this window continuously slides forward, incorporating new data points while discarding older ones. For each warehouse partition, a cache queue of location coordinates belonging to that partition is maintained in real time, with each point in the queue carrying its precise timestamp. When performing statistics, the cache queues of each partition are first cleared, removing all data points whose timestamps are earlier than the start time of the current sliding window. The number of remaining data points in the cache queue of each partition is counted. This count value represents the number of coordinates for that partition within the current time window. The number of coordinates for a partition is directly related to the frequency of material movement and dwell density within that partition. Partition activity calculation is then performed. Normalization is achieved by introducing the time window length, for example, by calculating the average number of coordinates per unit time (number of coordinates / time window length T), to eliminate numerical differences caused by different window length settings, ensuring consistency of the activity index across different configurations. After this calculation, each partition obtains an initial activity value.

[0033] The initial activity level is compared with the activity threshold in the warehouse partition information. If the initial activity level is higher than the activity threshold, a corresponding active status identifier is generated based on the warehouse partition; otherwise, a corresponding inactive status identifier is generated based on the warehouse partition. The real-time regional activity level is obtained by integrating all the active status identifiers, the inactive status identifiers, and the initial activity level.

[0034] Specifically, three elements are generated for each warehouse partition: a unique partition identifier, an initial activity value, and an active or inactive status identifier. This is done on a partition-by-partition basis. For each defined partition in the warehouse, a structured data object or record is created. This record contains at least three core fields: partition ID (as the primary key), initial activity value (floating-point or integer type), and region status identifier (string or enumeration type). All partitions are iterated through, and the corresponding three elements are filled into the corresponding fields of the record, thus generating a complete status record for that partition. All partition status records are collected to form a data set in the form of a list, map, or array; this set represents the real-time region activity.

[0035] The method provided in this embodiment acquires predefined warehouse partition information based on positioning tag signals, providing an accurate and consistent spatial reference framework for all subsequent area analyses, ensuring that the area activity calculation strictly corresponds to the actual physical layout of the warehouse. By identifying and classifying each coordinate point in the fused positioning trajectory to a specific warehouse partition, continuous movement trajectory data can be transformed into a discrete event sequence at the business area level, laying a data foundation for refined regional statistical analysis. By statistically analyzing the number of coordinates in each partition within a preset time window and performing activity calculations to obtain the initial activity level, the raw positioning data can be transformed into a quantitative indicator that dynamically reflects the frequency of material movement in each area, thereby objectively characterizing the real-time operational intensity of each part of the warehouse.

[0036] In one embodiment, the step of performing material access analysis by comparing the real-time regional activity with the historical access information of the stored materials to obtain dynamic access information includes: Based on the real-time regional activity and the historical access information, the target warehouse partition that is currently active and has historical access records is identified. Based on the target warehouse partition, extract the corresponding number of historical access events from the historical access information; The number of access events occurring in each of the target warehouse partitions in the fused positioning trajectory is counted according to the preset time window. Specifically, when the fused positioning trajectory shows that a material enters a specific sub-area of ​​a target partition (such as the coordinate range near a row of shelves) and remains stationary within that area for more than a brief threshold (such as 3 seconds), and then the trajectory starts moving again and leaves the area, this series of actions can be defined as a complete storage / retrieval event (which may be storage or retrieval). An event counter is set for each target warehouse partition, and incoming trajectory data is continuously monitored. For each material trajectory, it is determined whether its current coordinates fall within a target partition. If so, a preset event detection rule is applied to analyze the movement state (speed, position change) of the trajectory point in the current and preceding brief time periods. Once a trajectory segment is determined by the rule to have completed a valid storage / retrieval action, the event counter for that partition is incremented. The entire statistical process is performed within a continuous sliding time window. An event queue is maintained, recording the precise timestamp of each event. As the current time progresses, event records in the queue that occurred earlier than the start point of the current time window are periodically cleared. The number of storage / retrieval events for each target warehouse partition is output. This number is a real-time observation value that directly reflects the frequency of material handling operations actually completed in these key partitions within the most recent time window. The number of historical access events is completely consistent with the statistical criteria (time window length, event definition rules).

[0037] Calculate the ratio of the actual number of access events to the historical number of access events for each target warehouse partition to obtain the access change rate; The dynamic access information is generated by integrating all the access change rates and the number of access events.

[0038] The method provided in this embodiment identifies target warehouse partitions based on real-time regional activity and historical access information. This allows for precise focus of analysis on key areas with currently busy operations and comparable historical benchmarks, avoiding resource consumption caused by indiscriminate calculations across all warehouse partitions and improving the efficiency and relevance of the overall analysis process. By extracting the number of historical events for the target partition from historical access information and calculating the actual number of access events within the current time window for each partition, a quantitative comparison between historical norms and real-time observations can be established for each key partition, enabling time-varying tracking of operational activities. By calculating the ratio of actual to historical quantities to obtain the access change rate, the activity levels of partitions with different workloads can be transformed into a unified relative change indicator, clearly revealing which areas have experienced abnormal increases or decreases in operational intensity, providing a standardized and comparable decision-making basis for priority assessment.

[0039] In one embodiment, the trajectory offset analysis of the fused positioning trajectory and the real-time regional activity based on a preset expected storage coordinate set to obtain in-database correction information includes: Extract the current positioning coordinate sequence from the fused positioning trajectory, and extract the target expected coordinates corresponding to the stored materials from the expected storage coordinate set; The expected storage coordinate set is a pre-defined data table that stores the coordinates of the designated storage locations of all warehouse materials within the warehouse management system. The storage coordinates of each material are unique and static within this set.

[0040] Calculate the coordinate deviation between the current positioning coordinate sequence and the target desired coordinates; Specifically, the X-coordinate of the center point is obtained by taking the arithmetic mean of the X-coordinate values ​​of all coordinate points in the current positioning coordinate sequence; similarly, the Y-coordinate of the center point is obtained by taking the arithmetic mean of the Y-coordinate values ​​(and including the Z-coordinate if it is a three-dimensional position). The resulting center point of the observation location serves as the spatial representative of the entire recent observation sequence. The Euclidean distance between the center point of the observation location and the desired target coordinates is calculated. The straight-line distance between the two points in a two-dimensional plane is obtained by taking the square root of the sum of the squares of the coordinate differences between the two points, thus yielding the coordinate deviation. For example, if the coordinates of the observation center point are (102.5, 55.3) and the desired target coordinates are (100.0, 50.0), the calculated coordinate deviation is approximately 5.85 units.

[0041] The coordinate deviation is compared with a preset deviation threshold. If the coordinate deviation is higher than the deviation threshold, it is determined that the storage material has been misaligned, and an offset identifier is generated. If the coordinate deviation is not higher than the deviation threshold, the positioning is determined to be normal and the process ends. Based on the real-time regional activity, traverse the free storage coordinates around the stored materials and calculate the storage distance between each free storage coordinate and the target desired coordinate. Specifically, the area surrounding the stored materials is defined as a circular region centered on the target desired coordinates and bounded by a fixed search radius, or a logical range limited to the same aisle or shelf group. All storage coordinates and their current occupancy status within this range are retrieved from the warehouse storage location status table. Constraints are applied based on real-time area activity: if the partition containing the target desired coordinates or the partition containing candidate free coordinates is marked as highly active, avoidance or de-weighting is implemented in the search strategy. For example, all free coordinates located in highly active partitions can be pre-removed from the candidate set to avoid exacerbating congestion in busy areas during the correction task; or, a penalty coefficient can be added to the storage distance of these coordinates during subsequent weight calculations. After this filtering, a final list of candidate free storage coordinates is obtained. Then, for each free coordinate point in the list, distance calculation is performed to determine the straight-line distance between the free point and the original theoretical storage point of the material (target desired coordinates).

[0042] The free storage coordinate with the smallest storage distance is selected as the recommended correction coordinate. The offset identifier, the coordinate deviation, and the recommended correction coordinate are combined to generate the in-library correction information.

[0043] The method provided in this embodiment extracts the current coordinate sequence from the fused positioning trajectory and calculates the deviation by comparing it with the target coordinates of the desired coordinate set. This enables precise quantification and automated detection of material position offsets, providing an objective and reliable quantitative basis for correction decisions. By comparing the coordinate deviation with a preset threshold and automatically generating an offset flag or determining a normal process termination, the method can intelligently distinguish between abnormal states requiring processing and negligible normal fluctuations. This avoids invalid calculations for materials that do not require correction, significantly improving the overall processing efficiency and resource utilization of the analysis process. By introducing real-time regional activity as a constraint when searching for idle correction coordinates, the selection of recommended coordinates can intelligently avoid currently busy areas, preventing correction tasks from exacerbating local congestion and achieving dynamic coordination between correction scheduling and the overall warehouse operation status.

[0044] In one embodiment, the step of performing transport and storage planning based on the dynamic access information, the real-time regional activity, and the in-database correction information to obtain an in-database transport task includes: The dynamic access information is matched with a preset handling equipment status table to obtain available handling equipment information; Traverse each material identifier to be corrected in the correction information in the library, extract the current position coordinates corresponding to the material identifier to be corrected from the fused positioning trajectory, and search for the target storage coordinates corresponding to the material identifier to be corrected from the expected storage coordinate set; Specifically, each piece of information in the correction information database is processed one by one. For the currently processed single correction information, the core material identifier to be corrected is extracted. Two coordinate parsing branches are started in parallel. The first branch is responsible for determining the starting point of the transport. Using the material identifier as a query condition, the real-time maintained fused positioning trajectory service is accessed. The query request aims to obtain the coordinate point with the latest timestamp under that identifier, which is considered valid. The trajectory service returns this coordinate point, which is confirmed as the current position coordinate of the material, serving as the starting spatial position of the transport operation. The second branch is responsible for determining the expected destination of the transport. Using the same material identifier as the key query key, a request is sent to the static expected storage coordinate set. The coordinate set returns the theoretical storage location pre-registered for the material, i.e., the target storage coordinate. This coordinate represents the final target position that the correction action hopes to achieve, and is the spatial destination of the offset correction. After obtaining these two coordinates, the current loop processing for the current material ends, and the result is that a definite pair of (current position coordinates, target storage coordinates) is bound to the material. The process continues to iterate through the next material identifier to be corrected, repeating the above process until all information in the set has been processed.

[0045] Based on the current location coordinates and the target storage coordinates, a spatial offset vector is calculated. The spatial offset vector, the material identifier to be corrected, and the target storage coordinates are combined to obtain a basic handling instruction. Specifically, for each material to be corrected, its current position coordinates and target storage coordinates have been obtained. For each such coordinate pair, the spatial offset vector is calculated. The calculation process follows the geometric principle of vector subtraction: subtract the X-coordinate value of the current position coordinate from the X-coordinate value of the target storage coordinate to obtain the X component of the offset vector; similarly, subtract the Y-coordinate value of the current position coordinate from the Y-coordinate value of the target storage coordinate to obtain the Y component of the offset vector. If three-dimensional space is involved, the Z-coordinate is also processed in the same way. The resulting set of values ​​(ΔX, ΔY, (ΔZ)) constitutes the spatial offset vector. The magnitude of this vector is the square root of the sum of the squares of its components. After the calculation is completed, the process enters the data combination stage. A new data structure is created as a container for the basic handling instructions. This structure contains three core fields: the first field stores the identifier of the material to be corrected to clarify the task object; the second field stores the calculated spatial offset vector to describe the geometric characteristics of the movement; and the third field stores the target storage coordinates to define the target location of the task. Fill the three data items corresponding to the current material into the corresponding fields of the structure to complete a basic handling instruction.

[0046] Based on the real-time regional activity, all the basic transport instructions are prioritized to obtain a priority label sequence; Based on the priority tag sequence, the available handling equipment information and the basic handling instructions are used to plan the handling tasks, thereby obtaining the in-warehouse handling tasks.

[0047] The method provided in this embodiment obtains available equipment information by matching dynamic access information with the handling equipment status table, and binds the start and end coordinates of materials to the correction information. This enables real-time and precise locking of execution resources and spatial start and end points for task planning, laying a reliable input foundation for generating executable instructions. By calculating spatial offset vectors and combining them to generate basic handling instructions, abstract correction requirements can be transformed into geometrically quantified standard task units. By prioritizing basic handling instructions based on real-time area activity, the task execution order can dynamically respond to the real-time busy level of each area in the warehouse, ensuring that critical correction tasks with the greatest impact on the overall workflow are processed first, thereby optimizing the business rationality of scheduling decisions.

[0048] In one embodiment, the step of planning the handling task based on the available handling equipment information and the basic handling instructions according to the priority tag sequence to obtain the in-warehouse handling task includes: Based on the priority tag sequence, the basic handling instruction is bound to the available handling equipment information to obtain the equipment to be assigned; Based on the basic handling instructions, a path is planned for the equipment to be assigned, and the initial handling path obtained is compared with the preset planned task path for path conflict detection. If conflicting paths exist, route planning and conflict detection are performed again until the conflict is resolved. When there are no conflicting paths, all the basic transport instructions, the devices to be assigned, and the initial transport path are bound together to obtain the in-warehouse transport task.

[0049] The method provided in this embodiment, by binding devices according to a priority tag sequence, ensures that high-priority handling instructions receive the best or nearest available equipment resources first, thereby linking task urgency with resource allocation efficiency. By planning initial paths for the bound devices to be assigned and performing conflict detection, potential risks such as path intersections, congestion, or time conflicts that may occur during multi-device collaborative operations can be identified and warned in advance. By triggering a closed-loop process of path replanning and re-detection when conflicting paths are detected, the device assignment or path scheme can be adjusted until all conflicts are eliminated, thus ensuring the coordination and feasibility of the final generated handling task set in the spatiotemporal dimensions. By finally binding conflict-free instructions, devices, and paths, a set of in-warehouse handling tasks with complete information, clear responsibilities, and direct ability to drive physical equipment execution can be generated, achieving a seamless transformation from optimization decisions to executable work instructions, improving the reliability and overall efficiency of warehouse automation execution.

[0050] Reference Figure 2 As shown, the present invention also provides a real-time positioning and correction device for warehouse materials, applied to any of the above-described real-time positioning and correction methods for warehouse materials, comprising: The acquisition module is used to acquire the positioning tag signals and material images of the stored materials, perform trajectory calculation on the positioning tag signals and the material images, and generate a fused positioning trajectory. The analysis module is used to perform regional activity calculation on the fused positioning trajectory based on preset warehouse partition information to obtain real-time regional activity, and to perform material access analysis on the real-time regional activity and the historical access information of the stored materials to obtain dynamic access information. The association module is used to perform trajectory offset analysis on the fused positioning trajectory and the real-time regional activity based on a preset expected storage coordinate set, so as to obtain in-database correction information. The processing module is used to perform transport and storage planning based on the dynamic access information, the real-time regional activity, and the in-database correction information to obtain in-database transport tasks.

[0051] This invention provides a real-time positioning and correction device for warehouse materials. By fusing positioning tag signals and material images to generate a high-precision fused positioning trajectory, it effectively overcomes the interference problem of single wireless signals in complex environments, providing continuous and stable real-time position feedback, thus laying a reliable data foundation for subsequent accurate correction. By automatically comparing and analyzing real-time positioning data with preset desired coordinates, it can accurately and automatically identify whether there is a deviation in the actual position of the material, achieving a leap from simple positioning to intelligent diagnosis of material status. By comprehensively considering real-time area activity, dynamic access information, and deviation analysis results, the generation of correction and handling tasks can closely align with the dynamic workload of the warehouse and the real-time access needs of materials, realizing a transformation from static, fixed task scheduling to dynamic, adaptive resource planning.

[0052] Reference Figure 3 As shown, the present invention also provides a real-time positioning and correction system for warehouse materials, comprising: Memory, used to store programs; A processor is configured to execute the program to implement the steps of a real-time positioning and correction method for warehouse materials as described in any of the preceding claims.

[0053] In this embodiment, the processor and memory can be connected via a bus or other means. The memory may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive. The processor may be a general-purpose processor, such as a central processing unit, digital signal processor, application-specific integrated circuit, or one or more integrated circuits configured to implement embodiments of the present invention.

[0054] The present invention also provides a storage medium storing computer instructions for causing a computer to perform any of the methods described above.

[0055] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the system and each module described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0056] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for real-time positioning and correction of stored materials, characterized in that, include: The system collects location tag signals and material images of stored materials, performs trajectory calculations on the location tag signals and material images, and generates a fused location trajectory. Based on the preset warehouse partition information, the fused positioning trajectory is used to calculate the regional activity to obtain the real-time regional activity. The real-time regional activity is then combined with the historical storage and retrieval information of the stored materials to perform material storage and retrieval analysis to obtain dynamic storage and retrieval information. Based on a preset expected storage coordinate set, trajectory offset analysis is performed on the fused positioning trajectory and the real-time regional activity to obtain in-database correction information; Based on the dynamic access information, the real-time regional activity, and the in-database correction information, a transport and storage plan is developed to obtain the in-database transport task.

2. The real-time positioning and correction method for stored materials according to claim 1, characterized in that, The process of collecting location tag signals and material images of stored materials, performing trajectory calculations on the location tag signals and material images, and generating a fused location trajectory includes: Extract the first timestamp and the first coordinate sequence from the positioning tag signal, and identify the corresponding second coordinate sequence from the material image; Based on the first timestamp, the first coordinate sequence and the second coordinate sequence are aligned and paired to form a set of coordinate pairs; A fusion decision is made for each coordinate pair in the coordinate pair set based on a preset coordinate distance threshold. When the spatial distance between two coordinate points in the coordinate pair is less than the preset coordinate distance threshold, the coordinate pair is fused and calculated to obtain the positioning trajectory point; otherwise, the corresponding coordinate point in the first coordinate sequence is used as the interpolation trajectory point. By integrating all the positioning trajectory points and the interpolated trajectory points, the fused positioning trajectory is obtained.

3. The real-time positioning and correction method for stored materials according to claim 1, characterized in that, The step of calculating the regional activity of the fused positioning trajectory based on preset warehouse partition information to obtain real-time regional activity includes: Based on the location tag signal, obtain the warehouse partition information of the warehouse where the stored materials are located; Based on the warehouse partition information, identify the warehouse partition to which each positioning coordinate in the fused positioning trajectory belongs; The number of coordinates of the location coordinates existing in each of the warehouse partitions within a preset time window is counted, and the number of coordinates is used to calculate the partition activity to obtain the initial activity level. The initial activity level is compared with the activity threshold in the warehouse partition information. If the initial activity level is higher than the activity threshold, a corresponding active status identifier is generated based on the warehouse partition; otherwise, a corresponding inactive status identifier is generated based on the warehouse partition. The real-time regional activity level is obtained by integrating all the active status identifiers, the inactive status identifiers, and the initial activity level.

4. The real-time positioning and correction method for stored materials according to claim 3, characterized in that, The step of performing material access analysis by combining the real-time regional activity level with the historical access information of the stored materials to obtain dynamic access information includes: Based on the real-time regional activity and the historical access information, the target warehouse partition that is currently active and has historical access records is identified. Based on the target warehouse partition, extract the corresponding number of historical access events from the historical access information; The number of access events occurring in each of the target warehouse partitions in the fused positioning trajectory is counted according to the preset time window. Calculate the ratio of the actual number of access events to the historical number of access events for each target warehouse partition to obtain the access change rate; The dynamic access information is generated by integrating all the access change rates and the number of access events.

5. The real-time positioning and correction method for stored materials according to claim 1, characterized in that, The trajectory offset analysis based on the fused positioning trajectory and the real-time regional activity, using a preset expected storage coordinate set, yields in-database correction information, including: Extract the current positioning coordinate sequence from the fused positioning trajectory, and extract the target expected coordinates corresponding to the stored materials from the expected storage coordinate set; Calculate the coordinate deviation between the current positioning coordinate sequence and the target desired coordinates; The coordinate deviation is compared with a preset deviation threshold. If the coordinate deviation is higher than the deviation threshold, it is determined that the storage material has been misaligned, and an offset identifier is generated. If the coordinate deviation is not higher than the deviation threshold, the positioning is determined to be normal and the process ends. Based on the real-time regional activity, traverse the free storage coordinates around the stored materials and calculate the storage distance between each free storage coordinate and the target desired coordinate. The free storage coordinate with the smallest storage distance is selected as the recommended correction coordinate. The offset identifier, the coordinate deviation, and the recommended correction coordinate are combined to generate the in-library correction information.

6. The real-time positioning and correction method for stored materials according to claim 1, characterized in that, The process of planning the material movement and storage based on the dynamic access information, the real-time regional activity, and the in-database correction information to obtain the in-database material movement task includes: The dynamic access information is matched with a preset handling equipment status table to obtain available handling equipment information; Traverse each material identifier to be corrected in the correction information in the library, extract the current position coordinates corresponding to the material identifier to be corrected from the fused positioning trajectory, and search for the target storage coordinates corresponding to the material identifier to be corrected from the expected storage coordinate set; Based on the current location coordinates and the target storage coordinates, a spatial offset vector is calculated. The spatial offset vector, the material identifier to be corrected, and the target storage coordinates are combined to obtain a basic handling instruction. Based on the real-time regional activity, all the basic transport instructions are prioritized to obtain a priority label sequence; Based on the priority tag sequence, the available handling equipment information and the basic handling instructions are used to plan the handling tasks, thereby obtaining the in-warehouse handling tasks.

7. The real-time positioning and correction method for stored materials according to claim 6, characterized in that, The step of planning the in-warehouse handling task based on the available handling equipment information and the basic handling instructions according to the priority marker sequence includes: Based on the priority tag sequence, the basic handling instruction is bound to the available handling equipment information to obtain the equipment to be assigned; Based on the basic handling instructions, a path is planned for the equipment to be assigned, and the initial handling path obtained is compared with the preset planned task path for path conflict detection. If conflicting paths exist, route planning and conflict detection are performed again until the conflict is resolved. When there are no conflicting paths, all the basic transport instructions, the devices to be assigned, and the initial transport path are bound together to obtain the in-warehouse transport task.

8. A real-time positioning and correction device for stored materials, characterized in that, The real-time positioning and correction method for warehouse materials according to any one of claims 1-7 includes: The acquisition module is used to acquire the positioning tag signals and material images of the stored materials, perform trajectory calculation on the positioning tag signals and the material images, and generate a fused positioning trajectory. The analysis module is used to perform regional activity calculation on the fused positioning trajectory based on preset warehouse partition information to obtain real-time regional activity, and to perform material access analysis on the real-time regional activity and the historical access information of the stored materials to obtain dynamic access information. The association module is used to perform trajectory offset analysis on the fused positioning trajectory and the real-time regional activity based on a preset expected storage coordinate set, so as to obtain in-database correction information. The processing module is used to perform transport and storage planning based on the dynamic access information, the real-time regional activity, and the in-database correction information to obtain in-database transport tasks.

9. A real-time positioning and correction system for warehouse materials, characterized in that, include: Memory, used to store programs; A processor is configured to execute the program to implement the various steps of the real-time positioning and correction method for warehouse materials as described in any one of claims 1-7.

10. A storage medium, characterized in that, The computer contains computer instructions for causing the computer to perform the method according to any one of claims 1 to 7.

Images