An intelligent warehousing information management system based on Internet of Things

Abstract

The intelligent warehouse information management system based on the Internet of Things belongs to the technical field of Internet of Things information management systems, comprising a signal acquisition module for acquiring the detection signal strength received by each intelligent shelf unit from adjacent units; a first processing module for determining the physical distance between each intelligent shelf unit and adjacent units based on the detection signal strength and a preset logarithmic distance path loss model, and generating a physical neighbor set representing the current physical layout according to a preset distance threshold; and a second processing module for comparing the physical neighbor set with a preset logical topology in the warehouse management system, and calculating a topology inconsistency index according to a binary difference function. The present application significantly improves the accuracy and robustness of topology perception.

Description

An IoT-based intelligent warehouse information management system Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) information management systems, specifically to an IoT-based intelligent warehouse information management system. Background Technology

[0002] Traditional smart shelving systems rely on connecting independent IoT devices, such as RFID readers and sensors, directly to a central management server. The core of this system is to achieve single-point, real-time data collection and uploading of inventory information. The foundation of this technology lies in the assumption that the physical location of each IoT sensor and its logical address in the warehouse management system are statically configured and remain unchanged over a long period of time. The physical layout and topology of the entire warehouse are completely defined and maintained by the central server, while the sensor nodes themselves are unaware of their relative position in the entire network.

[0003] In modern warehousing environments requiring high agility and high-density operations, such as e-commerce forward warehouses, rack layouts need to be dynamically reconfigured frequently and on a large scale. In this case, every physical movement of the racks requires manual intervention to recalibrate the logical positions of hundreds or thousands of sensors; this is not only inefficient and costly, but also prone to errors, leading to inaccurate data and turning the system from intelligent to cumbersome.

[0004] The fundamental flaw in existing technology is that it treats the physical location of the shelf as a static attribute that needs to be input and configured from the external world. However, in highly dynamic scenarios, the physical location has evolved into a dynamic variable that the system itself needs to discover and perceive in real time.

[0005] The information disclosed in the background section above is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent warehouse information management system based on the Internet of Things to solve the problems mentioned in the background art.

[0007] The technical solution of the present invention includes:

[0008] The signal acquisition module is used to acquire the strength of the detection signals received by each smart shelf unit from neighboring units;

[0009] The first processing module is used to determine the physical distance between each smart shelf unit and its neighboring units based on the strength of the detected signal and a preset logarithmic distance path loss model, and to generate a set of physical neighbors representing the current physical layout according to a preset distance threshold.

[0010] The second processing module is used to compare the physical neighbor set with the preset logical topology in the warehouse management system, and calculate the topology inconsistency index based on the binary difference function.

[0011] The topology calibration module is used to identify target shelf units whose logical and physical locations do not match when the topology inconsistency index is greater than zero, and automatically update the logical location in the warehouse management system based on the main logical region to which the physical neighbor set of the target shelf unit belongs.

[0012] The path optimization module is used to determine the optimal picking path based on the updated logical location in the warehouse management system and by using the topological inconsistency index as a penalty.

[0013] Preferably, the first processing module determines the physical distance, including:

[0014] Invoke the detected signal strength, the preset reference distance, the reference signal strength at the reference distance, and the environmental propagation factor;

[0015] Based on the logarithmic distance path loss model, the physical distance between the smart shelf unit and its neighboring units is calculated.

[0016] Preferably, the first processing module generates a set of physical neighbors, including:

[0017] The determined physical distance is compared with the preset physical neighbor threshold distance;

[0018] If the physical distance is less than the physical neighbor threshold distance, the corresponding neighboring unit is determined as a physical neighbor, and all determined physical neighbors are aggregated to generate a physical neighbor set.

[0019] Preferably, the second processing module calculates the topology inconsistency index, including:

[0020] By using a binary difference function, we can identify all pairs of units whose neighborhood relationships are inconsistent between the physical neighbor set and the logical topology.

[0021] All identified cell pairs are weighted and summed to generate a topological inconsistency index.

[0022] Preferred weighted summation includes:

[0023] For any pair of units, the weighting coefficients are determined based on the corresponding physical distance;

[0024] The smaller the physical distance, the higher the weighting coefficient.

[0025] Preferably, the topology calibration module automatically updates the logical location, including:

[0026] The physical neighbors of the target shelf unit are identified within the logical regions of the warehouse management system.

[0027] When the number of physical neighbors belonging to a specific logical area exceeds the preset affiliation threshold, the logical location of the target shelf unit will be automatically updated to an available location within the specific logical area.

[0028] Preferably, the path optimization module determines the optimal picking path, including:

[0029] Obtain the logical picking distance of the items to be picked;

[0030] Get the topological inconsistency value of the shelf where the product is located;

[0031] Multiply the topology inconsistency value by a preset topology inconsistency penalty coefficient to generate the topology penalty cost;

[0032] The topology penalty cost is used as a correction to the logical distance of each edge on the path to obtain the total picking path cost;

[0033] The optimal picking path is determined based on the principle of minimizing the total picking path cost.

[0034] Preferably, the topological inconsistency value of the shelf where the goods are located is a sub-item of the topological inconsistency index.

[0035] This invention provides an improved IoT-based intelligent warehouse information management system, which, compared with existing technologies, has the following improvements and advantages:

[0036] 1. By acquiring the detection signal strength through signal acquisition modules deployed in each smart shelf unit, and then using the first processing module to accurately determine the physical distance between units based on the signal strength and a preset logarithmic distance path loss model, this approach abandons the coarse method of relying solely on the presence or absence of signals for adjacency judgment. By introducing an environmentally calibrated physical model, it can generate a set of physical neighbors that accurately represents the current physical layout. This design provides quantitative, high-fidelity underlying data for subsequent topology comparison and calibration, ensuring that the constructed physical topology map can realistically and accurately reflect the actual layout of the shelves in the physical world, significantly improving the accuracy and robustness of topology perception.

[0037] 2. The second processing module compares the real-time generated set of physical neighbors with the preset logical topology in the warehouse management system. This module identifies all pairs of units with inconsistent neighbor relationships using a binary difference function and performs a weighted summation to calculate the topology inconsistency index. The weighting coefficient is determined based on the physical distance between the unit pairs; the smaller the physical distance, the higher the weighting coefficient. This design not only determines whether there is topology inconsistency in the system but also accurately quantifies the severity of inconsistency through the index size. It prioritizes reflecting the nearest-distance topology errors that have the greatest impact on compact operations such as picking paths, providing a more accurate and directional triggering basis for subsequent calibration modules.

[0038] 3. The system's topology calibration module can automatically identify target shelf units whose logical and physical locations do not match when the topology inconsistency index is greater than zero. This module counts the physical neighbors of the target shelf unit in the various logical areas to which they belong in the warehouse management system. When the number of physical neighbors belonging to a specific logical area exceeds a preset threshold, the module automatically updates the logical location of the target shelf unit to an available location within that specific logical area. This mechanism replaces the inefficient and error-prone manual calibration steps in the traditional process with a reliable, group data-based inference algorithm, achieving plug-and-play physical deployment of shelves and 100% self-calibrated real-time mapping of physical and logical topology.

[0039] 4. The system's path optimization module seamlessly integrates the real-time status of topology awareness into upper-layer application decision-making. When determining the optimal picking path, this module not only obtains the logical picking distance based on the updated logical location, but also incorporates the topology inconsistency value of the shelf where the product is located as a dynamic penalty item into the calculation of the total picking path cost. During the brief period when the physical location of the shelf changes but the logical location has not yet been synchronized, this penalty item can automatically prevent the system from generating erroneous picking paths based on outdated data. After the topology self-calibration is completed, the penalty item disappears, and the path planning immediately returns to the optimal state based on the latest accurate topology. This design constructs a risk avoidance closed loop from topology awareness to operational decision-making, ensuring the maximization of both picking efficiency and accuracy in a highly dynamic environment, and greatly improving the automation level and operational agility of warehouse management. Attached Figure Description

[0040] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0041] Figure 1 is a flowchart of the Internet of Things-based intelligent warehouse information management system of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0043] Example 1

[0044] Please refer to Figure 1. This invention provides an IoT-based intelligent warehouse information management system, including:

[0045] The signal acquisition module is used to acquire the strength of the detection signals received by each smart shelf unit from neighboring units;

[0046] The first processing module is used to determine the physical distance between each smart shelf unit and its neighboring units based on the strength of the detected signal and a preset logarithmic distance path loss model, and to generate a set of physical neighbors representing the current physical layout according to a preset distance threshold.

[0047] The second processing module is used to compare the physical neighbor set with the preset logical topology in the warehouse management system, and calculate the topology inconsistency index based on the binary difference function.

[0048] The topology calibration module is used to identify target shelf units whose logical and physical locations do not match when the topology inconsistency index is greater than zero, and automatically update the logical location in the warehouse management system based on the main logical region to which the physical neighbor set of the target shelf unit belongs.

[0049] The path optimization module is used to determine the optimal picking path based on the updated logical location in the warehouse management system and by using the topological inconsistency index as a penalty.

[0050] The implementation of a binary difference function requires a more detailed definition of two types of inconsistency: physical neighbors but logically non-neighbors, and logical neighbors but physically non-neighbors. For example, for a shelf unit... and ,if ,Right now exist In the physical neighbor set, but In the logical topology of the warehouse management system, not The neighbors, Conversely, if In a warehouse management system, it is Logical neighbors, but ,but Both of these cases are considered topologically inconsistent.

[0051] This embodiment provides an intelligent warehouse information management system based on the Internet of Things. The core purpose of this system is to build an intelligent network that can automatically sense the physical topology of the shelving layout in a scenario of high frequency and large-scale dynamic reorganization, and synchronize with the logical topology in the warehouse management system in real time, thereby solving the cost, delay and error of manual calibration.

[0052] In specific implementation scenarios, such as e-commerce forward warehouses requiring highly agile operations, this system forms a self-organizing Internet of Things (IoT) through intelligent units deployed on each shelf; the system's operation process constitutes a complete and self-consistent technical closed loop, specifically reflected in:

[0053] In the initial stage, the various smart shelf units deployed in the warehouse environment are activated through the signal acquisition module. The purpose of this module is to acquire raw data characterizing the physical proximity relationships between the shelf units. In this embodiment, after each smart shelf unit is activated or its position changes, it actively broadcasts a low-power, short-range detection signal. At the same time, it acts as a receiver, continuously acquiring detection signals from its neighboring units and recording data such as the received signal strength indication. This process initially transforms the adjacency relationships in the physical world into digital signals that can be computed.

[0054] In the data processing stage, the first processing module processes the collected signal strength data. The definition and purpose of this module is to transform the raw signal strength data into structured physical topology information. In this embodiment, based on the detected signal strength and a preset logarithmic distance path loss model, the physical distance between each smart shelf unit and its neighboring units is accurately calculated. Subsequently, by comparing the calculated physical distance with a preset distance threshold, the true physical neighbors are selected, thereby generating a set of physical neighbors that can accurately represent the current physical layout.

[0055] In the topology comparison phase, the second processing module intervenes to quantify the discrepancies between the real physical world and the digital twin system. In this embodiment, this module compares the real-time physical neighbor set generated by the first processing module with the preset or previously recorded logical topology in the warehouse management system point by point. It uses a binary difference function to determine whether there is any inconsistency in each neighbor relationship, for example, physical adjacency but not logical adjacency. By weighted summation, the topology inconsistency index is calculated. This index is a key quantitative indicator; the higher its value, the greater the deviation between the physical layout and the system record.

[0056] When the topology inconsistency index is greater than zero, the topology calibration module is activated. The definition and purpose of this module is to automatically repair the inconsistency between the physical world and the digital world, and realize the self-healing of the topology. In this embodiment, based on the calculation results of the second processing module, target shelf units whose logical positions do not match their physical positions due to physical movement are accurately identified. The physical neighbor set of the target shelf unit is analyzed to determine which main logical area in the warehouse management system, such as picking area A, belongs to most of its physical neighbors. Based on this attribution determination, the module automatically executes instructions to update the logical position of the target shelf unit in the warehouse management system to an available position in that main logical area. The entire process requires no manual intervention.

[0057] At the application level, the path optimization module leverages self-calibrated topology information to empower upper-layer applications. The module's purpose is to ensure that all warehousing operations, such as picking path planning, are always based on the latest and 100% accurate topology data. In this embodiment, when the warehouse management system needs to generate picking paths, this module not only calculates the basic picking distance based on the updated logical shelf locations in the warehouse management system, but also innovatively incorporates the topology inconsistency index calculated by the second processing module as a dynamic penalty term into the total cost calculation. Before topology calibration is complete, the inconsistency index is high, and the penalty term significantly increases the picking cost of the relevant shelves, causing the path planning system to temporarily avoid shelves with unreliable location information. After calibration, the index returns to zero, the penalty term disappears, and the system can generate the optimal picking path based on the actual layout.

[0058] Through the collaborative work of the above modules, this system transforms the physical location of shelves from a static attribute requiring manual configuration and maintenance into a dynamic variable that can be discovered and perceived by the system in real time. It solves the problems of low deployment efficiency, inaccurate data, and high operation and maintenance costs caused by frequent changes in physical layout in traditional intelligent shelf systems in dynamic warehousing environments. It achieves plug-and-play physical deployment of shelves in seconds, 100% self-calibrated real-time mapping of physical-logical topology, and dynamic operation optimization based on high-fidelity digital twins, which greatly improves the automation level, data accuracy, and operational agility of warehouse management.

[0059] Example 2

[0060] The first processing module determines the physical distance, including:

[0061] Invoke the detected signal strength, the preset reference distance, the reference signal strength at the reference distance, and the environmental propagation factor;

[0062] Based on the logarithmic distance path loss model, the physical distance between the smart shelf unit and its neighboring units is calculated.

[0063] The first processing module generates a set of physical neighbors, including:

[0064] The determined physical distance is compared with the preset physical neighbor threshold distance;

[0065] If the physical distance is less than the physical neighbor threshold distance, the corresponding neighboring unit is determined to be a physical neighbor, and all determined physical neighbors are gathered to generate a physical neighbor set.

[0066] This embodiment is a specific implementation of the first processing module in Embodiment 1, aiming to illustrate how it accurately determines physical distance and generates a set of physical neighbors;

[0067] The first processing module determines the physical distance by invoking a radio signal propagation model based on physical laws to ensure the accuracy and reliability of the distance estimation. To achieve this, the module calls three core inputs:

[0068] The strength of the detection signal acquired in real time by the signal acquisition module ;

[0069] Preset reference distance and the reference signal strength measured at the reference distance. ;

[0070] Environmental propagation factor reflecting the signal attenuation characteristics of the current warehouse environment ;

[0071] Reference distance This refers to the physical distance used as a benchmark for calibrating the relationship between signal strength and distance. Its function is to provide a known starting point for the path loss model. In this embodiment, this value is set to 1 meter based on industry-known parameters.

[0072] Reference signal strength This refers to the reference distance. The average signal strength value obtained from actual measurements is used to calibrate the signal transmission capability of specific hardware at a reference distance; it originates from the calibration phase before system deployment, at a distance of [missing information]. Multiple signal strength measurements were performed at the location to obtain a set of calibrated signal strength measurement values. The arithmetic mean of the data set is used to determine the error caused by individual hardware differences.

[0073] Environmental transmission factors This refers to a dimensionless parameter describing the rate attenuation of signal energy with distance in a specific physical space. Its purpose is to enable the ranging model to adapt to different warehouse environments, such as variations in goods density and the amount of metal obstacles. It originates from the calibration phase before system deployment, where a set of known and different calibration distances are selected within the warehouse. For example, at distances of 1 meter, 2 meters, and 5 meters, signal strength measurements are performed to obtain the corresponding calibrated signal strength. Based on this group of multiple , The calibration dataset composed of data pairs is obtained by performing regression analysis on the logarithmic distance path loss model using the least squares method, thereby achieving accurate modeling of specific environments.

[0074] This module is based on a logarithmic distance path loss model and calculates the intelligent shelf unit using the following formula. With neighboring units physical distance between :

[0075]

[0076] in: Shelf unit The physical distance between them, in meters, is calculated using this formula;

[0077] Reference distance, in meters, obtained through preset;

[0078] : Signal strength at the reference distance, in decibels and milliwatts (dBm), obtained through prior calibration;

[0079] The received signal strength, measured in decibels and milliwatts, is provided in real time by the signal acquisition module.

[0080] Environmental propagation factor, dimensionless, obtained through prior calibration;

[0081] This formula is derived using the linear relationship of the logarithmic distance path loss model in the decibel domain; where, the numerator of the exponential part... This actually represents the signal attenuation from the reference distance to the actual distance, and its unit is decibels (dB). This value is calculated by dividing by... It then becomes a dimensionless pure number, thus keeping the formula consistent in terms of dimensions;

[0082] The technical motivation behind this formula is that it can quickly and cost-effectively estimate the relative distance between adjacent devices by using simple signal strength changes without relying on expensive and complex absolute coordinate positioning systems. This transforms the problem from obtaining absolute coordinates to identifying relative topological relationships, greatly reducing system deployment costs and improving flexibility.

[0083] After determining the physical distances, the first processing module generates a set of physical neighbors as follows: The physical distances calculated using the above formula are... Distance from preset physical neighbor threshold Compare;

[0084] For standard single-aisle picking, the threshold can be set to the shelf width plus the aisle width. Furthermore, to ensure the accuracy of the threshold, multiple signal tests can be conducted before system deployment in different aisle types, such as straight-through aisles, cross aisles, and areas with different shelf densities, to collect signal strength and actual distance data. Then, statistical methods, such as cluster analysis or confidence interval calculation, can be used to determine an optimal threshold. This value ensures that over 95% of physical neighbor pairs can be correctly identified;

[0085] Preset physical neighbor threshold distance For example, for standard single-aisle picking, it can be set to the shelf width plus the aisle width, i.e. The specific value can be determined through on-site testing before deployment, based on the actual warehouse layout; for example, it can be set at 3-5 meters. This threshold... The configuration needs to be dynamically adjusted based on the actual aisle width and shelf dimensions of the warehouse. Furthermore, to address different warehouse environments, the stable signal reception range at different distances can be determined experimentally before deployment, and this range should be used as a basis for decision-making. The reference value is set based on the principle that only closely adjacent shelves that are directly related in physical layout are considered as neighbor nodes in the topology network. The threshold is determined by considering the physical dimensions of the shelf and the standard aisle width, and taking into account reasonable layout tolerances. For example, it can be set as the sum of the shelf width and the maximum allowable spacing.

[0086] If the calculated physical distance less than the physical neighbor threshold distance Then the system determines the corresponding neighboring unit. It is a unit The module aggregates all cells identified as physical neighbors, forming a unit... Generate its own set of physical neighbors. ;

[0087] It's important to note that this model also exhibits physical robustness in extreme cases. When two shelf units are very close together, leading to variations in received signal strength... Much greater than the reference signal strength When the signal is completely lost or unreceived, a default distance value can be set or the shelf unit can be excluded from the physical neighbor set to avoid miscalculations.

[0088] Through the precise calculation of physical distance and reliable screening of physical neighbors, this embodiment not only provides quantitative, high-fidelity underlying data for subsequent topology comparison and calibration, but also ensures that the constructed physical topology map can truly and accurately reflect the actual layout of the shelves in the physical world. Compared with the coarse method of judging adjacency relationships by relying solely on the presence or absence of signals, this solution significantly improves the accuracy and robustness of topology perception by introducing an environmentally calibrated physical model.

[0089] The second processing module calculates the topology inconsistency index, including:

[0090] By using a binary difference function, we can identify all pairs of units whose neighborhood relationships are inconsistent between the physical neighbor set and the logical topology.

[0091] All identified cell pairs are weighted and summed to generate a topological inconsistency index;

[0092] Weighted summation, including:

[0093] For any pair of units, the weighting coefficients are determined based on the corresponding physical distance;

[0094] The smaller the physical distance, the higher the weighting coefficient.

[0095] This embodiment is a specific implementation of the second processing module in Embodiment 1, aiming to illustrate how it can accurately calculate the topological inconsistency index based on the physical neighbor set and logical topology;

[0096] The core of the second processing module's calculation of the topological inconsistency index lies in the introduction of a quantitative evaluation model, which concretizes the abstract concept of synchronization into a numerical value that can be processed by a computer. In this embodiment, the calculation process is as follows:

[0097] Using binary difference functions, the system systematically identifies all unit pairs in the physical neighbor set that are inconsistent with the logical topology and stored in the warehouse management system.

[0098] Bivariate function This refers to the method used to determine any two shelf units. and A function that determines whether the neighborhood relationships between neighbors are consistent at both the physical and logical levels;

[0099] The implementation of this function requires a more detailed definition of two types of inconsistency: physical neighbors but logical non-neighbors, and logical neighbors but physical non-neighbors; for example, when a shelf is moved to a new area, its original logical neighbor relationships are physically broken, and in this case... At the same time, it establishes new physical relationships with its neighbors in the new area, but these relationships are not yet logically synchronized. To differentiate the importance of these two cases, different weighting coefficients can be introduced into the weighted summation. For example, higher weights can be given to logical relationships that are physically close but do not match, so as to prioritize the handling of topological errors that directly affect the current picking path.

[0100] Its function is to serve as a fundamental operator for detecting topological differences; it is defined as: when a shelf unit and They are neighbors physically, that is... However, they are not neighbors in the logical topology, or conversely, they are logical neighbors but not physical neighbors. When the relationship between the two is completely consistent at both the physical and logical levels, ;

[0101] After identifying all inconsistent unit pairs, i.e. After the unit pairs are processed, the system performs a weighted summation to generate the final topology inconsistency index. The summation process follows the formula below:

[0102]

[0103] in:

[0104] Topological inconsistency index, dimensionless, the smaller the value, the better the physical and logical topology match, and it is the final output of this module;

[0105] The total number of smart shelf units in the warehouse, dimensionless, is a system configuration parameter;

[0106] :unit The set of physical neighbors is provided by the first processing module;

[0107] : Weighting coefficient, dimensionless, used to adjust the contribution of different neighbors to the inconsistency index, the source of which is detailed below;

[0108] : A binary difference function, with a value of 0 or 1, determined by the real-time comparison results;

[0109] In order to avoid duplicate calculations, this summation process usually calculates only once for each pair of inconsistent units;

[0110] Furthermore, the weighted summation process reflects a differentiated consideration of different physical relationships; in this embodiment, for any identified inconsistent unit pair... The system will base its decisions on the physical distance between them. The weighting coefficients are dynamically determined by the first processing module. The technical principle behind this design is that the closer two shelves are physically, the more defined and important their adjacency relationship is; therefore, the smaller the physical distance, the higher the weighting coefficient should be. For example, this can be determined using a normalized inverse proportional function, such as:

[0111]

[0112] in Weighting coefficients, dimensionless. Physical neighbor threshold distance: The maximum distance limit used to determine whether two shelf units constitute a direct neighbor relationship. Shelf unit and The physical distance between them, in meters. The preset weight index and This design ensures that the smaller the physical distance, the higher its corresponding dimensionless weight; this makes the system more sensitive to mismatches in the relationship between close neighbors, and prioritizes the handling of topology errors that are most likely to affect actual operation.

[0113] To avoid physical distance When the weighting coefficient approaches zero, For problems approaching infinity, a minimum physical distance threshold can be set. When the calculated physical distance At this time, the weighting coefficients can be limited to a maximum value, or... Considered equal to Perform calculations; for example, the formula can be modified as follows: ;

[0114] Weighting index The selection of [a specific parameter] directly affects the system's sensitivity to neighbors at different distances; when [a specific parameter] is selected, it directly affects the system's sensitivity to neighbors at different distances; When the value is large, the system will penalize nearby topological inconsistencies more strongly, thus prioritizing the repair of the most critical neighbor relationships; when When the distance is small, the system will handle inconsistencies across all distances relatively evenly; recommended... The value range is usually 100%. In practical applications, The value should be adjusted based on the density of picking operations and the average spacing between shelves;

[0115] This embodiment, by introducing a binary difference function and a weighted summation mechanism, can not only determine whether there is topological inconsistency in the system, but also, through exponential... The magnitude of the index precisely quantifies the severity of this inconsistency; in particular, the introduction of weighting coefficients makes the index more physically meaningful and business-oriented, and it can prioritize reflecting close-range topology errors that have the greatest impact on compact operations such as picking paths; this provides a more accurate and directional trigger for subsequent topology calibration modules, enabling them to more efficiently correct critical topology deviations.

[0116] The topology calibration module automatically updates logical locations, including:

[0117] The physical neighbors of the target shelf unit are identified within the logical regions of the warehouse management system.

[0118] When the number of physical neighbors belonging to a specific logical area exceeds the preset affiliation threshold, the logical location of the target shelf unit will be automatically updated to an available location within the specific logical area.

[0119] This embodiment is a specific implementation of the topology calibration module in Embodiment 1, aiming to illustrate how it automatically updates the logical position of the target shelf unit;

[0120] After determining the new logical area of ​​the target shelf unit, the system will further analyze the specific logical location of its physical neighbors in the warehouse management system, and find the available location within the logical area that is logically closest to the physical neighbor for allocation, so as to maintain a smooth transition between the old and new logical topologies.

[0121] The execution logic of the topology calibration module automatically updating the logical location is a decision-making process based on the majority rule, aiming to intelligently infer the correct logical location of the moved shelf without human intervention; when the system detects a topology inconsistency index... After identifying the target shelf unit, the automatic update process is as follows:

[0122] The system identifies the physical neighbors of a target shelf unit within the logical regions of the warehouse management system; for target shelf units identified as having a location mismatch... The module will traverse its set of physical neighbors. Each neighboring unit For each neighbor The module will query the warehouse management system database to obtain the logical region to which the logical location of the current record belongs. For example, A-01-01 belongs to A-region. The module will count the number of all physical neighbors that belong to A-region, B-region, C-region, etc.

[0123] The statistical results are compared with preset thresholds to make an attribution judgment; when the number of physical neighbors belonging to a specific logical area, such as area A, exceeds a preset attribution judgment threshold, the system determines the target shelf unit. It has most likely been moved to that specific logical region;

[0124] The attribution judgment threshold refers to a standard for the number of neighbors used to determine the new logical attribution of a target shelf unit. Its purpose is to ensure the robustness of the location update decision and avoid misjudgment caused by the accidental noise signals of a few neighbors. This threshold is a configurable system parameter, and its determination is based on the business's requirements for decision confidence. For example, it can be set to 50% of the total number of physical neighbors, which means that as long as more than half of the physical neighbors of the target shelf belong to the same logical area, the update will be triggered.

[0125] This threshold is a configurable system parameter, determined based on the business's requirements for decision confidence; for example, it can be set to 50% of the total number of physical neighbors.

[0126] This threshold should be considered a confidence parameter; if set to 50%, it may lead to decision uncertainty when the number of physical neighbors in two logical regions is similar; to enhance robustness, a confidence interval judgment based on statistics can be introduced; for example, automatic updates should only be triggered when the number of physical neighbors in a particular logical region is statistically significantly greater than that in other regions, such as exceeding 65% or reaching a certain absolute number, in order to avoid misjudgments caused by a few drifting shelf neighbors;

[0127] Execute the update; once the above conditions are met, the system will immediately execute a database command to update the target shelf unit. The logical location is automatically updated to an available location within that specific logical area; for example, if the shelf is determined to be... If a shelf belongs to area A, the system will search for an available location in the logical storage location map of area A and assign it to the shelf. This process is executed entirely in a closed loop by the system, requiring no manual review or confirmation.

[0128] The automatic update mechanism described in detail in this embodiment is a key element in realizing the core value of the system's plug-and-play capability and 100% self-calibration. It replaces the inefficient and error-prone manual calibration steps in the traditional process with a reliable inference algorithm based on population data. This makes warehouse layout adjustments a highly efficient physical operation. After the physical movement is completed, the digital twin system immediately and automatically synchronizes, thereby greatly improving the agility of warehouse operations and the accuracy of data.

[0129] Example 3

[0130] The path optimization module determines the optimal picking path, including:

[0131] Obtain the logical picking distance of the items to be picked;

[0132] Get the topological inconsistency value of the shelf where the product is located;

[0133] Multiply the topology inconsistency value by a preset topology inconsistency penalty coefficient to generate the topology penalty cost;

[0134] The topology penalty cost is used as a correction to the logical distance of each edge on the path to obtain the total picking path cost;

[0135] The optimal picking route is determined based on the principle of minimizing the total picking route cost;

[0136] The topological inconsistency value of the shelf where the product is located is a sub-item of the topological inconsistency index;

[0137] A preset topology inconsistency penalty coefficient, the value of which can be determined based on historical data analysis or simulation experiments. For example, the average extra walking distance caused by a single picking error can be used as its baseline value.

[0138] This embodiment is a specific implementation of the path optimization module in Embodiment 1, aiming to illustrate how it determines the optimal picking path and ensures the real-time accuracy of path planning;

[0139] The core innovation of the path optimization module in determining the optimal picking path lies in upgrading traditional path planning based on static logical distance to an intelligent planning method that can perceive and adapt to dynamic changes in the topology; the process of determining the optimal path is as follows:

[0140] To construct the cost function, the module obtains the logical picking distance of the items to be picked. This distance is calculated based on the current recorded shelf logical topology diagram in the warehouse management system, which is updated in real time by the topology calibration module; it represents the distance from the picker's current location to the goods under ideal conditions, i.e., when physical and logical synchronization is complete. The theoretical shortest distance to walk to the location of the shelf;

[0141] The module retrieves the product. Topological inconsistency value of the shelf Based on the further limitations mentioned above, the topological inconsistency value of the shelf where the goods are located is the topological inconsistency index. A sub-item;

[0142] commodity Shelf Topological inconsistency value It should be defined as being related to the shelf. The sum of inconsistency weights related to it and all its neighbors can be expressed as:

[0143]

[0144] in, It is a shelf The set of physical neighbors, These are weighting coefficients. It is a binary difference function; after topology calibration is completed, The value is 0 for all neighbors, therefore It will also return to zero;

[0145] Topological Inconsistency Index This refers to the topological inconsistency values ​​of all racking units in the warehouse. The result of summation is therefore the topological inconsistency value of a single product's shelf. It constitutes the topological inconsistency index A summation term;

[0146] ektopo refers to a product specifically designed for carrying goods. The local inconsistency measure of that shelf unit, with dimensionless physical dimensions; this value comes directly from the calculation results of the second processing module; when the shelf has just been moved and its position has not yet been confirmed by the topology calibration module. The value will be significantly greater than zero; once calibration is complete, the value will quickly converge to zero.

[0147] When calculating the total picking route cost, the route optimization module will traverse all edges on the route. If the two shelf units connected by an edge are inconsistent in physical and logical topology, a penalty cost will be added to this edge.

[0148] Topological inconsistency penalty coefficient It is a key adjustable parameter, measured in meters, and its function is to adjust dimensionless topological inconsistencies. As a correction factor for logical picking distance;

[0149] The quantification of this coefficient should be linked to the severity of topological inconsistency. For example, different levels of topological inconsistency values ​​can be determined by performing multiple regression analysis on historical data. By using the extra distance walked as the independent variable and the additional distance walked as the dependent variable, a more dynamic penalty function can be fitted. Instead of a static constant, this method can more accurately reflect the actual impact of different degrees of topological errors; for example, by analyzing picking tasks during periods of topological inconsistency, the average extra walking distance caused by path planning errors or detours can be statistically analyzed as a measure of the actual impact. The baseline value. This value can be obtained through the formula: actual distance to logical distance.

[0150]

[0151] Quantification is performed, among which This represents the number of incorrect picking tasks caused by inconsistencies. Actual distance. It is the first The actual distance traveled in the second incorrect picking task, while the logical distance. It is the first The theoretical shortest distance calculated based on logical topology in the sub-error picking task;

[0152] The established technical logic is to quantify the actual losses that may result from inaccurate topology information; the value of this coefficient can be determined based on historical data analysis or simulation experiments. For example, the average extra walking distance caused by a single picking error can be used as its baseline value; in extreme scenarios where shelf reorganization is extremely frequent, the value can be dynamically adjusted. The value is increased to more severely punish data inconsistency, forcing the path planning system to avoid relevant areas before the topology is fully calibrated;

[0153] The module adds the logical picking distance to the topology penalty cost to obtain the total picking path cost. For a total task containing M items to be picked, the total cost is calculated as follows:

[0154]

[0155] in: Total picking path cost, in meters, is the objective function for path optimization.

[0156] The total number of items to be picked, dimensionless, determined by the picking order;

[0157] :commodity The logical picking distance, in meters, is calculated in real time by the warehouse management system.

[0158] : Topology inconsistency penalty coefficient, in meters, is a preset adjustable parameter;

[0159] :commodity The topological inconsistency value of the shelf is dimensionless and comes from the real-time calculation of the second processing module.

[0160] This formula is a simplified cost function. The idea is to attach a penalty to each item to be picked that is inconsistent with the topology of the shelf it is located in, so as to guide the path planning algorithm to avoid these shelves as much as possible during periods of topology inconsistency.

[0161] The system is based on minimizing the total picking path cost. Based on the principle of using algorithms such as Dijkstra's or A* in graph theory, the optimal picking path is determined.

[0162] The total picking path cost is the sum of the costs of all edges on the path, where the cost of each edge is determined by both the logical distance and the topology penalty cost.

[0163] This embodiment achieves unexpected technical results by introducing a dynamic penalty term strongly correlated with the real-time state of the topology into the path cost model; it constructs a risk avoidance mechanism that automatically prevents the system from generating erroneous picking paths based on outdated data during the brief chaotic state when the physical location of the shelf changes but the logical location has not yet been synchronized; once the topology self-calibration is completed, Resetting the topology to zero eliminates the penalty, and path planning immediately reverts to the optimal state based on the latest and most accurate topology. This forms a seamless closed loop from topology awareness to operational decision-making, ensuring the maximization of both picking efficiency and accuracy in highly dynamic environments, and elevating warehouse operational agility to a new order of magnitude. This model is a simplified cost function whose main purpose is to address the problem of topology inconsistency. In practical applications, this model can be extended to incorporate more factors affecting picking efficiency, such as traffic congestion coefficients in different shelf areas, the weight or volume of picked goods, and the real-time location of pickers, thereby constructing a more comprehensive and accurate path cost model.

[0164] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent warehouse information management system based on the Internet of Things, characterized in that, include: The signal acquisition module is used to acquire the strength of the detection signals received by each smart shelf unit from neighboring units; The first processing module is used to determine the physical distance between each smart shelf unit and its neighboring units based on the strength of the detected signal and a preset logarithmic distance path loss model, and to generate a set of physical neighbors representing the current physical layout according to a preset distance threshold. The second processing module compares the physical neighbor set with the preset logical topology in the warehouse management system and calculates the topology inconsistency index based on the binary difference function. The topology calibration module identifies target shelf units whose logical and physical locations do not match when the topology inconsistency index is greater than zero, and automatically updates their logical location in the warehouse management system based on the main logical region to which the physical neighbor set of the target shelf unit belongs. The path optimization module determines the optimal picking path based on the updated logical location in the warehouse management system and the topology inconsistency index as a penalty. The first processing module determines the physical distance by: invoking the detection signal strength, the preset reference distance, the reference signal strength at the reference distance, and the environmental propagation factor; this module calculates the intelligent shelf unit based on the logarithmic distance path loss model using the following formula. With neighboring units physical distance between ： in: Shelf unit The physical distance between them, in meters, is calculated using this formula; Reference distance, in meters, obtained through preset; : Signal strength at the reference distance, in decibels and milliwatts (dBm), obtained through prior calibration; The received signal strength, measured in decibels and milliwatts, is provided in real time by the signal acquisition module. The environmental propagation factor is dimensionless and obtained through prior calibration. The second processing module calculates the topological inconsistency index, including: identifying all cell pairs with inconsistent neighbor relationships in the physical neighbor set and logical topology using a binary difference function; and performing a weighted summation on all identified cell pairs to generate the topological inconsistency index. The summation process follows the formula below: in: Topological inconsistency index, dimensionless, the smaller the value, the better the physical and logical topology match, and it is the final output of this module; The total number of smart shelf units in the warehouse, dimensionless, is a system configuration parameter; :unit The set of physical neighbors is provided by the first processing module; : Weighting coefficient, dimensionless, used to adjust the contribution of different neighbors to the inconsistency index; The binary difference function has a value of 0 or 1, determined by real-time comparison results; weighted summation includes: for any pair of units, determining the weighting coefficient based on the corresponding physical distance; the smaller the physical distance, the higher the weighting coefficient; the path optimization module determines the optimal picking path, including: obtaining the logical picking distance of the goods to be picked; obtaining the topology inconsistency value of the shelf where the goods are located; multiplying the topology inconsistency value by a preset topology inconsistency penalty coefficient to generate the topology penalty cost; goods Shelf Topological inconsistency value It should be defined as being related to the shelf. The sum of inconsistency weights related to it and all its neighbors can be expressed as: in, It is a shelf The set of physical neighbors, These are weighting coefficients. It is a binary difference function; after topology calibration is completed, The value is 0 for all neighbors, therefore It will also return to zero; Topological Inconsistency Index This refers to the topological inconsistency values ​​of all racking units in the warehouse. The result of summation is therefore the topological inconsistency value of a single product's shelf. It constitutes the topological inconsistency index An additive item; ektopo refers specifically to the item that carries the goods. The local inconsistency measure of that shelf unit uses the topology penalty cost as a correction to the logical distance of each edge on the path to obtain the total picking path cost; based on the principle of minimizing the total picking path cost, the optimal picking path is determined; the topology inconsistency value of the shelf where the product is located is a sub-item of the topology inconsistency index; the module adds the logical picking distance and the topology penalty cost to obtain the total picking path cost. For a total task containing M items to be picked, the total cost is calculated as follows: in: Total picking path cost, in meters, is the objective function for path optimization. The total number of items to be picked, dimensionless, determined by the picking order; :merchandise The logical picking distance, in meters, is calculated in real time by the warehouse management system. : Topology inconsistency penalty coefficient, in meters, is a preset adjustable parameter; :merchandise The topological inconsistency value of the shelf is dimensionless and comes from real-time calculation by the second processing module.

2. The IoT-based intelligent warehouse information management system according to claim 1, characterized in that, The first processing module generates a physical neighbor set, including: comparing the determined physical distance with a preset physical neighbor threshold distance; if the physical distance is less than the physical neighbor threshold distance, then the corresponding neighboring unit is determined as a physical neighbor, and all determined physical neighbors are aggregated to generate a physical neighbor set.

3. The IoT-based intelligent warehouse information management system according to claim 1, characterized in that, The topology calibration module automatically updates the logical location, including: counting the physical neighbors of the target shelf unit and assigning them to various logical areas in the warehouse management system; when the number of physical neighbors belonging to a specific logical area exceeds the preset assignment threshold, the logical location of the target shelf unit is automatically updated to an available location within that specific logical area.

Images