A fine intelligent warehouse safety management method and system

Abstract

This invention discloses a refined intelligent warehouse safety management method and system, relating to the field of warehouse management technology. The method includes: constructing a cargo storage strategy model based on cargo characteristic data; outputting a cargo storage plan through the model; and allocating storage racks to the cargo according to the plan; collecting dynamic load indicators of the storage racks; fusing these indicators with predefined environmental indicators to comprehensively determine high-risk rack locations prone to tipping events; identifying adjacent objects at high-risk rack locations; using a spatial collision detection model to predict the impact range of a tipping event on adjacent objects; and formulating warehouse safety early warning measures based on the impact range. This invention can accurately assess the health status of rack structures and dynamically identify potential high-risk rack locations. Based on the connectivity of microcracks and structural health assessment, it identifies strain concentration areas and predicts rack breakage risks in advance.

Description

Technical Field

[0001] This invention relates to the field of warehouse management technology, and more specifically, to a sophisticated intelligent warehouse safety management method and system. Background Technology

[0002] Warehousing refers to the management and control of materials and goods in both spatial and temporal dimensions. Its core function is to achieve the orderly storage, protection, and circulation of goods through facilities, ensuring the continuity and efficiency of the supply chain. The purpose of storing goods is to achieve orderly management and efficient circulation of materials. Reasonable warehousing extends the lifespan of goods, avoids damage, improves space utilization, ensures supply chain continuity, and enables rapid retrieval, distribution, or dispatch when needed.

[0003] Shelving is prone to breakage due to factors such as prolonged exposure to excessive loads, material aging, fatigue damage, poor structural design, or external impacts. Shelving breakage is usually a gradual process; if it cannot be predicted, it can lead to sudden breakage, causing goods to fall and personnel to be injured. Furthermore, a broken shelf can affect adjacent shelves or goods, even causing widespread warehouse chaos, thereby compromising the safety and stability of the warehouse environment.

[0004] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0005] In response to the problems in related technologies, this invention proposes a refined intelligent warehouse safety management method and system to overcome the aforementioned technical problems existing in the existing related technologies.

[0006] Therefore, the specific technical solution adopted by the present invention is as follows:

[0007] According to one aspect of the present invention, a refined intelligent warehouse security management method is provided, the method comprising:

[0008] A cargo storage strategy model is constructed based on cargo characteristic data. The cargo storage strategy model outputs a cargo storage plan, and storage shelves are allocated to the cargo according to the cargo storage plan.

[0009] Collect dynamic load indicators of storage shelves, integrate the dynamic load indicators with predefined environmental indicators, and comprehensively judge high-risk shelf locations that are prone to tipping events;

[0010] Identify adjacent objects at high-risk shelving locations and use a spatial collision detection model to predict the impact range of a tipping over at a high-risk shelving location on adjacent objects. Based on the impact range, formulate warehouse safety early warning measures.

[0011] Preferably, dynamic load indicators of the storage shelves are collected, and these indicators are integrated with predefined environmental indicators to comprehensively determine high-risk shelf locations prone to tipping events, including:

[0012] Construct the topology of the storage rack and introduce the embedded material parameters of the rack to generate a rack structure model that meets the strength constraints.

[0013] Finite element simulation technology was used to simulate the strain response behavior of a rack structure model under dynamic load and environmental conditions, and the microcrack propagation path of the rack structure model was identified based on the strain response behavior.

[0014] Based on the microcrack propagation path identification, strain concentration regions of the shelf structure model are identified, and the evolution process of microcracks in the strain concentration regions from initial propagation to complete fracture is tracked.

[0015] The yield strength of strain concentration regions during the evolution of microcracks is analyzed and compared with a preset threshold. If the yield strength is greater than the preset threshold, the current storage rack is marked as a high-risk rack location prone to tipping events; otherwise, the dynamic load index and environmental index are updated.

[0016] Preferably, the microcrack propagation path of the shelf structure model based on strain response behavior identification includes:

[0017] Microcrack features are extracted from the shelf structure model based on strain response behavior, and the microcrack features are integrated to obtain a microcrack topology tree. The connectivity of each microcrack in the microcrack topology tree is calculated.

[0018] In the shelf structure model, any microcrack point is selected as the starting point, and the selected microcrack point is used as the input. The improved slime mold path search algorithm is used to search for the set of crack connection paths.

[0019] Crack connectivity is used as the evaluation vector to determine the validity of each path in the set of crack connectivity paths, and crack connectivity paths that meet the validity conditions are taken as microcrack propagation paths.

[0020] Preferably, microcrack features are extracted from the shelf structure model based on strain response behavior, and the microcrack features are integrated to obtain a microcrack topology tree. The connectivity of each microcrack in the microcrack topology tree is calculated, including:

[0021] Microcrack features are extracted based on strain response behavior. Microcrack features with similar characteristics are grouped into the same group to obtain several sets of microcrack features. The connection relationship of microcrack points in the microcrack feature set is used as the edge to generate a microcrack topology tree.

[0022] The main microcrack branches are selected from all edges of the microcrack topology tree according to preset rules, and the main microcrack branches are encoded.

[0023] The perturbation combination of the main microcrack branches is analyzed sequentially according to the encoding using sequence derivation technology, and the microcrack failure event of the main microcrack branches is analyzed by introducing perturbation vectors.

[0024] Based on the microcrack failure event, the corresponding edges in the microcrack topology tree are removed to generate auxiliary microcrack branches. The connectivity probability function between the auxiliary microcrack branches is constructed, and the crack connectivity flow of the auxiliary microcrack branches is output as the microcrack connectivity.

[0025] Preferably, an arbitrary microcrack point is selected as the starting point in the shelf structure model, and the selected microcrack point is used as input. The improved slime mold path search algorithm is used to search for the set of crack connectivity paths, including:

[0026] Several microcrack points in the microcrack feature set were randomly selected as slime mold source points, and a slime mold population was generated.

[0027] The shrinkage behavior of each slime mold in the population at the slime mold source point is obtained, the fitness value of the slime mold source point after shrinkage is calculated, and the fitness values ​​are sorted to obtain the optimal slime mold source point;

[0028] Starting from the optimal slime mold source point, the expansion path from the optimal slime mold source point to the other slime mold source points is iteratively searched, and the correlation between the expansion path and the auxiliary microcrack branch is calculated.

[0029] Based on the correlation calculation results, duplicate paths with auxiliary microcrack branches in the expansion path are removed to obtain the crack connectivity path set.

[0030] Preferably, adjacent objects at high-risk shelving locations are identified, and a spatial collision detection model is used to predict the impact range of a tipping-over at a high-risk shelving location on adjacent objects. Based on the impact range, warehouse safety early warning measures are formulated, including:

[0031] The three-dimensional spatial layout of the warehouse environment is obtained from the database, and the spatial location of adjacent target objects at high-risk shelf locations is identified based on the three-dimensional spatial layout.

[0032] The finite element simulation technology is used to simulate the impact intensity data of each tipping point when the storage shelf is tipped over. The impact intensity data is then input into a pre-built spatial collision detection model, which outputs the range of impact of the tipping event on adjacent targets.

[0033] Based on the impact range and intensity of adjacent targets, the spatial location of adjacent targets is divided to obtain a multi-gradient target risk space, and warehouse safety early warning measures are formulated for each multi-gradient target risk space.

[0034] Preferably, the impact intensity data is input into a pre-built spatial collision detection model, and the spatial collision detection model outputs the spillover range of the tipping event on adjacent targets, including:

[0035] Based on the three-dimensional spatial layout of the warehouse environment, the bounding boxes of high-risk shelf locations and adjacent target objects are obtained respectively, and the intersection test of the two sets of bounding boxes is carried out.

[0036] If the bounding boxes of high-risk shelving locations and adjacent targets intersect, the spatial layout information of the high-risk shelving locations and adjacent targets is inserted into the hash table; otherwise, the time period for simulating shelving tipping is adjusted and the bounding box intersection test is re-executed.

[0037] The spatiotemporal hash perception algorithm is used to determine whether the indices of the spatial layout information in the hash table intersect. If they intersect, the collision target point is calculated; otherwise, the time period for simulating the tipping of the storage shelves is adjusted and the bounding box intersection test is re-executed.

[0038] By using a weighted fusion method, the collision target points and impact intensity data between high-risk shelf locations and adjacent targets are superimposed to obtain the impact range of a high-risk shelf location tipping over to adjacent targets.

[0039] Preferably, determining whether the indices of the spatial layout information in the hash table intersect using a spatiotemporal hash-aware algorithm includes:

[0040] Based on the spatial layout information of high-risk shelf locations and adjacent target objects, an initial spatial layout point is randomly selected from the hash table, and the hash value of the initial spatial layout point is calculated.

[0041] The context space layout information of the initial spatial point is stored as an index item in the hash table based on the hash value, and the neighborhood space of the index item in the hash table is traversed and queried.

[0042] The confidence level is calculated using the confidence level formula to determine the position with the highest confidence level for the index item in the neighborhood space, and the confidence level value of the current initial spatial point is updated.

[0043] Calculate the Euclidean distance between adjacent index items based on the confidence value of the current initial spatial point, and determine whether the Euclidean distance between adjacent index items meets the preset conditions;

[0044] If the preset conditions are met, it indicates that the indexes of the risk shelf locations and the spatial layout information of adjacent target objects in the hash table intersect; otherwise, the initial spatial layout points are reselected and the confidence calculation is repeated.

[0045] Preferably, the expression for the connectivity probability function between auxiliary microcrack branches is:

[0046] ;

[0047] In the formula, P conn ( i , j ) indicates the first i The and the first j The connectivity probability between auxiliary microcrack branches; C ( i , j ) indicates the first i The and the first j Geometric connectivity between individual microcrack branches; E ( i , j ) indicates the first i The and the first j Mechanical connectivity of individual microcrack branches; σ ( i , j ) indicates the first i The and the first j The relative values ​​of material strength for each auxiliary microcrack branch; R ( i ) indicates the first i The resistance to damage of individual microcrack branches; R ( j ) indicates the first j The resistance to damage of individual microcrack branches.

[0048] According to another aspect of the present invention, a sophisticated intelligent warehouse security management system is also provided, the system comprising:

[0049] The goods storage and allocation module is used to build a goods storage strategy model based on goods characteristic data, output a goods storage plan through the goods storage strategy model, and allocate storage shelves to the goods according to the goods storage plan.

[0050] The shelving tipping analysis module is used to collect dynamic load indicators of the stored shelving, integrate the dynamic load indicators with predefined environmental indicators, and comprehensively judge the high-risk shelving locations that are prone to tipping events.

[0051] The early warning analysis module is used to identify adjacent targets at high-risk shelving locations and use a spatial collision detection model to predict the impact range of a high-risk shelving location tipping over on adjacent targets. Based on the impact range, warehouse safety early warning measures are formulated.

[0052] The beneficial effects of this invention are as follows:

[0053] 1. This invention, by introducing microcrack propagation path identification, finite element simulation, and an improved slime mold path search algorithm, can accurately assess the health status of shelving structures and dynamically identify potentially high-risk shelving locations. Based on the connectivity of microcracks and structural health assessment, strain concentration areas are identified, and the risk of shelving fracture is predicted in advance. Furthermore, the improved slime mold path search algorithm of this invention enhances the accuracy and robustness of path identification through multi-source population and iterative search, accurately eliminating redundant paths, optimizing the judgment of crack connectivity paths, and improving the reliability of crack propagation prediction. This contributes to improving the safety and reliability of the warehousing system, thereby enhancing the intelligence level and decision support capabilities of warehousing management.

[0054] 2. This invention, by simulating the impact intensity data of a shelf tipping over and combining it with a spatial collision detection model, can assess the impact range of a tipping event on surrounding targets in real time. This allows for the development of reasonable early warning measures for areas with different risk gradients. Furthermore, by using a spatiotemporal hash perception algorithm to determine whether indexes intersect, the invention effectively improves the retrieval efficiency and precision of spatial layout.

[0055] 3. By integrating the three-dimensional spatial layout of the storage environment, finite element simulation, and spatial collision detection model, this invention can achieve accurate prediction and early warning management of the impact range of high-risk shelving locations being tipped over. By combining the impact intensity data generated at each tipping point for simulation and detection, it can not only clearly delineate the affected area, but also construct a multi-gradient risk space, thereby customizing targeted safety early warning measures for areas with different risk levels. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 This is a flowchart of a refined intelligent warehouse safety management method according to an embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of a refined intelligent warehouse safety management system according to an embodiment of the present invention;

[0059] Figure 3 This is a flowchart illustrating the analysis of the impact range of a tipping event on adjacent targets using a spatial collision detection model in a refined intelligent warehouse safety management method according to an embodiment of the present invention.

[0060] In the picture:

[0061] 1. Goods storage and allocation module; 2. Shelf tilting analysis module; 3. Early warning analysis module. Detailed Implementation

[0062] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0063] According to embodiments of the present invention, a refined intelligent warehouse safety management method and system are provided.

[0064] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, a refined intelligent warehouse security management method includes:

[0065] S1. Construct a cargo storage strategy model based on cargo characteristic data, output cargo storage plan through the cargo storage strategy model, and allocate storage shelves to the cargo according to the cargo storage plan.

[0066] It should be noted that the cargo storage strategy model is constructed based on cargo characteristic data. The model outputs a cargo storage plan, and storage shelves are allocated to the cargo according to the plan.

[0067] Collect characteristic data of all goods, including category, volume, weight, frequency of entry and exit from the warehouse, fragility, storage temperature and humidity requirements, hazard level, etc., and standardize them into a computable feature vector.

[0068] A cargo storage strategy model is constructed based on multi-objective optimization. Based on storage priority rules (such as ABC classification, frequency and path optimization association) and constraints (such as goods cannot be stored in the same warehouse or need to be managed in different areas), the warehouse space is discretized in the model to form simulated shelf units. The optimal mapping is performed according to the multi-dimensional attributes of different types of goods to generate cargo storage solutions.

[0069] The implementation of shelf allocation based on the goods storage plan includes: matching goods labels or codes with shelf numbers, updating the storage location mapping table, assigning operators or automated equipment (such as AGVs and forklifts) to perform the initial shelving operation, and setting up dynamic adjustment strategies to ensure that the location can be optimized according to the goods flow trend in the future.

[0070] S2. Collect dynamic load indicators of the storage shelves, integrate the dynamic load indicators with predefined environmental indicators, and comprehensively judge the high-risk shelf locations that are prone to tipping events.

[0071] This includes collecting dynamic load indicators from storage shelves, integrating these indicators with predefined environmental indicators, and comprehensively identifying high-risk shelf locations prone to tipping incidents.

[0072] Construct the topology of the storage rack and introduce the embedded material parameters of the rack to generate a rack structure model that meets the strength constraints.

[0073] It should be noted that the embedded material parameters of the shelving refer to the material properties of the shelving, including but not limited to the material type, structural strength parameters, and coefficient of thermal expansion of the shelving.

[0074] Finite element method (FEM) simulation technology was used to simulate the strain response behavior of a rack structure model under dynamic load and environmental conditions, and the microcrack propagation path of the rack structure model was identified based on the strain response behavior.

[0075] It should be noted that dynamic load indicators include the external dynamic loads borne by the rack during use, including the dynamic gravity of the goods, the vibration and impact loads of the goods, and the vibration of the equipment during operation; environmental indicators are the external environmental factors that affect the rack structure, including environmental conditions such as temperature changes, humidity, air flow, corrosive media, and light.

[0076] The strain response behavior of a rack structure model under dynamic load and environmental conditions was simulated using finite element method (FEM) technology, including:

[0077] Establish a geometric and material model of the shelving structure, including the various components and connections of the shelving, as well as their corresponding physical properties (such as elastic modulus, density, Poisson's ratio, etc.).

[0078] Simulated dynamic loads are applied, including periodic, impulsive, or random external forces and changes caused by environmental factors (such as temperature field, humidity field, etc.). The load time history is set in the finite element simulation software and dynamic analysis is performed.

[0079] The stress, strain, deformation, and displacement responses of the rack structure under different loads and environmental conditions are calculated using time-domain or frequency-domain solution methods. The strain response behavior of the rack structure model under dynamic load and environmental indices is then comprehensively analyzed by combining stress, strain, deformation, and displacement responses.

[0080] Among them, the microcrack propagation path based on the strain response behavior identification of the shelf structure model includes:

[0081] Microcrack features are extracted from the shelf structure model based on strain response behavior, and the microcrack features are integrated to obtain a microcrack topology tree. The connectivity of each microcrack in the microcrack topology tree is then calculated.

[0082] Specifically, microcrack features are extracted from the shelf structure model based on strain response behavior, and these features are integrated to obtain a microcrack topology tree. The connectivity of each microcrack in the microcrack topology tree is calculated, including:

[0083] Microcrack features are extracted based on strain response behavior. Microcrack features with similar characteristics are grouped into the same group to obtain several sets of microcrack features. The connection relationship of microcrack points in the microcrack feature set is used as the edge to generate a microcrack topology tree.

[0084] The main microcrack branches are selected from all edges of the microcrack topology tree according to preset rules, and the main microcrack branches are encoded.

[0085] It should be noted that the main microcrack branch is usually the dominant path during crack propagation. Therefore, the main branch can be determined based on the crack propagation direction during screening. For example, the path with the fastest crack growth rate and stable propagation direction can be selected as the main branch.

[0086] The perturbation combination of the main microcrack branch is analyzed sequentially according to the encoding using sequence derivation technology, and the microcrack failure event of the main microcrack branch is analyzed by introducing perturbation vector.

[0087] It should be noted that the perturbation combination of the main microcrack branches is analyzed sequentially according to the encoding using sequence derivation techniques, and the microcrack failure events of the main microcrack branches are analyzed by introducing perturbation vectors.

[0088] Step 1: Define the perturbation vector and, in conjunction with external factors during crack propagation, such as dynamic loads, environmental changes, and material degradation, establish a perturbation model to simulate the behavior of the main microcrack branches under different time and environmental conditions.

[0089] Step 2: Quantitatively analyze the perturbation effect on crack propagation using perturbation vectors, gradually deduce the crack development path, identify its key change points and turning points, and analyze the specific impact of these changes on the branch propagation of the main microcrack through perturbation combination analysis.

[0090] Step 3: Combine perturbation vectors to monitor microcrack failure events in real time, analyze the various stages of crack propagation and their impact on structural stability; by introducing the triggering conditions of microcrack failure events, simulate the crack propagation process under specific loads, and evaluate the failure probability of the main microcrack branch based on the simulation results of perturbation vectors and failure events.

[0091] Based on the microcrack failure event, the corresponding edges in the microcrack topology tree are removed to generate auxiliary microcrack branches. The connectivity probability function between the auxiliary microcrack branches is constructed, and the crack connectivity flow of the auxiliary microcrack branches is output as the microcrack connectivity.

[0092] It should be noted that the expression for the connectivity probability function between auxiliary microcrack branches is:

[0093] ;

[0094] In the formula, P conn ( i , j ) indicates the first i and the j The connectivity probability between auxiliary microcrack branches; C ( i , j ) indicates the first i and the j The geometric connectivity between microcrack branches depends on the geometry of the crack and the spatial distance between the crack branches; E ( i , j ) indicates the first i and the j Mechanical connectivity of microcrack branches is a factor that measures whether a crack can effectively transmit stress or deformation under mechanical load. σ ( i , j ) indicates the first i and the j The relative strength of materials with microcrack branches varies depending on the material's inhomogeneity or the type of crack. If the crack strength is low, it may be more prone to fracture or propagation. R ( i ) indicates the first i The resistance to damage of individual microcrack branches; R ( j ) indicates the first j The resistance to damage of individual microcrack branches.

[0095] In the shelf structure model, any microcrack point is selected as the starting point, and the selected microcrack point is used as input to search for the set of crack connection paths using the improved slime mold path search algorithm.

[0096] It should be noted that the improved slime mold path search algorithm has efficient optimization and network reconstruction capabilities in constructing a set of crack connected paths, and can adapt to the dynamic changes in crack topology and the uncertainty of paths. However, when applied to storage shelves made of common materials, the improved slime mold path search algorithm is preferentially applicable to materials with high connectivity and brittle propagation characteristics of crack behavior, such as ceramics, glass, and fiber-reinforced composites. For materials with plastic crack resistance mechanisms, such as metals, material crack resistance factors, crack tip energy consumption parameters, or microstructure interference terms should be introduced.

[0097] Taking metallic materials as an example, microcracks in metallic materials typically exhibit a "crack arrest effect" under stress, meaning that a plastic zone or stress-relieving zone appears at the crack tip, preventing further crack propagation or causing crack passivation. The slime mold algorithm essentially relies on a growth-contraction mechanism to simulate path expansion, but the crack arrest property of metals interrupts the continuous growth simulated by the slime mold algorithm, leading to problems such as path discontinuity, misidentification of cracks, and even path fabrication in the stress field.

[0098] Therefore, introducing material crack resistance factors, crack tip energy dissipation parameters, or microstructure interference terms is mainly to improve the adaptability and accuracy of the improved slime mold path search algorithm in tough materials such as metals. For example, the crack propagation resistance factor (including plastic energy dissipation density in the crack tip region, crack tip closing force, crack propagation threshold, etc.) of the material during crack propagation can be extracted through numerical simulation. In the construction stage of the microcrack topology tree, the material microstructure is transformed into a high-resistivity region in the microcrack topology tree. At the same time, the crack propagation resistance factor is embedded as a dynamic update term in the fitness function during the path search process. Under this mechanism, the improved slime mold path search algorithm can find the most likely crack propagation path under the premise of satisfying the physical constraints of the material, thereby effectively improving its applicability and reliability in the structural health monitoring of tough materials such as metals.

[0099] It should be noted that the core idea of ​​the improved slime mold path search algorithm is to simulate the mechanism by which slime molds achieve the optimal resource acquisition path during foraging by releasing mucus, adjusting pipe diameter and flow rate, and selecting paths. Specifically, this includes:

[0100] Step 1: Initialize the microcrack feature space, randomly select several microcrack points as slime mold source points, and initialize variables such as the number of slime mold individuals, pipe diameter flow rate, and behavioral parameters of the slime mold population.

[0101] Step 2: Perform path expansion and flow decay simulation for each slime mold individual at its source location. This involves simulating its crawling towards neighboring nodes and updating the slime flow along the crawling path. If the flow is below a threshold, simulate "shrinkage". Record the evolution trajectory and energy consumption information of each path during this process.

[0102] The fitness of each slime mold body on its current path is evaluated based on the fitness function, which can comprehensively consider multiple indicators such as path length, crack point density, fracture probability, and structural stability. The fitness of all slime mold bodies is ranked, and the current best individual is selected as the main search body.

[0103] Based on the path of the optimal individual, the global path search module is triggered. Through the biomimetic fluid guidance mechanism, multiple paths between the optimal source point and other slime mold source points are iteratively optimized. In this process, the geometric, topological and mechanical matching degree between the extended path and the auxiliary microcrack branch is calculated.

[0104] After path optimization is completed, a path filtering operation is performed. Based on the calculated correlation, extended segments that have a high degree of overlap with or share paths with auxiliary microcrack branches are removed, while non-redundant paths with differences and connectivity are retained.

[0105] Compared to traditional pathfinding algorithms, its key improvements are:

[0106] 1) Introducing biological behavioral mechanisms, combined with pipe diameter shrinkage and flow adjustment mechanisms, no longer relying on deterministic graph structures but simulating dynamic adaptation processes;

[0107] 2) Enhance the algorithm's ability to search for path diversity in complex topologies by using multi-source population co-evolution rather than single-path updates.

[0108] In this process, an arbitrary microcrack point is selected as the starting point in the shelf structure model, and the selected microcrack point is used as input. The improved slime mold path search algorithm is used to search for the set of connected crack paths, including:

[0109] Several microcrack points in the microcrack feature set were randomly selected as slime mold source points, and a slime mold population was generated.

[0110] The shrinkage behavior of each slime mold in the population at the slime mold source point is obtained, the fitness value of the slime mold source point after shrinkage is calculated, and the fitness values ​​are sorted to obtain the optimal slime mold source point;

[0111] Starting from the optimal slime mold source point, the expansion path from the optimal slime mold source point to the other slime mold source points is iteratively searched, and the correlation between the expansion path and the auxiliary microcrack branch is calculated.

[0112] Based on the correlation calculation results, duplicate paths with auxiliary microcrack branches in the expansion path are removed to obtain the crack connectivity path set.

[0113] It should be noted that using the improved slime mold path search algorithm to search for the set of connected crack paths significantly improves the adaptability and accuracy of path search in microcrack structures, enhances the crack propagation model's responsiveness to complex environmental disturbances, avoids the problems of paths getting stuck in local optima or lacking connectivity due to hard-coded cost functions in traditional algorithms, and provides a more bio-inspired and multi-source evaluation theoretical basis for crack controllability analysis and structural optimization design.

[0114] Crack connectivity is used as the evaluation vector to determine the validity of each path in the set of crack connectivity paths, and crack connectivity paths that meet the validity conditions are taken as microcrack propagation paths.

[0115] Based on the microcrack propagation path identification, strain concentration regions of the shelf structure model are identified, and the evolution process of microcracks in the strain concentration regions from initial propagation to complete fracture is tracked.

[0116] The yield strength of strain concentration regions during the evolution of microcracks is analyzed and compared with a preset threshold. If the yield strength is greater than the preset threshold, the current storage rack is marked as a high-risk rack location prone to tipping events; otherwise, the dynamic load index and environmental index are updated.

[0117] It should be noted that during the microcrack evolution process, finite element analysis is used to track the strain changes in the strain concentration area in real time, calculate the yield strength at each moment, and judge the stability of the shelf by comparing the yield strength with a preset threshold.

[0118] If the yield strength is greater than the preset threshold, it indicates that the material in that area is close to its limit and is prone to crack propagation and failure. This area is then marked as a high-risk shelving location, indicating a high risk of tipping over or structural failure.

[0119] If the yield strength is less than the preset threshold, it indicates that the bearing capacity of the area has not yet reached its limit and it can continue to be used. At this time, the dynamic load index and environmental index are updated in real time, and the trend and risk of crack propagation are reassessed based on new environmental or load changes.

[0120] S3. Identify adjacent targets at high-risk shelving locations and use a spatial collision detection model to predict the impact range of a high-risk shelving location tipping over on adjacent targets. Based on the impact range, formulate warehouse safety early warning measures.

[0121] This includes identifying adjacent objects at high-risk shelving locations, using a spatial collision detection model to predict the impact range of a tipping over at a high-risk shelving location on adjacent objects, and developing warehouse safety early warning measures based on the impact range, including:

[0122] The three-dimensional spatial layout of the warehouse environment is obtained from the database, and the spatial location of adjacent target objects at high-risk shelf locations is identified based on the three-dimensional spatial layout.

[0123] Finite element simulation technology is used to simulate the impact intensity data of each tipping point when the storage shelf is tipped over. The impact intensity data is then input into a pre-built spatial collision detection model, which outputs the impact range of the tipping event on adjacent targets.

[0124] Among them, such as Figure 3 As shown, the impact intensity data is input into a pre-built spatial collision detection model. The spatial collision detection model outputs the spillover range of the tipping event on adjacent targets, including:

[0125] Based on the three-dimensional spatial layout of the warehouse environment, the bounding boxes of high-risk shelf locations and adjacent target objects are obtained respectively, and the intersection test of the two sets of bounding boxes is carried out.

[0126] If the bounding boxes of high-risk shelving locations and adjacent targets intersect, the spatial layout information of the high-risk shelving locations and adjacent targets is inserted into the hash table; otherwise, the time period for simulating shelving tipping is adjusted and the bounding box intersection test is re-executed.

[0127] The spatiotemporal hash perception algorithm is used to determine whether the indices of the spatial layout information in the hash table intersect. If they intersect, the collision target point is calculated; otherwise, the time period for simulating the tipping of the storage shelves is adjusted and the bounding box intersection test is re-executed.

[0128] It should be noted that the spatiotemporal hash-aware algorithm is an algorithm that combines temporal and spatial information for efficient data indexing and querying, primarily used for spatial layout analysis in dynamic environments. It uses hashing technology to map each location in space to a hash table index, storing spatial layout information within this index. A spatiotemporal awareness mechanism is then used to dynamically perceive and update the spatial layout. The core idea of ​​the algorithm is to consider the temporal changes and relative positions of spatial points to determine and track in real time whether different spatial layouts intersect or interact.

[0129] Among them, using the spatiotemporal hash awareness algorithm to determine whether the indices of spatial layout information in a hash table intersect includes:

[0130] Based on the spatial layout information of high-risk shelf locations and adjacent target objects, an initial spatial layout point is randomly selected from the hash table, and the hash value of the initial spatial layout point is calculated.

[0131] The context space layout information of the initial spatial point is stored as an index item in the hash table based on the hash value, and the neighborhood space of the index item in the hash table is traversed and queried.

[0132] The confidence level is calculated using the confidence level formula to determine the position with the highest confidence level for the index item in the neighborhood space, and the confidence level value of the current initial spatial point is updated.

[0133] It should be noted that the confidence level is calculated using the following formula:

[0134] ;

[0135] In the formula, C ab Represents the initial space point a and adjacent index points b The confidence level between them; ab : Represents the initial space point a and adjacent spatial points b The Euclidean distance between them; η Sensitivity coefficient representing confidence level; δ The threshold used for confidence level calculation; exp This represents an exponential function.

[0136] Calculate the Euclidean distance between adjacent index items based on the confidence value of the current initial spatial point, and determine whether the Euclidean distance between adjacent index items meets the preset conditions;

[0137] If the preset conditions are met, it indicates that the indexes of the risk shelf locations and the spatial layout information of adjacent target objects in the hash table intersect; otherwise, the initial spatial layout points are reselected and the confidence calculation is repeated.

[0138] By using a weighted fusion method, the collision target points and impact intensity data between high-risk shelf locations and adjacent targets are superimposed to obtain the impact range of a high-risk shelf location tipping over to adjacent targets.

[0139] It should be noted that by introducing a spatiotemporal hashing awareness algorithm, dynamic and efficient indexing and intersection judgment of the spatial layout information of high-risk shelf locations and adjacent targets can significantly improve the accuracy and computational efficiency of spatial intersection detection. Specifically, using a hash function to map spatial points to index items can effectively reduce the data complexity brought by spatial dimensions; by traversing the neighborhood space and combining confidence calculation and Euclidean distance judgment, rapid identification of spatial information intersection status is achieved; when it is found that the preset conditions are not met, it has adaptive backtracking capability, which can continuously adjust the initial layout points and repeat the confidence evaluation to ensure the stability and robustness of detection.

[0140] Based on the sweep range and impact intensity of adjacent targets, the spatial locations of adjacent targets are divided to obtain a multi-gradient target risk space. Warehouse safety early warning measures are then formulated for each of these multi-gradient target risk spaces, specifically including:

[0141] Based on the spatial distribution of the affected areas, the space of adjacent targets is divided into multiple risk gradient zones according to the magnitude of the impact intensity and the range of the impact. Each gradient zone represents a different risk level. Potential security threats are assessed for each gradient zone. Based on the risk level of different gradient zones, corresponding warehouse security early warning measures are designed, such as increasing security isolation in high-risk areas, strengthening monitoring in medium-risk areas, and conducting routine inspections in low-risk areas. A dynamic monitoring mechanism is established to regularly update the risk gradient space and adjust early warning measures according to actual environmental changes to ensure the safety of the warehouse environment.

[0142] It should be noted that by comprehensively utilizing 3D spatial layout, hash tables, spatiotemporal awareness algorithms, and collision analysis, high-risk shelf locations in the warehouse environment and their potential impact on surrounding objects can be dynamically and accurately identified, allowing for early prediction and prevention of tipping events. Bounding box intersection tests accurately determine the spatial relationship between high-risk shelf locations and adjacent objects; if intersections exist, the relevant information is promptly stored in a hash table to provide data support for subsequent analysis.

[0143] Based on the impact range and intensity, the risk space of adjacent targets is divided into multiple gradients, and corresponding warehouse safety early warning measures are formulated for each gradient, forming a complete and dynamically updated warehouse safety management system, which effectively improves the predictive ability and response efficiency of shelf tipping risk in the warehouse environment.

[0144] According to another embodiment of the invention, such as Figure 2 As shown, a sophisticated intelligent warehouse security management system is also provided, which includes:

[0145] The goods storage and allocation module 1 is used to build a goods storage strategy model based on goods characteristic data, output a goods storage plan through the goods storage strategy model, and allocate storage shelves to the goods according to the goods storage plan.

[0146] The shelf tipping analysis module 2 is used to collect dynamic load indicators of the stored shelves, integrate the dynamic load indicators with predefined environmental indicators, and comprehensively judge the high-risk shelf locations that are prone to tipping events.

[0147] The early warning analysis module 3 is used to identify adjacent targets at high-risk shelf locations and use a spatial collision detection model to predict the impact range of a high-risk shelf location tipping over on adjacent targets. Based on the impact range, warehouse safety early warning measures are formulated.

[0148] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A refined intelligent warehouse safety management method, characterized in that, The method includes: S1. Assign storage shelves to the goods; S2. Construct the topology of the storage rack and introduce the embedded material parameters of the rack to generate a rack structure model that meets the strength constraints. Use finite element simulation technology to simulate the strain response behavior of the rack structure model under dynamic load and environmental indicators, and identify the microcrack propagation path of the rack structure model based on the strain response behavior. Identify the strain concentration area of ​​the rack structure model based on the microcrack propagation path, track the evolution process of microcracks in the strain concentration area from the beginning of propagation to complete fracture, and comprehensively judge the high-risk rack locations that are prone to tipping events. The embedded material parameters of the rack are the material properties of the storage rack, including at least the material type, structural strength parameters and thermal expansion coefficient of the storage rack. The dynamic load indicators include the external dynamic loads borne by the rack during use, including the dynamic gravity of the goods, the vibration and impact load of the goods, and the vibration of the equipment during operation. The environmental indicators include temperature changes, humidity, air flow, corrosive media and light. The rack structure model includes the geometric model and material model of the rack structure, including the various components, connection parts and corresponding physical properties of the rack. S3. Identify adjacent targets at high-risk shelving locations and use a spatial collision detection model to predict the impact range of a high-risk shelving location tipping over on adjacent targets. Based on the impact range, formulate warehouse safety early warning measures. The microcrack propagation path of the shelf structure model based on strain response behavior recognition includes: Microcrack features are extracted from the shelf structure model based on strain response behavior, and the microcrack features are integrated to obtain a microcrack topology tree. The connectivity of each microcrack in the microcrack topology tree is calculated. In the shelf structure model, any microcrack point is selected as the starting point, and the selected microcrack point is used as input to search for the set of crack connection paths using the improved slime mold path search algorithm. Crack connectivity is used as the evaluation vector to determine the validity of each path in the set of crack connectivity paths, and the crack connectivity paths that meet the validity conditions are taken as microcrack propagation paths. The process involves extracting microcrack features from the shelf structure model based on strain response behavior, integrating these features to obtain a microcrack topology tree, and calculating the connectivity of each microcrack in the topology tree, including: Microcrack features are extracted based on strain response behavior. Microcrack features with similar characteristics are grouped into the same group to obtain several sets of microcrack features. The connection relationship of microcrack points in the microcrack feature set is used as the edge to generate a microcrack topology tree. The main microcrack branches are selected from all edges of the microcrack topology tree according to preset rules, and the main microcrack branches are encoded. The perturbation combination of the main microcrack branches is analyzed sequentially according to the encoding using sequence derivation technology, and the microcrack failure event of the main microcrack branches is analyzed by introducing perturbation vectors. Based on the microcrack failure event, the corresponding edges in the microcrack topology tree are removed to generate auxiliary microcrack branches. The connectivity probability function between the auxiliary microcrack branches is constructed, and the crack connectivity flow of the auxiliary microcrack branches is output as the microcrack connectivity. The step of selecting any microcrack point in the shelf structure model as the starting point and using the selected microcrack point as input to search for the set of connected crack paths using the improved slime mold path search algorithm includes: Several microcrack points in the microcrack feature set were randomly selected as slime mold source points, and a slime mold population was generated. The shrinkage behavior of each slime mold in the population at the slime mold source point is obtained, the fitness value of the slime mold source point after shrinkage is calculated, and the fitness values ​​are sorted to obtain the optimal slime mold source point; Starting from the optimal slime mold source point, the expansion path from the optimal slime mold source point to the other slime mold source points is iteratively searched, and the correlation between the expansion path and the auxiliary microcrack branch is calculated. Based on the correlation degree calculation results, duplicate paths with auxiliary microcrack branches in the expansion path are removed to obtain the crack connectivity path set; The expression for the connectivity probability function between the auxiliary microcrack branches is: ； In the formula, P conn ( i , j ) indicates the first i The and the first j The connectivity probability between each auxiliary microcrack branch; C ( i , j ) indicates the first i The and the first j The geometric connectivity between individual microcrack branches is used to characterize the geometry of the crack and the spatial distance between crack branches; E ( i , j ) indicates the first i The and the first j The mechanical connectivity of individual microcrack branches is used to characterize factors that measure whether a crack can effectively transmit stress or deformation under mechanical load. σ ( i , j ) indicates the first i The and the first j The relative values ​​of material strength for each auxiliary microcrack branch; R ( i ) indicates the first i The resistance to damage of individual microcrack branches; R ( j ) indicates the first j The resistance to damage of individual microcrack branches.

2. The refined intelligent warehouse safety management method according to claim 1, characterized in that, The aforementioned shelves for allocating and storing goods include: A cargo storage strategy model is constructed based on cargo characteristic data. The cargo storage strategy model outputs a cargo storage plan, and storage shelves are allocated to the cargo according to the cargo storage plan. The high-risk shelving locations identified as prone to tipping incidents include: The yield strength of strain concentration regions during microcrack evolution was analyzed, and the yield strength was compared with a preset threshold. If the yield strength is greater than the preset threshold, the current storage shelf will be marked as a high-risk shelf location prone to tipping events; otherwise, the dynamic load index and environmental index will be updated.

3. The refined intelligent warehouse safety management method according to claim 1, characterized in that, The process of identifying adjacent objects at high-risk shelving locations and using a spatial collision detection model to predict the impact range of a tipping-over at a high-risk shelving location on adjacent objects, and then developing warehouse safety early warning measures based on the impact range, includes: The three-dimensional spatial layout of the warehouse environment is obtained from the database, and the spatial location of adjacent target objects at high-risk shelf locations is identified based on the three-dimensional spatial layout. The finite element simulation technology is used to simulate the impact intensity data of each tipping point when the storage shelf is tipped over. The impact intensity data is then input into a pre-built spatial collision detection model, which outputs the range of impact of the tipping event on adjacent targets. Based on the impact range and intensity of adjacent targets, the spatial location of adjacent targets is divided to obtain a multi-gradient target risk space, and warehouse safety early warning measures are formulated for each multi-gradient target risk space.

4. The refined intelligent warehouse safety management method according to claim 3, characterized in that, The step of inputting impact intensity data into a pre-built spatial collision detection model, and outputting the spillover range of the tipping event on adjacent targets through the spatial collision detection model, includes: Based on the three-dimensional spatial layout of the warehouse environment, the bounding boxes of high-risk shelf locations and adjacent target objects are obtained respectively, and the intersection test of the two sets of bounding boxes is carried out. If the bounding boxes of high-risk shelving locations and adjacent targets intersect, the spatial layout information of the high-risk shelving locations and adjacent targets is inserted into the hash table; otherwise, the time period for simulating shelving tipping is adjusted and the bounding box intersection test is re-executed. The spatiotemporal hash perception algorithm is used to determine whether the indices of the spatial layout information in the hash table intersect. If they intersect, the collision target point is calculated; otherwise, the time period for simulating the tipping of the storage shelves is adjusted and the bounding box intersection test is re-executed. By using a weighted fusion method, the collision target points and impact intensity data between high-risk shelf locations and adjacent targets are superimposed to obtain the impact range of a high-risk shelf location tipping over to adjacent targets.

5. The refined intelligent warehouse safety management method according to claim 4, characterized in that, The method of using a space-time hash awareness algorithm to determine whether the indices of spatial layout information in a hash table intersect includes: Based on the spatial layout information of high-risk shelf locations and adjacent target objects, an initial spatial layout point is randomly selected from the hash table, and the hash value of the initial spatial layout point is calculated. The context space layout information of the initial spatial point is stored as an index item in the hash table based on the hash value, and the neighborhood space of the index item in the hash table is traversed and queried. The confidence level is calculated using the confidence level formula to determine the position with the highest confidence level for the index item in the neighborhood space, and the confidence level value of the current initial spatial point is updated. Calculate the Euclidean distance between adjacent index items based on the confidence value of the current initial spatial point, and determine whether the Euclidean distance between adjacent index items meets the preset conditions; If the preset conditions are met, it indicates that the indexes of the risk shelf locations and the spatial layout information of adjacent target objects in the hash table intersect; otherwise, the initial spatial layout points are reselected and the confidence calculation is repeated.

6. A sophisticated intelligent warehouse safety management system, used to implement the intelligent warehouse safety management method according to any one of claims 1-5, characterized in that, The system includes: The goods storage and allocation module is used to allocate storage shelves for goods; The shelving tipping analysis module is used to construct the topology of the storage shelving and introduce the embedded material parameters of the shelving to generate a shelving structural model that meets the strength constraints. It uses finite element simulation technology to simulate the strain response behavior of the shelving structural model under dynamic load and environmental indices, and identifies the microcrack propagation path of the shelving structural model based on the strain response behavior. Based on the microcrack propagation path, it identifies the strain concentration area of ​​the shelving structural model, tracks the evolution process of microcracks in the strain concentration area from the beginning of propagation to complete fracture, and comprehensively judges the high-risk shelving locations that are prone to tipping events. The early warning analysis module is used to identify adjacent targets at high-risk shelving locations and use a spatial collision detection model to predict the impact range of a high-risk shelving location tipping over on adjacent targets. Based on the impact range, warehouse safety early warning measures are formulated.

Images