Warehouse goods in and out management method and system

Abstract

The application relates to the technical field of warehouse logistics management, in particular to a warehouse goods warehouse-in and warehouse-out management method and system. The method comprises the following steps: obtaining historical order data of warehousing; based on the historical order data weighted by a time decay function, calculating the weighted occurrence frequency of each commodity and the weighted co-occurrence frequency of a commodity pair to determine the correlation degree between the commodity pair; determining the adjacent storage value of the commodity pair according to the average value of the correlation degree between the commodity pair, the weighted co-occurrence frequency, the storage category and the weight difference; based on all available storage locations, the adjacent storage value of the commodity pair and the weighted occurrence frequency of each commodity, constructing an optimization model with the target of minimizing the total picking cost, solving the optimization model to obtain a storage location allocation scheme; and performing the warehouse-in and warehouse-out operation according to the storage location allocation scheme. The application can optimize the picking path and improve the warehousing operation efficiency by evaluating the commodity correlation and the physical constraint.

Description

Technical Field

[0001] This invention relates to the field of warehousing and logistics management technology, and in particular to a method and system for managing the inbound and outbound operations of stored goods. Background Technology

[0002] Warehouse Management Systems (WMS) are the core of modern logistics and supply chain management, and their operational efficiency directly determines a company's responsiveness to market demands and customer satisfaction. Within the current technological framework, WMS utilizes automated identification technologies such as barcodes, QR codes, or radio frequency identification (RFID) to achieve basic digital management of key processes such as goods receiving, shelving, storage, picking, and outbound shipments, providing transparency and accuracy for warehousing operations.

[0003] Among the many aspects of warehouse optimization, the location allocation strategy is a key factor in determining operational efficiency. The industry has long adopted the ABC classification method, which evolved from the Pareto principle. This method divides goods based on the single dimension of historical outbound frequency. It places A-class goods, which contribute the majority of outbound volume, in the location closest to the outbound area and most convenient for picking, while B-class and C-class goods are placed in relatively distant locations. In the traditional model with a relatively simple order structure, this strategy has indeed effectively shortened the single picking path of high-frequency best-selling products and improved the picking efficiency of individual items.

[0004] However, this strategy has limitations: it focuses only on the independent attributes of individual goods, while ignoring the common correlations between different goods in actual orders. In business scenarios, many different goods may be picked simultaneously in the same order due to customer consumption habits or production support needs. If these strongly related goods are scattered and stored in various parts of the warehouse using the ABC classification method, picking personnel or equipment must move extensively within the warehouse to complete an order. This not only increases the ineffective walking distance but also directly leads to a decrease in the overall order processing efficiency. Summary of the Invention

[0005] To address the technical problem that the ABC classification-based warehouse location allocation strategy relies solely on the historical outbound frequency of goods, neglecting the correlation between goods within an order, which leads to increased picking paths and decreased processing efficiency for complex orders, this invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for managing the inbound and outbound operations of stored goods, applicable to a warehouse containing several storage locations and a common picking exit, the method comprising the steps of: Obtain historical order data from the warehouse; select a first item and a second item from the warehouse, denoted as an item pair; based on the historical order data, weight the order completion time using a preset time decay weight, and calculate the weighted occurrence count of each individual item and the weighted co-occurrence count of the item pair; divide the weighted co-occurrence count of the item pair by the weighted occurrence count of the first item and the second item respectively to obtain a first correlation and a second correlation; determine the proximity storage value of the item pair based on the mean of the first correlation and the second correlation, the weighted co-occurrence count of the item pair, and the storage category and weight difference of the item pair; construct an optimization model with the objective of minimizing the total picking cost based on all available storage locations, the proximity storage value of the item pair, and the weighted occurrence count of each item, and solve the optimization model to obtain a storage location allocation scheme; execute inbound and outbound operations according to the storage location allocation scheme.

[0007] This invention first processes historical order data by introducing time decay weights to calculate the weighted occurrence frequency and weighted co-occurrence frequency, reflecting market timeliness, and determines the first and second correlation degrees based on this, thus uncovering the real and dynamic business correlations between goods. Next, it integrates business correlations with the physical storage constraints of goods to construct proximity storage value, realizing the mapping from abstract data to physical constraints. Finally, it constructs and solves a dual-objective total picking cost optimization model, which minimizes not only the common picking cost of high-frequency goods to the exit but also the internal movement cost between highly correlated goods. By combining dynamic business correlations with physical constraints for global optimization, this invention enables the warehouse location allocation scheme to simultaneously meet the needs of high-frequency goods being close to the exit and highly correlated goods being close to each other, reducing the total picking path and improving the overall efficiency of warehousing operations.

[0008] Preferably, determining the proximity storage value of the product pair based on the mean of the first correlation degree and the second correlation degree, the weighted co-occurrence frequency of the product pair, and the storage category and weight difference of the product pair includes: determining the physical compatibility function of the product pair based on the storage category and weight difference of the product pair; and multiplying the mean of the first correlation degree and the second correlation degree, the weighted co-occurrence frequency of the product pair, and the value of the physical compatibility function to determine the proximity storage value.

[0009] This invention ensures that only pairs of goods that simultaneously meet the requirements of strong business association and high picking frequency can obtain high proximity storage value by multiplying the mean correlation, weighted co-occurrence frequency, and physical compatibility function. When goods are physically incompatible, such as in storage categories, the proximity storage value is zero. This ensures the security and compliance of the warehouse location allocation scheme and avoids incorrectly placing incompatible goods together.

[0010] Preferably, obtaining the physical compatibility function includes: when the two goods in the goods pair have the same storage category, the category compatibility is recorded as 1, and when the storage categories are different, the category compatibility is recorded as 0; calculating the absolute weight difference between the two goods in the goods pair, multiplying the absolute weight difference by a preset weight difference sensitivity coefficient and taking the negative value to obtain the weight difference factor; and recording the product of the calculation result with the natural constant as the base and the weight difference factor as the power and the category compatibility as the physical compatibility function.

[0011] This invention constructs a two-layer physical constraint model. Category compatibility ensures that different categories of goods, such as refrigerated goods and room-temperature goods, are not stored in adjacent locations, thus meeting basic warehousing safety requirements. The weight difference factor models the weight difference between goods through a negative exponential function. When the weight difference between goods is large, the function value gradually approaches zero, effectively reducing the possibility of them being stored adjacent to each other. This mechanism, while ensuring warehousing safety, further optimizes the human-machine coordination and the rationality of the operation process in picking operations.

[0012] Preferably, the total picking cost is the weighted sum of internal movement cost and public picking cost; wherein, the internal movement cost is the sum of the products of the proximity storage value of all item pairs and the walking distance between the respective storage locations of the two items contained in the corresponding item pair, and the public picking cost is the sum of the products of the weighted number of occurrences of all items and the walking distance from the respective storage location of the corresponding item to the public picking exit.

[0013] This invention resolves long-standing warehouse layout conflicts by explicitly defining total picking cost as the weighted sum of internal movement cost and common picking cost. Internal movement cost brings closely related goods closer together to optimize picking efficiency for multi-category orders; common picking cost brings high-frequency goods closer to the exit to optimize picking efficiency for single-category orders. This invention allows for flexible adjustment of optimization objectives based on the actual business model of the warehouse by adjusting the weights of the two costs, thereby finding a balance between the two costs and achieving optimal warehouse location layout.

[0014] Preferably, the optimization model aimed at minimizing total picking costs includes constraints, which include: the sum of the weights of all goods assigned to any storage location does not exceed the maximum load-bearing capacity of the corresponding storage location; the sum of the volumes of all goods assigned to any storage location does not exceed the available space capacity of the corresponding storage location; and the storage category of any goods matches the allowed storage category of the assigned storage location.

[0015] Preferably, the preset time decay weight is a negative exponential function, and its independent variable is the product of the difference between the current time and the order completion time and the preset time decay coefficient.

[0016] Preferably, obtaining the weighted co-occurrence count of the product pair includes: obtaining orders from historical orders that simultaneously contain both products of the product pair, denoted as the target order set; for each target order in the target order set, obtaining the total number of product types in the target order and the preset time decay weight corresponding to the completion time of the target order; dividing the preset time decay weight by the difference between the total number of product types and 1 to obtain the contribution value of the target order; and summing the contribution values ​​of all target orders in the target order set to obtain the weighted co-occurrence count of the product pair.

[0017] This invention achieves timeliness weighting by accumulating contribution values ​​and relating them to a time decay function. The normalization of contribution values ​​reduces the noise weight caused by the accidental co-occurrence of any two goods in large orders, while amplifying the stable recurrence of associations in a large number of small orders. This enables the discovery of more realistic and stable product associations.

[0018] Preferably, the step of performing inbound and outbound operations according to the storage location allocation scheme includes: during inbound management, allocating goods to the target storage location determined by the storage location allocation scheme; during outbound management, generating a picking list sorted by storage location according to the storage location allocation scheme, and performing picking.

[0019] Preferably, the solution to the optimization model to obtain the storage location allocation scheme includes: solving the optimization model using a heuristic algorithm, wherein the heuristic algorithm includes one of genetic algorithm, simulated annealing algorithm, and tabu search algorithm.

[0020] In a second aspect, the present invention provides a warehouse goods inbound and outbound management system, which includes a memory and a processor. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a warehouse goods inbound and outbound management method of the first aspect of the present invention is implemented.

[0021] By adopting the above technical solution, a computer program for the warehouse goods entry and exit management method of the first aspect of the present invention is generated and stored in a memory so that it can be loaded and executed by a processor, thereby creating a terminal device based on the memory and the processor for convenient use.

[0022] The beneficial effects of this invention are as follows: First, starting from historical order data, this invention calculates the weighted co-occurrence frequency and correlation degree, reflecting market dynamics and actual purchase combinations, by introducing time decay weights and order size normalization factors. Based on this, it further integrates the physical storage compatibility between goods to construct a proximity storage value index. Finally, it establishes a dual-objective total picking cost optimization model. This model not only places high-frequency goods near the picking exit but also centrally arranges strongly correlated goods with high proximity storage value. In this way, this invention can assess and utilize the hidden correlation of goods in orders, making warehouse location allocation not limited to a single dimension, reducing the invalid walking distance caused by scattered storage of goods during the picking process, and improving the overall processing efficiency of complex orders. Attached Figure Description

[0023] Figure 1 A flowchart illustrating a method for managing the inbound and outbound operations of stored goods, provided as an embodiment of the present invention; Figure 2 This is a structural block diagram of a warehouse goods inbound and outbound management system provided in an embodiment of the present invention. Detailed Implementation

[0024] The first aspect of this invention provides a method for managing the inbound and outbound operations of stored goods, applicable to a warehouse containing several storage locations and a common picking exit, such as... Figure 1 As shown, the method includes steps S100-S400: Step S100: Obtain historical order data from the warehouse.

[0025] It's important to note that historical outbound orders, as a core record of customer demand and goods circulation, contain the patterns and characteristics of goods flow. This data reflects key information such as the frequency and combination patterns of different goods leaving the warehouse in actual business operations, serving as the original basis for subsequent layout optimization. Without systematic and complete historical order data, any solution will lack data support, potentially deviating from actual business needs and leading to poor optimization results. Therefore, the first step is to comprehensively collect historical order data.

[0026] Specifically, detailed records of all completed outbound orders within a preset time period are retrieved from the WMS database. The preset time period can be set according to business characteristics, such as the past six months or one year, to balance data timeliness and statistical representativeness. Each record must contain at least four key fields: a unique order identifier to distinguish different orders; a timestamp of order completion to record the order's time attribute; a list of unique codes for all goods included in the order to clarify the order's goods composition; and the total number of goods types in the order to reflect the order's scale characteristics. This structured information lays the data foundation for subsequent analysis and model building.

[0027] At this point, the historical order data has been obtained.

[0028] Step S200: Select the first item and the second item from the warehouse, and record them as an item pair. Based on the historical order data, use a preset time decay weight to weight the order completion time, and calculate the weighted occurrence count of each individual item and the weighted co-occurrence count of the item pair. Divide the weighted co-occurrence count of the item pair by the weighted occurrence count of the first item and the second item respectively to obtain the first correlation and the second correlation.

[0029] It's important to note that, considering WMS orders are presented as scattered records of product combinations, this type of data only reflects basic information about which products are included in a given order. It cannot intuitively show whether there are stable relationships between different products, nor can it assess the strength of those relationships, making it difficult to discover hidden patterns in product combinations within the order data. Graph network models, with their unique topological structure, can solve this problem, effectively compensating for the shortcomings of fragmented data. This model abstracts each type of product as a node, connects co-occurring products within the same order with edges, and then accurately assesses the strength of the relationships through edge weights, allowing hidden patterns in the original data to be discovered.

[0030] When constructing a correlation graph, two key factors need to be considered to ensure the accuracy and timeliness of the model. First, market demand is dynamic; seasonality, promotional activities, and other factors cause the correlation between products to fluctuate over time. If all historical data are treated equally, the analysis results will lag behind the current market environment. Therefore, a time decay weight needs to be introduced to assign differentiated importance to order data from different periods, ensuring that recent market behavior has a greater impact on the model. Second, the interference of order size differences on correlation analysis needs to be eliminated. In real data, large orders containing hundreds of products will generate thousands of product pairs, while small orders containing only two products will generate only one product pair. If co-occurrence frequency is directly counted, the model will amplify the occasional product combinations in large orders while ignoring the core correlation combinations that stably recur in a large number of small orders. Therefore, an order size factor needs to be introduced to correct this bias.

[0031] Specifically, the first step is to calculate two metrics needed for the graph construction process: the weighted frequency of occurrence of a single item and the weighted co-occurrence frequency of two items.

[0032] To assess the timeliness and popularity of goods in the current market environment, this invention utilizes an existing exponential decay model to assign weights to historical orders containing the goods, thereby realizing the characteristic that data value decreases over time. Based on the above logic, for a single product... The weighted frequency of occurrence satisfies the following relationship: ; in: Goods Weighted frequency of occurrence; It includes all goods. orders A set; It is the current time; It is an order The time of completion; It is a preset time decay coefficient used to control the rate at which order weight decays over time.

[0033] In this relation, Used for all containing Orders are assigned a time-sensitivity weight, which is essentially a time decay weight. The magnitude of the time decay weight depends on... , Order completion time Distance from current time The larger the time difference, the older the order, and the closer the value of the exponent is to zero, meaning the lower the weight of that order. Conversely, the more recent the order, the closer its weight is to zero. .final, By summing all these weights, the goods are obtained. The weighted frequency of occurrence reflects its timeliness and popularity.

[0034] After obtaining the weighted occurrence count of a single item, the weighted co-occurrence count of two items is further calculated to measure the essential correlation strength between their co-purchase. For this indicator, not only time decay must be considered, but statistical bias caused by differences in order size must also be eliminated. This invention improves upon the existing exponential decay model by introducing an order size normalization factor. Based on the above logic, the weighted co-occurrence count of the two items... and The weighted co-occurrence counts satisfy the following relationship: ; in, Goods and The weighted co-occurrence count; It includes all goods at the same time and orders A set; It is an order The total number of product categories included. ; It is the current time; It is an order The time of completion; It is the preset time decay coefficient.

[0035] In this relation, Based on, added As a normalization factor to eliminate statistical bias caused by differences in order size, this normalization factor divides the time-weighted contribution value of each order by its size factor. For example, a scale of The large orders contributed only to [their] contribution. ; and the scale is Small orders contribute to When the order size from Increase to At that time, normalization factor The value from Reduce to approximately This reduces the weight of the accidental co-occurrence of any two items in a large order. Ultimately, By summing up all these normalized contribution values, the calculation results can more accurately reflect the essential correlation of customer purchasing behavior.

[0036] It should be noted that the time decay coefficient The settings need to be determined based on the business scenario. If the market trends of the stored goods change extremely rapidly, such as fast-moving consumer goods or fresh produce, a higher setting should be used. Values, for example This allows the model to focus on recent data. If the correlation between goods is very stable, such as hardware tools and industrial spare parts, a lower setting should be used. Values, for example This allows the model to have a longer memory cycle. As a preferred implementation, it is preferable to... In order to strike a balance between responding to market changes and maintaining data stability.

[0037] Finally, calculate from the goods Pointing to goods correlation and from goods Pointing to goods correlation , for and The ratio is used to assess the effectiveness of an order when goods appear in an order, after excluding the effects of time decay and order size differences. At the same time, the order also included goods. The higher the value of the corrected conditional probability, the more it indicates that... arrive The stronger the picking guidance relationship; for and The ratio is used to assess the effectiveness of an order when goods appear in an order, after excluding the effects of time decay and order size differences. At the same time, the order also included goods. The higher the value of the corrected conditional probability, the more it indicates that... arrive The stronger the picking guidance relationship, the better.

[0038] In a preferred embodiment, it is necessary to screen valid item pairs to eliminate the interference of weakly associated combinations on the optimization results and reduce computational complexity. Specifically, this includes calculating the product pairs for all items. Then, a time-varying weighted picking association graph is constructed. ,in It is a set of nodes; each unique item in the warehouse corresponds to a node in the graph. It is an edge set, for any pair of nodes If its If the value is greater than the preset minimum association threshold, a slave node is created in the graph. Pointing to node The directed edges; It is a set of edges This is the set of weight values ​​for each edge in the graph. It should be noted that the minimum association threshold is used to filter weak associations that occur only occasionally. It is set based on the distribution characteristics of historical association data or the required business accuracy. If the association degree is generally low in historical data, it can be set to 0.03; if it is necessary to strengthen core associations, the threshold can be increased to 0.08. In this embodiment, it is preferably set to 0.05. Subsequent actions will only be based on the edge set of the association graph. Goods inside Calculate the value of nearby storage .

[0039] At this point, the first and second correlation degrees of the goods pair have been completed.

[0040] Step S300: Determine the proximity storage value of the product pair based on the mean of the first correlation degree and the second correlation degree, the weighted co-occurrence frequency of the product pair, the storage category and weight difference of the product pair.

[0041] It should be noted that the degree of relevance and goods and Weighted co-occurrence count This involves abstract business-related data. To transform this data into comprehensive decision-making indicators that directly guide physical layout, it's necessary to combine business relationships with physical constraints. Considering the high probability that another item will be picked simultaneously in the same order when one item is being picked, indicating a strong business relationship, storing them adjacently reduces cross-regional movement during picking and improves operational efficiency. Furthermore, if a pair of items frequently co-occurs in historical orders, its high-frequency co-occurrence characteristic will accumulate significant path-saving benefits over long-term operation, making prioritizing such combinations more valuable. In addition, item pairs must meet physical conditions such as consistent storage category and weight compatibility to ensure the safety and feasibility of warehousing operations. Based on these factors, this invention integrates relationship strength, co-occurrence frequency, and physical compatibility to construct a proximity storage value index for evaluating the overall storage priority of each item pair.

[0042] Based on the above analysis, this step aims to construct the value of proximity storage. It is important to note the correlation degree. It is asymmetrical, meaning that from the goods Arrived goods With goods Arrived goods The degree of correlation may differ, but the actual physical layout is symmetrical, and the goods... Neighboring goods With goods Neighboring goods Since it's the same layout, calculations are needed first. The asymmetric business correlation is converted into symmetric weights by taking the average value to match the characteristics of the physical layout. Based on this, the symmetric weights are combined with the weighted co-occurrence count. Multiplication ensures that only strongly correlated and frequently occurring combinations yield high values ​​through the multiplication property; finally, a physical compatibility function is introduced as a multiplier to filter out physically incompatible combinations.

[0043] Based on the above logic, the value of proximity storage satisfies the following relationship: ; in, Goods With goods The value of nearby storage; From goods Arrived goods The degree of correlation; From goods Arrived goods The degree of correlation; Goods and The weighted co-occurrence count; Goods and goods The physical compatibility function is a function whose range is in The scalar values ​​between these are explained in detail below.

[0044] In this relation, the symmetric weights The introduction of this feature resolved the contradiction between the asymmetry of business relationships and the symmetry of physical layout, due to the physical goods... Neighboring goods With goods Neighboring goods If the layout is the same, taking the average of the two-way correlation can ensure that the correlation strength assessment matches the spatial characteristics, avoid layout deviations caused by one-way correlation, and ensure that the layout decision is consistent with the actual spatial characteristics. As the weighted co-occurrence frequency of item pairs, multiplied by the symmetric weights, to ensure high... Values ​​are derived only from strongly correlated pairs of goods that are frequently picked simultaneously, ensuring that the algorithm prioritizes combinations that yield the greatest overall benefit; physical compatibility function. The correction is achieved through the range characteristic: if they are physically incompatible, such as dangerous goods and ordinary goods, then... , If there is slight incompatibility, such as storage temperature and humidity requirements being close but not completely consistent, the business value is discounted proportionally. This complies with safety regulations while retaining some related benefits, making the indicators more realistic.

[0045] Physical compatibility function The design is based on the hierarchical nature of physical constraints in warehousing scenarios: some constraints are insurmountable limitations, such as differences in storage categories, which, if violated, would lead to security risks or operational infeasibility; others are limitations that affect efficiency, such as weight differences, which, while not absolutely prohibiting adjacent storage, would increase operational difficulty if the differences were too large. Therefore, the physical compatibility function needs to handle these two types of constraints in a hierarchical manner, ensuring both security and operational rationality.

[0046] Based on the above logic, the physical compatibility function The following relationship must be satisfied: ; in, Goods and goods Physical compatibility functions; , These are goods , Storage categories, such as ambient temperature, refrigerated, and hazardous materials; It is Kronek Function, when storage class When, its value is Otherwise ; It is the weight difference sensitivity coefficient, and its unit is the reciprocal of the weight unit, which is a preset positive number; , These are goods , The weight; It is the absolute value symbol.

[0047] In this relation, through Inspect the goods and Storage category, if the categories do not match: for example, room temperature items and refrigerated items, ,lead to , Also for Firstly, from an economic perspective, proximity between the two is a no-fault solution; secondly, if the categories match... Then through To check for weight differences, use exponential decay to assess weight differences. Adjustments are made based on weight differences. As the value of the exponential decay term increases, the value of the exponential decay term decreases from... Towards Reduce proportionally , This decrease reflects the operational risks associated with excessive weight differences, enabling a more refined assessment of physical compatibility. Thus, storage category compatibility is a hard constraint, determining whether two items can be stored adjacently; while weight difference is a soft constraint. It doesn't absolutely prohibit adjacent storage, but rather smooths out the storage value through an exponential decay function, reflecting the increasing operational difficulty.

[0048] To ensure that the storage location allocation scheme conforms to the natural laws of warehousing safety and operations, the proximity storage value formula further introduces a physical compatibility function as a correction factor. This function reflects the physical law that, under a gravitational field, excessive mass deviation between adjacent storage locations increases uneven shelf stress and picking difficulty. When the storage categories of goods are inconsistent, such as ambient temperature goods and refrigerated goods, the formula directly resets the proximity value to zero using the Kronecker function, physically preventing the close stacking of incompatible goods. When the storage categories are consistent, a negative exponential function is used for smoothing adjustment based on the weight difference of the goods. As the weight difference increases, the proximity storage value decreases, thereby assessing the operational risks caused by weight deviation and optimizing the storage layout while ensuring physical safety.

[0049] It should be noted that the weight difference sensitivity coefficient The setting depends on the actual limitations imposed by warehouse operations on the weight variation of goods; this coefficient Controlled physical compatibility functions Differences in weight The rate of decay should be set higher if the warehouse has high requirements for matching weight differences, such as when using precision automated equipment or light-duty shelving. Values, such as 0.5, will lead to The value decreases rapidly as weight variance increases; conversely, if the warehouse has less stringent requirements for weight variance matching, such as manual picking or even shelving load distribution, a lower value should be set. A value, such as 0.05, makes the decay of the function term more gradual, allowing goods with slightly larger weight differences to still obtain a high physical compatibility score. This embodiment is preferred. This value strikes a balance between precisely controlling weight differences and maintaining reasonable flexibility, making it suitable for most general warehousing scenarios.

[0050] At this point, the proximity storage value between all pairs of goods has been obtained.

[0051] Step S400: Based on all available storage locations, the proximity storage value of product pairs, and the weighted occurrence frequency of each product, construct an optimization model with the objective of minimizing the total picking cost, solve the optimization model to obtain the storage location allocation scheme, and execute inbound and outbound operations according to the storage location allocation scheme.

[0052] It should be noted that this step, as the final decision-making stage, requires first calculating the weighted frequency of occurrences. and neighboring storage value Effectively applied to actual warehouse physical space, this system determines the optimal storage location allocation scheme by constructing an optimization model. Based on this, it standardizes the entire process of inbound shelving and outbound picking, balancing the reduction of picking costs with the guarantee of inbound and outbound efficiency. Considering high... Goods pairs with a high co-occurrence probability in orders can be placed in adjacent storage locations to reduce cross-area movement during picking and improve operational continuity; considering high... Goods that appear frequently in inbound and outbound operations should be assigned to storage locations near exits or main operational aisles to shorten the picking path and reduce overall travel time. Furthermore, storage location allocation must match the actual storage environment to ensure that the storage requirements for goods in terms of category and weight are met, avoiding weakened optimization effects due to physical incompatibility or process execution deviations. Based on these factors, this invention combines product correlation, inbound / outbound frequency, and physical storage feasibility to construct a storage location optimization model, achieving systematic planning of product storage locations and collaborative control of operational processes.

[0053] Specifically, first, obtain all available storage locations in the warehouse. 3D coordinate information Based on this, any two storage locations can be pre-calculated or obtained by querying the warehouse electronic map system. and The actual picking walking distance between Obtain detailed information for each storage location from the warehouse management system or racking equipment ledger. Physical property limitations, including its maximum allowable load. and maximum available space capacity At the same time, confirm each item to be stored. its own weight and volume Define a fixed public picking exit within the warehouse as the baseline reference point for global optimization, and obtain its accurate coordinates.

[0054] After data preparation, an optimization model is established to minimize the total picking cost, which is a weighted sum of internal movement costs and public round-trip costs. The objective function satisfies the following relationship: ; in, It is the function that minimizes the total picking cost; Goods With goods The value of nearby storage; , These are goods Goods The assigned storage location; Goods With goods The walking distance between the assigned storage locations; Goods Weighted frequency of occurrence; It refers to the storage locations and goods at the public picking exit. The assigned storage location The walking distance between them; , These are preset weighting coefficients; This is the total number of goods.

[0055] In this relation, the first term... This is the internal movement cost, which is calculated for each pair of goods. Proximity storage value Walking distance between them and their storage locations Minimizing the sum of the products of these terms will cause the optimization algorithm to place high-value goods pairs close to each other; the second term... Calculated for each item Weighted occurrence count Its storage location Distance to public picking exit Minimizing the sum of the products of these terms will cause the optimization algorithm to concentrate high-frequency items near the common picking exit. This relationship is achieved by minimizing... The goal is to determine the optimal layout under the two cost balance conditions, while simultaneously satisfying the requirements that high-frequency products are close to the export and highly related products are close to each other.

[0056] This relationship follows the definition of work in physics, meaning that handling costs are positively correlated with object displacement and operation frequency. This invention characterizes the basic energy consumption during the circulation process by weighting the frequency of goods occurrence with the distance to the warehouse exit. Simultaneously, it cleverly utilizes the clustering effect in spatial topology by weighting the negative correlation between proximity value and warehouse spacing, substantially reducing path redundancy by minimizing the physical displacement between highly correlated goods. This invention systematically aims to minimize physical power consumption, constrained by the natural attributes of warehouse space, and has clear basis and technical effectiveness.

[0057] It should be added that, and The setup depends on the business model. If orders are mostly multi-category, small-batch, such as in e-commerce retail, where consumers purchase multiple items like food, daily necessities, and small appliances in small quantities, the cost of moving between different items during picking is dominant, and therefore, the cost should be increased. ,For example If orders are mostly for a single product category in large quantities, such as wholesale business, where distributors purchase large quantities of the same brand of beverages at once, the round-trip export costs dominate, and improvements should be made. ,For example In this embodiment, preferably... This indicates that both costs are equally important.

[0058] The optimal layout, satisfying all physical constraints, is determined through a specific algorithm: hard constraints that must be met at all times during the solution process are defined; an optimization algorithm is used to obtain a list of suggested warehouse location adjustments, which includes the movement routes of goods within the warehouse. It should be noted that the hard constraints are: for any warehouse location... The sum of the weights of all goods on the storage location is less than or equal to the maximum allowable weight of the storage location. The sum of the volumes of all goods on the storage location is less than or equal to the maximum available space capacity of the location. The storage category of the goods must be compatible with the allowed storage category of the storage location.

[0059] The optimization algorithm, taking a greedy algorithm as an example, has the following solution process: traverse all possible item swapping or item movement operations; for each operation, if it violates the hard constraint, discard the operation; if it satisfies the hard constraint, calculate the corresponding objective function caused by the operation. Change Of all feasible operations, execute the one that reduces costs the most; repeat this process until the cost is reduced. The algorithm has converged and cannot be reduced further. Greedy algorithms are existing technology and will not be elaborated upon here. Besides greedy strategies, those skilled in the art can also employ other optimization algorithms, such as simulated annealing and genetic algorithms, to seek better global solutions.

[0060] At this point, the algorithm has completed its solution, ultimately producing a list of suggested warehouse location adjustments. This list details the movement routes of the goods within the warehouse, ensuring physical safety and minimizing total picking costs. To achieve optimal results.

[0061] Then, during inbound management, based on the suggested list of warehouse location adjustments and the storage categories of goods, newly arrived goods are directly assigned to the corresponding target warehouse locations. Simultaneously, the inbound time, batch, and warehouse location information are recorded in the warehouse management system to ensure consistency between inbound shelving and optimized layout. During outbound management, the system automatically generates a picking list sorted by warehouse location area based on the goods information included in the order and the optimized warehouse location distribution. After the pickers complete the picking according to the list, they verify the quantity and specifications of the goods next to the corresponding warehouse location. After verification, the system inventory is updated. Finally, all picked goods are gathered at the public picking exit to complete outbound verification and loading. At the same time, based on the latest inbound and outbound order data each quarter, the solution process of the first stage is repeated to update the warehouse location layout and simultaneously adjust the inbound and outbound operation standards to ensure that the layout optimization and inbound and outbound needs are dynamically adapted.

[0062] In one feasible implementation, the warehouse layout includes multiple storage locations for storing goods and at least one common picking exit, which refers to a pre-designated public area for collecting picked goods from different storage locations as the common endpoint of the picking path.

[0063] Thus, by combining the optimization model solution with the execution of inbound and outbound operations, this invention achieves dynamic and refined inbound and outbound management of stored goods.

[0064] The second aspect of this embodiment provides a warehouse goods inbound and outbound management system, such as... Figure 2 As shown, the warehouse goods inbound and outbound management system includes a memory and a processor. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a warehouse goods inbound and outbound management method according to the first aspect of the present invention is implemented.

[0065] The warehouse goods inbound and outbound management system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0066] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as resistive random access memory (DRAM), dynamic random access memory (DRAM), static random access memory (SRAM), enhanced dynamic random access memory (DRAM), high-bandwidth memory, hybrid memory cube, etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device.

[0067] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for managing the inbound and outbound operations of stored goods, applied to a warehouse containing several storage locations and a common picking exit, characterized in that, Including the following steps: Obtain historical order data from the warehouse; First and second items are selected from the warehouse, denoted as a product pair. Based on historical order data, the order completion time is weighted using a preset time decay weight. The weighted occurrence count of each individual item and the weighted co-occurrence count of the product pair are calculated. The weighted co-occurrence count of the product pair is divided by the weighted occurrence count of the first and second items respectively to obtain the first correlation and the second correlation. The acquisition of the weighted co-occurrence count of the product pair includes: obtaining orders that simultaneously contain both items in the product pair from historical orders, denoted as the target order set; for each target order in the target order set, obtaining the total number of product types in the target order and the preset time decay weight corresponding to the target order completion time; dividing the preset time decay weight by the difference between the total number of product types and 1 to obtain the contribution value of the target order; summing the contribution values ​​of all target orders in the target order set to obtain the weighted co-occurrence count of the product pair. The proximity storage value of a pair of goods is determined based on the mean of the first and second correlation degrees, the weighted co-occurrence frequency of the goods pair, the storage category of the goods pair, and the weight difference. This includes: determining the physical compatibility function of the goods pair based on the storage category and weight difference; multiplying the mean of the first and second correlation degrees, the weighted co-occurrence frequency of the goods pair, and the value of the physical compatibility function to determine the proximity storage value; wherein, the acquisition of the physical compatibility function includes: when the two goods in the goods pair have the same storage category, the category compatibility is recorded as 1, and when the storage categories are different, the category compatibility is recorded as 0; calculating the absolute weight difference between the two goods in the goods pair, multiplying the absolute weight difference by a preset weight difference sensitivity coefficient and taking the negative value to obtain the weight difference factor; and recording the product of the calculation result with the weight difference factor raised to the power of the natural constant and the category compatibility as the physical compatibility function; wherein, the setting of the weight difference sensitivity coefficient depends on the actual restrictions on the weight difference of the goods in warehouse operations. Based on all available storage locations, the proximity storage value of product pairs, and the weighted occurrence frequency of each product, an optimization model is constructed with the objective of minimizing the total picking cost. The storage location allocation scheme is obtained by solving the optimization model. Inbound and outbound operations are then performed according to the storage location allocation scheme.

2. The method for managing the inbound and outbound operations of stored goods according to claim 1, characterized in that, The total picking cost is the weighted sum of internal movement cost and public picking cost; wherein, the internal movement cost is the sum of the products of the proximity storage value of all item pairs and the walking distance between the respective storage locations of the two items contained in the corresponding item pair; and the public picking cost is the sum of the products of the weighted number of occurrences of all items and the walking distance from the respective storage location of the corresponding item to the public picking exit.

3. The method for managing the inbound and outbound operations of stored goods according to claim 1, characterized in that, The optimization model aimed at minimizing the total picking cost includes constraints, which include: The total weight of all goods allocated to any storage location shall not exceed the maximum load-bearing capacity of the corresponding storage location; The total volume of all goods allocated to any storage location shall not exceed the available space capacity of the corresponding storage location; The storage category of any item matches the allowed storage category of the assigned storage location.

4. The method for managing the inbound and outbound storage of goods according to claim 1, characterized in that, The preset time decay weight is a negative exponential function, and its independent variable is the product of the difference between the current time and the order completion time and the preset time decay coefficient.

5. The method for managing the inbound and outbound operations of stored goods according to claim 1, characterized in that, The process of performing inbound and outbound operations according to the storage location allocation scheme includes: During warehousing management, goods are allocated to the target storage locations determined by the storage location allocation scheme. During outbound management, a picking list sorted by storage location is generated based on the storage location allocation scheme, and picking is then performed.

6. The method for managing the inbound and outbound operations of stored goods according to claim 1, characterized in that, The method of solving the optimization model to obtain the storage location allocation scheme includes: solving the optimization model using a heuristic algorithm, wherein the heuristic algorithm includes one of the following: genetic algorithm, simulated annealing algorithm, and tabu search algorithm.

7. A warehouse goods inbound and outbound management system, characterized in that, The warehouse goods inbound and outbound management system includes a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a warehouse goods inbound and outbound management method according to any one of claims 1-6 is implemented.

Images