Modularized three-dimensional warehousing system for heterogeneous goods and self-adaptive sorting method

Abstract

The invention discloses a modularized three-dimensional warehousing system for heterogeneous goods and a self-adaptive sorting method, and relates to the technical field of three-dimensional warehousing, the modularized three-dimensional warehousing system comprises a storage module, the storage module comprises an adjustable goods shelf and a special-shaped goods fixing frame, and the heterogeneous goods are classified and stored; the intelligent optimization module is used for reading the information of the goods according to the RFID tags to identify the turnover frequency of the goods and partitioning the warehouse according to the Manhattan distance between the warehouse and the entrance and the exit, and the intelligent control module is used for adjusting the floor height data of the adjustable goods shelf; high-frequency goods are stored in a high-frequency area, and a Manhattan distance partitioning principle is combined, so that the length of a goods picking path is remarkably shortened, the order processing efficiency is improved, and the order processing capacity in the peak period is greatly improved; the dynamic partition strategy optimizes the goods storage layout, so that the high-turnover goods are preferentially stored in the core area, the storage space occupied by the unsalable goods is reduced, and the inventory turnover efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional warehousing, and more specifically, to a modular three-dimensional warehousing system for heterogeneous goods and an adaptive sorting method. Background Art

[0002] Traditional warehousing systems use fixed-height shelves, do not consider the dynamic size differences of high-turnover goods, lack the ability to dynamically adjust the shelf height according to the size of the goods, and cannot adapt to the real-time storage needs of high-frequency turnover goods. Existing warehousing mostly relies on historical data to divide cold and high-frequency areas and cannot capture the periodic fluctuations of the turnover frequency of goods in real time;

[0003] During e-commerce promotions or festival peak seasons, the turnover mode of goods changes every hour. Due to the lack of real-time data analysis capabilities, the existing system leads to an increase in the mismatch rate in the high-frequency area. Moreover, the turnover of goods is analyzed in isolation, ignoring the impact of combined orders. The existing technology only analyzes the turnover rate of single goods and does not identify the relevance of combined orders. Low-frequency but highly correlated goods are assigned to the low-frequency area due to their low independent turnover rate, resulting in a decrease in the picking efficiency of combined orders. Summary of the Invention

[0004] To solve the above problems, the present invention provides a modular three-dimensional warehousing system for heterogeneous goods and an adaptive sorting method.

[0005] The present invention provides a modular three-dimensional warehousing system for heterogeneous goods, including a storage module, the storage module includes adjustable shelves and fixed racks for special-shaped goods, and classifies and stores heterogeneous goods;

[0006] The goods are packed in boxes and placed on adjustable shelves and fixed racks for special-shaped goods. The boxes are integrated with RFID tags, and the information of the goods can be read according to the RFID tags;

[0007] An intelligent optimization module, which is used to identify the turnover frequency of the goods according to the information of the goods read by the RFID tags, partition the warehouse according to the Manhattan distance from the entrance and exit, and classify goods with different frequencies into different partitions according to the turnover frequency;

[0008] An intelligent control module, which is used to adjust the layer height data of the adjustable shelves in the partition closest to the warehouse entrance and exit according to the partition result of the intelligent optimization module.

[0009] Preferably, the specific working steps of the intelligent optimization module are as follows:

[0010] Obtain the inbound and outbound records and order correlation data of the goods according to the RFID tags integrated on the boxes;

[0011] Set a time period T in advance, and calculate and obtain the real-time turnover frequency f of the goods according to the formula every time period ; i

[0012] where N i is the number of times the goods i are shipped out within the previous time period T, m i is the current inventory of the goods i, and τ is the comprehensive weight coefficient of the orders associated with the goods i;

[0013] Divide the warehouse into three-level storage areas according to the Manhattan distance. Set a distance range threshold in advance. Mark the area within the distance range from the entrance and exit as the medium-frequency area, mark the area with a distance from the entrance and exit less than the distance range as the high-frequency area, and mark the area with a distance from the entrance and exit greater than the distance range as the low-frequency area;

[0014] Set a threshold range of the real-time turnover frequency f i of the goods in advance. Store the goods with a real-time turnover frequency f i greater than this threshold range in the high-frequency area, store the goods within this threshold range in the medium-frequency area, and store the goods below this threshold range in the low-frequency area.

[0015] Preferably, the specific working steps of the intelligent optimization module further include the following:

[0016] Sort the number of other goods shipped out together with this good in the same order in the previous cycle of the current time period in advance, and mark the top z goods with the largest sorted quantity as associated goods;

[0017] According to the formula calculate the optimized turnover frequency fx of the goods i , where f total represents the total turnover frequency of all goods in the warehouse in the previous cycle, f k is the real-time turnover frequency of the kth associated good, where sz is the weight coefficient of the associated good;

[0018] Classify and store this good according to the optimized turnover frequency fx i of this good.

[0019] Preferably, the intelligent optimization module further includes a migration unit. The migration unit generates a migration task according to the change of the real-time turnover frequency of the goods and moves the goods to other partitions. The specific steps further include the following:

[0020] Set a threshold Δf for the change of the turnover frequency in advance th . If within two adjacent cycles T, the optimized turnover frequency fx i ​The change value is greater than the threshold Δf th , then mark it as a goods to be migrated, and according to its optimized goods turnover frequency fx i transport it to the corresponding partition;

[0021] Before transportation, according to the formula calculate and obtain the priority P of the goods to be migrated i , where f max is the highest turnover frequency in the warehouse in the previous cycle, and D current is the distance required for this goods to be transported to the corresponding partition;

[0022] Preset a threshold range for the priority.

[0023] If the priority P of the goods to be migrated i is greater than this threshold range, the handling task is immediately executed by the stacker. If the priority P of the goods to be migrated i is within this threshold range, the handling task is scheduled and executed by the AGV. If the priority P of the goods to be migrated i is less than this threshold range, the handling task is executed when both the stacker and the AGV have no working targets;

[0024] For the goods to be migrated of the same classification, they are transported in the order of priority according to the size of the priority P i .

[0025] Preferably, the specific working steps of the migration unit further include the following:

[0026] During the process of moving the goods to other partitions, according to the RFID tag integrated on the cargo box, obtain the remaining valid time of the goods, that is, the remaining shelf life time T remain ;

[0027] According to the formula calculate and obtain the priority G of the goods to be migrated i ;

[0028] According to the calculated priority G of the goods to be migrated i execute the migration task.

[0029] Preferably, the specific working steps of the migration unit further include the following:

[0030] During the process of executing the migration task according to the calculated priority G of the goods to be migrated i ;

[0031] The warehouse is divided into three-level storage areas according to the Manhattan distance. Each storage area contains a refrigerated area and a normal / medium-frequency area. The refrigerated area is marked as an isolation area, and the calculation of the refrigerated area path penalty is added based on the AGV path passing through the isolation area.

[0032] According to the formula Gx i = G i * C base *(1 + 0.5 * |T in - T out |), the total priority Gx is calculated when carrying both near-expired goods and goods that need to pass through the refrigerated area. i The migration of goods is sorted according to the total priority Gx when carrying both near-expired goods and goods that need to pass through the refrigerated area. i

[0033] Among them, C base is the basic path cost, and |T in - T out | is the absolute value of the temperature difference between outside and inside the refrigerated area.

[0034] When the remaining shelf life T remain is less than 2h, skip the calculation of the refrigerated area path penalty and directly select the shortest path.

[0035] Preferably, the specific working steps of the intelligent control module are as follows:

[0036] The adjustable shelf is integrated with a laser rangefinder, which is used to obtain the length, width, and height data of the goods. The goods are packaged in a container, and the container is integrated with an RFID tag. The RFID tag is read to obtain the weight data m.

[0037] Then, every period, according to the formula the target layer spacing H of the adjustable shelf is calculated. adj Among them, γ is the compensation coefficient, and H0 is the layer height of the adjustable shelf in the previous period. is the stress margin ratio.

[0038] According to the obtained target layer spacing H of the adjustable shelf. adj the layer height of the adjustable shelf is adjusted.

[0039] Preferably, the specific working steps of the compensation coefficient γ are as follows:

[0040] The volume v of the goods is obtained by multiplying the length, width, and height of the goods.

[0041] According to the formula

[0042] the compensation coefficient γ is calculated and obtained.

[0043] Preferably, the specific steps of the intelligent control module further include the following:

[0044] According to the formula calculate and obtain the space utilization rate of the adjustable shelf, where l j w j h j are the length, width, and height of the j-th item;

[0045] L is the length of the adjustable shelf, and W is the width of the adjustable shelf;

[0046] If θ is less than 0.7, trigger an adjustment suggestion and adjust the layer height H of the adjustable shelf i until θ is greater than or equal to 0.7.

[0047] The present invention also proposes a modular three-dimensional warehousing adaptive sorting method for heterogeneous goods, including the following steps:

[0048] Step 1: Through the modularly designed adjustable shelf and the special-shaped goods fixing rack, realize the physical classification storage of goods with different forms, adopt the standardized container packaging integrated with RFID tags, and record core information such as the weight of goods, inbound and outbound records, order-related data, and shelf life in real time;

[0049] Step 2: Based on the RFID tag data, use the Manhattan distance algorithm to dynamically partition the warehouse, give priority to allocating high-frequency turnover goods to the area near the entrance and exit, and automatically adjust the storage location of medium- and low-frequency goods according to the peak and valley periods of logistics;

[0050] Step 3: The intelligent control module dynamically adjusts the layer height parameters of the adjustable shelf for the limited space constraint in the entrance and exit area, analyzes the safety threshold of the current stacking density of goods through the pressure sensor and the visual recognition system, automatically reconstructs the shelf layer height configuration according to the change of turnover frequency, and combines the characteristics of special-shaped shelves such as shockproof bases to ensure the stability during the adjustment process.

[0051] Beneficial effects: By storing high-frequency goods in the high-frequency area and combining the Manhattan distance zoning principle, the picking path length is significantly shortened, the order processing efficiency is improved, and the order processing capacity during the peak period is greatly enhanced; the dynamic zoning strategy optimizes the storage layout of goods, enables high-turnover goods to be preferentially stored in the core area, reduces the occupation of storage space by slow-moving goods, and thus improves the inventory turnover efficiency;

[0052] By dynamically superimposing the contribution values of goods turnover, a storage strategy of "driving low frequency with high frequency" is realized, solving the core pain points of fragmented combined order response and passive lag of long-tail demand in traditional technologies. After testing, this mechanism improves the overall shipping frequency of the warehouse, especially for scenarios with a large number of SKUs and complex order combinations, such as new retail and pharmaceutical warehousing, which has optimization value. Brief Description of the Drawings

[0053] Figure 1 is a flowchart of the system of the present invention. Detailed Embodiments

[0054] As Figure 1 shown: A modular three-dimensional warehousing system for heterogeneous goods, including a storage module. The storage module includes adjustable shelves and fixed racks for special-shaped goods, and classifies and stores heterogeneous goods; it should be noted that the adjustable shelves can support dynamic adjustment of the layer height. The fixed rack for special-shaped goods is designed with a card slot + magnetic attraction device for cylindrical items to prevent rolling, and a shock-proof base is added to the shelf for heavy parts;

[0055] The goods are packaged in boxes and placed on adjustable shelves and fixed racks for special-shaped goods. RFID tags are integrated on the boxes, and the information of the goods can be read according to the RFID tags; it should be noted that the information specifically includes the weight of the goods, the inbound and outbound records of the goods and the order-related data, and the shelf life of the goods;

[0056] An intelligent optimization module, which is used to identify the turnover frequency of the goods according to the information of the goods read by the RFID tags, partition the warehouse according to the Manhattan distance from the entrance and exit, and classify the goods with different frequencies into different partitions according to the turnover frequency.

[0057] It should be noted that in large logistics warehouses, due to the large variety of goods, it is necessary to place the goods with high turnover frequency on the adjustable shelves close to the warehouse entrance and exit. However, the existing placement methods are relatively fixed. With the development speed of the logistics express delivery, the turnover frequency of the goods will change at different times of the day. Therefore, it is necessary to optimize the placement of the goods in real time;

[0058] The step that takes the longest time for the goods to be transferred out of the warehouse is to carry and pack them in the warehouse. In this way, the search and waiting time during the handling of the goods can be greatly reduced;

[0059] An intelligent control module, which is used to adjust the layer height data of the adjustable shelves closest to the warehouse entrance and exit according to the zoning result of the intelligent optimization module. It should be noted that adjusting the layer height of the adjustable shelves can increase the space utilization rate. Normally, the goods in the shelves are classified and placed according to their sizes, and no layer height adjustment is required. However, in a large logistics warehouse, since the turnover frequencies of different types of goods are different, it is necessary to place the goods with high turnover frequencies on the adjustable shelves close to the warehouse entrance and exit. The sizes and heights of these goods are different, so on the adjustable shelves close to the entrance and exit, it is necessary to adjust the layer height of the shelves so that the adjustable shelves at the entrance and exit can store more goods with high turnover frequencies, thereby improving the space utilization rate;

[0060] Based on the intelligent optimization module, the goods with the highest turnover frequency have been placed on the adjustable shelves close to the warehouse entrance and exit. However, since the area close to the entrance and exit in the warehouse is limited, in order to maximize the utilization of the adjustable shelves in the area close to the entrance and exit, the layer height is adjusted.

[0061] As an optional embodiment: The specific working steps of the intelligent optimization module are as follows:

[0062] According to the RFID tags integrated on the cargo boxes, obtain the inbound and outbound records and order correlation data of the goods;

[0063] Preset a time period T, and calculate and obtain the real-time turnover frequency f of the goods every time period according to the formula ; It should be noted that the time period T is preset, and the unit is hours. In this embodiment, T is one day, that is, 24 hours; i

[0064] Where N i is the number of outbound times of goods i in the previous time period T, m i is the current inventory of goods i, and τ is the weighted comprehensive coefficient of the order associated with goods i. It should be noted that the way to obtain the weighted comprehensive coefficient of the order associated with goods i is to obtain all the associated orders of goods i in the previous time period T. The associated order j refers to the order that has a combined outbound relationship with the current goods i. The specific definition is:

[0065] The order composed of other goods that are co-outbound with goods i in the same order. Example: A certain takeaway order contains both a hamburger (goods A) and a cola (goods B) at the same time, then order j is an associated order for both A and B

[0066] ​Mark the orders with the noted outbound time shorter than the average outbound time among all associated orders as urgent orders, and mark the remaining orders as normal orders. The weight coefficient of urgent orders is 1.5, and the weight coefficient of normal orders is 1. Add up the weight coefficients of all associated orders to obtain the comprehensive weight coefficient τ of the associated orders of item i.

[0067] Divide the warehouse into three-level storage areas according to the Manhattan distance. Set a distance range threshold in advance. Mark the area within the distance range from the entrance and exit as the medium-frequency area, mark the area with a distance less than the distance range from the entrance and exit as the high-frequency area, and mark the area with a distance greater than the distance range from the entrance and exit as the low-frequency area. It should be noted that the distance range threshold is set according to the scope of the warehouse. In this embodiment, the distance range threshold is greater than 10m and less than 25m.

[0068] Set a threshold range of the real-time turnover frequency f of the goods in advance i of the goods, and place the goods with a real-time turnover frequency f i greater than this threshold range in the high-frequency area, place the goods within this threshold range in the medium-frequency area, and place the goods below this threshold range in the low-frequency area.

[0069] It should be noted that by storing high-frequency goods in the high-frequency area and combining the Manhattan distance zoning principle, the picking path length is significantly shortened, the order processing efficiency is improved, and the order processing capacity during peak periods is greatly enhanced; the dynamic zoning strategy optimizes the storage layout of goods, enables high-turnover goods to be preferentially stored in the core area, reduces the occupancy of storage space by slow-moving goods, and thus improves the inventory turnover efficiency.

[0070] The emergency order processing mechanism based on the weight coefficient realizes fast response through priority determination and storage location optimization, effectively shortening the fulfillment cycle; combined with RFID automatic collection and zoning strategy, it reduces the dependence on manual inventory and the energy consumption of ineffective handling, realizes the dual optimization of labor and energy costs, and shows significant advantages in high-frequency inbound and outbound scenarios such as e-commerce warehousing and cold chain logistics.

[0071] As an optional embodiment: The specific working steps of the intelligent optimization module further include the following:

[0072] Sort in advance the quantities of other goods that were co-outbound with this item in the same order in the previous cycle of the current time period, and mark the top z goods with the sorted quantities as associated goods.

[0073] According to the formula calculate the optimized turnover frequency fx of the goods i , where f total represents the total turnover frequency of all goods in the warehouse in the previous cycle, f kis the real - time turnover frequency of the k - th associated item, where sz is the weight coefficient of the associated item;

[0074] It should be noted that traditional warehouse optimization only calculates the turnover rate of a single item, ignoring the relevance of order combinations. For example, power banks (low - frequency) and mineral waters (high - frequency) often appear in the same take - out order, but in the traditional solution, power banks are stored in the low - frequency area, resulting in pickers frequently traveling back and forth between the cold and high - frequency areas, causing path waste;

[0075] Due to the lack of independent turnover data for low - frequency associated items, storage strategy adjustments cannot be triggered in the existing technology. In an e - commerce warehouse, for example, mobile phone cases (low - frequency) and mobile phones (high - frequency) are stored separately for a long time without correlation analysis, resulting in an increase in the manual replenishment rate for mobile phone orders;

[0076] It should be noted that the specific method for obtaining the weight coefficient of associated items is as follows: if the turnover frequency of the k - th associated item is greater than the average turnover frequency of all items in the warehouse, take 1, otherwise take 0. By adding up the values of z associated items, the weight coefficient sz of the associated items is obtained;

[0077] According to the optimized item turnover frequency fx of this item i , this item is classified and stored. It should be noted that by dynamically identifying the associated items that co - occur frequently at a high frequency in the same order (such as the combination of instant noodles and ham sausages), the low - turnover but strongly associated items (such as ham sausages) are intelligently bound to the high - turnover items (instant noodles). Even if the item's own turnover rate is low, its storage location will be adjusted to the high - frequency area due to the high - frequency demand of the associated item, thus shortening the pick - up path for combined orders;

[0078] The traditional solution only relies on the historical data of the item itself, resulting in low - frequency but strongly associated items (such as decorative accessories for holiday gift boxes) staying in the low - frequency area for a long time. In this solution, through the weight coefficient of associated items (such as the weight of the main gift - box item = 1), the turnover rate of low - frequency items is dynamically superimposed with the contribution value of highly associated items, avoiding pick - up delays caused by remote storage locations;

[0079] The clustering analysis of associated items can identify potential combined demands in advance (such as midnight instant noodles and ham sausages), and achieve "covering multiple orders with one pick - up" through adjacent storage in the high - frequency area.

[0080] As an optional embodiment: The intelligent optimization module further includes a migration unit. The migration unit generates a migration task according to the change in the real - time turnover frequency of the item and moves the item to other partitions. The specific steps further include the following:

[0081] A threshold Δf for the change in turnover frequency is set in advance th , if within two adjacent periods T, the change value of the optimized item turnover frequency fx i of this item is greater than the threshold Δfth If so, mark it as a goods to be migrated, and according to its optimized goods turnover frequency fx i Transport it to the corresponding partition;

[0082] Before transportation, according to the formula Calculate and obtain the priority P of the goods to be migrated i , where f max Is the highest turnover frequency in the warehouse in the previous cycle, and D current Is the distance required for the goods to be transported to the corresponding partition;

[0083] Set a threshold range for the priority in advance. It should be noted that in this embodiment, the threshold range can be from 0.5 to 0.8;

[0084] If the priority P of the goods to be migrated i Is greater than this threshold range, the transportation task is immediately executed by the stacker. If the priority P of the goods to be migrated i Is within this threshold range, the transportation task is scheduled and executed by the AGV. If the priority P of the goods to be migrated i Is less than this threshold range, the transportation task is executed when both the stacker and the AGV have no working targets;

[0085] For the goods to be migrated of the same classification, they are transported according to the size of the priority P i For priority sorting.

[0086] It should be noted that based on the dynamic threshold monitoring and priority calculation mechanism of the real-time turnover frequency, the changes in the liquidity of the goods can be accurately identified (such as the sudden increase in the turnover rate of midnight food), and different migration strategies can be generated according to the priority formula (combining the change in the turnover frequency, the transportation distance, and the equipment efficiency). For high-priority tasks (such as above the threshold), the stacker immediately executes the emergency goods migration to ensure that the high-frequency turnover goods are quickly adjusted to the high-frequency area and the picking path is shortened; medium-priority tasks (within the threshold range) are dynamically scheduled by the AGV to balance the equipment load and response speed and reduce the energy consumption of cross-area transportation; low-priority tasks are intelligently executed during the idle period of the equipment to avoid waste of resources.

[0087] As an optional embodiment: The specific working steps of the migration unit further include the following:

[0088] During the process of moving the goods to other partitions, according to the RFID tag integrated on the cargo box, obtain the remaining valid time of the goods, that is, the remaining shelf life time T remain ;

[0089] According to the formula Calculate and obtain the priority G of the goods to be migratedi ;

[0090] Obtain the priority G of the goods to be migrated according to the calculation i Execute the migration task. It should be noted that when the goods are approaching expiration, the priority G of the migrated goods i is 11 times the priority of ordinary goods. Except in special cases, this non-linear amplification effect forces the system to prioritize the processing of high-risk goods.

[0091] As an optional embodiment: The specific working steps of the migration unit further include the following:

[0092] During the process of obtaining the priority G of the goods to be migrated according to the calculation i and executing the migration task;

[0093] The warehouse is divided into three-level storage areas according to the Manhattan distance. Each storage area contains a refrigerated area and a normal intermediate frequency area. The refrigerated area is marked as an isolation area, and the refrigerated area path penalty calculation is added to the AGV path passing through the isolation area;

[0094] According to the formula Gx i = G i * C base *(1 + 0.5 * |T in - T out |), calculate and obtain the total priority Gx when carrying both near-expired goods and goods that need to pass through the refrigerated area, i and sort the migration of the goods according to the total priority Gx when carrying both near-expired goods and goods that need to pass through the refrigerated area i ;

[0095] where C base is the basic path cost, and |T in - T out | is the absolute value of the temperature difference between outside and inside the refrigerated area; it should be noted that the basic path cost is calculated through the distance and time of goods handling, and can be specifically obtained by equally weighted calculation of the time cost and energy consumption cost of the AGV or stacker moving a unit distance in the warehouse;

[0096] When the remaining shelf life time T remain is less than 2h, skip the refrigerated area path penalty calculation and directly select the shortest path. It should be noted that by introducing the refrigerated area path penalty and the two-factor priority calculation mechanism, the operation efficiency and goods safety in the cold chain warehousing scenario are significantly improved. In the three-level storage area divided based on the Manhattan distance, the refrigerated area is marked as an isolation area and the path penalty coefficient (such as the absolute value of the temperature difference weight) is increased, forcing the AGV to preferentially select the normal intermediate frequency area path and reducing the energy consumption of cold chain equipment;

[0097] Meanwhile, for the priority superposition calculation of near-term goods (such as fresh food and drugs) (total priority = basic path cost + temperature difference × near-term weight), it is ensured that high-priority goods (such as food expiring on the same day) are preferentially scheduled and the path is minimized, so as to shorten the ineffective exposure time in the refrigerated area and reduce the loss rate of goods;

[0098] For example, after a certain pharmaceutical warehouse is applied, due to the drive of the priority formula, the number of times of AGV crossing the refrigerated area for vaccine goods is reduced, the temperature control stability is improved, and the response speed of automatically migrating near-term inventory to the fast picking area is increased;

[0099] This breaks through the extensive management defects of the traditional warehousing system for the mixing and adjustment of cold chain paths and near-term goods, solves the problems of high energy consumption of equipment and waste of the shelf life of goods caused by ignoring the temperature impact in path planning, and is especially suitable for high-value and high-timeliness scenarios such as fresh food e-commerce and pharmaceutical cold chain, achieving a double breakthrough in warehousing efficiency and goods safety.

[0100] As an optional embodiment: The specific working steps of the intelligent control module are as follows:

[0101] The adjustable shelf is integrated with a laser rangefinder, and the laser rangefinder is used to obtain the length, width and height data of the goods. The goods are packaged in a cargo box, and the cargo box is integrated with an RFID tag. Read the RFID tag to obtain the obtained weight data m;

[0102] Then, according to the formula every one time period Calculate the target layer spacing H of the adjustable shelf adj , where γ is the compensation coefficient, H0 is the layer height of the adjustable shelf in the previous time period, is the stress margin ratio; it should be noted that σ max is the maximum allowable stress of the adjustable shelf material, which is obtained by looking up the table according to the material type and temperature conditions with reference to GB / T28576 (shelf standard), GB12337 (pressure vessel design code), etc. σ y is the actual stress of the current shelf of the adjustable shelf, which is specifically obtained by stress sensors installed near the mid-span and support points of the adjustable shelf crossbeam. The time of one time period is a fixed set time period, which is 3 minutes in this embodiment;

[0103] It should also be noted that the stress margin ratio reflects the proportion of the current stress to the maximum allowable stress, where Converts the real-time structural safety state of the adjustable shelf into a quantitative ratio of layer height adjustment to achieve dynamic response, and then multiplies by the compensation system and the layer height of the adjustable shelf in the previous time period to obtain the target layer spacing H of the adjustable shelf adj ;

[0104] According to the obtained target layer spacing H of the adjustable shelfadj to adjust the layer height of the adjustable shelf. It should be noted that the layer height adjustment is achieved through a screw lifting device driven by a servo motor;

[0105] It should also be noted that by real-time monitoring the actual stress at the mid-span and support points of the shelf beam and comparing it with the allowable stress determined by looking up the table based on national standards, the algorithm quantifies the structural safety margin of the shelf, thereby dynamically adjusting the layer height to prevent overload risks. This can not only avoid mechanical fatigue caused by frequent adjustments but also flexibly adapt to the space utilization rate according to the characteristics of the goods distribution, automatically reduce the layer height to disperse the load, prevent the deformation or collapse of the shelf structure, and dynamically compress the ineffective space.

[0106] As an optional embodiment: The specific working steps of the compensation coefficient γ are as follows:

[0107] Obtain the volume v of the goods by multiplying the length, width, and height of the goods;

[0108] According to the formula

[0109] Calculate and obtain the compensation coefficient γ. It should be noted that among them is the normalized density value of the goods, represents non-linear amplification, enhancing the sensitivity of the high-density area with an exponential function.

[0110] As an optional embodiment: The specific steps of the intelligent control module further include the following:

[0111] According to the formula Calculate and obtain the space utilization rate of the adjustable shelf, where l j w j h j are the length, width, and height of the j-th good;

[0112] L is the length of the adjustable shelf, and W is the width of the adjustable shelf;

[0113] If θ is less than 0.7, then trigger an adjustment suggestion to adjust the layer height H of the adjustable shelf i until θ is greater than or equal to 0.7; It should be noted that the layer height of the adjustable shelf can be adjusted adaptively to improve the space utilization rate of the adjustable shelf.

[0114] The present invention also proposes a modular three-dimensional warehousing adaptive sorting method for heterogeneous goods, including the following steps:

[0115] Step 1: Through the modular adjustable shelves and special-shaped goods fixing racks, realize the physical classification storage of goods in different forms. Adopt standardized bins with integrated RFID tags to record core information such as the weight of goods, inbound and outbound records, order-related data, and shelf life in real time;

[0116] Step 2: Based on the RFID tag data, use the Manhattan distance algorithm to dynamically partition the warehouse. Prioritize the allocation of high-frequency turnover goods to the areas near the entrances and exits, and automatically adjust the storage locations of medium- and low-frequency goods according to the peak and valley periods of logistics;

[0117] Step 3: The intelligent control module dynamically adjusts the layer height parameters of the adjustable shelves for the limited space constraints in the entrance and exit areas. Through the pressure sensor and visual recognition system, analyze the safety threshold of the current goods stacking density, and automatically reconstruct the shelf layer height configuration according to the change of turnover frequency. Combine the characteristics of special-shaped shelves such as shockproof bases to ensure the stability during the adjustment process.

[0118] The above is only the preferred implementation mode of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements should also be regarded as within the protection scope of this template.

Claims

1. A modular three-dimensional storage system for heterogeneous goods, characterized in that, It includes a storage module, which includes adjustable shelves and special-shaped goods fixing racks for classifying and storing heterogeneous goods; The goods are packed in a container and placed on the adjustable shelves and special-shaped goods fixing racks. The container is integrated with an RFID tag, and the information of the goods can be read according to the RFID tag; An intelligent optimization module, which is used to identify the turnover frequency of the goods according to the information of the goods read by the RFID tag, partition the warehouse according to the Manhattan distance from the entrance and exit, and classify goods with different frequencies into different partitions according to the turnover frequency; An intelligent control module, which is used to adjust the layer height data of the adjustable shelves closest to the entrance and exit of the warehouse according to the partitioning result of the intelligent optimization module.

2. The modular three-dimensional storage system for heterogeneous goods according to claim 1, characterized in that The specific working steps of the intelligent optimization module are as follows: According to the RFID tag integrated on the container, obtain the inbound and outbound record domain order association data of the goods; A time period T is set in advance, and the real-time turnover frequency f of the goods is calculated and obtained every other time period according to the formula ; i ​ where N i is the number of times of outbound of product i in the previous time period T, m i is the current inventory of product i, and τ is the comprehensive weight coefficient of the order associated with product i; Divide the warehouse into three-level storage areas according to the Manhattan distance. Set a distance range threshold in advance. Mark the area within the distance range from the entrance and exit as the medium-frequency area, mark the area with a distance less than the distance range from the entrance and exit as the high-frequency area, and mark the area with a distance greater than the distance range from the entrance and exit as the low-frequency area; Set a threshold range for the real-time turnover frequency f of goods in advance i for the real-time turnover frequency f of goods i Goods with a real-time turnover frequency f greater than this threshold range are stored in the high-frequency area, goods within this threshold range are stored in the medium-frequency area, and goods below this threshold range are stored in the low-frequency area.

3. The modular three-dimensional storage system for heterogeneous goods according to claim 2, characterized in that, The specific working steps of the intelligent optimization module also include the following: Sort in advance the quantity of other goods that were co-shipped with this good in the same order in the previous cycle of the current time period, and mark the top z goods with the largest sorted quantity as associated goods; According to the formula the optimized inventory turnover frequency fx is calculated i , where f total represents the total turnover frequency of all goods in the warehouse in the previous cycle, and f k is the real-time turnover frequency of the kth associated good, where sz is the weight coefficient of the associated good; According to the optimized inventory turnover frequency fx of the goods i , the goods are classified and stored.

4. The modular three-dimensional storage system for heterogeneous goods according to claim 3, wherein The intelligent optimization module also includes a migration unit, which generates a migration task according to the change of the real-time turnover frequency of the goods and moves the goods to other partitions. The specific steps also include the following: Set a threshold value Δf for the change in turnover frequency in advance th , if within two adjacent cycles T, the change value of the optimized turnover frequency fx of the goods i is greater than the threshold value Δf th , then mark it as a goods to be migrated, and move it to the corresponding partition according to its optimized turnover frequency fx i ; Before handling, according to the formula calculate and obtain the priority P of the goods to be migrated i , where f max is the highest turnover frequency in the warehouse in the previous cycle, and D current is the distance required to move the goods to the corresponding partition; Set a threshold range of priority in advance. If the priority P of the goods to be relocated i is greater than this threshold range, the handling task is immediately executed by the stacker. If the priority P of the goods to be relocated i is within this threshold range, the handling task is scheduled and executed by the AGV. If the priority P of the goods to be relocated i is less than this threshold range, the handling task is executed when both the stacker and the AGV have no working targets; For the goods to be migrated with the same classification, they are carried according to the priority P i in descending order of size for priority sorting and handling.

5. The modular three-dimensional storage system for heterogeneous goods according to claim 4, characterized in that, The specific working steps of the migration unit also include the following: During the process of moving the goods to other partitions, the remaining valid time of the goods, i.e., the remaining shelf life time T, is obtained according to the RFID tag integrated on the container remain ; According to the formula calculate and obtain the priority G of the goods to be migrated i ; Obtain the priority G of the goods to be migrated according to the calculation i Execute the migration task.

6. The modular three-dimensional storage system for heterogeneous goods according to claim 5, wherein The specific working steps of the migration unit also include the following: Obtaining the priority G of the goods to be migrated according to the calculation i During the process of performing the migration task; The warehouse is divided into three-level storage areas according to the Manhattan distance. Each storage area includes a refrigerated area and a medium- and high-frequency area. Mark the refrigerated area as an isolation area, and add a refrigerated area path penalty calculation on the basis of the AGV path passing through the isolation area; According to the formula Gx i = G i * C base *(1 + 0.5 * |T in - T out |), calculate and obtain the total priority Gx when carrying both near - term goods and goods that need to pass through the refrigerated area i , and sort the migration of goods according to the total priority Gx when carrying both near - term goods and goods that need to pass through the refrigerated area i ; where C base is the basic path cost, |T in -T out | is the absolute value of the temperature difference between outside and inside the refrigerated area; When the remaining shelf life time T remain is less than 2h, skip the calculation of the refrigerated area path penalty and directly select the shortest path.

7. A modular three-dimensional storage system for heterogeneous goods according to claim 1, characterized in that, The specific working steps of the intelligent control module are as follows: The adjustable shelf is integrated with a laser rangefinder, which is used to obtain the length, width and height data of the goods. The goods are packed in a container, and the container is integrated with an RFID tag. Read the RFID tag to obtain the weight data m obtained; Then, in each time period, according to the formula calculate the target layer spacing H of the adjustable shelf adj , where γ is the compensation coefficient, H0 is the shelf height of the adjustable shelf in the previous time period, is the stress margin ratio; According to the obtained target layer spacing H of the adjustable shelf adj , the layer height of the adjustable shelf is adjusted.

8. The modular three-dimensional storage system for heterogeneous goods according to claim 7, characterized in that The specific working steps of the compensation coefficient γ are as follows: Obtain the volume v of the goods by multiplying the length, width and height of the goods; According to the formula Calculate and obtain the compensation coefficient γ.

9. The modular three-dimensional storage system for heterogeneous goods according to claim 8, characterized in that, The specific steps of the intelligent control module also include the following: According to the formula calculate and obtain the space utilization rate of the adjustable shelf, where l j w j h j are the length, width, and height of the j-th item; L is the length of the adjustable shelf, and W is the width of the adjustable shelf; If θ is less than 0.7, trigger an adjustment recommendation to adjust the height H of the adjustable shelf i until θ is greater than or equal to 0.

7.

10. A modular three-dimensional storage adaptive sorting method for heterogeneous goods, applicable to a modular three-dimensional storage system for heterogeneous goods described in any one of claims 1 to 9, characterized in that, It includes the following steps: Step 1: Through the modular-designed adjustable shelves and special-shaped goods fixing racks, realize the physical classification storage of goods with different forms. Adopt standardized containers integrated with RFID tags to record core information such as the weight of goods, inbound and outbound records, order association data and shelf life in real time; Step 2: Based on the RFID tag data, use the Manhattan distance algorithm to dynamically partition the warehouse. Prioritize the allocation of high-frequency turnover goods to the areas near the entrances and exits, and automatically adjust the storage locations of medium- and low-frequency goods according to the peak and valley periods of logistics. Step 3: In response to the limited space constraints in the entrance and exit areas, the intelligent control module dynamically adjusts the layer height parameters of the adjustable shelves. Through the pressure sensors and the visual recognition system, analyze the safety threshold of the current stacking density of goods, automatically reconstruct the layer height configuration of the shelves according to the change in turnover frequency, and combine the characteristics of special-shaped shelves such as shock-proof bases to ensure stability during the adjustment process.

Images