A warehouse layout generation method and device and electronic equipment

Abstract

The application provides a warehouse layout generation method and device and electronic equipment, and relates to the technical field of warehouse logistics automation. The method receives a natural language requirement description of a user, determines warehouse layout parameters according to contained layout structure information, and can reduce the pre-requisites and communication costs of scheme generation. In addition, space reasoning is performed in combination with preset layout constraint rules and warehouse layout parameters to automatically determine the size parameters and arrangement parameters of warehouse entities in a target layout space. This can eliminate the omissions and rework that are prone to occur due to long chains when manually calculating coupled relationships. The quality of the scheme can also be independent of individual experience fluctuations, and the optimization level of space utilization and logistics efficiency can be stably maintained. Furthermore, a three-dimensional model of the warehouse entity is automatically constructed according to the size parameters and arranged in the target space according to the arrangement parameters. The appearance and size of the model are dynamically driven by the parameters, and the overall layout scheme can be completed in a very short time, and the engineering design iteration efficiency can be substantially improved.

Description

Technical Field

[0001] This invention relates to the field of warehouse logistics automation technology, and in particular to a warehouse layout generation method, apparatus and electronic equipment. Background Technology

[0002] The three-dimensional layout planning of an Automated Storage and Retrieval System (ASRS) is a core aspect of logistics warehousing engineering design. A typical ASRS warehouse includes various types of equipment, such as racks, stacker cranes, conveyors, lifters, inbound and outbound platforms, with the number of equipment ranging from dozens to hundreds. The goal of layout planning is to rationally arrange the three-dimensional position, size, and orientation of all equipment within a limited warehouse space to achieve optimal space utilization, storage capacity, and logistics efficiency.

[0003] Existing technologies typically employ manual layout planning or fixed-parameter template generation. The manual method involves engineers manually positioning, adjusting dimensions, and setting orientations for each piece of equipment in 3D modeling software. The fixed-parameter template method automatically generates the layout based on preset rules. However, there are deep parametric coupling relationships between factors such as warehouse dimensions, rack configuration, conveyor line routing, and elevator positions. For example, increasing the number of rack groups directly changes the total layout width, which in turn affects a series of chain reactions, such as whether the conveyor line loop exceeds the warehouse boundary and whether the elevator position needs to be offset accordingly.

[0004] In a manual approach, modifying any parameter requires calculating and adjusting the positions of dozens of related devices one by one. This is not only tedious and involves a huge amount of rework, but also prone to oversights due to the lengthy coupling chain, often revealing infeasibility issues only in the later stages of the design process. While template-based methods can improve initial generation efficiency, their fixed rules cannot dynamically respond to changes in the linkage between parameters. When faced with diverse warehouse constraints, the generated solutions still require secondary manual corrections, and the coupling effects remain difficult to handle systematically. Furthermore, the selection of key spatial parameters such as shelf spacing and aisle depth relies heavily on engineers' personal experience, lacking systematic optimization support. This leads to significant fluctuations in solution quality, making it difficult to consistently achieve optimal space utilization and logistics efficiency. These combined shortcomings result in prominent problems such as lengthy design cycles, frequent rework, and insufficient overall layout reliability. Summary of the Invention

[0005] The purpose of this invention is to provide a warehouse layout generation method, apparatus and electronic device that can avoid rework and omissions caused by manual calculation of parameters, and can improve overall design efficiency and engineering reliability.

[0006] In a first aspect, the present invention provides a warehouse layout generation method, the method comprising: responding to receiving a user's natural language requirement description, determining warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description; determining the size parameters and arrangement parameters of the warehouse entity to be arranged in the target layout space based on preset layout constraint rules and warehouse layout parameters; constructing a physical model of the warehouse entity based on the size parameters, and arranging the physical model in the target layout space based on the arrangement parameters to form a warehouse layout scheme.

[0007] In an optional implementation, the step of determining the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description includes: constructing prompt words based on the natural language requirement description according to a predefined data structure; inputting the prompt words into a large language model, so that the large language model parses the prompt words to determine the layout structure information contained in the natural language requirement description; and converting the layout structure information into warehouse layout parameters according to a preset warehouse layout parameter structure.

[0008] In an optional implementation, the above-mentioned layout structure information includes the storage entities to be arranged, and the method further includes: determining the entity parameter group whose parameter content has been filled in the storage layout parameters; and determining the storage entity type corresponding to the entity parameter group as the storage entity to be arranged according to the preset correspondence between the entity parameter group and the preset storage entity type.

[0009] In an optional implementation, the steps of determining the size parameters and arrangement parameters of the storage entity in the target layout space based on preset layout constraint rules and storage layout parameters include: determining the size parameters of the storage entity based on preset layout constraint rules and storage layout parameters; determining the arrangement parameters of the storage entity in the target layout space based on the boundary conditions of the target layout space, preset layout constraint rules, and storage layout parameters; the arrangement parameters include the three-dimensional position coordinates of the storage entity and the relative positional relationship between multiple storage entities.

[0010] In an optional implementation, the step of determining the size parameters of a storage entity based on preset layout constraint rules and storage layout parameters includes: obtaining preset layout constraint rules; the layout constraint rules include boundary size restrictions of the target layout space and spacing constraints of the storage entities within the target layout space; determining the initial size parameters of the storage entities according to the entity description parameters and / or cargo parameters in the storage layout parameters; and adapting and adjusting the initial size parameters according to the boundary size restrictions and spacing constraints so that the adjusted size parameters satisfy the boundary size restrictions and spacing constraints, thereby obtaining the final size parameters.

[0011] In an optional implementation, the steps of constructing a physical model of the storage entity based on size parameters and arranging the physical model in the target layout space based on layout parameters to form a storage layout scheme include: calling a model generator, submitting the type and size parameters of the storage entity as a generation task to the model generator to obtain the physical model corresponding to the storage entity; and placing the physical model in the corresponding position in the target layout space according to the three-dimensional position coordinates and relative position relationships in the layout parameters to form a storage layout scheme.

[0012] In an optional implementation, the step of submitting the type and size parameters of the storage entity as a generation task to the model generator to obtain the entity model corresponding to the storage entity includes: submitting the type and size parameters of the storage entity as text prompts to the model generator to obtain the corresponding entity model; or, submitting a preset reference image to the model generator to obtain the corresponding entity model.

[0013] In an optional implementation, the method further includes: establishing a connection relationship between each storage entity based on the connector type and spatial location of each storage entity in the storage layout scheme to form a logistics topology network; determining at least one transportable cargo path based on the logistics topology network; and controlling the transmission of preset cargo along the transportable cargo path to simulate the flow process of preset cargo in the storage layout scheme.

[0014] In an optional implementation, after the step of controlling the preset goods to be transported along the transportable path, the method further includes: recording the cumulative processing time and cumulative waiting time of each logistics operation equipment involved in the goods transport in the warehousing entity during the transport process, so as to calculate the operating utilization rate of each logistics operation equipment; and evaluating the capacity throughput level of the warehousing entity based on the cumulative processing time, cumulative waiting time, and total simulation time.

[0015] Secondly, the present invention provides a warehouse layout generation device, which includes: a data acquisition module, used to respond to receiving a user's natural language requirement description and determine the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description; a parameter determination module, used to determine the size parameters and arrangement parameters of the warehouse entity to be arranged in the target layout space based on preset layout constraint rules and warehouse layout parameters; and an execution module, used to construct a physical model of the warehouse entity based on the size parameters and arrange the physical model in the target layout space based on the arrangement parameters to form a warehouse layout scheme.

[0016] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method of any of the above embodiments.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of any of the above embodiments.

[0018] The embodiments of this invention bring the following beneficial effects: This invention provides a warehouse layout generation method, apparatus, and electronic device. By receiving users' natural language requirement descriptions, it automatically determines warehouse layout parameters based on the layout structure information contained therein. This directly transforms non-professional, colloquial requirements into calculable, structured input, allowing designers to quickly start layout planning without needing to master complex modeling software operations, significantly reducing the pre-concept threshold and communication costs for solution generation. Furthermore, this invention also combines preset layout constraint rules and warehouse layout parameters for spatial reasoning, automatically determining the size and arrangement parameters of each warehouse entity in the target layout space. When any parameter such as the number of shelving groups, aisle width, or conveyor line direction changes, the system can dynamically recalculate the spatial position and geometric dimensions of the corresponding related equipment, ensuring that the layout scheme always meets physical boundaries and engineering specifications. This eliminates omissions and rework that are easily caused by the lengthy chain when manually calculating coupling relationships, while also freeing the solution quality from fluctuations in individual experience, and maintaining a stable level of optimization in space utilization and logistics efficiency. Furthermore, this invention automatically constructs a 3D model of the storage entity based on size parameters and arranges it in the target space according to the layout parameters. The appearance and size of the model are completely driven by the parameters dynamically, eliminating the need for manual production by professional modelers. The customized construction cycle of a single device is shortened from several days to minutes, enabling the overall layout scheme to complete the complete closed loop from demand input to 3D visualization in a very short time, and substantially improving the efficiency of engineering design iteration.

[0019] Other features and advantages of the invention will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above.

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0021] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1A flowchart illustrating a warehouse layout generation method provided in an embodiment of the present invention; Figure 2 A flowchart of another warehouse layout generation method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a warehouse layout generation device provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0024] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0025] To facilitate understanding, a warehouse layout generation method provided in an embodiment of the present invention will first be described, referring to... Figure 1 The method includes the following steps: Step S102: In response to receiving the user's natural language requirement description, determine the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description.

[0026] The user's natural language requirement description is used to characterize the user's desired configuration of the warehouse layout, including the warehouse's spatial dimensions, equipment type and quantity, arrangement, and logistics channel requirements. In one embodiment, natural language processing technology can be used to perform semantic parsing and key information extraction on the user-input description text to obtain structured warehouse layout parameters, which indicate specific quantitative indicators such as warehouse boundary dimensions, racking arrangement, number and width of aisles, conveyor line routing, and the location of inbound and outbound platforms. The natural language requirement description can take the form of colloquial or engineering expressions, such as "I need a warehouse that is 80 meters long and 50 meters wide, with 5 rows of racks, 12 columns per row, a stacker crane aisle width of 1.2 meters, and a conveyor line arranged along the north wall," or it can use a more structured description closer to an engineering document, such as "Warehouse dimensions 100m × 60m, double-aisle stacker crane configuration, and double-depth racking layout." This type of description typically includes layout structure information such as basic warehouse dimensions, equipment configuration, and spatial orientation, used to characterize the user's functional requirements and spatial constraints for the final layout scheme. Based on the corresponding layout structure information, the system can determine warehouse layout parameters such as warehouse length, width, number of rack rows, number of racks per row, aisle width, and conveyor line orientation, so as to clarify which equipment needs to be arranged in the warehouse, the quantity of each piece of equipment, and the specific dimensions of the warehouse boundary.

[0027] Step S104: Based on the preset layout constraint rules and warehouse layout parameters, determine the size parameters and arrangement parameters of the warehouse entities to be arranged in the target layout space.

[0028] Layout constraints are used to characterize the physical boundary conditions, engineering specifications, and logistics design principles that storage equipment must meet in its spatial arrangement. These include, but are not limited to, warehouse inner wall boundary restrictions, minimum safe distances between equipment, the matching relationship between aisle width and stacker crane model, requirements for conveyor line turning radius and passage width, reserved dimensions for fire lanes and maintenance lanes, and connection specifications between inbound / outbound platforms and conveyor lines. Storage entities to be arranged characterize the specific physical objects that need to be actually deployed in the warehouse, including but not limited to racks, stacker cranes, various sections of conveyor lines, lifting and transfer machines, inbound platforms, outbound platforms, hoists, and turntables, among other core logistics equipment. Furthermore, storage entities can also include scene auxiliary elements, including but not limited to fire-fighting equipment, signage, lighting fixtures, aisle railings, paint buckets, and bird models. In one implementation, the equipment type and quantity configuration can be directly obtained based on the warehouse layout parameters determined above, and the corresponding physical objects can be identified as the warehouse entities to be arranged. For example, when a user inputs "arrange 5 rows of shelves, 12 columns per row, configure 2 stacker cranes, 1 set of conveyor loops along the north wall, and 1 inbound / outbound platform each," the system can clearly identify the warehouse entities to be arranged based on this configuration information, including 5 rows of shelves, 2 stacker cranes, 1 set of conveyor loops, 1 inbound platform, and 1 outbound platform. Furthermore, users can also supplement the arrangement requirements of scene auxiliary elements in natural language requirements, such as "place fire-fighting equipment at both ends of the shelves," "arrange signs along both sides of the aisle," or "appropriately scatter paint buckets in the warehouse to enhance the scene realism," or "add bird models along the roof steel frame to enrich the visual layers." The system also incorporates these requirements into the warehouse entities to be arranged according to the description of the requirements, so that the final generated 3D layout scheme not only meets the functional requirements of logistics operations but also has a complete scene realism and engineering demonstration effect.

[0029] After obtaining the warehouse layout parameters, this embodiment of the invention combines these parameters with layout constraint rules to determine the size and arrangement parameters of each warehouse entity to be arranged in the target layout space. Size parameters include the three-dimensional geometric dimensions of each entity, such as length, width, and height. Arrangement parameters include the three-dimensional coordinates of each entity's reference point in the target layout space and its spatial orientation angle. For example, the width of each row of shelves can be determined based on the warehouse length and the number of rows, and the spacing between shelves can be determined based on the aisle width and stacker crane operating requirements, thus determining the specific coordinate position of each row of shelves in the warehouse. For conveyor lines, the remaining space can be calculated based on the warehouse boundary and the area occupied by the shelves, thus determining the length, direction, and coordinate position of the conveyor line in the warehouse. For decorative elements such as paint buckets, the system can automatically and randomly distribute them according to the user-specified area and determine their coordinate positions. When any input parameter, such as the number of shelf groups, aisle width, or warehouse boundary, changes, the system can dynamically recalculate the size and position of all entities globally, ensuring that the layout scheme always meets all engineering constraints.

[0030] Step S106: Construct a physical model of the warehouse entity based on the size parameters, and arrange the physical model in the target layout space based on the layout parameters to form a warehouse layout scheme.

[0031] The target layout space refers to the usable three-dimensional space within the warehouse specified by the user. This includes the physical boundaries in the length, width, and height directions, as well as restrictions on unoccupied areas such as column grids, doorways, and equipment rooms. For example, a rectangular warehouse space 80 meters long, 50 meters wide, and 12 meters high, with 0.8m x 0.8m structural columns at each corner that need to be avoided. Once the dimensional parameters of each storage entity are determined, the system can automatically generate a corresponding three-dimensional solid model based on the specific length, width, and height data of each entity using parametric modeling technology. The solid model represents the geometric shape, spatial occupancy, and key structural features of the storage entity, realistically reflecting the size proportions and outline of the equipment in a three-dimensional virtual environment. In one implementation, a parametric model template matching the entity type can be called, with the dimensional parameters as driving variables. The model template automatically adjusts the values ​​of each dimension of the geometry and generates complete three-dimensional mesh data, eliminating the need for manual modeling of each entity. The generation of a single entity's model can be reduced from several days to seconds.

[0032] After determining the physical models, this embodiment of the invention further assembles all physical models one by one into their corresponding positions in the target layout space according to the arrangement parameters of each entity (including its precise coordinate position and orientation angle in the target layout space), forming a complete warehouse layout scheme. This embodiment of the invention not only provides interactive viewing and space occupancy verification from any perspective in the 3D scene, but also allows direct export of engineering drawings or import into a logistics simulation environment for subsequent verification, without the need for additional secondary modeling, which can significantly shorten the cycle from scheme determination to project delivery.

[0033] Furthermore, based on the above embodiments, this invention also provides another method for generating warehouse layouts, referring to... Figure 2 The method includes the following steps: Step S202: In response to receiving the user's natural language requirement description, determine the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description.

[0034] In one implementation, prompt words can be constructed based on a natural language requirement description according to a predefined data structure. These prompt words are then input into a large language model, which parses them to determine the layout structure information contained in the natural language requirement description. The layout structure information is then converted into warehouse layout parameters according to a preset warehouse layout parameter structure. The prompt words are used to clearly indicate the type of task to be completed and the output format requirements to the large language model. During construction, they include both system prompt words and user messages. System prompt words are used to define the model's role, output specifications, and constraint rules, such as the JSON data structure definition of warehouse layout parameters and the description of spatial layout rules.

[0035] In one implementation, the JSON data structure definition of the warehouse layout parameters is shown in Table 1 below.

[0036] Table 1. Data structure definition for warehouse layout parameters:

[0037] Referring to Table 1, the above data structure definition includes warehouse parameters, rack parameters, and layout parameters. It may also include parameter groups such as conveyor parameters and lifter parameters to clarify the field names, data types, and value ranges of the model output results, ensuring that the output results have a unified and parsable format. The spatial layout rules describe the basic topological constraints and relative positional relationships between various devices within the warehouse, such as racks being placed against the back wall, conveyor lines arranged in a rectangular closed loop in front of the racks, each aisle having two lifting and transferring machines and two platforms connected to the racks, and all equipment located inside the warehouse walls and at least 1 meter away from the walls. This allows the large language model to reason and output according to the basic logic of warehouse engineering design when parsing user requirements. User messages carry the natural language description of the user's input requirements and the existing layout parameter context.

[0038] After receiving a complete prompt, the large language model can semantically parse the content of the user message according to the roles and specifications defined in the system prompt. It can identify and extract layout structure information such as warehouse dimensions, equipment configuration, and spatial orientation from unstructured natural language descriptions, and output it in the JSON data structure required by the system prompt. Furthermore, after obtaining the layout structure information, the system can convert the layout structure information into standardized warehouse layout parameters according to a preset warehouse layout parameter structure. In one optional implementation, the input-output relationship of the large language model can be represented as: P json =f LLM (S prompt U message ,T). Among them, S promptU is the system suggestion word. message This is the user message (containing current parameters and user requirements), where T is the temperature parameter (temperature, set to 0.7), and P... json The layout parameters are output in JSON format for the model.

[0039] Considering that the model output may contain unstructured content such as Markdown code block markers, a JSON extraction function can be used to extract valid JSON substrings from the returned content: P raw =extract(content,"{","}"); where the extract function searches for the substring between the first left curly brace and the last right curly brace in content.

[0040] Furthermore, each parameter obtained from the analysis can be range-bound to ensure that its value falls within a preset minimum and maximum allowable range. For example, the number of rack rows should not be less than one row and should not exceed the upper limit that the warehouse width can accommodate; the aisle width should not be less than the minimum safe clearance required for stacker crane operation, etc. In one implementation, the following formula can be used as a reference: p_i=clamp(p_i_raw,p_i_min,p_i_max)=max(p_i_min,min(p_i_max,p_i_raw)) Where p_i is the clamped value of the i-th parameter, p_i_raw is the original value of the model output, and p_i_min and p_i_max are the minimum and maximum allowed values ​​of this parameter. The clamping range is shown in Table 2 below: Table 2 Clamping Range

[0041] After the above processing, the system can determine warehouse layout parameters such as warehouse length, width, number of rack rows, number of racks per row, aisle width, and conveyor line orientation.

[0042] Corresponding to the data structure definition above, warehouse layout parameters may include warehouse parameter groups (warehouse length, width, height), rack parameter groups (number of rows, number of columns, number of layers, rack dimensions), conveyor line parameter groups (direction, length, width), elevator parameter groups (quantity, location), and layout parameter groups (overall layout method), etc. The correspondence between each entity parameter group and the warehouse entity type can be predefined. After parsing and obtaining the warehouse layout parameters, the system determines the entity parameter groups whose parameter content has been filled in each parameter group, and based on the predefined correspondence, determines the warehouse entity type corresponding to the filled entity parameter group as the warehouse entity to be arranged. For example, when the number of rows and columns in the rack parameter group is filled, the system determines that the rack is the warehouse entity type to be arranged; when the length and direction in the conveyor line parameter group are filled, the system determines that the conveyor line is the warehouse entity type to be arranged; when the quantity and location in the elevator parameter group are filled, the system determines that the elevator is the warehouse entity type to be arranged. In this way, the system can automatically determine which entity types and their corresponding quantities need to be deployed based on the parameter filling status, without requiring the user to specify an additional list of equipment to be deployed.

[0043] Step S204: Determine the size parameters of the storage entity based on the preset layout constraint rules and storage layout parameters.

[0044] Layout constraint rules characterize the physical boundary conditions and engineering specifications that storage equipment must meet in spatial arrangement. These include boundary dimensional restrictions on the target layout space, such as the physical boundaries of the warehouse's length, width, and height, as well as restrictions on unoccupied areas such as column grids, equipment rooms, and doorways. Alternatively, they may include spacing constraints on storage entities within the target layout space, such as the minimum distance between racks and side walls, the minimum width of stacker crane aisles, the safe clearance between conveyor lines and racks, the reserved dimensions for fire exits and maintenance access routes, and the connection spacing between inbound / outbound platforms and conveyor lines. The dimensional parameters of storage entities refer to the geometric dimensions of each piece of equipment in three-dimensional space, including the length, width, and height of racks, the width and length of conveyor lines, the area and height of platforms, and the dimensions of stacker cranes and their track spans.

[0045] Unlike solutions that directly use user-specified dimensions as the final dimensions for modeling, this invention also considers the spatial adaptability and engineering feasibility of warehouse entities in actual deployment. Specifically, the initial dimensions of each entity can be determined based on the equipment configuration and size expectations indicated by the warehouse layout parameters. These initial dimensions are then adapted and verified for engineering compliance according to the spatial boundaries and spacing requirements defined by the layout constraint rules, resulting in final dimension parameters that can be actually deployed within the target warehouse space. It is worth noting that the dimension information contained in the user's natural language requirement description may be a directional expected value or constraint boundary. For example, a user might specify "shelf height not less than 3 meters," indicating a lower limit for the expected height. This invention also considers factors such as warehouse net height, stacker crane travel range, and maximum cargo access height to comprehensively determine the actual height. Similarly, if a user specifies "shelf width not exceeding 5 meters," this only sets an upper limit constraint. The actual width may be optimized to 4.8 meters based on pallet size or adapted to 4.6 meters based on steel specifications, as long as it does not exceed 5 meters. In summary, the embodiments of the present invention use the size information extracted from user requirements as the initial size parameters, and adapt and adjust the initial size according to warehouse boundary conditions, equipment spacing constraints and equipment specification standardization requirements. Under the premise of satisfying all constraints, the final size parameters of each entity can be determined, so that the final solution not only meets the user's intention, but also has engineering feasibility.

[0046] In one implementation, preset layout constraint rules can be obtained, including boundary dimension restrictions of the target layout space, such as the physical boundaries of the warehouse in the length, width, and height directions; and spacing constraints of storage entities within the target layout space, such as the minimum spacing between racks and side walls, the minimum width of stacker crane aisles, the safety clearance between conveyor lines and racks, and the reserved dimensions for fire exits and maintenance aisles. After obtaining the storage layout parameters, the system determines the initial size parameters of each storage entity based on the entity description parameters and / or cargo parameters in the storage layout parameters. Specifically, the storage layout parameters include rack type (such as double-extension racks, single-extension racks, or shuttle racks), estimated length and height, number of left and right depth bays, aisle width, and other configuration information. The system can query the corresponding storage compartment size configuration based on the rack type. Different rack types correspond to different storage compartment lengths, widths, heights, and structural parameters such as column widths and beam thicknesses. Furthermore, by combining cargo parameters (such as pallet size) and configuration information in the warehouse layout parameters, the actual number of columns, layers, and left and right widths of the shelving can be calculated, thereby determining the actual total width, length, and height of each group of shelving. For example, the actual number of columns is calculated based on the estimated shelf length and shelf unit length; the actual number of layers is calculated based on the estimated height and shelf unit height; and the width of the left and right shelves is calculated based on the number of left and right depth positions and shelf unit width. The left and right widths are then added to the aisle width to obtain the actual total width of the shelving. Based on this, the system can also calculate the total number of storage locations for a single group of shelving based on the number of layers, columns, and left and right depth positions, and compare and verify this with the target number of storage locations.

[0047] After determining the initial dimensional parameters of each storage entity, these parameters are adjusted to fit the boundary dimensions and spacing constraints. In one implementation, multi-dimensional geometric constraint verification can be performed on key dimensional parameters in the storage layout. For example, in the width dimension, the total width of the shelving is calculated based on the width of each shelving group and the aisle clearance. The total width of the conveyor loop is then calculated by combining the loop edge distance, and it is determined whether it exceeds the usable width of the warehouse (total warehouse width minus the distance between the two side walls). If it does, the loop edge distance is automatically reduced for adaptive adjustment.

[0048] Calculate the total width of the shelving based on the width of each shelving group and the aisle clearance, referring to the following formula: W_rack_total=Σ(i=1toN_rack)[W_rack_i(type_i)+G_rack_i]-G_rack_{N_rack} Where N_rack is the number of rack groups (dimensionless); W_rack_i(type_i) is the actual width of the i-th rack group (in meters), determined by the rack type type_i (TwoRack / SingleRack / ShuttleRack correspond to different rack sizes); G_rack_i is the aisle clearance between the i-th and (i+1)-th rack groups (in meters); G_rack_{N_rack} is the redundant clearance after the last rack group (in meters), which is deducted in the summation to ensure dimensional consistency. Both the summation and deduction terms on the right side of the equation are in length dimension (meters), consistent with the length dimension of W_rack_total on the left.

[0049] Calculate the total width of the conveyor loop by combining the loop edge distance, referring to the following formula: W_loop = W_rack_total + 2 * M_loop; where M_loop is the loop margin.

[0050] The available width of the warehouse can be calculated using the following formula: W_avail = W_warehouse - 2 * M_wall; where W_warehouse is the width of the warehouse and M_wall is the distance between the walls (default 1.0 meter).

[0051] The width constraint is: W_loop <= W_avail. When this condition is not met, the system can automatically reduce the loop margin for adaptive adjustment, referring to the following formula: M_loop=max(1.0,(W_avail-W_rack_total) / 2) In terms of length, the total layout length is calculated based on the shelf length, distance from the shelf to the conveyor line, loop depth, platform length, wall distance, and shelf end allowance, referring to the following formula: L_total=L_rack+D_rack_conv+L_loop+L_station+M_wall+M_rack_end Where L_rack is the actual length of the rack, D_rack_conv is the distance from the rack to the conveyor line, L_loop is the loop depth, L_station is the station length, M_wall is the distance from the wall, and M_rack_end is the end allowance of the rack (default 5.0 meters). It can also determine whether it does not exceed the warehouse length, with the length constraint being: L_total <= L_warehouse.

[0052] In terms of height, it can be determined whether the actual height of the rack does not exceed the net height of the warehouse, referring to the following formula: H_rack <= H_warehouse. Where H_rack is the actual height of the rack, and H_warehouse is the net height of the warehouse. In terms of the number of storage locations, it can be determined whether the total number of storage locations has reached the target number of storage locations, referring to the following formula: N_total = N_loc_per_rack * N_rack; where N_loc_per_rack is the number of storage locations per rack, and N_rack is the number of rack groups. The storage location constraint is: N_total >= N_target.

[0053] If any constraint on width, length, or height is not met, the verification result is marked as invalid, and the system automatically reduces the corresponding size or prompts for an adjustment to the layout. If there are insufficient storage spaces, a warning is output but no invalidation is indicated. After the dimensional parameters pass the above verification, the system determines the final dimensional parameters of each storage entity based on the verified parameters.

[0054] Furthermore, regarding the determination of rack dimensions, the corresponding storage compartment size configuration can be queried based on the rack type (TwoRack, SingleRack, or ShuttleRack). Different rack types correspond to different storage compartment lengths (L_cell), widths (W_cell), heights (H_cell), upright widths (W_upright), and beam thicknesses (T_beam). Specific storage compartment size configurations are shown in Table 3 below. Table 3 Storage compartment size configuration

[0055] Based on the above storage compartment size configuration, the actual number of columns and layers can be calculated using the following formula: N_bays=floor(L_estimated / L_cell) N_levels=floor(H_estimated / H_cell) Where L_estimated and H_estimated are the estimated length and height.

[0056] Left / right shelf width calculation: W_left=W_cell*D_left+W_upright*2 W_right=W_cell*D_right+W_upright*2 Where D_left and D_right are the left deep bit and the right deep bit, respectively.

[0057] Calculation of the actual total width of the rack: W_rack = W_left + W_right + W_aisle Where Waisle is the width of the alley.

[0058] Calculation of actual length and height of the shelving: L_rack=L_cell*N_bays+W_upright H_rack=H_cell*N_levels+T_beam*(N_levels+1) Calculation of the number of cargo spaces per rack and the total number of cargo spaces: N_loc_per_rack=N_levels*N_bays*(D_left+D_right) N_total=N_loc_per_rack*N_rack When the 3D engine is unavailable (e.g., due to network connection failure), the system can automatically degrade to local calculation mode, using the built-in rack size configuration (see table above) to complete the same calculations, ensuring system reliability. In conjunction with the above embodiments, the maximum number of rack groups that can be accommodated can also be automatically calculated based on warehouse size constraints, referring to the following formula: N_rack_max=floor((W_avail+G_rack) / (W_rack+G_rack)) The system randomly generates parameters such as warehouse size, number of shelving groups, shelving length and height, and automatically checks the feasibility. If it is not feasible, it regenerates the parameters.

[0059] Step S206: Based on the boundary conditions of the target layout space, the preset layout constraint rules, and the warehouse layout parameters, determine the arrangement parameters of the warehouse entities in the target layout space.

[0060] After determining the dimensional parameters of each storage entity, the system further determines the layout parameters of each storage entity in the target layout space. The layout parameters include the three-dimensional position coordinates of each storage entity (i.e., the X, Y, and Z coordinates of each entity's reference point in the warehouse space, for example, the coordinates of the lower left corner of a row of shelves are (2.0m, 0.5m, 0), where the X-axis is along the width of the warehouse, the Y-axis is along the length of the warehouse, and the Z-axis is along the height) and the relative positional relationships between multiple storage entities (such as the spacing between shelves, the relative orientation of shelves and conveyor lines, the connection position of platforms and conveyor lines, the centering relationship of stacker crane tracks relative to shelves, the alignment relationship and spacing constraints between various devices, etc.).

[0061] In one implementation, the layout parameters of each warehouse entity can be automatically calculated by a rule engine based on the boundary conditions of the target layout space (i.e., the length, width, and height range of the available space inside the warehouse, as well as restrictions on unoccupied areas such as column grids and doorways), preset layout constraints (including equipment spacing standards and safety clearances), and warehouse layout parameters (such as the number of rack rows, the number of racks per row, aisle width, and conveyor line routing). For example, after determining the width of each rack row based on the warehouse length and the number of racks per row, the system determines the spacing between racks based on the aisle width and stacker crane operating requirements, thereby determining the specific coordinate position of each rack row in the warehouse. Assuming an aisle width of 1.2m, the center-to-center distance between two adjacent rows of shelves is the width of a single row of shelves (including storage depth) plus the aisle width. Based on this, the coordinate sequence of each row of shelves along the warehouse width is calculated as 0.5m, 1.7m, 3.2m... until the last row does not exceed the warehouse boundary. For conveyor lines, the system calculates the remaining space dimensions based on the warehouse boundary and the area occupied by the shelves, and then determines the length, direction, and coordinate position of the conveyor line within the warehouse. For platforms, the system determines the specific placement and spacing of the platforms on the side of the conveyor line based on the location of the conveyor line and the platform length and the number of inlet / outlet points in the layout parameters. When any input parameter, such as the number of shelf groups, aisle width, or warehouse boundary, changes, the system can dynamically recalculate the position and relative orientation of all entities globally, ensuring that the layout scheme always meets all engineering constraints.

[0062] The system can automatically calculate the 3D position and rotation angle of all devices through a rule engine. This rule engine can also generate the aforementioned size parameters, and further generate a device list. A 3D Cartesian coordinate system can be used: the X-axis represents the column direction (aisle direction), the Y-axis represents the height direction, and the Z-axis represents the row direction (depth direction). For example, assuming the warehouse width W_warehouse is 50 meters and the length L_warehouse is 80 meters, the system can calculate the total loop width W_loop based on the total rack width W_rack_total and the loop margin M_loop. If the loop width exceeds the available width, the margin is adaptively adjusted. Then, the left and right boundaries X_loop_left and X_loop_right of the loop are calculated to determine the left boundary X_rack_left of the rack area. Based on this, starting from the left boundary of the rack area, the system can sequentially accumulate the width of the preceding rack and the aisle clearance, and combine this with the current rack's own half-width to determine the precise coordinate position of each row of racks along the X-axis.

[0063] The calculation of the total loop width and the formula for adaptively adjusting the edge distance if it exceeds the available width are referenced in step S204, the constraint verification section above. The left and right boundaries of the loop are calculated using the following formula: X_loop_left=(W_warehouse-W_loop) / 2 X_loop_right=X_loop_left+W_loop The left boundary of the shelving area is calculated using the following formula: X_rack_left=X_loop_left+M_loop; Where X_rack_left is the coordinate of the left boundary of the shelving area (unit: meters).

[0064] The X-coordinate of the i-th rack (i=0,1,...,N_rack-1) is calculated using the following formula: X_rack_i=X_rack_left+Σ(k=0toi-1)[W_rack_k(type_k)+G_rack_k]+W_rack_i(type_i) / 2 Where Σ(k=0toi-1)[W_rack_k(type_k)+G_rack_k] represents the cumulative occupied width (in meters) of the first i groups of shelves, where W_rack_k(type_k) is the actual width (in meters) of the k-th shelf group determined by its type type_k, and G_rack_k is the aisle clearance (in meters) behind the k-th shelf group; W_rack_i(type_i) / 2 is the half-width (in meters) of the current i-th shelf group. The X-axis reference position of the i-th shelf group is determined by the starting boundary of the shelf area, the cumulative occupied width (including aisle clearance) of the first i-1 shelves, and the half-width of the current shelf group. All terms on the right side of the equation are in the dimension of length (meters), consistent with the dimension of length of X_rack_i on the left side.

[0065] The Z-axis coordinate of the shelving is calculated using the following formula: Z_back=L_warehouse-M_wall Z_front=Z_back-L_rack Z_rack = (Z_front + Z_back) / 2 When the shelf length exceeds the maximum usable length, it will automatically shrink, referring to the following formula: L_rack_max=L_avail-D_rack_conv-L_loop-L_station-M_wall-M_rack_end-2 If L_rack > L_rack_max and L_rack_max > 10, then L_rack = L_rack_max.

[0066] Further, the coordinates of the conveyor line loop can be calculated by the following formula: Calculation of Z coordinates for the front conveyor line and the rear conveyor line: Z_conv_front=Z_front-D_rack_conv Z_conv_back=Z_conv_front-L_loop Calculation of the Z coordinate at the front end of the station: Z_station_front=Z_conv_back-L_station When Z_station_front<M_wall, perform backward offset compensation: delta=M_wall-Z_station_front Z_conv_back=Z_conv_back+delta;Z_conv_front=Z_conv_back+L_loop Wherein, each aisle on the front conveyor line can correspond to two jacking translators (warehousing side and outbound side), with an offset of station_offset, and the corresponding coordinates can be calculated by the following formula: X_lifter_in_i=X_rack_i-station_offset X_lifter_out_i=X_rack_i+station_offset Turning jacking translators are arranged at both ends of the loop, and the rotation angle R is calculated as: R=90(if X==X_loop_left);R=270(if X==X_loop_right) The rotation angle R at the non-turning position is 0 (front section) or 180 (rear section).

[0067] Further, for the horizontal conveyor line, it can be automatically segmented according to a fixed segment length L_seg (default 1.5 meters), with reference to the following formula: N_seg=ceil(L_segment / L_seg)-N_avoid Where L_segment is the actual usable length of the conveyor line segment (in meters), i.e., the net length after deducting the space occupied by equipment; L_seg is the standard segment length of the conveyor line (in meters, default 1.5 meters); ceil(L_segment / L_seg) is the theoretical number of segments (dimensionless) divided according to the standard segment length; N_avoid is the number of equipment (such as lifting and transferring machines) that need to be avoided on the conveyor line segment (dimensionless), each of which occupies one segment length of space and therefore needs to be deducted. Both terms on the right side of the equation are dimensionless, consistent with the dimensionless N_seg on the left side.

[0068] The X-coordinate of the center of segment j (j=0,1,...,N_seg-1) is calculated using the following formula: X_seg_j = X_start + (j + 0.5) * L_seg The number of sections and section lengths of the side conveyor lines (longitudinal connecting sections on the left and right sides of the loop) are calculated based on the actual spacing using the following formula: N_side=max(1,round(L_side / L_seg)) L_side_seg=L_side / N_side Furthermore, the X-coordinate of the i-th platform is the same as that of the corresponding shelf. The platform type can be alternately assigned according to the aisle number, and the calculation formula is as follows: type_i=outlet(ifimod2==0);type_i=inlet(ifimod2==1) The Z-coordinate of the platform is calculated using the following formula: Z_station=Z_conv_back-W_conv / 2-L_station / 2 Where W_conv is the width of the conveyor line.

[0069] The dimensions of all storage entities can be uniformly converted to millimeters (multiplied by 1000) during output. The final generated device list includes fields for each device such as type, object type, position coordinates (position:x,y,z), size (size:length,height,width), and rotation angle (rotation:x,y,z).

[0070] Step S208: Construct a physical model of the warehouse entity based on the size parameters, and arrange the physical model in the target layout space based on the layout parameters to form a warehouse layout scheme.

[0071] After determining the size and layout parameters of each storage entity, the system proceeds to the model building and scene assembly stage. In one implementation, a model generator can be invoked, submitting the type and size parameters of the storage entities as generation tasks to obtain corresponding 3D solid models for each entity. In another implementation, the type and size parameters of the storage entities can be submitted as text prompts to the model generator. For example, submitting descriptions such as "double-extension rack, length 14.4 meters, width 2.4 meters, height 7.5 meters" will allow the model generator to generate the corresponding 3D model based on the text prompts. In yet another implementation, the system can submit preset reference images (such as reference photos or design drawings of standard racks or conveyor lines) to the model generator. The model generator will then generate the corresponding 3D model based on the image content, enabling rapid customization of personalized appearances without the need for manual modeling by professional modelers. Solid models are used to represent the geometric shape, spatial occupancy, and key structural features of storage entities, realistically reflecting the size proportions and outlines of the equipment in a 3D virtual environment.

[0072] When submitting a generation task, it can be executed in an asynchronous submission mode, where the system immediately returns the task ID after submission, and the generation progress and results can be queried and retrieved later using the task ID. Alternatively, it can be executed in a synchronous waiting mode, where, after submitting a 3D generation task, the system blocks on the server until the task completes, with a maximum waiting time of 600 seconds. Upon completion, it directly returns the access URL of the model file and a preview image, suitable for scenarios requiring immediate response time. Using these methods, model generation for a single device can be reduced from several days to seconds.

[0073] Before submitting the model generation task, the system can again invoke the large language model to simplify the user-inputted equipment text description into a model name of no more than six Chinese characters, serving as the identifier for the generated model. The system prompt for this invocation sets the model's role as a "model naming assistant," requiring it to extract core information from the user's description and condense it into a concise name based on an understanding of the equipment type and specifications. Simultaneously, the temperature parameter is set to 0.3 (lower than the 0.7 used during layout parameter parsing) to obtain a more stable and concise output, avoiding redundant or off-topic naming. For example, if the user's original description is "double-extension rack, length 30.12 meters, width 2.48 meters, height 6.5 meters," the system, after invoking the large language model, returns "double-extension rack" as the model name. Through this naming simplification process, the generated model file and 3D assets have standardized and identifiable tags, facilitating subsequent retrieval, management, and display. After determining the simplified model name, the system submits the warehouse entity's type, size parameters, and model name as a generation task to the model generator. Building upon this foundation, the system further persists task information such as model name, task identifier, prompt word, mode, status, and creation time to an SQLite database, supporting subsequent asynchronous queries and status updates. The data structure of task records can include fields such as task_id (primary key), job_id, status, model_url, model_format, preview_image, prompt, model_name, mode, created_at, and completed_at, enabling traceable management of the entire lifecycle of generated tasks.

[0074] After obtaining the physical models of each warehouse entity, the system places each entity model in its corresponding position in the target layout space based on the three-dimensional position coordinates and relative positional relationships in the layout parameters, forming a complete warehouse layout scheme. This invention not only provides interactive viewing and spatial occupancy verification from any perspective in a three-dimensional scene, but also allows for direct export of engineering drawings or import into a logistics simulation environment for subsequent verification, without the need for additional secondary modeling.

[0075] In one implementation, the above-mentioned layout calculation and model generation process can be functionally divided into two core modules: a planner and a model generator. The planner is responsible for parsing warehouse layout parameters, verifying constraints, and calculating equipment coordinates. It encompasses all functions from natural language requirement parsing and precise size calculation to determining layout parameters. Internally, it can further include a natural language parsing submodule (constructing system prompts and calling a large language model to convert user descriptions into structured layout parameters), a parameter verification submodule (performing four-dimensional constraint verification based on width, length, height, and number of storage locations), and an equipment layout submodule (automatically calculating the three-dimensional coordinates, dimensions, and rotation angles of all equipment using a rule engine and generating an equipment list). The model generator is responsible for generating corresponding three-dimensional solid models based on size parameters and equipment types. Its task submission submodule supports two generation methods: text-based 3D (submitting type and size parameters as text prompts) and image-based 3D (submitting preset reference images), and provides both asynchronous submission and synchronous waiting task modes. The task persistence submodule stores task information in a database to support status tracking and querying. The module composition is shown in Table 4 below.

[0076] Table 4 Module Composition

[0077] The planner and model generator are decoupled through a standardized data interface. The size and layout parameters output by the planner serve as input to the model generator. The 3D assets returned by the model generator are then uniformly assembled by the planner into the corresponding positions in the target layout space, forming a complete warehouse layout scheme. In one implementation, the planner and model generator can be deployed as independent services, allowing layout calculation and model generation to be executed in parallel without blocking each other. When the user only needs to adjust the layout parameters, there is no need to regenerate the 3D assets, thus significantly improving overall design efficiency and system response speed. In another implementation, they can also be integrated and deployed in the same process to simplify the system architecture and deployment complexity, which can be flexibly selected according to the actual application scenario. The planner adopts an abstract base class design, and different subclasses can be implemented to adapt to various warehouse layout scenarios. The model generator connects to different 3D generation services through standardized interfaces, facilitating the replacement or upgrading of the underlying generation model. This functional division method gives the system good scalability.

[0078] Building upon this foundation, this embodiment of the invention further employs a planner registry and factory mechanism to achieve dynamic selection and flexible expansion of planners. Specifically, the system maintains a planner registry (PLANNER_REGISTRY), and dynamically creates corresponding planner instances based on type and mode using the factory method get_planner(planner_type, mode). Planner types can include asrs_unity (relies on Unity to calculate shelf dimensions, with parameters in meters) and asrs_web (calculates shelf dimensions on the front end, with parameters in millimeters). The planner registry employs a lazy loading strategy, creating planner instances upon first call, avoiding unnecessary resource consumption. When adding a new planner type, it is only necessary to implement the abstract methods of BasePlanner (e.g., parameter validation, device layout generation, AI parameter parsing, AI response generation, random parameter generation) and register them with PLANNER_REGISTRY, without modifying the service orchestration module code, thus achieving good scalability.

[0079] The data flow relationship between the above modules is as follows: The service orchestration module, as the system entry point, is responsible for coordinating the calling order and data flow of each module. Specifically, after receiving a user request, the service orchestration module calls the planner module's `ai_generate_params` method to parse the natural language requirements into structured layout parameters, calls the `validate` method to perform multi-dimensional geometric constraint verification, calls the size calculation module to obtain precise shelf dimensions, calls the `generate_devices` method to generate a device list, calls the simulation verification module to verify the logistics path, and finally returns the layout data and simulation results to the front end or 3D engine. The model generator module runs independently of the main process. Users can submit 3D generation requests, query task status, and obtain generated model assets at any time through asynchronous interfaces without blocking the main process.

[0080] It should be noted that the functional division and registration mechanism of the planner and model generator described above is only one optional implementation of the present invention. In specific implementations, layout calculation and model generation can also be executed by the same engine, without being limited thereto.

[0081] In one implementation, after forming a warehouse layout plan, the system automatically constructs a discrete event simulation environment to verify the connectivity of logistics paths and equipment utilization. Specifically, based on the connector type and spatial location of each warehouse entity in the warehouse layout plan, a connection relationship is established between each warehouse entity to form a logistics topology network; based on the logistics topology network, at least one transportable route for goods is determined; and preset goods are controlled to be transported along the transportable route to simulate the flow process of the preset goods in the warehouse layout plan.

[0082] Connectors are interfaces on equipment used for connecting and transferring materials. Each connector has a specific directional attribute. Connector types include outgoing connectors (e.g., outbound, chain output, roller conveyor output) and incoming connectors (e.g., incoming, chain input, roller conveyor input). Spatial position refers to the connector's three-dimensional coordinates in the world coordinate system. When establishing a connection, the direction matching rule requires that the outgoing and incoming connectors have complementary orientations (e.g., if the outgoing direction of one device is +X, then the incoming direction of another device should be -X). Spatial proximity judgment requires that the distance difference between the two connectors in the X and Z directions is less than a preset threshold (e.g., 3000 mm). A connection is established only when both conditions are met simultaneously. Spatial proximity judgment is performed using the following formula: connect(C_i,C_j)iffdir_match(C_i.out,C_j.in)AND |Δx(C_i,C_j)| <D_thresholdAND|Δz(C_i,C_j)|<D_threshold Here, `dir_match` is the direction matching function (returning a dimensionless Boolean value), which determines whether the orientations of the direction connector C_i.out and the in direction connector C_j.in are complementary (e.g., if the outward direction of C_i is +X, then the inward direction of C_j should be -X); Δx(C_i,C_j) and Δz(C_i,C_j) are the differences in the world coordinates of the two connectors in the X and Z directions (unit: millimeters, length dimension); D_threshold is the proximity threshold (unit: millimeters, default 3000 millimeters, length dimension). A connection is established only when the orientations match and the distance difference in the X and Z directions is less than the threshold. Both sides of the equal sign are dimensionless Boolean judgments with consistent dimensions. For roller conveyor equipment, adjacent rollers arranged along the same coordinate line are grouped by coordinate (tolerance 500 mm), and within the same group, they are sorted by coordinate and connected sequentially to form the longitudinal spine of the roller conveyor (maximum connection distance 15000 mm, covering approximately 2 to 3 aisle spacings). The specific grouping method (grouping adjacent roller conveyor equipment arranged along the Z-axis on the same X-coordinate line according to the X-coordinate) refers to the following formula: group_key=round(X / tolerance_x)*tolerance_x Within the same group, sort by Z-coordinate and connect sequentially, with the connection conditions referring to the following formula: connect(roller_i,roller_j)iff|dz| <D_spine_max Where D_spine_max is the maximum connection distance of the spine (default 15000 mm, covering approximately 2 to 3 aisle spacings). Using the above method, a directed graph with equipment as nodes and connection relationships as edges can be established to represent the flow path of goods between various devices in the warehouse.

[0083] For example, a directed graph can be represented by G=(V,E).

[0084] Where V represents the set of unique identifiers for all warehouse entities, as shown in the following formula: V={device_id|device in devices}; E represents the set of established connections between devices, as shown in the following formula: E={(u,v)|connected(u.connector_out,v.connector_in)}.

[0085] Here, `devices` is the set of all warehouse entities, `device_id` is the unique identifier of each entity, `u` and `v` are the warehouse entities corresponding to the start and end points of the directed edges, respectively, and `connected()` is the connection determination function. It returns true when the outgoing connector of device `u` and the incoming connector of device `v` satisfy both directional matching and spatial proximity conditions, indicating the existence of a directed path from `u` to `v`. Based on the above directed graph construction, the system uses a breadth-first search algorithm to find the transportable path of goods from the inbound station node to the stacker crane node.

[0086] Since the warehouse layout scheme has established the precise location coordinates and relative orientations of each warehouse entity within the target layout space, and the equipment forms a complete connection network in space through connectors, the system can determine at least one transportable path for goods from the directed graph corresponding to this layout. In one implementation, a breadth-first search (BFS) algorithm can be used to find the shortest path from the receiving station to the stacker crane. Refer to the following formula: path = BFS(G, start_id, end_id). Where G is the directed graph, start_id is the receiving station node, and end_id is the stacker crane node. Path finding supports two modes: searching for a specified endpoint and searching for any node in the endpoint set. The former searches for a path from the receiving station to a specific stacker crane, while the latter searches for a path from the receiving station to any available stacker crane, adapting to the path selection requirements of different warehousing operation scenarios. For example, in a directed graph containing an inbound platform G1, a conveyor line C1, a lifting and transferring machine L1, and a stacker crane S1, with the connection relationship between the equipment being G1→C1→L1→S1, then the path G1-C1-L1-S1 is a transportable route for goods. Through the above path search, the system can automatically verify whether the logistics path is connected before the plan is implemented, effectively avoiding logistics blockages caused by missing equipment connections or unreasonable layout.

[0087] After determining the transportable paths, the preset goods can be controlled to move along these paths to simulate the flow of the preset goods in the warehouse layout scheme. In one embodiment, an intelligent goods generator can be created to continuously generate goods at preset time intervals (e.g., 20 seconds), as shown in the following formula: T_interval = 20s (goods generation interval). Each time a path is generated, one can be randomly selected from all reachable paths (i.e., randomly selected from all possible combinations of inbound stations, stacker cranes, and paths), as shown in the following formula: (gate, stacker, path) = random_choice(all_options); where all_options is the set of all (inbound station, stacker crane, path) triples, allowing the goods to be transferred sequentially between the selected devices along the selected path, simulating the complete flow of goods from inbound to storage. If no reachable path is found, the simulation can be downgraded to a traditional simulation mode, creating an independent goods generator for each inbound station, where goods only move within the aisle corresponding to that station and are not transported across aisles.

[0088] Furthermore, during the simulation of controlling the transport of pre-set goods along the transportable path, various operational data can be continuously collected and recorded in a time series, such as system-level flow indicators. Each data point includes indicators such as time, number of goods generated, number of goods processed, number of goods stored, and number of goods retrieved. These are used to evaluate the logistics operation status of the layout scheme in different time periods, identify periods of goods backlog and peak equipment processing times, and thus discover potential logistics congestion points before actual operation. Simultaneously, equipment-level performance indicators can also be recorded, such as the cumulative processing time (i.e., the time actually occupied by the equipment in processing goods, such as the time for stacker cranes to perform storage and retrieval operations, and the time for conveyor lines to transport goods) and cumulative waiting time (i.e., the time goods wait in line in front of the equipment) of each logistics operation equipment involved in goods transport in each warehousing entity. Based on this, the operating utilization rate of each logistics operation equipment can be calculated (utilization rate = cumulative processing time / total simulation time, with an upper limit of 100%). The utilization rate is calculated using the following formula: U_k = (T_busy_k + T_wait_k) / T_total * 100%. Where T_busy_k is the cumulative processing time of the k-th device (unit: seconds, time dimension), i.e., the time the device is performing goods handling or storage operations; T_wait_k is the cumulative waiting time (unit: seconds, time dimension), i.e., the time the device has tasks queued but has not yet started processing; T_total is the total simulation time (unit: seconds, time dimension). The numerator (T_busy_k + T_wait_k) is in time dimension, the denominator T_total is in time dimension, the ratio is a dimensionless ratio, multiplied by 100% to get a percentage form, consistent with the dimension of U_k on the left. The utilization rate upper limit can be set to 100%, referring to the following formula: U_k=min(1.0,T_busy_k / T_total)*100% By combining the above two types of indicators, it can be determined whether the equipment configuration in the layout scheme is reasonable and whether there are bottleneck or idle equipment. For example, a stacker crane with a utilization rate of 95% indicates that it is constantly busy and may become a system bottleneck, while a utilization rate of less than 10% indicates that the equipment configuration is excessive, and the number of equipment or the tasks can be optimized or the tasks can be reallocated. At the same time, the system also evaluates the capacity throughput level of the warehouse entity (such as the number of pallets processed per hour) based on the processing capacity of each logistics operation equipment and the actual processing volume during the simulation period, in order to assess whether the overall logistics efficiency of the layout scheme meets the warehouse operation requirements. The simulation process runs continuously and collects data until the preset simulation time is reached or the user manually terminates it. In one implementation, a maximum of 500 historical data points can be saved.

[0089] Through the above-mentioned automatic simulation verification, designers can comprehensively evaluate the logistics efficiency and equipment utilization rationality of the layout scheme before the scheme is implemented, without having to import the data into external simulation software for secondary verification, which significantly shortens the design verification cycle.

[0090] After simulation verification is completed, the results of the aforementioned steps can be integrated and output, including equipment layout data (i.e., an equipment list, containing complete attribute fields such as type identifier, 3D position coordinates, geometric dimensions, and spatial rotation angles for each storage entity), verification results (including dimensional verification information such as actual number of storage locations, total layout length, and total width), simulation results (including statistical indicators such as the operating utilization rate and cargo throughput of each logistics operation equipment), and customized 3D assets (if any). The above output data can be directly sent to the 3D rendering engine for visualization or sent to the front-end web interface for designers to preview interactively.

[0091] Based on the above system embodiments, this invention also provides a warehouse layout generation device, such as... Figure 3 As shown, the device includes: a data acquisition module 10, used to respond to receiving a user's natural language requirement description and determine the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description; a parameter determination module 20, used to determine the size parameters and arrangement parameters of the warehouse entity to be arranged in the target layout space based on preset layout constraint rules and warehouse layout parameters; and an execution module 30, used to construct a physical model of the warehouse entity based on the size parameters and arrange the physical model in the target layout space based on the arrangement parameters to form a warehouse layout scheme.

[0092] The warehouse layout generation device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0093] This invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described... Figures 1-2 The steps of any of the methods shown. Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the above-described steps. Figures 1-2 The steps of any of the methods shown. Embodiments of the present invention also provide a structural schematic diagram of an electronic device, such as... Figure 4The diagram shows the structure of the electronic device, which includes a processor 101 and a memory 100. The memory 100 stores computer-executable instructions that can be executed by the processor 101. The processor 101 executes the computer-executable instructions to implement the above-mentioned... Figures 1-2 Any of the methods shown.

[0094] exist Figure 4 In the illustrated embodiment, the electronic device further includes a bus 102 and a communication interface 103, wherein the processor 101, the communication interface 103, and the memory 100 are connected via the bus 102. The memory 100 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk drive. Communication between this system network element and at least one other network element is achieved through at least one communication interface 103 (which can be wired or wireless), using the Internet, wide area network, local area network, metropolitan area network, etc. Bus 102 can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, or an AMBA (Advanced Microcontroller Bus Architecture) bus. AMBA defines three types of buses: APB (Advanced Peripheral Bus), AHB (Advanced High-performance Bus), and AXI (Advanced eXtensible Interface). Bus 102 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4The diagram uses only a single double-headed arrow, but this does not imply a single bus or a single type of bus. Processor 101 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 101 or by instructions in software form. Processor 101 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this invention can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor 101 reads information from the memory and, in conjunction with its hardware, completes the aforementioned tasks. Figures 1-2 Any of the methods shown.

[0095] The computer program product of the warehouse layout generation method, apparatus, and electronic device provided in this invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system described above can be referred to the corresponding process in the preceding method embodiments, which will not be repeated here. If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0096] In the description of this invention, it should be noted that the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Finally, it should be noted that the above embodiments are merely specific implementations of this invention, used to illustrate the technical solutions of this invention, and not to limit it. The scope of protection of this invention is not limited thereto. Although this invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in this invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this invention, and should all be covered within the scope of protection of this invention. Therefore, the scope of protection of this invention should be determined by the scope of the claims.

Claims

1. A method for generating warehouse layouts, characterized in that, The method includes: In response to receiving a user's natural language requirement description, the warehouse layout parameters corresponding to the natural language requirement description are determined based on the layout structure information contained in the natural language requirement description. Based on the preset layout constraint rules and the warehouse layout parameters, the size parameters and arrangement parameters of the warehouse entities to be arranged in the target layout space are determined. Based on the size parameters, a physical model of the warehouse entity is constructed, and based on the layout parameters, the physical model is arranged in the target layout space to form a warehouse layout scheme.

2. The method according to claim 1, characterized in that, The step of determining the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description includes: Based on the natural language requirement description, prompt words are constructed according to a predefined data structure; The prompt words are input into a large language model, which then parses the prompt words to determine the layout structure information contained in the natural language requirement description. According to the preset warehouse layout parameter structure, the layout structure information is converted into the warehouse layout parameters.

3. The method according to claim 2, characterized in that, The layout structure information includes the storage entities to be arranged, and the method further includes: Identify the entity parameter group whose parameter content has been filled in the warehouse layout parameters; Based on the preset correspondence between the entity parameter group and the preset storage entity type, the storage entity type corresponding to the entity parameter group is determined as the storage entity to be deployed.

4. The method according to claim 1, characterized in that, The steps for determining the size parameters and arrangement parameters of the storage entity in the target layout space based on preset layout constraint rules and the storage layout parameters include: Based on the preset layout constraint rules and the warehouse layout parameters, the size parameters of the warehouse entity are determined; Based on the boundary conditions of the target layout space, the preset layout constraint rules, and the warehouse layout parameters, the arrangement parameters of the warehouse entities in the target layout space are determined; the arrangement parameters include the three-dimensional position coordinates of the warehouse entities and the relative positional relationships between multiple warehouse entities.

5. The method according to claim 4, characterized in that, The step of determining the size parameters of the storage entity based on the preset layout constraint rules and the storage layout parameters includes: Obtain the preset layout constraint rules; the layout constraint rules include the boundary size limit of the target layout space and the spacing constraint of the storage entities within the target layout space; Based on the entity description parameters and / or cargo parameters in the warehouse layout parameters, determine the initial size parameters of the warehouse entity; Based on the boundary size constraints and the arrangement spacing constraints, the initial size parameters are adapted and adjusted so that the adjusted size parameters satisfy the boundary size constraints and the arrangement spacing constraints, thus obtaining the final size parameters.

6. The method according to claim 1, characterized in that, The steps of constructing a physical model of the warehouse entity based on the size parameters, and arranging the physical model in the target layout space based on the layout parameters to form a warehouse layout scheme include: Call the model generator, submit the type of the warehouse entity and the size parameters as generation tasks to the model generator to obtain the entity model corresponding to the warehouse entity; Based on the three-dimensional position coordinates and relative position relationships in the arrangement parameters, the entity model is placed at the corresponding position in the target layout space to form the warehouse layout scheme.

7. The method according to claim 6, characterized in that, The steps of submitting the type and size parameters of the storage entity as generation tasks to the model generator to obtain the entity model corresponding to the storage entity include: The type and size parameters of the warehouse entity are submitted as text prompts to the model generator to obtain the corresponding entity model; or, a preset reference image is submitted to the model generator to obtain the corresponding entity model.

8. The method according to claim 1, characterized in that, The method further includes: Based on the connector type and spatial location of each warehousing entity in the warehousing layout scheme, establish the connection relationship between each warehousing entity to form a logistics topology network; Based on the logistics topology network, at least one transportable route for goods is determined; Control the transport of pre-set goods along the transportable path to simulate the flow of the pre-set goods in the warehouse layout scheme.

9. The method according to claim 8, characterized in that, After the step of controlling the transport of the preset goods along the transportable path of the goods, the method further includes: The cumulative processing time and cumulative waiting time of each logistics operation equipment involved in the transfer of goods in the warehousing entity are recorded during the transfer process in order to calculate the operating utilization rate of each logistics operation equipment. The capacity throughput level of the warehouse entity is evaluated based on the cumulative processing time, the cumulative waiting time, and the total simulation time.

10. A warehouse layout generation device, characterized in that, The device includes: The data acquisition module is used to respond to receiving a user's natural language requirement description and determine the warehouse layout parameters corresponding to the natural language requirement description based on the layout structure information contained in the natural language requirement description. The parameter determination module is used to determine the size parameters and arrangement parameters of the storage entity to be arranged in the target layout space based on the preset layout constraint rules and the storage layout parameters. The execution module is used to construct a physical model of the warehouse entity based on the size parameters, and arrange the physical model in the target layout space based on the layout parameters to form a warehouse layout scheme.

11. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in any one of claims 1-9.

12. A computer-readable storage medium, characterized in that, The device contains a computer program that, when executed by a processor, implements the method described in any one of claims 1-9.

Images