Intelligent algorithm driven material usage optimization method and system

Abstract

This invention relates to the field of building materials management technology, and in particular to an intelligent algorithm-driven method and system for optimizing material usage. The method generates candidate material usage schemes by analyzing the structural design model, constructs a constraint network for surplus material reuse and encodes the material sharing constraints between components, aims to minimize the total material procurement, uses an intelligent search algorithm to find the optimal solution, generates a batch material procurement list and performs batch optimization. This invention can effectively improve material utilization and reduce procurement costs and construction waste.

Description

Technical Field

[0001] This invention relates to the field of building materials management technology, and in particular to a method and system for optimizing material usage driven by intelligent algorithms. Background Technology

[0002] In engineering fields such as construction and manufacturing, accurate calculation and optimization of material usage is a key link in controlling costs and reducing waste. Traditional material usage planning usually relies on manual experience or simple calculations based on two-dimensional drawings. Designers manually count the dimensions and quantities of various components according to structural design drawings, and then estimate the theoretical amount of materials needed by combining the standard specifications of materials with empirical formulas or table lookup. This method largely depends on the planner's personal experience and familiarity with the process, and the calculation process is cumbersome and prone to errors.

[0003] In the actual material processing, cutting standard materials to meet specific component dimensions will inevitably generate leftover material. Traditional methods or existing software usually treat this leftover material as waste and fail to systematically consider how to use it for other suitable components during the planning stage. This leads to a decrease in material utilization and increases the total amount of unnecessary material procurement and waste disposal costs.

[0004] Conventional methods often rely on theoretical calculations of material usage, with the optimization goal simply being to meet the dimensional requirements of all components. This static planning fails to adequately consider dynamic construction constraints, such as the impact of processing sequence and batch procurement on material utilization. More importantly, it lacks a closed-loop feedback and learning mechanism. Processing losses during actual construction, deviations between actual material consumption and theoretical values, and the actual generation of surplus materials cannot be effectively collected and used to correct and optimize subsequent procurement and material cutting plans. This leads to a disconnect between planning and execution, making continuous optimization and improvement difficult. Summary of the Invention

[0005] This invention provides a method and system for optimizing material usage driven by intelligent algorithms, which can solve the problems in the prior art.

[0006] A first aspect of the present invention provides an intelligent algorithm-driven method for optimizing material usage, comprising:

[0007] Obtain structural design model data, construction process constraint data, and material physical property data for the target project;

[0008] Perform component geometry analysis and material requirement mapping on the structural design model data, establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, and generate a set of candidate material usage schemes;

[0009] Based on the set of candidate material usage schemes, and combined with the material physical property data and construction process constraint data, a material surplus reuse constraint network is constructed to identify reusable surplus material segments and encode the surplus material reuse relationship as material sharing constraints between components.

[0010] With minimizing the total material procurement quantity as the objective function and the constraints of component geometric parameter satisfaction and material sharing as the constraints, the optimal material usage scheme that maximizes material utilization is found in the set of candidate material usage schemes through an intelligent search algorithm.

[0011] Based on the optimal material usage plan, a batch material procurement list is generated, and the actual material consumption data and surplus material generation data of the previous batch are collected. The processing loss parameters are calculated, and the next batch material procurement list is optimized.

[0012] The structural design model data is subjected to component geometry analysis and material requirement mapping to establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, generating a set of candidate material usage schemes, including:

[0013] The geometric entity model of the component is parsed from the structural design model data, and the boundary contour of the component geometric entity model is extracted to obtain the contour vertex coordinate sequence and boundary curve equation of the component.

[0014] Based on the sequence of vertex coordinates of the contour, the area ratio of the minimum bounding rectangle to the maximum inscribed rectangle of the component is calculated to obtain the shape complexity coefficient of the component.

[0015] The shape complexity coefficient is correlated with the material elastic modulus and material density in the material physical property data to determine the material deformation compensation amount when the component is cut, and the coordinates of the contour vertex coordinate sequence are corrected according to the material deformation compensation amount.

[0016] Based on the corrected contour vertex coordinate sequence, the material standard plate size library is traversed and two-dimensional packing calculations are performed to generate multiple layout schemes for each component under different plate sizes.

[0017] The various plate arrangement schemes for each component are summarized, and a mapping relationship between component identifiers and plate arrangement schemes is established to form a set of candidate material usage schemes.

[0018] Based on the corrected contour vertex coordinate sequence, the system traverses all sheet metal specifications in the material standard sheet metal size library and performs two-dimensional packing calculations to generate multiple layout schemes for each component under different sheet metal specifications, including:

[0019] Obtain the length and width dimensions of the board from the standard board size library, establish a two-dimensional coordinate system for the board, and set the lower left corner vertex of the board as the origin of the coordinate system;

[0020] Based on the corrected contour vertex coordinate sequence, calculate the long side direction and short side direction of the minimum bounding rectangle, and generate the initial placement posture angle and the candidate placement posture angle after rotation transformation of the component.

[0021] For each candidate placement posture angle, the modified contour vertex coordinate sequence is rotated and transformed, the envelope size of the component contour after rotation and transformation is calculated, and it is determined whether the envelope size meets the size constraint conditions of the plate.

[0022] For candidate placement posture angles that meet the size constraints, the starting coordinates for placement of the component outline are determined in the two-dimensional coordinate system of the plate. An edge-fitting strategy is used to adjust the position of the component outline along the plate boundary, generating multiple candidate placement positions of the component on the plate.

[0023] For each candidate placement position, calculate the area occupied by the component outline on the board, record the occupied area and the corresponding candidate placement posture angle, and combine the board size to form multiple board layout schemes for each component.

[0024] Based on the set of candidate material usage schemes, and combined with the material physical property data and construction process constraint data, a material surplus reuse constraint network is constructed to identify reusable surplus material segments and encode the surplus material reuse relationships as material sharing constraints between components, including:

[0025] Based on the set of candidate material usage schemes, the plate cutting path is analyzed and the leftover material polygons are identified. The vertex coordinates and topological adjacency relationships of the leftover material polygons are extracted.

[0026] Perform a shape decomposition operation on the remaining polygon to decompose the irregular remaining material into multiple regular sub-remaining material units, and calculate the maximum inscribed rectangle size and shape regularity index of each sub-remaining material unit.

[0027] Extract the geometric contour and size requirements of the component to be processed, establish a multi-dimensional matching space between the component's geometric feature vector and the feature vector of the sub-scrap unit, perform nearest neighbor search in the multi-dimensional matching space, calculate the Euclidean distance between the feature vectors, identify the pairing relationship between the component and the sub-scrap where the Euclidean distance is less than the matching threshold, and combine the cutting accuracy loss parameter in the construction process constraint data to perform geometric feasibility verification on the pairing relationship, and screen out geometrically feasible scrap reuse pairings.

[0028] For the geometrically feasible material reuse pairing, a multi-level material reuse constraint network is constructed. The network includes three levels of nodes: plate layer, material reuse layer, and component layer. Cross-level reuse dependency edges are established. The reuse dependency edges are traversed to extract cross-level material flow paths. The plate identifier, material reuse unit identifier, and component identifier are encoded into material sharing constraints in the form of triples.

[0029] Based on the cutting accuracy loss parameters in the construction process constraint data, the geometric feasibility of the pairing relationships is verified, and geometrically feasible leftover material reuse pairings are selected, including:

[0030] The cutting accuracy loss parameters are obtained from the construction process constraint data. The cutting accuracy loss parameters include cutting positioning error, cutting path offset and edge burr compensation. The geometric feature vectors of the components and sub-residual material units in the pairing relationship are extracted.

[0031] The uncertainty range of the sub-residual material unit boundary position is calculated based on the cutting positioning error, the systematic deviation of the cutting size is calculated based on the cutting path offset, the removal amount of edge trimming is calculated based on the edge burr compensation amount, and the overall size deviation of the sub-residual material unit is calculated and the conservative usable size of the sub-residual material unit is determined by superimposing the results according to the most unfavorable combination principle.

[0032] Length and width requirements are extracted from the component geometric feature vector to form a size requirement vector. The conservative available size is constructed into a size supply vector. Vector subtraction is performed between the size supply vector and the size requirement vector to obtain the size margin vector.

[0033] Determine whether the length and width components of the dimensional margin vector are both greater than the safety margin threshold. If both are greater, the pairing relationship passes the geometric feasibility check. The pairing relationships that pass the geometric feasibility check are filtered out and output to form geometrically feasible scrap material reuse pairings.

[0034] Using minimizing the total material procurement quantity as the objective function and constraining the satisfaction of component geometric parameters and material sharing, an intelligent search algorithm is used to find the optimal material usage scheme that maximizes material utilization from the set of candidate material usage schemes, including:

[0035] Based on the set of candidate material usage schemes, the required plate specifications and quantities for each scheme are statistically analyzed, and a mapping table is established from candidate scheme codes to total material procurement quantity, with the total material procurement quantity as the objective function.

[0036] Based on the material sharing constraints in the form of triplets, the conflict relationships between multiple components sharing the same surplus material unit are identified, and the material sharing constraints are converted into mutually exclusive logical expressions between candidate solution variables.

[0037] Extract the dependency sequence between scrap units and components from the material sharing constraints, and establish the pre-dependency constraints between candidate solutions based on the order of plate cutting and the temporal relationship of scrap generation.

[0038] The initialization of the intelligent search algorithm's decoding method maps each candidate solution to a decision variable, generating an initial candidate solution population. The constraint violation degree is calculated for each candidate solution in the population, and the number of violations of mutual exclusion logic expressions and pre-dependent constraints is counted. The comprehensive evaluation value of the candidate solution is obtained by weighted summation of the objective function value and the constraint violation degree.

[0039] An elite retention strategy is adopted to iteratively evolve the candidate solution population. High-quality candidate solutions are selected based on the comprehensive evaluation value. New candidate solutions are generated through crossover and mutation operations. The search is terminated when the optimal solution of the population remains stable for several consecutive generations or reaches the maximum number of iterations. The candidate solution with the best comprehensive evaluation value is output as the optimal material usage scheme.

[0040] Based on the optimal material usage plan, a batch material procurement list is generated, and the actual material consumption data and surplus material generation data of the previous batch are collected. Processing loss parameters are calculated, and the next batch material procurement list is optimized, including:

[0041] Extract the material requirements list of the components from the optimal material usage scheme, divide the components into multiple sub-lists evenly according to the batch quantity of the construction plan and the construction time, and generate a batch material procurement list by statistically analyzing the plate specifications and quantities of each sub-list.

[0042] After the previous batch is completed, record the actual number of boards used for each material specification and query the corresponding theoretical number of boards. Calculate the ratio to obtain the actual material consumption data.

[0043] Shape scanning and size measurement are performed on the leftover material fragments generated in the previous batch. The subsequent component identifiers associated with each leftover material fragment are queried from the optimal material usage scheme. It is determined whether the actual size of the leftover material meets the requirements. If not, the subsequent component identifiers and their required plate specifications are extracted, and a mapping table from component identifiers to plate specifications is established and recorded as leftover material generation data.

[0044] Based on the actual material consumption data, the quantity of the next batch of material procurement list is incrementally calculated. The mapping table of surplus material generation data is traversed and the number of components corresponding to each board specification is extracted. The number of components is added to the corresponding board specification to generate the optimized next batch of material procurement list.

[0045] A second aspect of the present invention provides an intelligent algorithm-driven material usage optimization system, comprising:

[0046] The data acquisition unit is used to acquire structural design model data, construction process constraint data, and material physical property data of the target project.

[0047] The candidate material usage scheme generation unit is used to perform component geometry analysis and material requirement mapping on the structural design model data, establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, and generate a set of candidate material usage schemes.

[0048] The material sharing constraint unit is used to construct a material surplus reuse constraint network based on the candidate material usage scheme set, combined with the material physical property data and construction process constraint data, identify reusable surplus material segments, and encode the surplus material reuse relationship as material sharing constraints between components.

[0049] The optimal solution search unit is used to find the optimal material usage scheme that maximizes material utilization from the set of candidate material usage schemes, with the objective function of minimizing the total material procurement quantity and the constraints of component geometric parameter satisfaction and material sharing.

[0050] The feedback optimization unit is used to generate a batch material procurement list based on the optimal material usage plan, collect the actual material consumption data and surplus material generation data of the previous batch, calculate the processing loss parameters, and optimize the next batch material procurement list.

[0051] A third aspect of the present invention provides an electronic device, comprising:

[0052] processor;

[0053] Memory used to store processor-executable instructions;

[0054] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0055] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0056] This solution enables precise optimization and dynamic adjustment of material usage, significantly improving material utilization and reducing procurement costs. Through component geometric analysis and material requirement mapping, the design model is automatically transformed into multiple candidate cutting schemes, providing a rich feasible solution space for subsequent optimization. A material surplus reuse constraint network is constructed, transforming the surplus sharing relationship between different components into mathematical constraints, enabling surplus materials to be systematically identified and utilized, effectively reducing material waste.

[0057] Intelligent optimization, aimed at minimizing total procurement volume, automatically selects the globally optimal solution with the highest material utilization rate while meeting the geometric requirements of all components and constraints on reusing surplus materials. This method overcomes the shortcomings of traditional manual material cutting, which relies on experience and struggles to coordinate the overall situation. It achieves automated and intelligent decision-making from design data to the procurement list, significantly improving optimization efficiency and solution reliability.

[0058] By collecting actual consumption and surplus material data and calculating dynamic loss parameters, the system can provide feedback and optimization for subsequent batches of procurement plans. This closed-loop mechanism enables material usage prediction to continuously approach the actual demand as construction progresses, thereby reducing the risk of material shortages or surpluses caused by inaccurate forecasts and enhancing the adaptability and refinement of project material management. Attached Figure Description

[0059] Figure 1 This is a flowchart illustrating the intelligent algorithm-driven material usage optimization method according to an embodiment of the present invention.

[0060] Figure 2 This is a flowchart illustrating the method for generating a set of candidate material dosage schemes according to an embodiment of the present invention. Detailed Implementation

[0061] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0063] Figure 1 This is a flowchart illustrating the intelligent algorithm-driven material usage optimization method according to an embodiment of the present invention, as shown below. Figure 1 As shown, the intelligent algorithm-driven material usage optimization method includes:

[0064] Obtain structural design model data, construction process constraint data, and material physical property data for the target project;

[0065] Perform component geometry analysis and material requirement mapping on the structural design model data, establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, and generate a set of candidate material usage schemes;

[0066] Based on the set of candidate material usage schemes, and combined with the material physical property data and construction process constraint data, a material surplus reuse constraint network is constructed to identify reusable surplus material segments and encode the surplus material reuse relationship as material sharing constraints between components.

[0067] With minimizing the total material procurement quantity as the objective function and the constraints of component geometric parameter satisfaction and material sharing as the constraints, the optimal material usage scheme that maximizes material utilization is found in the set of candidate material usage schemes through an intelligent search algorithm.

[0068] Based on the optimal material usage plan, a batch material procurement list is generated, and the actual material consumption data and surplus material generation data of the previous batch are collected. The processing loss parameters are calculated, and the next batch material procurement list is optimized.

[0069] Figure 2 This is a flowchart illustrating a method for generating a set of candidate material usage schemes according to an embodiment of the present invention. In one optional implementation, the structural design model data undergoes component geometry analysis and material requirement mapping to establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, generating a set of candidate material usage schemes, including:

[0070] The geometric entity model of the component is parsed from the structural design model data, and the boundary contour of the component geometric entity model is extracted to obtain the contour vertex coordinate sequence and boundary curve equation of the component.

[0071] Based on the sequence of vertex coordinates of the contour, the area ratio of the minimum bounding rectangle to the maximum inscribed rectangle of the component is calculated to obtain the shape complexity coefficient of the component.

[0072] The shape complexity coefficient is correlated with the material elastic modulus and material density in the material physical property data to determine the material deformation compensation amount when the component is cut, and the coordinates of the contour vertex coordinate sequence are corrected according to the material deformation compensation amount.

[0073] Based on the corrected contour vertex coordinate sequence, the material standard plate size library is traversed and two-dimensional packing calculations are performed to generate multiple layout schemes for each component under different plate sizes.

[0074] The various plate arrangement schemes for each component are summarized, and a mapping relationship between component identifiers and plate arrangement schemes is established to form a set of candidate material usage schemes.

[0075] In this specific embodiment, when extracting component information from the BIM structural design model, the three-dimensional geometric entity model is parametrically analyzed, projecting the three-dimensional spatial shape of the component onto a two-dimensional plane to form a geometric reference for material cutting calculations. For each component geometric entity, a slicing method is used to perform cross-sectional analysis along its main stress direction to extract the boundary contour information of each cross-section. During the boundary contour extraction process, a B-spline curve fitting algorithm is used to accurately describe the component edges. The sampling density is adaptively adjusted according to the curvature changes of the component. The sampling point spacing is set to 2-5 mm in areas with large curvature, while the sampling spacing for straight segments with small curvature can be extended to 20 mm. The extracted contour vertex coordinate sequence is arranged counterclockwise to form a closed polygon vertex set. Each vertex is represented by a coordinate pair in a Cartesian coordinate system. It means that, among them The vertex index ranges from 1 to 1. , This represents the total number of vertices in the component's outline. For components containing curved boundaries, the curve equations of each boundary segment are recorded in addition to the vertex coordinate sequence. For circular arc segments, the three parameters of center coordinates, radius, and central angle are used for description. For free curve segments, the sequence of control points and the curve order are stored.

[0076] In the process of quantifying the complexity of component shapes, the minimum bounding rectangle is calculated based on the extracted sequence of contour vertex coordinates. The calculation of the minimum bounding rectangle employs a rotating caliper algorithm, traversing all possible directions of the component contour. For each direction, the projection length of the contour in that direction and its perpendicular direction is calculated, and the direction angle that minimizes the area of ​​the rectangle is found. Let the length and width of the minimum bounding rectangle be... and Then its area is The calculation of the maximum inscribed rectangle employs dynamic programming. The component outline is divided into a regular grid with a side length of 1 mm. It is determined whether each grid cell is completely within the component outline, and the largest rectangular region is found within the internal grid cells. Let the length and width of the maximum inscribed rectangle be... and Its area is Shape complexity coefficient Defined as This coefficient reflects the degree of irregularity in the shape of the component; for a standard rectangular component... Approaching 1, for irregularly shaped components Significantly greater than 1, typically between 1.2 and 3.5.

[0077] Determining the amount of material deformation compensation requires comprehensive consideration of the material's physical properties and the component's shape characteristics. The elastic modulus of the material to be cut should be retrieved from the material physical property database. (unit: GPa) and material density (Unit: kg / m³) 3 For steel plate materials, the elastic modulus is typically in the range of 190-210 GPa, and the density is approximately 7850 kg / m³. 3 For aluminum alloy sheets, the elastic modulus is in the range of 68-73 GPa, and the density is approximately 2700 kg / m³. 3 The calculation of deformation compensation takes into account the deformation caused by the material's own weight and cutting stress during the component cutting process, and the compensation coefficient... Determined in the following manner: ,in The reference elastic modulus is taken as 200 GPa. The reference density is taken as 7800 kg / m³ 3 For components with a large shape complexity coefficient, shape changes are difficult to control; therefore, the compensation coefficient is proportional to the shape complexity coefficient. After calculating the compensation coefficient, the coordinates of the contour vertices are corrected by expanding each vertex outward along the normal direction of the component contour by a distance... ,in The baseline compensation distance is set according to the component size; for components with a side length of less than 1 meter... Select components with a thickness of 0.5 mm and a side length of 1-3 meters. Select a component with a diameter of 1.0 mm and a side length exceeding 3 meters. Take 1.5 mm. During coordinate correction, for each vertex... Calculate the normal vectors of its two adjacent sides, bisect the two normal vectors at an angle, and offset the vertex coordinates along the angle bisector. The corrected vertex coordinates are... .

[0078] In the sheet metal layout generation stage, various commonly used sheet metal specifications are read from the material standard sheet metal size library. For steel sheets, these typically include specifications such as 1000×2000 mm, 1220×2440 mm, 1500×3000 mm, and 1500×6000 mm. For aluminum alloy sheets, these specifications include specifications such as 1000×2000 mm, 1220×2440 mm, and 1500×3000 mm. For each sheet metal specification, a two-dimensional bin packing algorithm is used to calculate the layout scheme of components on that specification of sheet metal. The bin packing algorithm adopts an optimization strategy based on a genetic algorithm, aiming to maximize the utilization rate of the sheet metal while satisfying the minimum spacing constraint between components. The minimum spacing between components is determined by the cutting process. For laser cutting, the spacing is usually required to be no less than 3 mm; for plasma cutting, the spacing is required to be no less than 5 mm; and for waterjet cutting, the spacing is required to be no less than 2 mm. During the packing calculation process, all components to be arranged are sorted from largest to smallest area. The lowest horizontal line algorithm is used to place each component in sequence. For each component, all placement positions on the current board are traversed, and the utilization value of the remaining board at each position is calculated. The position that makes it easiest to place subsequent components on the remaining board is selected. Four rotations are allowed when placing components: 0 degrees, 90 degrees, 180 degrees, and 270 degrees. The rotation angle that maximizes the utilization of the board is selected. For irregularly shaped components, a concave-convex matching strategy is also used to detect the concave areas formed by the placed components and prioritize filling the concave areas with the protruding parts of subsequent components to improve the utilization rate of board space.

[0079] For a single component, multiple layout schemes can be generated under different sheet metal specifications. Each scheme corresponds to different sheet metal utilization rates and material consumption. Arranging the same component alone under a specific sheet metal specification is one scheme, while combining multiple components on the same sheet metal is another. When combining components, components with complementary shapes are prioritized, as this significantly improves sheet metal utilization. For each layout scheme, information such as sheet metal specifications, component coordinates on the sheet metal, rotation angle, total sheet metal utilization rate, and material consumption is recorded. The formula for calculating sheet metal utilization rate is as follows: ,in This represents the total area of ​​all components on the board. This represents the total area of ​​the boards. Material consumption is calculated based on the number of boards. For example, if a project requires the use of 3 boards of 1220×2440 mm, the material consumption is recorded as 3 boards of that specification.

[0080] All component layout schemes generated under various sheet metal specifications are compiled and organized, establishing a mapping data structure from component identifiers to layout schemes. This mapping relationship is stored using a multi-level nested dictionary structure. The first level key is the component identifier number, the second level key is the sheet metal specification number, and the third level is a list of multiple layout schemes for that component under that sheet metal specification. Each layout scheme is stored in structured data format, including fields such as scheme number, component coordinates, rotation angle, list of associated components (if it is a combined layout), sheet metal utilization rate, and material consumption. This mapping relationship forms a set of candidate material consumption schemes, providing a rich scheme space for subsequent optimization searches, enabling the optimization algorithm to find the global optimum among a large number of candidate schemes. The size of the candidate scheme set is typically exponentially related to the number of components and sheet metal specifications. For a project containing 50 components, under 5 commonly used sheet metal specifications, the total number of candidate schemes can reach tens of thousands to hundreds of thousands.

[0081] In one optional implementation, based on the corrected sequence of contour vertex coordinates, the various sheet metal specifications in the material standard sheet metal size library are traversed and two-dimensional packing calculations are performed to generate multiple arrangement schemes for each component under different sheet metal specifications, including:

[0082] Obtain the length and width dimensions of the board from the standard board size library, establish a two-dimensional coordinate system for the board, and set the lower left corner vertex of the board as the origin of the coordinate system;

[0083] Based on the corrected contour vertex coordinate sequence, calculate the long side direction and short side direction of the minimum bounding rectangle, and generate the initial placement posture angle and the candidate placement posture angle after rotation transformation of the component.

[0084] For each candidate placement posture angle, the modified contour vertex coordinate sequence is rotated and transformed, the envelope size of the component contour after rotation and transformation is calculated, and it is determined whether the envelope size meets the size constraint conditions of the plate.

[0085] For candidate placement posture angles that meet the size constraints, the starting coordinates for placement of the component outline are determined in the two-dimensional coordinate system of the plate. An edge-fitting strategy is used to adjust the position of the component outline along the plate boundary, generating multiple candidate placement positions of the component on the plate.

[0086] For each candidate placement position, calculate the area occupied by the component outline on the board, record the occupied area and the corresponding candidate placement posture angle, and combine the board size to form multiple board layout schemes for each component.

[0087] In this specific embodiment, after obtaining the corrected contour vertex coordinate sequence, it is necessary to solve the problem of how to efficiently arrange the components on the standard-sized sheet metal. Available sheet metal specification information is extracted from a preset standard sheet metal size library, which stores common standard sheet metal sizes, such as lengths of 2440mm×1220mm and 3000mm×1500mm. With width dimension For each sheet material specification, an independent two-dimensional Cartesian coordinate system is established, with the lower left corner vertex of the sheet material as the origin of the coordinate system. The horizontal direction to the right is The positive direction of the axis, the vertical upward direction is The positive axis direction is used to represent the usable area of ​​the plate as a rectangular region. .

[0088] Obtain the corrected contour vertex coordinate sequence of the component Next, calculate the minimum bounding rectangle of the component's outline, specifically by traversing the coordinates of all vertices. Minimum value of coordinates With the maximum value ,as well as Minimum value of coordinates With the maximum value This gives the length of the circumscribed rectangle. With the shorter side dimension To fully utilize the board space, the arrangement of components at different placement angles needs to be considered. Let the initial placement angle of the components be denoted as... That is, the component outline maintains the original coordinate system orientation; at the same time, a set of candidate placement posture angles after rotation transformation is generated, which typically includes , , Discrete angles, as well as other angles dynamically generated based on the geometric features of the components. For example, for rectangular components, only two orientations, 0° and 90°, need to be considered, while for irregular polygons, more rotation angles need to be considered to find a more compact arrangement.

[0089] For each candidate placement angle The corrected contour vertex coordinate sequence is rotated and transformed to change the coordinates of each vertex. Transform into new coordinates using a rotation matrix. The transformation formula is: , After completing the rotation transformation, recalculate the envelope dimensions of the rotated component profile, that is, find the dimensions of all vertices after rotation. coordinate range and coordinate range The length of the rotated envelope is obtained. With envelope width To determine whether the envelope dimensions meet the dimensional constraints of the sheet material, a specific inspection is required. and If the condition is met, it means that the component can be placed on the plate of this specification under the current attitude angle; otherwise, the attitude angle is excluded and other angles are checked.

[0090] For candidate placement angles selected based on dimensional constraints, the starting coordinates for placing the component outline are determined in the 2D coordinate system of the sheet material. An edge-fitting strategy is used to progressively search for suitable placement positions. The core idea of ​​this strategy is to prioritize placing the component close to the left and bottom boundaries of the sheet material to reduce fragmentation of the sheet material space. In practice, the lower left corner vertex of the component outline is initially placed in the coordinate system. At this point, the coordinates of all vertices of the component's contour need to be translated to minimize the size of the rotated contour. Coordinates and minimum The coordinate point is aligned with the origin, and this initial placement position is recorded as the first candidate placement position.

[0091] After obtaining the initial placement position, other candidate placement positions are further generated along the plate boundary. In the axial direction, with a certain step size Starting point for placement of incremental components Coordinates, for example It can be set to 10mm or a percentage of the sheet length, shifting the component outline to the right each time. Check whether the right boundary of the component outline after translation exceeds the right boundary of the plate. If the position is within the specified range, the translated position is recorded as a new candidate placement position. Similarly, along... In the axial direction with step size Starting point for placement of incremental components Using coordinates, translate the component outline upwards and check if the upper boundary exceeds the upper boundary of the plate. If the position is within the specified range, the same location is recorded. This gridded search method can generate multiple discrete candidate placement position coordinates in the two-dimensional coordinate system of the board. .

[0092] For each candidate placement location, it is necessary to accurately calculate the area occupied by the component outline on the plate, and then translate the coordinate sequence of the rotated outline vertices to align the lower left corner of the component outline with the placement location. This yields the final vertex coordinate sequence of the component in the plate coordinate system. ,in , Calculate the bounding rectangle of the region enclosed by the final vertex coordinate sequence. The coordinates of the lower left corner of the bounding rectangle are... The coordinates of the upper right corner are The bounding rectangle represents the area occupied by the component on the plate. This occupied area is then compared with the corresponding candidate placement angle. Place the starting coordinates All data is recorded together to form a layout scheme data structure, which includes the field: board specifications. Posture angle Place coordinates Occupied Area .

[0093] The algorithm iterates through all sheet metal specifications in the standard sheet metal size library, repeating the two-dimensional packing calculation process for each specification to generate multiple layout schemes for each component under different sheet metal specifications. Due to size differences, different sheet metal specifications can accommodate different component orientation angles and placement positions. Larger sheet metal specifications can provide more candidate placement positions, while smaller sheet metal specifications can only accommodate components with specific orientation angles. The layout schemes generated under all sheet metal specifications are summarized to form a complete set of layout schemes for the component. This set provides basic data support for the subsequent construction of the material surplus reuse constraint network and the search for the optimal material usage scheme. By evaluating the sheet metal utilization rate, surplus size and shape, and other indicators under different layout schemes, the layout combination that meets the component's geometric requirements and maximizes material utilization can be selected.

[0094] In generating various layout schemes, it is also necessary to consider the impact of sheet metal processing technology on the layout. For example, some sheets metal have anisotropy in texture or mechanical properties in specific directions, and it is necessary to ensure that the stress direction of the components is consistent with the texture direction of the sheet metal during layout. Some cutting processes require that the edges of the components maintain a certain safety distance from the edges of the sheet metal to avoid edge chipping or deformation due to cutting stress. In this case, an additional safety margin needs to be added when calculating the occupied area. This expands the actual occupied area to Furthermore, when determining dimensional constraints, it simultaneously checks whether the expanded area exceeds the board boundary. Through this refined layout generation mechanism, the actual construction process constraints can be fully considered at the algorithm level, ensuring that the generated material usage plan is not only theoretically optimal but also feasible in actual processing.

[0095] In one optional implementation, based on the set of candidate material usage schemes, and combining the material physical property data and construction process constraint data, a material surplus reuse constraint network is constructed to identify reusable surplus material segments and encode the surplus material reuse relationship as material sharing constraints between components, including:

[0096] Based on the set of candidate material usage schemes, the plate cutting path is analyzed and the leftover material polygons are identified. The vertex coordinates and topological adjacency relationships of the leftover material polygons are extracted.

[0097] Perform a shape decomposition operation on the remaining polygon to decompose the irregular remaining material into multiple regular sub-remaining material units, and calculate the maximum inscribed rectangle size and shape regularity index of each sub-remaining material unit.

[0098] Extract the geometric contour and size requirements of the component to be processed, establish a multi-dimensional matching space between the component's geometric feature vector and the feature vector of the sub-scrap unit, perform nearest neighbor search in the multi-dimensional matching space, calculate the Euclidean distance between the feature vectors, identify the pairing relationship between the component and the sub-scrap where the Euclidean distance is less than the matching threshold, and combine the cutting accuracy loss parameter in the construction process constraint data to perform geometric feasibility verification on the pairing relationship, and screen out geometrically feasible scrap reuse pairings.

[0099] For the geometrically feasible material reuse pairing, a multi-level material reuse constraint network is constructed. The network includes three levels of nodes: plate layer, material reuse layer, and component layer. Cross-level reuse dependency edges are established. The reuse dependency edges are traversed to extract cross-level material flow paths. The plate identifier, material reuse unit identifier, and component identifier are encoded into material sharing constraints in the form of triples.

[0100] In this specific embodiment, after obtaining the set of candidate material usage schemes, it is necessary to systematically analyze the leftover material generated after cutting the board. For the board cutting path in each candidate scheme, the geometric contour of the remaining area after cutting is identified by a boundary tracking algorithm. A two-dimensional Cartesian coordinate system is established on the board plane, with the lower left corner of the board as the origin. The trajectory points of the cutting path form a closed polygon in a clockwise or counterclockwise direction. For a standard plywood with dimensions of 2440mm × 1220mm, after completing the cutting of the main components, the vertex coordinates of the leftover material polygon are obtained by scanning the intersection of the start and end points of the cutting path with the board boundary. For example, the vertex coordinates of a certain leftover material area are (320, 0), (820, 0), (820, 450), (620, 450), (620, 280), and (320, 280). These coordinate points form the geometric boundary of the leftover material in the order of connection. At the same time, the topological adjacency relationship between adjacent vertices is recorded, and an association matrix between vertex numbers and edges is established to provide a data foundation for subsequent shape analysis.

[0101] After identifying the leftover polygons, it is necessary to determine their shape regularity. For concave polygons or irregular leftovers with narrow protrusions, direct use in subsequent component processing results in low material utilization. A strategy combining convex hull decomposition and rectangle fitting is adopted to decompose the irregular leftovers into multiple regular sub-leftover units. The convex hull of the leftover polygon is calculated, and the area difference between the original polygon and the convex hull is determined. If the area difference exceeds 15% of the original polygon's area, it is identified as an irregular leftover that needs to be decomposed. By identifying the concave vertex position, the original polygon is divided into several convex polygons along the perpendicular direction between the concave vertex and the opposite side. For each convex polygon sub-region, the rotating caliper algorithm is used to calculate its maximum inscribed rectangle. This algorithm enumerates each edge of the convex polygon as an edge of the rectangle, rotates the coordinate system so that the edge is parallel to the coordinate axis, and solves for the length and width of the maximum inscribed rectangle in the new coordinate system. The maximum inscribed rectangle size of each sub-leftover unit is recorded, and the shape regularity index is calculated simultaneously. This index is defined as the ratio of the area of ​​the maximum inscribed rectangle to the actual area of ​​the sub-leftover unit. The closer the shape regularity index is to 1, the closer the shape of the scrap material is to a regular rectangle, and the more suitable it is for secondary processing of standard components.

[0102] For geometric feature extraction of components to be processed, the three-dimensional geometric model of each component is parsed from the structural design model, projected onto the main processing plane to obtain the two-dimensional contour, and the basic geometric parameters such as length, width, area, and perimeter of the component, as well as shape feature parameters such as centroid position and moment of inertia, are extracted. For rectangular components, the feature vector includes length. ,width ,area Aspect Ratio Four dimensions. For a sub-waste cell, its eigenvector includes the length of the largest inscribed rectangle. ,width ,area Shape regularity Four dimensions. To achieve multi-dimensional matching between components and scrap materials, these feature parameters are normalized to the range of 0 to 1, and the component feature vector is represented as follows: The feature vector of the scrap material is represented as The superscript z denotes the normalized value. In the four-dimensional feature space, the Euclidean distance between the component feature vector and the scrap feature vector is calculated. A matching threshold of 0.3 is set. When the Euclidean distance is less than this threshold, the component and the scrap are considered to be initially matched in terms of dimensional characteristics, forming a candidate pairing relationship.

[0103] After initial matching, geometric feasibility verification is required. Precision parameters of the cutting equipment are extracted from the construction process constraint data. For example, the kerf width of a laser cutter is 0.2mm, the positioning error is ±0.15mm, and the sheet metal clamping deformation does not exceed 0.5mm. These error sources are considered, and the cumulative cutting precision loss parameter is calculated, set as a 1.0mm machining allowance per side. For candidate paired components and scrap materials, the 2.0mm machining allowance on both sides is deducted from the maximum inscribed rectangle size of the scrap material to determine if the effective usable size of the scrap material still meets the component's size requirements. Specifically, the effective length of the scrap material must be no less than the component length plus 2.0mm, and the effective width of the scrap material must be no less than the component width plus 2.0mm, ensuring that the scrap material can still be completely processed into the target component after considering cutting losses. Simultaneously, the material physical properties of the scrap material are checked to ensure consistency with the component requirements, including the matching of parameters such as sheet metal thickness, material grade, and surface treatment state. Only pairings that pass both size and property verification are considered geometrically feasible for scrap material reuse.

[0104] Based on the selected geometrically feasible pairings, a multi-level material scrap reuse constraint network is constructed. This network adopts a three-layer topology: the top layer consists of sheet metal layer nodes, each representing a piece of original sheet metal, with attributes including sheet metal number, original dimensions, and material information. The middle layer consists of scrap layer nodes, each representing a sub-scrape material unit, with attributes including scrap material number, the sheet metal number to which it belongs, geometric coordinates, available dimensions, and shape regularity. The bottom layer consists of component layer nodes, each representing a component to be processed, with attributes including component number, size requirements, and priority. Affiliation edges are established between the sheet metal layer and the scrap layer, indicating that a scrap material originates from the cutting residue of a sheet metal. Reuse edges are established between the scrap layer and the component layer, indicating that a scrap material can be used to process a component. The weight of the reuse edge is set to the size matching degree between the scrap material and the component; the higher the matching degree, the greater the weight. In subsequent optimization, reuse paths with higher weights are prioritized.

[0105] A depth-first traversal of the constructed reuse constraint network is performed to extract complete material flow paths. Starting from the sheet metal layer nodes, the path extends along the ownership relationship edges to the scrap material layer nodes, and then along the reuse relationship edges to the component layer nodes, forming a three-level flow chain: "original sheet metal → cutting scrap material → secondary component". For each valid flow path, the sheet metal identifier, scrap material unit identifier, and component identifier are extracted and encoded into triples. For example, the triple (P001, R005, C023) indicates that after the first cutting of the sheet metal numbered P001, scrap material numbered R005 is generated, and this scrap material is designated for processing component numbered C023. This triple encoding method clarifies the source, intermediate form, and final destination of the material, constituting an explicit expression of the material sharing constraint. All valid triples are summarized to form a material sharing constraint set. This set serves as the constraint condition for subsequent optimization models, ensuring that the identified scrap material reuse relationships are strictly enforced in the global material usage optimization, avoiding the omission of reusable scrap material or the repeated purchase of equivalent materials. By constructing this systematic constraint network for the reuse of surplus materials, the entire life cycle of material circulation is tracked, providing a refined constraint management mechanism for maximizing material utilization.

[0106] In one optional implementation, the geometric feasibility of the pairing relationships is verified by combining the cutting accuracy loss parameters in the construction process constraint data, and geometrically feasible pairs for reusing surplus materials are selected, including:

[0107] The cutting accuracy loss parameters are obtained from the construction process constraint data. The cutting accuracy loss parameters include cutting positioning error, cutting path offset and edge burr compensation. The geometric feature vectors of the components and sub-residual material units in the pairing relationship are extracted.

[0108] The uncertainty range of the sub-residual material unit boundary position is calculated based on the cutting positioning error, the systematic deviation of the cutting size is calculated based on the cutting path offset, the removal amount of edge trimming is calculated based on the edge burr compensation amount, and the overall size deviation of the sub-residual material unit is calculated and the conservative usable size of the sub-residual material unit is determined by superimposing the results according to the most unfavorable combination principle.

[0109] Length and width requirements are extracted from the component geometric feature vector to form a size requirement vector. The conservative available size is constructed into a size supply vector. Vector subtraction is performed between the size supply vector and the size requirement vector to obtain the size margin vector.

[0110] Determine whether the length and width components of the dimensional margin vector are both greater than the safety margin threshold. If both are greater, the pairing relationship passes the geometric feasibility check. The pairing relationships that pass the geometric feasibility check are filtered out and output to form geometrically feasible scrap material reuse pairings.

[0111] In this specific embodiment, when screening for the reuse and pairing of leftover materials, the cutting accuracy loss parameter from the construction process constraint data needs to be introduced into the geometric feasibility verification of the pairing relationship. The cutting accuracy loss parameter is a key factor affecting the actual usable size of the leftover material, directly determining whether the leftover material can meet the geometric size requirements of the target component. The cutting accuracy loss parameter is extracted from the construction process constraint database. This parameter contains three core components: cutting positioning error, cutting path offset, and edge burr compensation. The cutting positioning error reflects the positional offset of the clamping device of the cutting equipment when fixing the material. This offset causes a spatial difference between the actual cutting starting point and the theoretical cutting starting point. The cutting path offset describes the trajectory deviation of the cutting tool during its movement due to mechanical wear, thermal deformation, or control system response lag, resulting in a systematic deviation between the actual cutting contour and the design contour. The edge burr compensation corresponds to the burr protrusions on the material cross-section after cutting, which need to be removed by grinding before subsequent use. The actual removal amount will reduce the usable size of the material.

[0112] For each pairing relationship, the geometric feature vector of the target component is extracted from the component data, including key geometric parameters such as the component's length, width, thickness, and diagonal dimensions. At the same time, the geometric feature vector of the sub-scrap unit is extracted from the scrap material library, including attributes such as the sub-scrap unit's actual length, actual width, usable area, and shape integrity.

[0113] After obtaining the geometric feature vectors, it is necessary to evaluate the uncertainty of the boundary position of the sub-residue unit. Based on the cutting positioning error, uncertainty intervals are established for the four boundaries of the sub-residue unit. Assuming the theoretical length boundary position of the sub-residue unit is... The cutting positioning error is Then the actual location of the boundary falls within the interval For rectangular scrap, each of the four boundaries has an independent positioning error. After considering the uncertainties of the four boundaries, the effective length range of the sub-scraping unit becomes the theoretical length minus twice the cutting positioning error.

[0114] Cutting path offset introduces systematic dimensional deviations. When the cutting tool moves along the designed path, due to the response delay of the control system and mechanical transmission backlash, the actual cutting path deviates from the theoretical path in a constant direction. Assuming the cutting path offset is... This offset has an effect on the two opposite boundaries of the cut, causing the actual size of the sub-residue unit to decrease relative to the theoretical size. This systematic deviation has a cumulative effect when materials are cut in batches, and it needs to be deducted during dimensional verification.

[0115] The calculation of edge burr compensation needs to consider the material type and cutting method. For the heat-affected zone generated by laser cutting, an oxide layer and tiny slag protrusions will form at the edges. For the plastic deformation zone generated by mechanical cutting, metal chips and burrs will remain at the edges. These edge defects must be removed by grinding before component assembly. The actual removal depth is denoted as... For the four edges of the leftover material, edge trimming is required. Therefore, twice the edge burr compensation amount needs to be deducted in both the length and width directions.

[0116] Based on the worst-case combination principle, the overall dimensional deviation of the sub-piece material unit is calculated by superimposing the cutting positioning error, cutting path offset, and edge burr compensation. It is assumed that the theoretical length of the sub-piece material unit is... The theoretical width is Then the conservative usable length and conservative available width The calculation method is as follows: subtract twice the cutting positioning error, twice the cutting path offset, and twice the edge burr compensation from the theoretical length, i.e. The same calculation logic is used in the width direction. This conservative estimation method ensures that even under the most unfavorable processing conditions, the sub-residue unit can still meet the dimensional requirements of the component.

[0117] Extract the length and width requirements of the component from its geometric feature vector to form a size requirement vector. Assume the minimum required length of the component is... Minimum width is The size demand vector can be represented as a two-dimensional vector containing length and width requirements. The conservative available length and conservative available width calculated above are used to construct a size supply vector, which also contains two components: length supply and width supply.

[0118] Perform vector subtraction on the size supply vector and size demand vector to obtain the size margin vector. The length component of the size margin vector equals the conservative available length minus the component length requirement, and the width component equals the conservative available width minus the component width requirement. The physical meaning of this vector is the remaining space of the sub-scrap unit relative to the component size requirement after deducting all processing errors. If a component of the size margin vector is negative, it indicates that the available size of the sub-scrap unit in that direction is insufficient to meet the component requirement.

[0119] To ensure the reliability of the pairing relationship, a safety margin threshold is introduced as a judgment criterion. The safety margin threshold is set according to the material type and component precision requirements. For high-precision components, the safety margin threshold is larger, typically 2% to 5% of the component size. For components with lower precision requirements, the safety margin threshold can be appropriately reduced. The judgment is made on whether the length component of the dimensional margin vector is greater than the safety margin threshold, and simultaneously whether the width component is greater than the safety margin threshold. Only when both components are greater than their respective safety margin thresholds is the pairing relationship considered to have passed the geometric feasibility check.

[0120] For pairings that pass the geometric feasibility check, they are marked as geometrically feasible and output to the screening result set. The screening result set contains all scrap material reuse pairs that meet the size requirements. These pairings can serve as effective material sharing solutions in subsequent material usage optimization. For pairings that fail the check, the reasons for failure are recorded, including whether the length or width is insufficient, as well as the specific size difference, to provide a reference for subsequent scrap material classification and processing.

[0121] In practical applications, geometric feasibility verification also needs to consider the anisotropic characteristics of materials. For materials with texture direction, such as wood or fiber-reinforced composite materials, the length direction of the component needs to be aligned with the main texture direction of the material. In this case, directional constraint parameters need to be added to the pairing relationship. The verification process needs to verify whether the texture direction of the sub-residual material unit is consistent with the texture direction required by the component. Only pairing relationships with matching directions and dimensional requirements can pass the final verification. Through a multi-level geometric feasibility verification mechanism, it is ensured that the selected reusable material pairings are operable in actual construction, effectively avoiding material waste and rework losses due to processing errors.

[0122] In one optional implementation, with minimizing the total material procurement quantity as the objective function and the constraints of component geometric parameter satisfaction and material sharing as conditions, an intelligent search algorithm is used to find the optimal material usage scheme that maximizes material utilization from the set of candidate material usage schemes, including:

[0123] Based on the set of candidate material usage schemes, the required plate specifications and quantities for each scheme are statistically analyzed, and a mapping table is established from candidate scheme codes to total material procurement quantity, with the total material procurement quantity as the objective function.

[0124] Based on the material sharing constraints in the form of triples, the conflict relationships between multiple components sharing the same scrap unit are identified, and the material sharing constraints are converted into mutually exclusive logical expressions between candidate solution variables. The dependency sequence between scrap units and components is extracted from the material sharing constraints, and the pre-dependency constraints between candidate solutions are established according to the order of plate cutting and the temporal relationship of scrap generation.

[0125] The initialization of the intelligent search algorithm's decoding method maps each candidate solution to a decision variable, generating an initial candidate solution population. The constraint violation degree is calculated for each candidate solution in the population, and the number of violations of mutual exclusion logic expressions and pre-dependent constraints is counted. The comprehensive evaluation value of the candidate solution is obtained by weighted summation of the objective function value and the constraint violation degree.

[0126] An elite retention strategy is adopted to iteratively evolve the candidate solution population. High-quality candidate solutions are selected based on the comprehensive evaluation value. New candidate solutions are generated through crossover and mutation operations. The search is terminated when the optimal solution of the population remains stable for several consecutive generations or reaches the maximum number of iterations. The candidate solution with the best comprehensive evaluation value is output as the optimal material usage scheme.

[0127] In this specific embodiment, after obtaining the set of candidate material usage schemes and the corresponding material sharing constraint network, it is necessary to select the optimal scheme that maximizes material utilization from the candidate scheme set. This process requires a comprehensive statistical analysis of the candidate scheme set, traversing each candidate scheme, and extracting all plate specifications required by the scheme, including attributes such as the length, width, thickness, and material type of the plates. For steel structure projects, if a candidate scheme involves three types of components: H-beams, channel steel columns, and steel plate shear walls, then the required Q345B steel plates are calculated as follows: 15 sheets of 12mm×2440mm×6000mm steel, 8 sheets of 16mm×2440mm×6000mm steel, and 20 H-beams of 400mm×200mm×8mm×13mm steel, each 12m long. By performing this type of statistical analysis on all candidate schemes, a direct mapping relationship between the candidate scheme code and the total material procurement quantity is established, providing a clear objective function definition for subsequent optimization algorithms. The total material procurement quantity includes not only the quantity of main materials, but also the processing allowance and safety reserve that must be reserved to meet the construction process requirements.

[0128] Handling material sharing constraints is the core challenge of this optimization process. When multiple components plan to reuse the same scrap unit, it is necessary to clearly identify the conflict relationships between these components. Assume that components A1, A2, and A3 can all be cut from scrap unit R5 to obtain the required material, but the actual size of scrap unit R5 can only satisfy the simultaneous use of two of these components. In this case, a mutual exclusion relationship is formed among the three components. By analyzing the triplet data in the material sharing constraint network, constraint records in the form of "component A1 - scrap unit R5 - component A2" are extracted to determine whether the available area of ​​scrap unit R5 is sufficient to simultaneously satisfy the material requirements of A1 and A2. If the area of ​​the scrap unit is 0.8 square meters, A1 needs 0.5 square meters, and A2 needs 0.4 square meters, then neither can obtain material from R5 simultaneously, and this conflict needs to be converted into a logical expression. Let decision variable x1 represent the candidate scheme of choosing A1 to use R5, and x2 represent the candidate scheme of choosing A2 to use R5. Then the mutual exclusion constraint can be expressed as x1 + x2 ≤ 1, ensuring that only one component occupies the scrap unit at any given time.

[0129] Plate cutting and scrap generation have strict temporal logic. When the cutting operation of component B1 generates scrap unit R8, and component B2 plans to use R8 as raw material, it must be ensured that the processing of B1 is performed before that of B2. This dependency relationship is extracted from the material sharing constraint network to identify the pairing relationship between scrap-generating components and scrap-using components. Assume that in a steel structure floor slab project, the cutting of main beam M1 generates scrap unit R12 with dimensions of 600mm × 1200mm. The web of secondary beam M2 can be cut from R12. At this time, a pre-dependency relationship is formed between M1 and M2. Let decision variable y1 represent choosing a specific cutting scheme for M1 to generate R12, and y2 represent choosing the scheme for M2 to use R12. Then the pre-dependency constraint is expressed as y2 ≤ y1, meaning that the dependent scheme of M2 is only feasible when the scheme of M1 is selected. By traversing the material sharing constraint network, the generation nodes and usage nodes of all scrap units are extracted, constructing a complete set of pre-dependency constraints.

[0130] The intelligent search algorithm employs a genetic algorithm framework to efficiently screen candidate solutions. Each candidate solution is considered a gene encoding unit. If the candidate solution set contains 120 different material cutting schemes, the individual's code is a 120-bit binary string, with each position representing whether the corresponding scheme is selected. During initialization, a population of 50 individuals is randomly generated, each representing a combination of candidate solution selections. For each individual in the population, constraint violation calculation is performed. All mutually exclusive logical expressions are traversed, checking for instances where conflicting schemes are selected simultaneously in the individual's code. For each conflict found, the constraint violation score is incremented by 1. Pre-dependency constraints are also checked; if an individual selects the scheme using surplus material R15 but does not select the pre-dependency scheme that generates R15, the constraint violation score increases accordingly. Let the total material purchase quantity of an individual be Q tons, and the number of constraint violations be V. The comprehensive evaluation value F = Q + λV is defined, where λ is the penalty coefficient, typically the product of the material unit price and the average purchase quantity, ensuring sufficient penalty effect for constraint violations.

[0131] The population iterative evolution process combines an elite preservation strategy with a roulette wheel selection mechanism. Before each generation begins, all individuals in the population are sorted according to their comprehensive evaluation value. The top 10% of individuals with the best evaluation value are directly selected for the next generation to avoid losing high-quality solutions during the iteration process. For the remaining individuals, roulette wheel selection is performed based on their fitness. The probability of an individual being selected is proportional to the reciprocal of its comprehensive evaluation value. Individuals with lower evaluation values ​​are more likely to be selected as parents. Selected parent individuals undergo a single-point crossover operation, randomly selecting a position in the encoding string as the crossover point and exchanging the gene segments of the two parents after that point to generate two new offspring individuals. The crossover probability is set to 0.8 to ensure sufficient diversity in the population. The mutation operation randomly flips certain bits in the individual's encoding string with a probability of 0.05, changing the originally selected scheme to unselected or activating the unselected scheme, introducing a new search direction.

[0132] Newly generated offspring individuals require recalculation of constraint violation degree and comprehensive evaluation value. For individuals after crossover mutation, new constraint conflicts may occur or existing constraint violations may be repaired. Through a fast constraint checking algorithm, for gene positions that have changed in the encoding string, only the constraint expressions related to these positions are recalculated, avoiding full duplication of calculations. If a mutation operation flips the 35th gene position, only the mutual exclusion constraints and prerequisite dependency constraints involving the 35th candidate scheme are checked, while the violation status of other irrelevant constraints remains unchanged. Incremental constraint checking significantly improves the execution efficiency of the algorithm, keeping the single-generation evolution time within an acceptable range.

[0133] The iteration termination condition is set with a dual-judgment mechanism. When the comprehensive evaluation value of the best individual in the population remains unchanged for 20 consecutive generations with a change of less than 0.1%, the algorithm is considered to have converged to the vicinity of a local optimum, and the search is terminated early. Simultaneously, the maximum number of iterations is set to 500 generations; even if the convergence condition is not met, the algorithm is forcibly stopped after 500 generations to avoid getting bogged down in a long period of ineffective searching. In practical engineering applications, for a medium-sized steel structure project containing 300 components and 80 surplus material units, the algorithm typically converges around 150 generations. At this point, the candidate solution combination corresponding to the output optimal individual code is the optimal material usage scheme. This scheme explicitly specifies the specific material cutting strategy for each component, the surplus material reuse path, and the specifications and quantity of plate procurement, providing precise decision-making basis for subsequent batch procurement and construction organization. Compared with traditional manual material preparation methods, material utilization is improved by 8% to 15%, significantly reducing material waste and engineering costs.

[0134] In one optional implementation, a batch material procurement list is generated based on the optimal material usage plan, and the actual material consumption data and surplus material generation data of the previous batch are collected. Processing loss parameters are calculated, and the next batch material procurement list is optimized, including:

[0135] Extract the material requirements list of the components from the optimal material usage scheme, divide the components into multiple sub-lists evenly according to the batch quantity of the construction plan and the construction time, and generate a batch material procurement list by statistically analyzing the plate specifications and quantities of each sub-list.

[0136] After the previous batch is completed, record the actual number of boards used for each material specification and query the corresponding theoretical number of boards. Calculate the ratio to obtain the actual material consumption data.

[0137] Shape scanning and size measurement are performed on the leftover material fragments generated in the previous batch. The subsequent component identifiers associated with each leftover material fragment are queried from the optimal material usage scheme. It is determined whether the actual size of the leftover material meets the requirements. If not, the subsequent component identifiers and their required plate specifications are extracted, and a mapping table from component identifiers to plate specifications is established and recorded as leftover material generation data.

[0138] Based on the actual material consumption data, the quantity of the next batch of material procurement list is incrementally calculated. The mapping table of surplus material generation data is traversed and the number of components corresponding to each board specification is extracted. The number of components is added to the corresponding board specification to generate the optimized next batch of material procurement list.

[0139] In this specific embodiment, after obtaining the optimal material usage plan, it needs to be converted into an executable procurement plan. All component objects are extracted from the data structure of the optimal material usage plan. Each component object contains a unique component identifier, geometric dimension parameters, required sheet metal specifications, and a scheduled processing time. The project construction plan document is read to obtain the total number of construction batches. and the planned start time for each batch. According to the predetermined processing time attribute of the components, all components are assembled. Divided into Subset subset Including processing time in the first All components within the batch time window.

[0140] For each subset Iterate through the component objects and extract the board specifications, which typically include length. ,width and thickness Three-dimensional parameters are used to establish a dictionary data structure. ternary group based on sheet material specifications The key is the required quantity of sheet metal of that specification. Traverse the subset. For each component, read its sheet metal specification attributes and store them in the dictionary. The dictionary is searched for the corresponding key. If the key exists, the value is incremented by 1; otherwise, a new key is created and initialized with a value of 1. After completing the traversal, the dictionary... That is, including the first All sheet metal specifications and corresponding quantities required for the batch. (Dictionary) Convert the data into a structured purchase order document. The document includes batch numbers, planned purchase dates, and a detailed list of sheet materials. Each row in the sheet material list includes the length, width, and height specifications of the sheet materials, as well as the quantity to be purchased. The documents are generated in batch order. Each procurement list document constitutes a complete batch of material procurement lists.

[0141] In the After the batch of construction is completed, the material consumption data collection process is initiated. The actual quantity of each board specification used in this batch is retrieved from the material management system at the construction site and recorded as follows: ,in An index indicating the sheet metal specifications. Starting from the... Find the theoretical quantity of the corresponding sheet material specifications in the batch's theoretical purchase list. Calculate the actual consumption ratio .like A value greater than 1.05 indicates that there is excessive consumption of this specification of sheet material, and redundancy needs to be added in subsequent batches; if A value less than 0.95 indicates that the theoretical calculation is conservative. The purchase quantity can be appropriately reduced in subsequent batches, and the actual consumption ratio of all sheet metal specifications can be stored as a vector. ,in This vector represents the total number of sheet metal specifications involved and serves as the actual material consumption data.

[0142] For the The leftover materials generated in the batch are processed. These leftover materials usually exist in the form of irregularly shaped fragments. A 3D scanning device is used to acquire the shape of the leftover material fragments, obtaining point cloud data of the leftover material. Boundary extraction and fitting are performed on the point cloud data to calculate the effective usable size of the leftover material, which is recorded as the length of the leftover material object. ,width and thickness From the surplus material reuse constraint network of the optimal material usage scheme, query the source component identifier and the subsequent component identifier planned for reuse for the surplus material segment. Based on the subsequent component identifier, retrieve the required sheet metal geometry for the component from the optimal material usage scheme. , as well as Perform size matching judgment and check conditions. and and Does the condition meet all the requirements? If the condition is met, it means that the leftover material can be reused as originally planned without additional procurement; if the condition is not met, it means that the actual size of the leftover material does not meet the reuse requirements and new sheet material needs to be procured for subsequent components.

[0143] For cases where dimensions are not met, extract the unique identifier for subsequent components. and the complete sheet metal specifications required for this component. Establish a mapping table for waste material generation data. This table uses a ternary key of sheet metal specifications and a list of component identifiers requiring additional procurement as values. The sheet metal specifications... As a key in the mapping table If the key already exists, then... Add to the corresponding component identifier list; if the key does not exist, create a new key and initialize the list. Traversing the first Complete the mapping table for all residual material fragments generated in the batch. The mapping table fully records the additional sheet material requirements due to the failure of scrap material reuse, serving as scrap material generation data.

[0144] For the The batch material procurement list has been optimized and adjusted. The theoretical quantity of each sheet specification is retrieved from the original procurement list. Based on actual material consumption data, the theoretical quantity is adjusted using the consumption ratio, and the adjusted quantity is calculated. ,in This indicates a rounding up operation to ensure that the purchase quantity is an integer.

[0145] To further consider the impact of surplus material generation data, the mapping table is traversed. Each key-value pair in the data has a key representing the board material specification. The value is a list of component identifiers. The length of this list is calculated to determine the number of components needed to replace the sheet metal of this specification. In the first Find the relevant information in the batch purchase list. If a matching sheet metal specification is found, increase the purchase quantity for that item. ,Right now If no matching entry is found, a new record will be added to the purchase list with the specified specifications. The quantity is After traversing all entries in the mapping table, an optimized procurement list is obtained that takes into account actual consumption deviations and compensation for leftover material failures.

[0146] Generate the optimized first This is a batch material procurement list document. Based on the original list, the quantity fields for each sheet material specification have been updated, and new sheet material specification entries not included in the original theoretical plan have been added. The list also includes explanations of the optimization basis, including the consumption ratio data referenced. In addition, statistics on supplementary needs caused by the failure of surplus materials will be compiled, and the optimized procurement list will be submitted to the procurement department for execution. At the same time, the parameter data generated during the optimization process will be stored in the project database for continuous optimization reference in subsequent batches.

[0147] Through the aforementioned batch processing and closed-loop optimization mechanism, dynamic adaptation between material procurement plans and actual construction conditions is achieved, avoiding the material backlog or shortage problem caused by the disconnect between theoretical plans and actual needs in traditional methods, and significantly improving the accuracy and cost-effectiveness of material procurement.

[0148] A second aspect of the present invention provides an intelligent algorithm-driven material usage optimization system, comprising:

[0149] The data acquisition unit is used to acquire structural design model data, construction process constraint data, and material physical property data of the target project.

[0150] The candidate material usage scheme generation unit is used to perform component geometry analysis and material requirement mapping on the structural design model data, establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, and generate a set of candidate material usage schemes.

[0151] The material sharing constraint unit is used to construct a material surplus reuse constraint network based on the candidate material usage scheme set, combined with the material physical property data and construction process constraint data, identify reusable surplus material segments, and encode the surplus material reuse relationship as material sharing constraints between components.

[0152] The optimal solution search unit is used to find the optimal material usage scheme that maximizes material utilization from the set of candidate material usage schemes, with the objective function of minimizing the total material procurement quantity and the constraints of component geometric parameter satisfaction and material sharing.

[0153] The feedback optimization unit is used to generate a batch material procurement list based on the optimal material usage plan, collect the actual material consumption data and surplus material generation data of the previous batch, calculate the processing loss parameters, and optimize the next batch material procurement list.

[0154] A third aspect of the present invention provides an electronic device, comprising:

[0155] processor;

[0156] Memory used to store processor-executable instructions;

[0157] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0158] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0159] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing material usage driven by intelligent algorithms, characterized in that, include: Obtain structural design model data, construction process constraint data, and material physical property data for the target project; Perform component geometry analysis and material requirement mapping on the structural design model data, establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, and generate a set of candidate material usage schemes; Based on the set of candidate material usage schemes, and combined with the material physical property data and construction process constraint data, a material surplus reuse constraint network is constructed to identify reusable surplus material segments and encode the surplus material reuse relationship as material sharing constraints between components. With minimizing the total material procurement quantity as the objective function and the constraints of component geometric parameter satisfaction and material sharing as the constraints, the optimal material usage scheme that maximizes material utilization is found in the set of candidate material usage schemes through an intelligent search algorithm. Based on the optimal material usage plan, a batch material procurement list is generated, and the actual material consumption data and surplus material generation data of the previous batch are collected. The processing loss parameters are calculated, and the next batch material procurement list is optimized.

2. The method according to claim 1, characterized in that, The structural design model data is subjected to component geometry analysis and material requirement mapping to establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, generating a set of candidate material usage schemes, including: The geometric entity model of the component is parsed from the structural design model data, and the boundary contour of the component geometric entity model is extracted to obtain the contour vertex coordinate sequence and boundary curve equation of the component. Based on the sequence of vertex coordinates of the contour, the area ratio of the minimum bounding rectangle to the maximum inscribed rectangle of the component is calculated to obtain the shape complexity coefficient of the component. The shape complexity coefficient is correlated with the material elastic modulus and material density in the material physical property data to determine the material deformation compensation amount when the component is cut, and the coordinates of the contour vertex coordinate sequence are corrected according to the material deformation compensation amount. Based on the corrected contour vertex coordinate sequence, the material standard plate size library is traversed and two-dimensional packing calculations are performed to generate multiple layout schemes for each component under different plate sizes. The various plate arrangement schemes for each component are summarized, and a mapping relationship between component identifiers and plate arrangement schemes is established to form a set of candidate material usage schemes.

3. The method according to claim 2, characterized in that, Based on the corrected contour vertex coordinate sequence, the system traverses all sheet metal specifications in the material standard sheet metal size library and performs two-dimensional packing calculations to generate multiple layout schemes for each component under different sheet metal specifications, including: Obtain the length and width dimensions of the board from the standard board size library, establish a two-dimensional coordinate system for the board, and set the lower left corner vertex of the board as the origin of the coordinate system; Based on the corrected contour vertex coordinate sequence, calculate the long side direction and short side direction of the minimum bounding rectangle, and generate the initial placement posture angle and the candidate placement posture angle after rotation transformation of the component. For each candidate placement posture angle, the modified contour vertex coordinate sequence is rotated and transformed, the envelope size of the component contour after rotation and transformation is calculated, and it is determined whether the envelope size meets the size constraint conditions of the plate. For candidate placement posture angles that meet the size constraints, the starting coordinates for placement of the component outline are determined in the two-dimensional coordinate system of the plate. An edge-fitting strategy is used to adjust the position of the component outline along the plate boundary, generating multiple candidate placement positions of the component on the plate. For each candidate placement position, calculate the area occupied by the component outline on the board, record the occupied area and the corresponding candidate placement posture angle, and combine the board size to form multiple board layout schemes for each component.

4. The method according to claim 1, characterized in that, Based on the set of candidate material usage schemes, and combined with the material physical property data and construction process constraint data, a material surplus reuse constraint network is constructed to identify reusable surplus material segments and encode the surplus material reuse relationships as material sharing constraints between components, including: Based on the set of candidate material usage schemes, the plate cutting path is analyzed and the leftover material polygons are identified. The vertex coordinates and topological adjacency relationships of the leftover material polygons are extracted. Perform a shape decomposition operation on the remaining polygon to decompose the irregular remaining material into multiple regular sub-remaining material units, and calculate the maximum inscribed rectangle size and shape regularity index of each sub-remaining material unit. Extract the geometric contour and size requirements of the component to be processed, establish a multi-dimensional matching space between the component's geometric feature vector and the feature vector of the sub-scrap unit, perform nearest neighbor search in the multi-dimensional matching space, calculate the Euclidean distance between the feature vectors, identify the pairing relationship between the component and the sub-scrap where the Euclidean distance is less than the matching threshold, and combine the cutting accuracy loss parameter in the construction process constraint data to perform geometric feasibility verification on the pairing relationship, and screen out geometrically feasible scrap reuse pairings. For the geometrically feasible material reuse pairing, a multi-level material reuse constraint network is constructed. The network includes three levels of nodes: plate layer, material reuse layer, and component layer. Cross-level reuse dependency edges are established. The reuse dependency edges are traversed to extract cross-level material flow paths. The plate identifier, material reuse unit identifier, and component identifier are encoded into material sharing constraints in the form of triples.

5. The method according to claim 4, characterized in that, Based on the cutting accuracy loss parameters in the construction process constraint data, the geometric feasibility of the pairing relationships is verified, and geometrically feasible leftover material reuse pairings are selected, including: The cutting accuracy loss parameters are obtained from the construction process constraint data. The cutting accuracy loss parameters include cutting positioning error, cutting path offset and edge burr compensation. The geometric feature vectors of the components and sub-residual material units in the pairing relationship are extracted. The uncertainty range of the sub-residual material unit boundary position is calculated based on the cutting positioning error, the systematic deviation of the cutting size is calculated based on the cutting path offset, the removal amount of edge trimming is calculated based on the edge burr compensation amount, and the overall size deviation of the sub-residual material unit is calculated and the conservative usable size of the sub-residual material unit is determined by superimposing the results according to the most unfavorable combination principle. Length and width requirements are extracted from the component geometric feature vector to form a size requirement vector. The conservative available size is constructed into a size supply vector. Vector subtraction is performed between the size supply vector and the size requirement vector to obtain the size margin vector. Determine whether the length and width components of the dimensional margin vector are both greater than the safety margin threshold. If both are greater, the pairing relationship passes the geometric feasibility check. The pairing relationships that pass the geometric feasibility check are filtered out and output to form geometrically feasible scrap material reuse pairings.

6. The method according to claim 1, characterized in that, Using minimizing the total material procurement quantity as the objective function and constraining the satisfaction of component geometric parameters and material sharing, an intelligent search algorithm is used to find the optimal material usage scheme that maximizes material utilization from the set of candidate material usage schemes, including: Based on the set of candidate material usage schemes, the required plate specifications and quantities for each scheme are statistically analyzed, and a mapping table is established from candidate scheme codes to total material procurement quantity, with the total material procurement quantity as the objective function. Based on the material sharing constraints in the form of triplets, the conflict relationships between multiple components sharing the same surplus material unit are identified, and the material sharing constraints are converted into mutually exclusive logical expressions between candidate solution variables. Extract the dependency sequence between scrap units and components from the material sharing constraints, and establish the pre-dependency constraints between candidate solutions based on the order of plate cutting and the temporal relationship of scrap generation. The initialization of the intelligent search algorithm's decoding method maps each candidate solution to a decision variable, generating an initial candidate solution population. The constraint violation degree is calculated for each candidate solution in the population, and the number of violations of mutual exclusion logic expressions and pre-dependent constraints is counted. The comprehensive evaluation value of the candidate solution is obtained by weighted summation of the objective function value and the constraint violation degree. An elite retention strategy is adopted to iteratively evolve the candidate solution population. High-quality candidate solutions are selected based on the comprehensive evaluation value. New candidate solutions are generated through crossover and mutation operations. The search is terminated when the optimal solution of the population remains stable for several consecutive generations or reaches the maximum number of iterations. The candidate solution with the best comprehensive evaluation value is output as the optimal material usage scheme.

7. The method according to claim 1, characterized in that, Based on the optimal material usage plan, a batch material procurement list is generated, and the actual material consumption data and surplus material generation data of the previous batch are collected. Processing loss parameters are calculated, and the next batch material procurement list is optimized, including: Extract the material requirements list of the components from the optimal material usage scheme, divide the components into multiple sub-lists evenly according to the batch quantity of the construction plan and the construction time, and generate a batch material procurement list by statistically analyzing the plate specifications and quantities of each sub-list. After the previous batch is completed, record the actual number of boards used for each material specification and query the corresponding theoretical number of boards. Calculate the ratio to obtain the actual material consumption data. Shape scanning and size measurement are performed on the leftover material fragments generated in the previous batch. The subsequent component identifiers associated with each leftover material fragment are queried from the optimal material usage scheme. It is determined whether the actual size of the leftover material meets the requirements. If not, the subsequent component identifiers and their required plate specifications are extracted, and a mapping table from component identifiers to plate specifications is established and recorded as leftover material generation data. Based on the actual material consumption data, the quantity of the next batch of material procurement list is incrementally calculated. The mapping table of surplus material generation data is traversed and the number of components corresponding to each board specification is extracted. The number of components is added to the corresponding board specification to generate the optimized next batch of material procurement list.

8. A smart algorithm-driven material usage optimization system, used to implement the method as described in any one of claims 1-7, characterized in that, include: The data acquisition unit is used to acquire structural design model data, construction process constraint data, and material physical property data of the target project. The candidate material usage scheme generation unit is used to perform component geometry analysis and material requirement mapping on the structural design model data, establish a many-to-many mapping relationship between component geometric parameters and material cutting schemes, and generate a set of candidate material usage schemes. The material sharing constraint unit is used to construct a material surplus reuse constraint network based on the candidate material usage scheme set, combined with the material physical property data and construction process constraint data, identify reusable surplus material segments, and encode the surplus material reuse relationship as material sharing constraints between components. The optimal solution search unit is used to find the optimal material usage scheme that maximizes material utilization from the set of candidate material usage schemes, with the objective function of minimizing the total material procurement quantity and the constraints of component geometric parameter satisfaction and material sharing. The feedback optimization unit is used to generate a batch material procurement list based on the optimal material usage plan, collect the actual material consumption data and surplus material generation data of the previous batch, calculate the processing loss parameters, and optimize the next batch material procurement list.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images