A user demand-oriented supplier fabric intelligent screening method and system

Abstract

The application discloses a user demand-oriented supplier fabric intelligent screening method and system, relates to the technical field of data intelligent screening, and comprises the following steps: acquiring the calibration girth parameters and corresponding calibration single-piece material meters of a target style and constructing a girth adjacent reference system; identifying the neighborhood landing points of candidate fabric girth parameters based on the girth adjacent reference system and constructing a girth offset identification set according to the identification result; identifying the material fluctuation range of the calibration single-piece material meters based on the girth offset identification set, obtaining a material correction adjacent domain set, and performing the same-domain expansion screening on the material correction adjacent domain set to obtain a candidate fabric domain set containing a correction state marker, generate a differentiated fabric screening output instruction, and distribute the output instruction to a user interaction interface, thereby solving the problems that the single-piece material fluctuation caused by girth changes cannot be automatically quantified, corrected and classified screened in the fabric screening stage, and the replacement fabric identification is insufficient and the recommended result is omitted.

Description

Technical Field

[0001] This invention relates to the field of intelligent data screening technology, and more specifically, to a method and system for intelligent screening of supplier fabrics based on user needs. Background Technology

[0002] In the current garment production and supply chain management process, fabric selection is usually based on the width specifications provided by the supplier and the number of meters of material used per piece obtained from historical sampling or experience statistics. By simply comparing the width specifications of different fabrics with the standard material usage of the target style, a preliminary screening and recommendation of fabric compatibility is carried out, thereby determining the range of candidate fabrics.

[0003] However, in the fabric selection stage, the lack of structured modeling of the coupling effect between the constraints of the pattern layout process and the adjacency relationship of fabric width makes it impossible to automatically and continuously quantify the fluctuations in material usage per piece when the fabric width changes. As a result, the differences in material usage between different fabric width specifications remain at the level of empirical estimation or discrete comparison, making it difficult to form correction results with calculable boundaries. At the same time, the lack of a hierarchical screening and mapping mechanism based on the degree of fabric width offset makes it impossible to effectively distinguish and rank the compatibility relationships of candidate fabrics in different fabric width ranges. This makes it easy to miss non-standard fabric widths with potential substitution value during the screening process, which limits the coverage and matching accuracy of the final recommendation results, affecting the completeness of fabric substitution selection and the reliability of decision-making. Summary of the Invention

[0004] To overcome the aforementioned shortcomings of existing technologies, a supplier fabric intelligent screening method and system oriented towards user needs is provided. By constructing a width adjacency reference system based on the constraints of the cutting piece layout process, and combining width offset indicators to perform hierarchical quantitative correction and same-domain expansion screening of the fluctuation range of material usage per piece, this method solves the problem that the fabric screening stage cannot automatically quantify and hierarchically screen the fluctuation of material usage per piece caused by width changes, resulting in insufficient identification of alternative fabrics and omission of recommendation results.

[0005] To achieve the above objectives, the present invention provides the following technical solution: Firstly, a method for intelligent supplier fabric selection based on user needs is provided, comprising the following steps: obtaining the calibrated width parameters and corresponding calibrated single-piece material usage meters of the target style, and constructing a width adjacency reference system based on preset pattern layout process constraints; identifying the neighborhood landing points of candidate fabric width parameters based on the width adjacency reference system, and constructing a width offset identifier set based on the identification results; identifying the material usage fluctuation range of the calibrated single-piece material usage meters in the width adjacency reference system based on the width offset identifier set, and obtaining a material usage correction adjacency domain set; performing same-domain expansion screening on the material usage correction adjacency domain set in a preset supplier fabric database to obtain a candidate fabric domain set containing correction status markers; generating differentiated fabric selection output instructions based on the candidate fabric domain set containing correction status markers, and distributing the output instructions to the user interface.

[0006] In a preferred embodiment, the step of obtaining the calibration width parameters and corresponding calibration single-piece material usage in meters for the target style, and constructing a width adjacency reference system based on preset pattern layout process constraints, specifically involves: obtaining the style identifier of the target style uploaded by the user, and retrieving the corresponding pattern layout diagram from a preset process database based on the style identifier; parsing the geometric contours of all patterns in the pattern layout diagram, generating a pattern set, and obtaining the width direction dimension and thread direction constraint of each pattern in the pattern set; grouping the patterns according to the width direction dimension and thread direction constraint of each pattern to obtain several pattern groups; for each pattern group, extracting the largest width direction dimension as the calibration width constraint value of the pattern group, and generating calibration width parameters based on the calibration width constraint values ​​of all pattern groups, while simultaneously extracting the calibration single-piece material usage in meters for the style from the preset process database.

[0007] In a preferred embodiment, the construction of the width adjacency reference system based on the preset fabric layout process constraints specifically involves: using the calibrated width parameters as the origin, generating several adjacent width nodes on both sides of the origin with a preset width step size; obtaining the minimum spacing constraint and rotation constraint of the fabric pieces in the preset fabric layout process constraints, calculating the fabric layout density that can accommodate a group of fabric pieces for each adjacent width node, and obtaining a density mapping set; according to the density mapping set, marking the adjacent width nodes whose fabric layout density exceeds a preset density threshold as valid adjacent nodes; sorting all valid adjacent nodes according to the width value and connecting adjacent valid adjacent nodes to obtain the width adjacency reference system.

[0008] In a preferred embodiment, the step of identifying the neighborhood landing points of candidate fabric width parameters based on the width adjacency reference system and constructing a width offset identifier set based on the identification results specifically involves: extracting the fabric width parameters of all candidate fabrics from a preset supplier fabric database, and comparing the values ​​of each fabric width parameter with the effective adjacent nodes in the width adjacency reference system; selecting the fabric width parameter whose value is equal to any effective adjacent node as the benchmark width parameter, and the rest as non-benchmark width parameters; for each non-benchmark width parameter, calculating the value of its value with two adjacent effective adjacent nodes in the width adjacency reference system. The system calculates the numerical difference and identifies the weighted adjacent nodes of non-benchmark width parameters based on the numerical difference. It generates an offset direction identifier based on the direction of the difference between the width value of the weighted adjacent node and the corresponding non-benchmark width parameter, and generates an offset magnitude identifier based on the ratio of the absolute value of the difference to the width step size. It marks the offset direction identifiers corresponding to all benchmark width parameters as zero offset identifiers and combines and encodes the offset direction identifiers and offset magnitude identifiers corresponding to all non-benchmark width parameters to obtain the width offset identifier for each candidate fabric. Finally, it associates and stores the fabric identifiers of all candidate fabrics with their corresponding width offset identifiers to obtain a width offset identifier set.

[0009] In a preferred embodiment, the step of calculating the numerical difference between each non-reference width parameter and two adjacent valid adjacent nodes in the width adjacency reference system, and identifying the weighted adjacent nodes of the non-reference width parameter based on the numerical difference, specifically involves: identifying the numerical range of the non-reference width parameter in the width adjacency reference system, and determining the left and right adjacent nodes of the numerical range; calculating the first numerical difference between the non-reference width parameter and the left adjacent node, and calculating the second numerical difference between the right adjacent node and the non-reference width parameter; obtaining the discharge density corresponding to the left and right adjacent nodes in the width adjacency reference system, and calculating the ratio of the discharge density of the left adjacent node to the discharge density of the right adjacent node; correcting the first and second numerical differences based on the ratio of the discharge density of the left adjacent node to the discharge density of the right adjacent node; and identifying the weighted adjacent nodes of the non-reference width parameter based on the correction results.

[0010] In a preferred embodiment, the step of identifying and calibrating the material usage range of a single piece in the adjacent reference system based on the width offset identifier set to obtain a material usage correction adjacency domain set specifically involves: extracting all candidate fabrics with non-zero offset level identifiers from the width offset identifier set, and grouping the candidate fabrics into a first offset group and a second offset group according to their offset direction identifier and offset level identifier; in the first offset group, sequentially accumulating the preset material usage increment value corresponding to each offset level according to the numerical value of the offset level identifier to obtain a first material usage fluctuation sequence; in the second offset group, according to the offset... The numerical value of the displacement level identifier is successively reduced by the preset material reduction value corresponding to each offset level to obtain the second material fluctuation sequence; the maximum value in the first material fluctuation sequence is added to the calibrated material usage in meters to obtain the upper limit of material fluctuation, and the minimum value in the second material fluctuation sequence is added to the calibrated material usage in meters to obtain the lower limit of material fluctuation; starting from the lower limit of material fluctuation and ending from the upper limit of material fluctuation, several continuous material usage intervals are divided according to the preset material usage step size, and each material usage interval is used as a material correction adjacency domain. All material usage intervals are summarized to obtain the material correction adjacency domain set.

[0011] In a preferred embodiment, the step of performing same-domain expansion screening on the material usage correction adjacency domain set in a preset supplier fabric database to obtain a candidate fabric domain set containing correction status markers specifically involves: extracting the actual single-piece material usage meters of each candidate fabric from the supplier fabric database, and determining the inclusion relationship between the actual single-piece material usage meters and each material usage interval in the material usage correction adjacency domain set; identifying candidate fabrics within each material usage interval based on the determination result and obtaining the number of candidate fabrics within each material usage interval; if the number of candidate fabrics within the interval is less than a preset first threshold, merging the candidate fabrics within the adjacent material usage intervals of the current material usage interval into the current material usage interval to obtain an expanded candidate fabric set; and generating an expansion identifier for each candidate fabric in the expanded candidate fabric set corresponding to each material usage interval based on the distance between its original material usage interval and the current material usage interval to obtain a candidate fabric domain set containing correction status markers.

[0012] In a preferred embodiment, the step of generating differentiated fabric screening output instructions based on a candidate fabric domain set containing correction status markers and distributing the output instructions to the user interface specifically involves: traversing the candidate fabric domain set containing correction status markers, extracting the expansion identifier from the correction status marker of each candidate fabric, and assigning candidate fabrics with an expansion identifier of zero to a first matching group, and assigning candidate fabrics with a non-zero expansion level identifier to a second matching group; generating a first type of output instruction for the candidate fabrics in the first matching group based on their corresponding material range identifier; sorting the candidate fabrics in the second matching group according to the value of the expansion identifier from smallest to largest, and generating a second type of output instruction in sequence according to the sorting result; and distributing the first type of output instruction and the second type of output instruction to the user interface in sequence.

[0013] Secondly, a supplier fabric intelligent screening system oriented towards user needs is provided, comprising the following modules: a reference construction module, used to obtain the calibrated width parameters of the target style and the corresponding calibrated single-piece material usage in meters, and to construct a width adjacency reference system based on preset pattern layout process constraints; a landing point identification module, used to identify the neighborhood landing points of the candidate fabric width parameters based on the width adjacency reference system, and to construct a width offset identifier set based on the identification results; a material usage correction module, used to identify the range of material usage fluctuation in the calibrated single-piece material usage in meters in the width adjacency reference system based on the width offset identifier set, and to obtain a material usage correction adjacency domain set; an expansion screening module, used to perform same-domain expansion screening on the material usage correction adjacency domain set in a preset supplier fabric database, and to obtain a candidate fabric domain set containing correction status markers; and a fabric screening module, used to generate differentiated fabric screening output instructions based on the candidate fabric domain set containing correction status markers, and to distribute the output instructions to the user interface.

[0014] This invention has at least the following technical effects and advantages: This invention obtains the calibrated width parameters and corresponding calibrated single-piece material usage meters of the target style, and constructs a width adjacency reference system based on preset pattern layout process constraints. This transforms the width variation relationship from discrete empirical judgment to a structured reference basis supported by process constraints. By identifying the neighborhood landing points of candidate fabric width parameters based on the width adjacency reference system and constructing a width offset identifier set, the offset state of different candidate fabrics in the width space can be quantitatively expressed. Furthermore, by identifying the material usage fluctuation range of the calibrated single-piece material usage meters in the width adjacency reference system based on the width offset identifier set, a material usage correction adjacency domain set is formed, thereby achieving interval-based modeling and connection of the influence of width variation on material usage. The process involves further mapping; by expanding and filtering the adjacent domain set of material usage correction in the supplier fabric database, a candidate fabric domain set containing correction status markers is obtained, allowing potential alternative fabrics under non-standard width conditions to be included in the screening scope; and by generating differentiated fabric screening output instructions based on this candidate fabric domain set and distributing them to the user interface, the hierarchical expression and orderly output of fabric recommendation results are realized. This solves the problem that the fabric screening stage cannot automatically quantify and correct and classify the material usage fluctuations of individual pieces caused by width changes, resulting in insufficient identification of alternative fabrics and omissions in recommendation results, thus improving the coverage completeness of fabric screening, the accuracy of material usage assessment, and the reliability of alternative recommendations. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a user-demand-oriented intelligent supplier fabric selection method according to the present invention.

[0016] Figure 2 This is a schematic diagram of the structure of a supplier fabric intelligent screening system oriented towards user needs according to the present invention. Detailed Implementation

[0017] 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.

[0018] Example 1, Figure 1 This invention presents a smart supplier fabric selection method tailored to user needs, comprising the following steps: S1, obtain the standard width parameters of the target style and the corresponding standard single piece material usage in meters, and construct the width adjacency reference system based on the preset cutting piece layout process constraints; In this embodiment, the specified width parameters of the target style and the corresponding specified material usage per piece in meters are obtained, and a width adjacency reference system is constructed based on the preset pattern layout process constraints, specifically: Obtain the style identifier of the target style uploaded by the user, and retrieve the corresponding cutting pattern from the preset process database based on the style identifier; Analyze the geometric contours of all cut pieces in the cut piece layout diagram, generate a cut piece set, and obtain the width dimension and thread direction constraint of each cut piece in the cut piece set; The cut pieces are grouped according to the width and grain direction constraints of each piece to obtain several cut piece groups; For each cut piece group, the largest width dimension is extracted as the calibration width constraint value of the cut piece group, and calibration width parameters are generated based on the calibration width constraint values ​​of all cut piece groups. At the same time, the calibration single piece material meterage corresponding to the style is extracted from the preset process database.

[0019] It should be noted that the style identifier of the target garment uploaded by the user is a structured identification information used to uniquely identify the target garment style within the process system. This identification information is used not only to distinguish different product categories, but also to distinguish different pattern structures and process versions of the same type of product, thereby ensuring the uniqueness and consistency of subsequent process data retrieval. For example, when a user uploads a shirt style, its style identifier can be "SH-2026-A01", where "SH" indicates the garment category and "2026-A01" represents the specific pattern structure and process version, making it a unique index in the process data system.

[0020] Furthermore, the pre-built process database refers to a dataset pre-constructed to store production process data for different styles. It includes at least pattern layout data, pattern structure parameter data, process constraint relationship data between pattern pieces, and the standard material usage per piece for each style. This database uses style identifiers as its index, enabling a traceable correspondence between different styles in terms of pattern layout and material usage standards.

[0021] After obtaining the style identifier uploaded by the user, the identifier is used as the sole retrieval criterion to match and locate the corresponding process record unit in the preset process database. Once a match is successful, the corresponding cut pattern data is extracted from that process record unit.

[0022] Among them, the cut piece layout data is used to characterize the spatial arrangement and relative positional relationship of each cut piece in the fabric width direction under standard process conditions. For example, when the style is identified as "SH-2026-A01", the layout structure formed under standard production conditions can be extracted by searching and locating the corresponding process record unit, and then used for subsequent cut piece analysis processing.

[0023] Furthermore, after obtaining the cut piece layout diagram, it undergoes structured graphic analysis to identify the boundary contour information of each independent cut piece in the layout diagram. This contour information is formed by identifying the closure of continuous boundary lines and extracting boundary nodes, thereby obtaining the complete geometric shape of each cut piece. After completing the contour analysis, all cut pieces are treated as independent objects and uniformly summarized to form a cut piece set for subsequent process attribute calculations. Based on this, the width dimension and fiber direction constraint information of each cut piece in the cut piece set are extracted. The width dimension is used to characterize the maximum unfolded length of the cut piece in the transverse direction of the fabric width, while the fiber direction constraint is used to characterize the fixed requirements of the fiber direction of the cut piece during the layout process, such as non-rotation, allowing 90-degree rotation, or limiting unidirectional layout, thereby forming a cut piece-level process attribute description.

[0024] After obtaining the width and grain direction constraints of each piece, the pieces are grouped based on the consistency of their processing techniques and structural compatibility. Specifically, firstly, using the grain direction constraint as the primary constraint, pieces with the same or equivalent grain restrictions are grouped into the same candidate category to ensure consistency in the layout direction. Subsequently, within the same candidate category, a second grouping is performed based on the distribution characteristics of the width dimensions, so that pieces with similar dimensions and similar layout space requirements are grouped into the same piece group. Through this grouping method, each piece group simultaneously satisfies both directional constraint consistency and dimensional structural compatibility, thereby forming a piece set structure with unified layout constraint characteristics.

[0025] Furthermore, after grouping the cut pieces, statistical processing is performed on the cut pieces within each group. The dimensional data of all cut pieces in the width direction within that group are extracted, and the maximum value is selected as the standard width constraint value for that group. This value characterizes the maximum required fabric width boundary during the layout process. For example, if the cut piece widths in a certain group are 46cm, 48cm, and 44cm, then 48cm is used as the standard width constraint value for that group to ensure that layout conflicts do not occur due to excessively wide cut pieces in certain areas during the layout process.

[0026] Finally, the calibrated width constraint values ​​corresponding to all cut pieces are summarized to form a set of calibrated width parameters for the target style, reflecting the width requirement distribution characteristics of the style at the overall structural level. Simultaneously, based on the style identifier, the standard single-piece material usage in meters is extracted from a pre-set process database. This material usage in meters is determined based on process calculations under standard layout conditions; for example, the standard single-piece material usage for a certain shirt is 1.65 meters. Through the above processing, a correspondence is established between the width constraint structure and the material usage benchmark data within the same process system. This provides a basic constraint basis for subsequent material usage correction calculations based on width offsets and ensures that material usage changes under different width conditions are calculable and traceable.

[0027] In this embodiment, a width adjacency reference system is constructed based on preset fabric layout process constraints, specifically as follows: Using the calibrated width parameters as the origin, several adjacent width nodes are generated on both sides of the origin with a preset width step size. Obtain the minimum spacing constraint and rotation constraint of the cut pieces in the preset cut piece layout process constraints, calculate the layout density that can accommodate the cut piece group for each adjacent width node, and obtain the density mapping set; Based on the density mapping set, adjacent span nodes whose discharge density exceeds the preset density threshold are marked as valid adjacent nodes; Sort all valid adjacent nodes according to their width values ​​and connect adjacent valid adjacent nodes to obtain the width adjacency reference system.

[0028] It should be noted that defining the calibration width parameter as the width reference origin means using the calibration width value of the target style as a baseline. Based on this baseline width value, the width is expanded at equal intervals to both sides through a preset width step, thereby forming multiple candidate width nodes with a sequential relationship. For example, when the calibration width parameter is 150cm and the preset width step is 5cm, nodes of 145cm, 140cm, etc. are generated sequentially in the smaller direction from 150cm, and nodes of 155cm, 160cm, etc. are generated sequentially in the larger direction, thus forming a discrete set of width nodes around the calibration width.

[0029] Among them, the minimum spacing constraint of cut pieces refers to the minimum safe distance that must be maintained between adjacent cut pieces during the layout process. This distance is used to avoid structural overlap caused by cutting errors, fabric offset and tool path interference. For example, in actual layout, it may be required that at least a spacing area of ​​0.5cm to 1cm be maintained between cut pieces.

[0030] In addition, the rotation constraint of the cut pieces refers to the process restrictions that allow or prohibit the rotation of the cut pieces within the fabric width. For example, some cut pieces with requirements for the grain direction can only be arranged in a fixed direction and are not allowed to rotate; while some cut pieces with lower requirements for texture are allowed to be arranged by rotation within a 90-degree range. This constraint directly affects the feasibility of arranging the cut pieces under different width conditions.

[0031] After obtaining each adjacent width node, the virtual arrangement for each width node is not a simple position superposition. Instead, it is based on the piece set and piece group, combined with the width dimension and grain direction constraints of the pieces, to perform regularized arrangement of the pieces under the corresponding width conditions: First, the width value corresponding to the current width node is used as the horizontal arrangement boundary, and a continuous arrangement space is constructed along the width direction; then, according to the division order of the piece groups, pieces within the same piece group are selected for arrangement first. During the arrangement process, the placement direction of each piece is constrained and determined. For pieces with a fixed grain direction, only those pieces with a fixed grain direction are allowed to be placed in a specific direction. The fabric pieces are placed in a preset orientation. For pieces that can be rotated, the placement method that minimizes lateral space occupancy is selected while satisfying rotation constraints. During actual arrangement, the pieces are placed sequentially from left to right and top to bottom, with a minimum spacing constraint maintained between adjacent pieces. When the current row cannot accommodate a new piece, the process moves to the next row until the current width limit prevents the complete arrangement of the piece group. During this process, the number of successfully arranged pieces, the number of layers, and the remaining empty space are recorded to characterize the compactness of the arrangement at that width node. For example, when the width node is 150cm, three sets of piece structures can be arranged completely with minimal remaining space, while when the width node is 145cm, only two sets of piece structures can be arranged completely with significant corner waste. This method quantifies the utilization of the arrangement at different width nodes, forming a description of the fabric density corresponding to each width node. Subsequently, all width nodes and their corresponding material discharge density results are summarized and recorded to form a density mapping set, which is used to reflect the correspondence between width changes and material discharge utilization efficiency.

[0032] Furthermore, after forming the density mapping set, the consistency of the fabric layout density corresponding to each adjacent width node is determined, and nodes with a fabric layout density higher than a preset density threshold are identified as valid adjacent nodes. This preset density threshold is used to characterize the minimum space utilization efficiency requirement of the cut pieces under reasonable layout conditions. For example, if a width node can accommodate cut pieces but there is too much waste of empty fabric, its corresponding fabric layout density is lower than the threshold and is therefore not included in the valid range. In addition, the preset density threshold is not arbitrarily set, but is determined in combination with the constraints of the cut piece layout process and the utilization characteristics of the actual layout results. Specifically, it can be based on the fabric layout density of the target style under the calibrated width parameters as a benchmark reference value, and combined with the density distribution formed by adjacent width nodes during the layout simulation process, to limit the acceptable range of the fabric layout density; for example, under the layout state corresponding to the calibrated width, the cut pieces are arranged compactly and the corner allowance is small, this state can be... The corresponding fabric density is used as the upper limit of the preferred reference. At the same time, combined with the lower limit of fabric utilization allowed by the process (such as allowing a certain proportion of corner loss space), a lower limit critical value is selected as the preset density threshold within this range. Furthermore, the influence of the minimum spacing constraint and rotation constraint of the cut pieces on the compactness of the fabric layout can be combined to compare and analyze the fabric density under different width nodes. The density corresponding to the nodes that are significantly lower than the reasonable utilization level, have large fabric layout gaps, or have discontinuous cut piece layouts is excluded. In this way, a density threshold that can reflect the feasibility of fabric layout and ensure the efficiency of material utilization is determined. This threshold is used to screen effective adjacent nodes, so that the constructed width adjacency reference system retains only the width intervals that have actual fabric layout significance.

[0033] For example, if the material density is high and the utilization rate is sufficient at a 150cm node, then that node is marked as a valid adjacent node; while at a 135cm node, although material can be arranged, the utilization rate is significantly insufficient, so that node does not meet the threshold condition and is therefore rejected. In this way, the validity of span nodes is screened, ensuring that subsequent adjacency relationships are established only between nodes with high utilization rates.

[0034] Finally, after selecting all valid adjacent nodes, the nodes are sorted in ascending order of width value to form a node sequence with an ordered relationship. Based on this, the nodes are connected according to their numerical adjacency, that is, valid nodes that are numerically continuous or approximately continuous are connected sequentially to form a width variation path with structural continuity. For example, if the valid adjacent nodes are 140cm, 150cm, 155cm, and 165cm, then after arranging them in numerical order, 140cm is connected to 150cm, then 150cm is connected to 155cm, and finally connected to 165cm, forming a continuous width reference path structure. This path structure constitutes the width adjacency reference system, used to characterize the continuous mapping relationship between different width conditions in terms of fabric layout feasibility.

[0035] S2, based on the width adjacency reference system, the neighborhood landing point of the candidate fabric width parameter is identified, and the width offset identifier set is constructed according to the identification result; In this embodiment, the neighborhood landing point of the candidate fabric width parameter is identified based on the width adjacency reference system, and a width offset identifier set is constructed according to the identification result, specifically as follows: Extract the fabric width parameters of all candidate fabrics from the preset supplier fabric database, and compare the values ​​of each fabric width parameter with the effective adjacent nodes in the width adjacency reference system. Based on the comparison results, the fabric width parameter that is equal to the value of any valid adjacent node is selected as the benchmark width parameter, and the rest are selected as non-benchmark width parameters. For each non-benchmark width parameter, calculate the numerical difference between it and the two adjacent valid adjacent nodes in the width adjacency reference system, and identify the weighted adjacent nodes of the non-benchmark width parameter based on the numerical difference. Based on the difference direction between the weighted adjacent node's width value and the corresponding non-baseline width parameter, an offset direction identifier is generated, and an offset magnitude identifier is generated based on the ratio of the absolute value of the difference to the width step size. Mark the offset direction identifiers corresponding to all reference width parameters as zero offset identifiers, and combine and encode the offset direction identifiers and offset magnitude identifiers corresponding to all non-reference width parameters to obtain the width offset identifier for each candidate fabric. All candidate fabric identifiers are associated with and stored with their corresponding width offset identifiers to obtain a width offset identifier set.

[0036] It should be noted that, from a pre-set supplier fabric database, each candidate fabric record is read one by one, and the corresponding fabric width parameter is parsed out. This parameter represents the actual width specification of the fabric in production, such as common values ​​like 145cm, 150cm, and 160cm. Then, the width parameter of each candidate fabric is numerically matched and compared with the valid adjacent nodes in the width adjacency reference system. The valid adjacent nodes are the set of width nodes that have already passed the fabric layout density screening. The comparison process is based on numerical consistency and adjacency relationships to identify whether the candidate fabric width falls within the reference node set. For example, when the valid nodes in the reference system include 150cm, 155cm, and 160cm, if the candidate fabric width is 150cm, it is determined that it directly corresponds to a valid node; if it is 148cm, it is determined that it is not in the node set and proceeds to the subsequent non-benchmark processing flow.

[0037] After numerical comparison, candidate fabric width parameters that are completely consistent with the values ​​of any valid adjacent node in the width adjacency reference system are classified as baseline width parameters. These parameters indicate that their width specifications have accurately fallen into the preset discrete node system. For example, if a 150cm fabric has a corresponding node in the node set, it is classified as a baseline width parameter. For other width parameters that fail to form a numerical consistency relationship with any valid adjacent node, they are uniformly classified as non-baseline width parameters, such as width specifications like 148cm and 153cm. These belong to a continuous offset state between nodes, thus providing an object set for subsequent offset analysis.

[0038] For each non-reference width parameter, its position in the width adjacency reference system is first determined, i.e., its corresponding left and right adjacent valid nodes are identified. For example, when the non-reference width is 148cm, its left adjacent node is 145cm, and its right adjacent node is 150cm. Then, the numerical difference between the non-reference width and its left and right adjacent nodes is calculated, with one side reflecting the distance from the lower bound and the other reflecting the distance from the upper bound. This difference is then adjusted based on the weighted difference in material density between the two adjacent nodes. If one adjacent node has better material distribution suitability, it is identified as a weighted adjacent node, i.e., given a higher structural reference priority. For example, if the material distribution density of the 150cm node is significantly higher than that of the 145cm node, even if 148cm is closer to 145cm, 150cm may still be identified as a weighted adjacent node.

[0039] Furthermore, after determining the weighted adjacent nodes, the offset direction identifier is determined based on the relationship between the width value of the weighted adjacent node and the non-reference width parameter. This identifier characterizes whether the non-reference width is offset towards the larger width side or the smaller width side relative to the reference node. For example, if the non-reference width is 148cm and the weighted adjacent node is 150cm, then it is determined to be a downward offset direction.

[0040] Based on this, the degree of offset is classified into different levels according to the numerical difference between the non-baseline width and the weighted adjacent nodes, and in conjunction with the preset width step size, to form an offset level label. For example, when the width step size is 5cm, if the difference between 148cm and 150cm is small, it is classified as a low-level offset; if the difference between 142cm and 150cm is large, it is classified as a high-level offset.

[0041] Finally, for all baseline width parameters, since they are already completely consistent with their effective adjacent nodes, their offset direction identifiers are uniformly assigned a zero offset identifier and are not included in the offset magnitude calculation, thus representing that they are in the standard width baseline state. For all non-baseline width parameters, their offset direction identifiers and offset magnitude identifiers are combined and encoded to form a unique corresponding width offset identifier.

[0042] In this embodiment, for each non-reference width parameter, the numerical difference between it and two adjacent valid adjacent nodes in the width adjacency reference system is calculated, and the weighted adjacent nodes of the non-reference width parameter are identified based on the numerical difference. Specifically: Identify the numerical range of non-reference gate width parameters in the gate width adjacency reference system, and determine the left and right adjacent nodes of this numerical range; Calculate the first numerical difference between the non-reference gate width parameter and the left adjacent node, and calculate the second numerical difference between the right adjacent node and the non-reference gate width parameter; Obtain the material discharge density corresponding to the left and right adjacent nodes in the span adjacency reference system, and calculate the ratio of the material discharge density of the left adjacent node to the material discharge density of the right adjacent node; The first and second numerical differences are corrected based on the ratio of the discharge density of the left neighbor node to the discharge density of the right neighbor node. Identify weighted adjacent nodes for non-benchmark gate parameters based on the correction results.

[0043] It should be noted that for each non-baseline width parameter, it is first sorted and matched with all valid adjacent nodes in the width adjacency reference system according to their numerical values ​​to determine the range of the width parameter's position in the overall width sequence. The numerical range refers to the width span range formed between two adjacent valid adjacent nodes of the non-baseline width parameter.

[0044] For example, if the effective nodes in the width adjacency reference system include 145cm and 150cm, and a non-reference width parameter is 148cm, then it is determined that it falls within the interval between 145cm and 150cm, where 145cm is the left adjacent node and 150cm is the right adjacent node. These left and right adjacent nodes together constitute the structural envelope boundary of this non-reference width parameter.

[0045] After determining the left and right adjacent nodes, the difference between the non-reference width parameter and the left and right adjacent nodes is calculated. The first numerical difference characterizes the deviation distance between the non-reference width parameter and the left adjacent node, and the second numerical difference characterizes the deviation distance between the right adjacent node and the non-reference width parameter. For example, when the non-reference width is 148cm, the left adjacent node is 145cm, and the right adjacent node is 150cm, the first numerical difference between 148cm and 145cm reflects its offset relative to the lower boundary node; the second numerical difference between 150cm and 148cm reflects its offset relative to the upper boundary node.

[0046] It should be noted that the difference between the first and second numerical differences is that the first numerical difference uses the left adjacent node as the reference benchmark to reflect the degree of downward boundary offset; the second numerical difference uses the right adjacent node as the reference benchmark to reflect the degree of upward boundary offset. Both are used together to characterize the relative positional distribution characteristics of the non-reference span parameter within the interval.

[0047] Furthermore, after determining the left and right adjacent nodes, the fabric density values ​​of the left and right adjacent nodes are extracted from the fabric density mapping results corresponding to the width adjacency reference system. Fabric density is used to characterize the space utilization and compactness of the cut piece group under the given width condition. After obtaining the fabric density relationship between the left and right adjacent nodes, the first numerical difference and the second numerical difference are weighted and corrected according to the density ratio. Specifically, the comparison of the discharge density of the left and right adjacent nodes is used as the basis for weight allocation. When the discharge density of one side of the adjacent node is significantly higher than that of the other side, it indicates that the side is more suitable for actual discharge. In the correction process, the difference corresponding to the side with the lower discharge density is compressed, that is, its numerical influence is appropriately reduced. The difference corresponding to the side with the lower discharge density is maintained or moderately amplified, so that the corrected difference can better reflect the priority of discharge suitability. For example, when the discharge density of the right adjacent node is higher than that of the left adjacent node, the second numerical difference is reduced by convergence, making it more advantageous in subsequent comparisons, while the first numerical difference is maintained at its original scale or slightly amplified, further highlighting its disadvantage. This correction process is essentially a segmented proportional adjustment mechanism based on the division of density ratio intervals. Its core is to guide the difference weight shift through density differences, so as to make the weighted result more in line with the preferred trend in the actual discharge process while maintaining the stability of the calculation.

[0048] For example, given a non-standard width parameter of 148cm, its left adjacent node in the width adjacency reference system is 145cm, and its right adjacent node is 150cm. Therefore, the first numerical difference relative to 145cm is 3cm, and the second numerical difference relative to 150cm is 2cm. Geometrically, this width is closer to the 150cm node. Further analysis of material layout density data shows that the 145cm node corresponds to a material layout density of 0.76, while the 150cm node corresponds to a material layout density of 0.91. This demonstrates that the 150cm node is significantly superior to the 145cm node in terms of material layout compactness and utilization efficiency. When correcting the difference based on this, instead of simply multiplying 3cm or 2cm, the difference is adjusted in segments according to the density hierarchy: since the 150cm node has a higher density, its corresponding second numerical difference (2cm) is compressed, for example, adjusted to about 1.5cm, to further highlight the advantage of the "superior node" in the comparison; while the first numerical difference (3cm) corresponding to the 145cm node is maintained or slightly amplified, for example, adjusted to about 3.3cm, thereby strengthening its disadvantage in material layout adaptation. After the above correction, the right adjacent node 150cm is more obvious in the corrected comparison, so 150cm is finally identified as the weighted adjacent node of the 148cm width parameter.

[0049] Furthermore, if a non-standard width is 152cm, its left adjacent node is 150cm (discharge density 0.92), and its right adjacent node is 155cm (discharge density 0.80), then the first numerical difference is 2cm, and the second numerical difference is 3cm. Since the 150cm node has a higher density, the first numerical difference is compressed, for example, adjusted to about 1.4cm, while the second numerical difference is maintained or amplified, for example, adjusted to about 3.4cm. This ensures that 150cm retains its advantage after correction and is ultimately determined as the weighted adjacent node.

[0050] Finally, after the difference correction is completed, the corrected first numerical difference is compared with the second numerical difference to determine the adjacent node with a stronger correlation to the non-benchmark width parameter. This node is then identified as the weighted adjacent node. A weighted adjacent node refers to the node that, after considering the geometrically closest adjacent node and the influence weight of material density, ultimately has a dominant constraint effect on the non-benchmark width parameter. For example, in the case of a 148cm width, although it is geometrically closer to 145cm, if the material density corresponding to the 150cm node is significantly higher, then in the corrected judgment, 150cm may be identified as the weighted adjacent node, thus serving as the main reference benchmark for subsequent offset direction and offset intensity determination. Through this weighting mechanism, the adjacency relationship reflects not only numerical distance but also the degree of material arrangement process adaptation, thereby enhancing the process consistency and stability of width offset analysis.

[0051] S3, based on the width offset identifier set, identify the range of material fluctuation in the width adjacent reference system for the meter of material used in a single piece, and obtain the material correction adjacent domain set; In this embodiment, based on the width offset identifier set, the range of material usage fluctuation in meters of a single piece within the width adjacent reference system is identified, and the material usage correction adjacent domain set is obtained, specifically: Extract all candidate fabrics with non-zero offset magnitude from the width offset marker set, and group the candidate fabrics into the first offset group and the second offset group according to their offset direction marker and offset magnitude marker. In the first offset group, the preset material increment value corresponding to each offset level is accumulated sequentially according to the numerical value of the offset level identifier to obtain the first material fluctuation sequence; In the second offset group, the preset material reduction value corresponding to each offset level is subtracted sequentially according to the numerical value of the offset level identifier to obtain the second material fluctuation sequence; Add the maximum value in the first material usage fluctuation sequence to the calibrated material usage in meters to obtain the upper limit of material usage fluctuation; add the minimum value in the second material usage fluctuation sequence to the calibrated material usage in meters to obtain the lower limit of material usage fluctuation. Starting from the lower limit of material fluctuation and ending at the upper limit of material fluctuation, several consecutive material usage intervals are divided according to the preset material usage step size. Each material usage interval is used as a material usage correction adjacency domain. All material usage intervals are summed to obtain a set of material usage correction adjacency domains.

[0052] It should be noted that the offset identifier information corresponding to each candidate fabric is read one by one from the width offset identifier set. From this identifier information, a set of candidate fabrics with non-zero offset magnitude identifiers is selected to exclude fabrics in the baseline width state, retaining only those with width offset characteristics. Subsequently, based on the offset direction identifier carried by each candidate fabric, they are classified. Candidate fabrics with offset directions pointing towards the width increase direction are assigned to the first offset group, and candidate fabrics with offset directions pointing towards the width decrease direction are assigned to the second offset group. When the width increases, the available lateral layout space increases, and cut pieces that originally needed to be laid out in rows can now be laid out in the same row, thus reducing the longitudinal extension length of the layout diagram and causing a decrease in the number of meters of fabric used per piece. Conversely, when the width decreases, lateral space is limited, and to ensure the complete placement of cut pieces, the number of rows needs to be increased or the layout path lengthened, resulting in an increase in the overall number of meters of fabric used. However, it should be noted that this relationship is not absolutely linear. The final material usage is still affected by factors such as the combination structure of the cut pieces, the constraint of the grain direction, and the minimum spacing constraint. For example, when the width of some cut pieces is close to the critical value of the fabric width, a small increase in the fabric width may not change the number of layers, thus having a limited impact on the material usage.

[0053] The first offset group represents the set of material usage changes corresponding to the offset of the width relative to the reference value in an increasing direction, typically indicating an increasing trend in material usage. The second offset group represents the set of material usage changes corresponding to the offset of the width in a decreasing direction, typically indicating a decreasing trend in material usage. The difference between the two groups is that the offset directions are opposite, thus corresponding to different paths of material usage change.

[0054] It should be noted that within the first offset group, candidate fabrics are sorted in ascending order of offset magnitude to ensure the continuity of the progressive relationship of material usage changes. Based on this sorting, the preset material usage increment value corresponding to each offset magnitude is accumulated sequentially, so that the material usage increase effect brought about by each offset level is gradually superimposed, forming a material usage change sequence with hierarchical progressive characteristics. For example, when the offset magnitudes are level one, level two, and level three, they correspond to different magnitudes of material usage increase contributions. Therefore, the increments corresponding to level two and level three are gradually superimposed on level one, thus forming the first material usage fluctuation sequence. This sequence is used to characterize the cumulative material usage growth trend when the fabric width changes in the direction of increase.

[0055] It should be noted that the preset material increment value is not arbitrarily set, but is obtained by comparing and analyzing the virtual layout results of cut pieces under different width nodes, based on the mapping results of the previous width adjacent reference system and the material density. Specifically, the material layout result corresponding to the calibrated width parameter is used as the baseline state. On this basis, each adjacent node in the width increasing direction is selected, and the virtual layout of the cut piece group is completed according to the aforementioned layout rules. The material length of the single piece formed by the corresponding layout is recorded. Then, the material usage result corresponding to each width node is compared with the calibrated single piece material length under the baseline width. The material usage change range produced under different width increment conditions is extracted, and it is classified and merged according to the width step size to make the same or similar width offsets The material usage changes corresponding to each level are uniformly organized into a set of representative incremental values. For example, in the process of increasing the width from 150cm to 155cm and 160cm, the material usage can be observed to decrease by about 0.02 meters and 0.05 meters respectively through the layout results. The material usage change range corresponding to the offset level can be organized into preset material usage increment values ​​for different levels. Furthermore, the above increment values ​​can be smoothed or range representative values ​​can be taken by combining historical layout records or statistical results of multiple styles to avoid individual layout exceptions from interfering with the overall pattern. This forms preset material usage increment values ​​that can reflect the trend of width expansion and have stable applicability. These preset material usage increment values ​​are used for step-by-step accumulation calculation in the first offset group to characterize the material usage fluctuation pattern under the condition of widening.

[0056] Secondly, within the second offset group, the offset levels are also arranged in ascending order, ensuring that the change process conforms to a progressive relationship from slight to severe offset. Then, based on the preset material reduction value corresponding to each offset level, the material usage baseline is progressively reduced to reflect the material reduction effect caused by the shrinking fabric layout space when the fabric width decreases. For example, as the fabric width gradually decreases from the baseline value, a first-level offset corresponds to a slight reduction in material usage, while a second-level offset corresponds to a more significant reduction. This progressive reduction forms a second material usage fluctuation sequence, used to characterize the material usage change path under conditions of fabric width shrinkage.

[0057] It should be noted that the process basis for obtaining the preset material reduction value and the preset material increase value is consistent. This is derived from the comparative analysis of the cut piece layout results when the width changes in the direction of reduction: Specifically, the layout result corresponding to the calibrated width parameter is used as the baseline state. Based on this, each adjacent node in the direction of width reduction is selected, and the cut piece groups are arranged one by one according to the aforementioned virtual layout rules. The layout length required to complete the same cut piece combination under different width shrinkage conditions is recorded. Subsequently, the material usage result corresponding to each width node is compared with the calibrated single piece material usage in meters. The increase in material usage caused by the width reduction is extracted, and it is graded and merged according to the width step size, so that different offsets The material usage changes corresponding to each level form a set of reduction values ​​with hierarchical relationships. For example, when the door width decreases from 150cm to 145cm and 140cm, the material usage increases by approximately 0.03 meters and 0.07 meters respectively, as observed from the layout results. These changes can be organized into preset material reduction values ​​corresponding to different offset levels, and recorded in absolute terms for subsequent cumulative reduction processing. Furthermore, the reduction values ​​can be smoothed or ranged by combining multiple layout results or historical samples to weaken the fluctuations caused by individual extreme layout situations, thereby forming stable and representative preset material reduction values ​​for cumulative reduction by level in the second offset group.

[0058] After obtaining the first and second material usage fluctuation sequences, extreme value features are extracted from the two sequences to define the overall material usage fluctuation range. The material usage fluctuation does not change linearly with the width, but is affected by the piece combination structure and critical layout conditions, exhibiting a phased change characteristic: During the width expansion process, when the width increment does not reach the critical condition for changing the layout structure, it does not have a substantial impact on the number of rows or the piece combination method, and the material usage remains unchanged; however, when the width expansion reaches a certain critical value, it can enable the piece combination that originally needed to be arranged in rows to be arranged in the same row, which will cause a sudden change in the layout structure, resulting in a step decrease in material usage per piece; similarly, during the width contraction process, when the width reduction does not trigger a change in the layout structure, the material usage remains stable; however, when the width reduction is to the point that the original layout method cannot be maintained, it is necessary to increase the number of rows or adjust the combination structure, resulting in a step increase in material usage. Based on the aforementioned nonlinear variation characteristics, by performing discrete statistics and hierarchical accumulation on the material usage changes corresponding to different offset levels, the first and second material usage fluctuation sequences can reflect the aforementioned structural abrupt change process. The maximum and minimum change values ​​are extracted respectively and superimposed with the calibrated single-piece material usage in meters to form the upper and lower limits of material usage fluctuation. For example, when the calibrated single-piece material usage in meters is 1.65 meters, an initial increase in width may have no significant impact on material usage, but after reaching a certain critical width, the material usage decreases to approximately 1.58 meters. Conversely, during the process of width reduction, it may rise to approximately 1.72 meters at a certain critical point, thus forming a material usage fluctuation range with staged jump characteristics.

[0059] Finally, after determining the lower and upper limits of material usage fluctuation, the range is divided into equal-interval sections. The lower limit is used as the starting boundary, and the upper limit as the ending boundary. The range is further divided into several consecutive material usage intervals according to a preset material usage step size, ensuring that each interval corresponds to a certain range of material usage changes. For example, within the 1.58-meter to 1.72-meter range, dividing the range with a step size of 0.02 meters can create multiple adjacent material usage intervals, each corresponding to a different level of material usage change. Subsequently, each material usage interval is treated as a material usage correction adjacency domain to characterize the consistency of material usage correction for candidate fabrics within that interval. Finally, all material usage correction adjacency domains are aggregated to form a material usage correction adjacency domain set, thus constructing a continuous mapping structure for material usage changes based on width offset.

[0060] S4. In the preset supplier fabric database, perform same-domain expansion screening on the adjacent domain set of material modification to obtain a candidate fabric domain set containing the modification status marker. In this embodiment, within a preset supplier fabric database, the adjacent domain set for material usage correction is expanded and filtered to obtain a candidate fabric domain set containing correction status markers, specifically: Extract the actual single-piece material usage in meters of each candidate fabric from the supplier fabric database, and determine the inclusion relationship between the actual single-piece material usage in meters and each material usage interval in the material usage correction adjacency set. Based on the judgment results, identify the candidate fabrics within each material usage interval and obtain the number of candidate fabrics within each material usage interval. If the number of candidate fabrics in the domain is less than the preset first threshold, then the candidate fabrics in the domain in the adjacent material usage intervals of the current material usage interval are merged into the current material usage interval to obtain an expanded set of candidate fabrics. For each candidate fabric in the expanded candidate fabric set corresponding to each material usage interval, an expansion identifier is generated based on the distance between its original material usage interval and the current material usage interval, resulting in a candidate fabric domain set containing the correction status marker.

[0061] It should be noted that the actual fabric consumption per garment is extracted from the supplier's fabric database for each candidate fabric. This fabric consumption per garment corresponds to the material consumption value per unit under actual production or standard process conditions. For example, the actual fabric consumption per garment for a certain fabric might be 1.63 meters, 1.67 meters, or 1.71 meters. Subsequently, this actual fabric consumption per garment is compared one by one with each fabric consumption interval contained in the fabric consumption correction adjacency domain set. Each fabric consumption interval is formed by the preceding steps and is used to characterize the fabric consumption fluctuation within a specific range. When the actual fabric consumption per garment for a candidate fabric falls within the upper and lower boundaries of a fabric consumption interval, it is determined that the candidate fabric and that fabric consumption interval form an inclusion relationship. For example, if the fabric consumption interval is 1.62 meters to 1.66 meters, and the actual fabric consumption of a candidate fabric is 1.63 meters, then it is determined that it belongs to that interval.

[0062] After determining the inclusion relationship, each material usage interval is categorized and summarized. All candidate fabrics that satisfy the inclusion relationship of that interval are assigned to the corresponding candidate set within that interval, thus forming a fabric aggregation structure based on intervals. Subsequently, the number of candidate fabrics within each material usage interval is counted to obtain the distribution density of candidate fabrics within that interval. For example, in the 1.62-meter to 1.66-meter interval, if the collected candidate fabrics include multiple fabric models corresponding to 1.63-meter, 1.64-meter, and 1.65-meter lengths, they are all uniformly assigned to that interval, and the number of candidate fabrics within that interval is counted to reflect the degree of supply concentration in that interval.

[0063] Furthermore, after completing the statistical analysis of the number of candidate fabrics in each material usage interval, the number of candidate fabrics in each interval is compared with a preset first threshold. When the number of candidate fabrics in a certain material usage interval is lower than the threshold, it is determined that the supply density of that interval is insufficient and it does not have independent screening stability. At this time, the candidate fabrics in the adjacent material usage intervals of that interval are introduced into the current interval for merging and expansion processing, so that the originally scattered low-density intervals are formed into an expanded unified candidate set.

[0064] It should be noted that the preset first threshold is not a fixed constant, but is determined by combining the aforementioned material usage interval division results and the distribution characteristics of candidate fabrics in each material usage interval. Its purpose is to determine whether a certain material usage interval has a stable basis for independent screening. Specifically, the number of candidate fabrics in all material usage intervals can be statistically analyzed first to obtain the quantity distribution of each interval, and the overall average level or median distribution level can be used as a reference benchmark. On this basis, the minimum sample size that can support effective screening and ranking output is set by combining the granularity of the material usage interval division and the scale characteristics of the supplier fabric library, thereby determining the range of the first threshold value.

[0065] For example, in a certain application scenario, if the number of candidate fabrics in most material usage intervals is concentrated between 5 and 12, the lower limit to ensure screening stability can be set to 3 or 4. When the number of candidate fabrics in a certain interval is lower than this threshold, it is determined that the sample in that interval is insufficient, and it is necessary to expand and supplement by introducing candidate fabrics from adjacent material usage intervals. Furthermore, the output stability under different thresholds can be compared with historical screening results, and the threshold that can ensure interval discrimination while avoiding the candidate results being too sparse can be selected as the final setting. Thus, the first threshold reflects the actual distribution density of candidate fabrics in the material usage interval and can support the continuity and effectiveness of the subsequent expansion screening process.

[0066] After completing the interval expansion process, an identifier generation process is performed on each candidate fabric within the expanded candidate fabric set corresponding to each material usage interval. Specifically, based on the offset distance between the original material usage interval to which the candidate fabric belonged and the current expanded target material usage interval, an expansion identifier is generated to characterize the degree to which it has migrated from the original interval to the current interval. For example, if a fabric originally belonged to the 1.60-meter to 1.62-meter interval, but after expansion it is classified into the 1.62-meter to 1.66-meter interval, its expansion identifier corresponds to a lower migration level; if the migration spans multiple intervals, it corresponds to a higher migration level.

[0067] Finally, all candidate fabrics are associated with and stored with their expansion status identifiers, ultimately forming a candidate fabric domain set containing corrected status markers. This candidate fabric domain set with corrected status markers refers to a structured set of fabrics formed by further introducing interval expansion behavior and migration degree identifiers on top of the material usage interval division. This set not only contains the correspondence between fabrics and usage intervals but also information on the state changes of the fabrics during interval adjustment, thus enabling the fabric selection results to simultaneously reflect the original distribution characteristics and the adaptation relationship after corrected expansion.

[0068] S5 generates differentiated fabric filtering output instructions based on the candidate fabric domain set containing correction status markers, and distributes the output instructions to the user interface.

[0069] In this embodiment, differentiated fabric filtering output instructions are generated based on the candidate fabric domain set including the correction status flag, and the output instructions are distributed to the user interface, specifically as follows: Traverse the candidate fabric domain set containing the correction status flag, extract the expansion flag from the correction status flag of each candidate fabric, and assign candidate fabrics with an expansion flag of zero to the first matching group, and assign candidate fabrics with a non-zero expansion flag to the second matching group. For the candidate fabrics in the first matching group, generate the first type of output instruction based on their corresponding material usage range identifier; For the candidate fabrics in the second matching group, sort them according to the value of the expansion identifier from smallest to largest, and generate the second type of output instructions in sequence according to the sorting results; The first type of output command and the second type of output command are distributed to the user interface in sequence.

[0070] It should be noted that the candidate fabric domain set containing the correction status marker is traversed one by one, and the correction status marker information carried by each candidate fabric is read sequentially. The correction status marker includes at least an expansion identifier, which is used to characterize the migration state of the candidate fabric during the expansion of the material application range. Then, the expansion identifier is determined. When the expansion identifier is zero, it means that the candidate fabric has not undergone range migration or expansion adjustment, and its material application range maintains the original matching state. Therefore, it is classified into the first matching group. When the expansion identifier is not zero, it means that the candidate fabric has undergone varying degrees of migration or adsorption adjustment during the range expansion. In this case, it is classified into the second matching group.

[0071] The first matching group is used to characterize the set of fabrics that maintain the original correspondence between the material usage range and the actual material usage of the fabric. Its characteristic is that the matching relationship is stable and has not been adjusted by range expansion. The second matching group is used to characterize the set of fabrics that have undergone range expansion or cross-region migration. Its characteristic is that the material usage attribution relationship has been corrected, so priority needs to be distinguished in the output order.

[0072] After grouping, the candidate fabrics in the first matching group are structured according to their corresponding material usage range identifiers. These material usage range identifiers characterize the standard material usage fluctuation range to which the fabric belongs, such as 1.62 meters to 1.66 meters or 1.66 meters to 1.70 meters. Then, using the material usage range as the classification criterion, candidate fabrics within the same range are merged into the same output unit, generating a corresponding first-type output instruction. The first-type output instruction characterizes the recommended candidate fabrics without expansion correction. Its content structure emphasizes the stable matching attribute of the material usage range. For example, when a 1.63-meter fabric is classified into the 1.62-meter to 1.66-meter range, all fabrics in the first matching group within that range will be output in a range-priority manner, thus forming a stable matching result set.

[0073] For candidate fabrics in the second matching group, they are sorted according to the value of their expansion identifiers. The expansion identifier represents the migration level the fabric experiences during interval expansion; a smaller value indicates a lighter expansion, and a larger value indicates a stronger expansion. After sorting, output content is generated sequentially from low expansion level to high expansion level, forming a second type of output instruction. For example, if candidate fabric A has an expansion identifier of level 1, fabric B is level 2, and fabric C is level 3, then the output is generated in the order of A, B, and C, thus prioritizing fabrics with lower expansion levels. The second type of output instruction represents the candidate fabric recommendation results after interval correction or cross-regional adsorption, and its output logic reflects the relationship between correction priority and migration intensity ranking.

[0074] After generating the first and second types of output instructions, the two types of instructions are pushed to the user interface for display in a preset order. Typically, the first type of output instruction is output first to ensure stable matching results are presented first, followed by the second type of output instruction to supplement the expanded and corrected candidate results. The first type of output instruction refers to the recommended instruction generated based on the original material range matching relationship. Its characteristic is that it does not contain range expansion and migration information, only reflecting stable matching results within the standard range. The second type of output instruction refers to the recommended instruction generated based on range expansion and migration correction. Its characteristic is that it includes expansion identifiers and migration level information, reflecting the adaptation results of the fabric during dynamic range adjustment.

[0075] Example 2, Figure 2 This invention presents a smart supplier fabric screening system tailored to user needs, comprising a reference construction module, a landing point recognition module, a material correction module, an expanded screening module, and a fabric screening module. The reference construction module is used to obtain the standard width parameters of the target style and the corresponding standard single piece material usage in meters, and to construct the width adjacency reference system based on the preset cutting piece layout process constraints; The landing point recognition module is used to identify the neighborhood landing point of the candidate fabric width parameters based on the width adjacency reference system, and to construct a width offset identifier set based on the recognition results; The material usage correction module is used to identify the range of material usage fluctuation in the adjacent reference system of the width based on the width offset identifier set, and to obtain the material usage correction adjacent domain set. The expanded filtering module is used to perform same-domain expanded filtering on the adjacent domain set of material modification in the preset supplier fabric database to obtain a candidate fabric domain set containing the modification status flag. The fabric filtering module is used to generate differentiated fabric filtering output instructions based on a set of candidate fabric domains containing correction status markers, and to distribute the output instructions to the user interface.

[0076] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0077] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0078] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0079] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0080] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent selection of supplier fabrics based on user needs, characterized in that, Includes the following steps: Obtain the standard width parameters and corresponding standard single-piece material usage in meters for the target style, and construct a width adjacency reference system based on the preset pattern layout process constraints, specifically: Using the calibrated width parameter as the origin, several adjacent width nodes are generated on both sides of the origin with a preset width step size; the minimum spacing constraint and rotation constraint of the cut pieces in the preset cut piece layout process constraints are obtained, and the layout density that can accommodate the cut piece group is calculated for each adjacent width node to obtain a density mapping set; according to the density mapping set, the adjacent width nodes whose layout density exceeds the preset density threshold are marked as valid adjacent nodes; all valid adjacent nodes are sorted according to the width value and connected to adjacent valid adjacent nodes to obtain the width adjacency reference system; Based on the width adjacency reference system, the neighborhood landing point of the candidate fabric width parameter is identified, and a width offset identifier set is constructed according to the identification results, specifically: The fabric width parameters of all candidate fabrics are extracted from a pre-defined supplier fabric database. Each fabric width parameter is then numerically compared with the effective adjacent nodes in the width adjacency reference system. Based on the comparison results, the fabric width parameter with the same value as any effective adjacent node is selected as the baseline width parameter, and the rest are designated as non-baseline width parameters. For each non-baseline width parameter, the numerical difference between it and two adjacent effective adjacent nodes in the width adjacency reference system is calculated. Based on the numerical difference and the fabric density corresponding to the two adjacent effective adjacent nodes, the weighted adjacent nodes of the non-baseline width parameter are identified. An offset direction identifier is generated based on the direction of the difference between the width value of the weighted adjacent node and the corresponding non-baseline width parameter. An offset magnitude identifier is generated based on the ratio of the absolute value of the difference to the width step size. All offset direction identifiers corresponding to the baseline width parameters are marked as zero offset identifiers. The offset direction identifiers and offset magnitude identifiers corresponding to all non-baseline width parameters are combined and encoded to obtain the width offset identifier for each candidate fabric. The fabric identifiers of all candidate fabrics are associated and stored with their corresponding width offset identifiers to obtain a width offset identifier set. Based on the width offset identifier set, the range of material usage fluctuation in meters per piece within the width adjacent reference system is identified, resulting in a material usage correction adjacent domain set, specifically: Extract all candidate fabrics with offset magnitude indicators from the width offset indicator set, and group the candidate fabrics into the first offset group and the second offset group according to their offset direction indicator and offset magnitude indicator. In the first offset group, the preset material increment value corresponding to each offset level is accumulated sequentially according to the value of the offset level identifier to obtain the first material usage fluctuation sequence; in the second offset group, the preset material decrement value corresponding to each offset level is subtracted sequentially according to the value of the offset level identifier to obtain the second material usage fluctuation sequence; the maximum value in the first material usage fluctuation sequence is added to the calibrated material usage in meters to obtain the upper limit of material usage fluctuation; and the minimum value in the second material usage fluctuation sequence is added to the calibrated material usage in meters to obtain the lower limit of material usage fluctuation. Starting from the lower limit of material fluctuation and ending at the upper limit of material fluctuation, several consecutive material usage intervals are divided according to the preset material usage step size. Each material usage interval is used as a material usage correction adjacency domain. All material usage intervals are summed to obtain a set of material usage correction adjacency domains. In the pre-defined supplier fabric database, the adjacent domain set for material usage correction is expanded and filtered to obtain a candidate fabric domain set containing correction status markers, specifically: The actual single-piece material usage meters of each candidate fabric are extracted from the supplier fabric database, and the inclusion relationship between the actual single-piece material usage meters and each material usage interval in the material usage correction adjacency set is determined. Based on the determination result, candidate fabrics within each material usage interval are identified and the number of candidate fabrics within each material usage interval is obtained. If the number of candidate fabrics within the interval is less than a preset first threshold, the candidate fabrics within the adjacent material usage intervals of the current material usage interval are merged into the current material usage interval to obtain an expanded candidate fabric set. For each candidate fabric in the expanded candidate fabric set corresponding to each material usage interval, an expansion identifier is generated based on the distance between its original material usage interval and the current material usage interval to obtain a candidate fabric domain set containing the correction status marker. Differentiated fabric filtering output instructions are generated based on the candidate fabric domain set containing correction status markers, and the output instructions are distributed to the user interface.

2. The intelligent supplier fabric selection method based on user needs according to claim 1, characterized in that, The process of obtaining the standard width parameters of the target style and the corresponding standard single-piece material usage in meters, and constructing a width adjacency reference system based on preset cutting pattern layout process constraints, specifically involves: Obtain the style identifier of the target style uploaded by the user, and retrieve the corresponding cutting pattern from the preset process database based on the style identifier; Analyze the geometric contours of all cut pieces in the cut piece layout diagram, generate a cut piece set, and obtain the width dimension and thread direction constraint of each cut piece in the cut piece set; The cut pieces are grouped according to the width and grain direction constraints of each piece to obtain several cut piece groups; For each cut piece group, the largest width dimension is extracted as the calibration width constraint value of the cut piece group, and calibration width parameters are generated based on the calibration width constraint values ​​of all cut piece groups. At the same time, the calibration single piece material meterage corresponding to the style is extracted from the preset process database.

3. The intelligent supplier fabric selection method based on user needs as described in claim 2, characterized in that, For each non-reference width parameter, the numerical difference between it and two adjacent valid adjacent nodes in the width adjacency reference system is calculated. Based on the numerical difference and the discharge density corresponding to the two adjacent valid adjacent nodes, the weighted adjacent nodes of the non-reference width parameter are identified. Specifically: Identify the numerical range of non-reference gate width parameters in the gate width adjacency reference system, and determine the left and right adjacent nodes of this numerical range; Calculate the first numerical difference between the non-reference gate width parameter and the left adjacent node, and calculate the second numerical difference between the right adjacent node and the non-reference gate width parameter; Obtain the material discharge density corresponding to the left and right adjacent nodes in the span adjacency reference system, and calculate the ratio of the material discharge density of the left adjacent node to the material discharge density of the right adjacent node; The first and second numerical differences are corrected based on the ratio of the discharge density of the left neighbor node to the discharge density of the right neighbor node. Identify weighted adjacent nodes for non-benchmark gate parameters based on the correction results.

4. The intelligent supplier fabric selection method based on user needs according to claim 3, characterized in that, The process of generating differentiated fabric filtering output instructions based on a candidate fabric domain set including corrected status markers, and distributing these output instructions to the user interface, specifically involves: Traverse the candidate fabric domain set containing the correction status flag, extract the expansion flag from the correction status flag of each candidate fabric, and assign candidate fabrics with an expansion flag of zero to the first matching group, and assign candidate fabrics with a non-zero expansion flag to the second matching group. For candidate fabrics in the first matching group, a first type of output instruction is generated based on their corresponding material usage range identifier, wherein the material usage range identifier is used to characterize the standard material usage fluctuation range to which the fabric belongs; For the candidate fabrics in the second matching group, sort them according to the value of the expansion identifier from smallest to largest, and generate the second type of output instructions in sequence according to the sorting results; The first type of output command and the second type of output command are distributed to the user interface in sequence.

5. A system using the intelligent supplier fabric screening method based on user needs as described in any one of claims 1-4, characterized in that, This includes a reference construction module, a landing point recognition module, a material correction module, an expanded screening module, and a fabric screening module: The reference construction module is used to obtain the standard width parameters of the target style and the corresponding standard single piece material usage in meters, and to construct the width adjacency reference system based on the preset cutting piece layout process constraints; The landing point recognition module is used to identify the neighborhood landing point of the candidate fabric width parameters based on the width adjacency reference system, and to construct a width offset identifier set based on the recognition results; The material usage correction module is used to identify the range of material usage fluctuation in the adjacent reference system of the width based on the width offset identifier set, and to obtain the material usage correction adjacent domain set. The expanded filtering module is used to perform same-domain expanded filtering on the adjacent domain set of material modification in the preset supplier fabric database to obtain a candidate fabric domain set containing the modification status flag. The fabric filtering module is used to generate differentiated fabric filtering output instructions based on a set of candidate fabric domains containing correction status markers, and to distribute the output instructions to the user interface.

Images