Gift design scheme optimization method and system fusing multi-modal user demand perception

Abstract

This invention discloses a method and system for optimizing gift design schemes by integrating multimodal user demand perception, relating to the field of data processing technology. The method includes: acquiring a set of gift size demand features; calculating the weight coefficients of each gift size demand feature through sensitivity analysis combined with historical gift design schemes; constructing a collaborative optimization objective function for gift size, using size design parameters as independent variables and maximizing the overall matching degree as the optimization objective, while setting cost constraint thresholds and material size matching thresholds as optimization constraints; acquiring the gift size design range, iteratively solving the collaborative optimization objective function to obtain multiple candidate size design schemes that satisfy all constraints, and the overall matching degree corresponding to each candidate scheme; selecting the candidate size design scheme with the highest overall matching degree and single-dimensional matching degree to obtain the optimal gift size design scheme set. This invention effectively improves the overall matching degree of gift size design under multimodal user demand.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method and system for optimizing gift design schemes by integrating multimodal user demand perception. Background Technology

[0002] In bulk purchasing scenarios, it is often necessary to choose from multiple packaging specifications to simultaneously meet various requirements such as display effect, cost control, and convenient transportation. Currently, gift size design largely relies on manual experience or simple judgment based on a single dimension.

[0003] However, existing technologies cannot coordinate and integrate multimodal requirements such as display size, cost, and transportation, which can easily lead to conflicts between different dimensions of requirements. This makes it difficult to efficiently select candidate solutions that meet the requirements, resulting in insufficient overall matching degree of gift size design under multimodal user needs. Consequently, it is impossible to provide enterprises with the best overall gift size design solution for bulk procurement. Summary of the Invention

[0004] This invention provides a method and system for optimizing gift design schemes by integrating multimodal user needs perception, aiming to solve the technical problem of insufficient comprehensive matching degree of existing gift size designs under multimodal user needs.

[0005] In view of the above problems, the present invention provides a method and system for optimizing gift design schemes by integrating multimodal user demand perception.

[0006] In a first aspect, the present invention provides a method for optimizing gift design schemes by integrating multimodal user demand perception, including: Obtain a gift size requirement feature set, wherein the gift size requirement feature set includes display size requirements, cost requirements, and transportation requirements; By combining sensitivity analysis with historical gift design schemes, the weighting coefficients of the size requirements for each gift were calculated. Based on the gift size requirement feature set and the weight coefficient, a collaborative optimization objective function for gift size is constructed, with size design parameters as independent variables and maximizing the comprehensive matching degree as the optimization objective. At the same time, cost constraint threshold and material size matching threshold are set as optimization constraints. Obtain the gift size design range, iterate through and solve the collaborative optimization objective function to obtain multiple candidate size design schemes that satisfy all constraints, and the comprehensive matching degree corresponding to each candidate scheme; Select the candidate size design scheme with the highest overall matching degree and single-dimensional matching degree to obtain the optimal size design scheme set for the gift.

[0007] Secondly, this invention provides a gift design optimization system that integrates multimodal user demand perception, including: The demand feature acquisition module is used to acquire a gift size demand feature set, wherein the gift size demand feature set includes display size requirements, cost requirements, and transportation requirements. The weighting coefficient calculation module is used to calculate the weighting coefficients of the size requirements of each gift by combining sensitivity analysis with historical gift design schemes. The objective function construction module is used to construct a collaborative optimization objective function for gift size based on the gift size requirement feature set and the weight coefficients. The size design parameters are used as independent variables, and the optimization objective is to maximize the comprehensive matching degree. At the same time, cost constraint threshold and material size matching threshold are set as optimization constraints. The candidate solution acquisition module is used to obtain the gift size design range, iterate through and solve the collaborative optimization objective function, obtain multiple candidate size design schemes that satisfy all constraints, and the comprehensive matching degree corresponding to each candidate scheme; The optimal solution selection module is used to select the candidate size design scheme with the highest comprehensive matching degree and single-dimensional matching degree to obtain the optimal size design scheme set for gifts.

[0008] One or more technical solutions provided in this invention have at least the following technical effects or advantages: This invention provides a method and system for optimizing gift design schemes by integrating multimodal user needs perception. By acquiring a multimodal feature set including display size, cost, and transportation requirements, and using sensitivity analysis combined with historical schemes to calculate weight coefficients, it achieves accurate quantification of the differences in the importance of multiple needs, avoiding biases from subjective weighting. A collaborative optimization objective function is constructed with the goal of maximizing the overall matching degree, and dual constraints of cost and material size are set, transforming the multi-objective conflict problem into a solvable single-objective optimization problem. Through traversal optimization, multiple candidate schemes satisfying all constraints are obtained, ensuring sufficient exploration of the design space and diversity of the solution set. By selecting the scheme with the highest overall matching degree and the highest matching degree in each single dimension, a unified output of global and local optima is achieved. This invention solves the technical problems of difficulty in coordinating multiple needs, strong subjectivity in weight setting, and low efficiency in candidate scheme selection in gift size design, effectively improving the overall matching degree and decision reliability of gift size design under multimodal user needs. Attached Figure Description

[0009] Figure 1 A flowchart illustrating the gift design optimization method that integrates multimodal user demand perception provided in an embodiment of the present invention; Figure 2 A schematic diagram of the structure of the gift design optimization system that integrates multimodal user demand perception provided in an embodiment of the present invention; The components represented by each number in the attached diagram are explained below: The module includes: 11 for obtaining demand features, 12 for calculating weight coefficients, 13 for constructing objective functions, 14 for obtaining candidate solutions, and 15 for selecting the optimal solution. Detailed Implementation

[0010] This invention provides a method and system for optimizing gift design schemes by integrating multimodal user needs perception, which is used to address the technical problem of insufficient comprehensive matching degree of existing gift size designs under multimodal user needs.

[0011] Example 1, as Figure 1 As shown, this invention provides a method for optimizing gift design schemes by integrating multimodal user demand perception, the method comprising: S100: Obtain the gift size requirement feature set, wherein the gift size requirement feature set includes display size requirements, cost requirements, and transportation requirements.

[0012] In this embodiment of the invention, a gift size requirement feature set is obtained, which includes display size requirements, cost requirements, and transportation requirements. In the prior art, gift size design often relies on the designer's experience and judgment, lacking quantitative data collection on actual user scenarios, budget constraints, and logistics conditions. Different users have varying requirements for gift display space, different budget ranges, and different transportation loading conditions depending on the logistics channel. Therefore, it is necessary to quantify the three types of requirements—display size requirements, cost requirements, and transportation requirements—into calculable size requirement ranges to provide data support for subsequent weight calculations and objective function construction.

[0013] Step S100 in the method provided in this embodiment of the invention includes: Collect data on user gift usage scenarios and display space requirements to determine display size needs; Obtain the user's gift budget, and combine it with the unit price of gift materials and processing fee rates to obtain cost requirements; Quantitatively analyze historical gift procurement transportation and loading adaptation data to obtain transportation requirements; The display size requirements, cost requirements, and transportation requirements are integrated to form the gift size requirement feature set, wherein the display size requirements, cost requirements, and transportation requirements include the gift size requirement range.

[0014] First, data on user gift usage scenarios and display space is collected to determine display size requirements. Display size requirements refer to the range of space dimensions that gift packaging must fit in actual use and display scenarios, quantified by ranges of length, width, and height. These are the size requirements to ensure the effective display of the gift. Data on the gift usage scenarios specified by the company, such as reception area displays, customer desktop placement, and supermarket booths, is collected, and the minimum and maximum accommodated dimensions of the corresponding display spaces are measured on-site to establish a range of display sizes.

[0015] For example, a company purchases Mid-Autumn Festival mooncake gift boxes in bulk for use as front desk displays and gifts to customers. The single-layer space of the front desk display rack is 380mm long, 320mm wide, and 180mm high. The minimum suitable size for customer desktop placement is 300mm long, 240mm wide, and 100mm high. Therefore, the required display size range is determined to be: length 300mm-380mm, width 240mm-320mm, and height 100mm-180mm.

[0016] Secondly, obtain the user's gift budget, and combine it with the unit price of gift materials and processing fee rates to obtain cost requirements. Cost requirements refer to the cost range that the company can afford to purchase gift packaging in bulk. This is calculated by combining the material cost and processing cost of a single package, and serves as a constraint to control the total procurement cost. Collect the company's total budget or the upper limit of the budget per package for gift packaging, and combine it with the unit price of the packaging materials and processing fee rates to calculate the cost range for a single package.

[0017] For example, the company plans to purchase 5,000 mooncake gift boxes, with a budget limit of 18 yuan per box; the packaging uses 350g white cardboard, with a material unit price of 0.8 yuan / m³. 2 The processing fee for hot stamping and die-cutting is 3 yuan per piece. For example, the reasonable cost range for a single gift box is calculated to be 10-18 yuan, which is the cost requirement.

[0018] Secondly, we obtain and quantitatively analyze historical gift procurement transportation and loading adaptation data to identify transportation requirements. Transportation requirements refer to the size of the gift packaging needing to be compatible with the loading specifications of standard transport vehicles. This is quantified by size compatibility range to improve loading efficiency and reduce transportation losses. By retrieving historical gift procurement transportation data from the company and statistically analyzing it, we determine the size range with the optimal loading effect, thus forming a transportation demand range.

[0019] For example, we retrieve the company's transportation data for Mid-Autumn Festival gift purchases over the past three years. The data shows that when the gift boxes are between 320mm and 360mm in length, 260mm and 300mm in width, and 120mm and 160mm in height, they can be neatly stacked in standard 600mm x 400mm logistics cartons, with each carton holding 8 boxes. This results in a space utilization rate of over 92%, and a transportation damage rate of less than 0.5%. When the dimensions exceed this range, the number of cartons loaded decreases or gaps appear, leading to increased transportation costs. Based on this, the generated transportation requirements are: length range 320mm-360mm, width range 260mm-300mm, and height range 120mm-160mm.

[0020] Finally, the display size requirements, cost requirements, and transportation requirements are integrated to form the gift size requirement feature set, which includes gift size requirement ranges. The gift size requirement feature set is a standardized dataset formed by integrating the quantified ranges of display size, cost, and transportation requirements; it serves as the foundation for subsequent weight calculations and collaborative optimization. Integrating the display size requirement range, cost requirement range, and transportation requirement range forms complete multi-dimensional requirement feature data.

[0021] For example, after integration, the size requirement feature set of this mooncake gift box is obtained: Display size requirement: length 300mm-380mm, width 240mm-320mm, height 100mm-180mm; Cost requirement: 10-18 yuan / piece; Transportation requirement: length 320mm-360mm, width 260mm-300mm, height 120mm-160mm.

[0022] In this embodiment of the invention, vague subjective needs are transformed into standardized interval data in three dimensions: display, cost, and transportation. This completes the quantification and integration of multimodal user needs, clarifies the core constraints and objectives of gift size design, and provides accurate and reliable data support for subsequent weight coefficient calculation and collaborative optimization objective function construction. This avoids the problem of unreasonable size design caused by unclear needs from the source.

[0023] S200: Calculate the weighting coefficients of the size requirements for each gift by combining sensitivity analysis with historical gift design schemes.

[0024] In this embodiment of the invention, sensitivity analysis is combined with historical gift design schemes to calculate the weight coefficients of each gift size requirement feature. In the size optimization design of gifts, size requirements, cost requirements, and transportation requirements have varying degrees of influence on the final size design scheme. If equal weights or weights are set based on subjective experience, subsequent optimization schemes will deviate from actual needs, failing to achieve a synergistic balance of multimodal requirements. Therefore, it is necessary to combine the actual adaptability of historical gift design schemes, quantify the influence of each requirement feature through sensitivity analysis, and scientifically calculate the weight coefficients to provide an objective and quantifiable weight basis for constructing the subsequent collaborative optimization objective function.

[0025] Step S200 in the method provided in this embodiment of the invention includes: Obtain historical gift design schemes and extract display size adaptation data, cost adaptation data, and transportation adaptation data from each scheme. Among them, the display size adaptation data is the quantified value of the deviation between the historical scheme size and the display size requirement, the cost adaptation data is the quantified value of the deviation between the historical scheme actual cost and the cost requirement, and the transportation adaptation data is the quantified value of the deviation between the historical scheme size and the transportation requirement. Based on the adaptation data of various dimensions of historical gift design schemes, the historical adaptation contribution of each gift size requirement characteristic is calculated and normalized to obtain the historical average weight. Based on the gift size demand feature set, sensitivity analysis is performed on each demand feature, and the sensitivity coefficient of each gift size demand feature is calculated. The sensitivity coefficients of each demand feature are normalized to obtain the sensitivity weights of each gift size demand feature. The weighted sum of the historical average weight and the sensitivity weight is used to obtain the weight coefficients of each gift size requirement feature.

[0026] First, obtain historical gift design schemes and extract display size adaptation data, cost adaptation data, and transportation adaptation data from each scheme. Among them, the display size adaptation data is the quantified value of the deviation between the historical scheme size and the display size requirement, the cost adaptation data is the quantified value of the deviation between the historical scheme actual cost and the cost requirement, and the transportation adaptation data is the quantified value of the deviation between the historical scheme size and the transportation requirement.

[0027] Among them, historical gift design schemes refer to packaging size design schemes that have been implemented when the company purchased similar mooncake gift boxes in the past, including complete data such as actual size, actual cost, and transportation compatibility. Display size compatibility data is the quantified value of the deviation between the actual size of the historical scheme and the display size requirements in S100; the smaller the deviation, the higher the compatibility. Cost compatibility data is the quantified value of the deviation between the actual cost of the historical scheme and the cost requirements in S100; the smaller the deviation, the higher the compatibility. Transportation compatibility data is the quantified value of the deviation between the actual size of the historical scheme and the transportation size requirements in S100; the smaller the deviation, the higher the compatibility.

[0028] Specifically, retrieve the historical design schemes of mooncake gift boxes implemented by the company in the past three years, extract the actual length, width and height dimensions and the actual cost per unit for each scheme; use the median demand value as the benchmark and the demand range as the denominator to calculate the relative deviation value of each dimension, which is the corresponding adaptation data. The calculation formula is: adaptation data for a certain dimension = |historical actual value - median demand value for that dimension| / demand range for that dimension, where median demand value = (minimum demand value + minimum demand value) / 2, and demand range = minimum demand value - minimum demand value.

[0029] For example, based on the demand range of S100, the display size length requirement is 300mm-380mm, with a median of 340mm and a range of 80mm; the cost requirement is 10-18 yuan, with a median of 14 yuan and a range of 8 yuan; the transportation size length requirement is 320mm-360mm, with a median of 340mm and a range of 40mm. We retrieved several historical mooncake gift box design schemes implemented by the company over the past three years. For instance, the first scheme had an actual length of 360mm, a cost of 16 yuan, and a transportation-compatible length of 350mm. The corresponding display size adaptation data is |360-340| / 80=0.25, the cost adaptation data is 0.25, and the transportation adaptation data is 0.25. The width and height dimensions were calculated using the same method and then averaged to obtain the complete three-dimensional adaptation data for each scheme.

[0030] Secondly, based on the adaptation data of various dimensions of historical gift design schemes, the historical adaptation contribution of each gift size requirement feature is calculated and normalized to obtain the historical average weight. The historical adaptation contribution refers to the degree to which a certain requirement feature contributes to the overall adaptation effect across all historical schemes. The historical average weight is the normalized historical adaptation contribution, reflecting the average importance of each requirement feature in the historical schemes.

[0031] Specifically, first calculate the average fit value of each requirement feature in all historical solutions, then calculate the historical fit contribution: historical fit contribution of a feature = average fit value of the feature / sum of the average fit values ​​of all features. Finally, normalize the historical fit contribution to obtain the historical average weight.

[0032] For example, based on statistical calculations of historical gift design schemes, the average adaptation value for display size requirements is 0.30, the average adaptation value for cost requirements is 0.20, and the average adaptation value for transportation requirements is 0.10. The historical adaptation contribution of display size requirements is 0.30 / 0.60 = 0.5, the historical adaptation contribution of cost requirements is 0.20 / 0.60 ≈ 0.3333, and the historical adaptation contribution of transportation requirements is 0.10 / 0.60 ≈ 0.1667. After normalization, the historical average weights are: display size requirement weight 0.5, cost requirement weight 0.3333, and transportation requirement weight 0.1667, with the sum of the three being 1.

[0033] Secondly, based on the gift size requirement feature set, sensitivity analysis is performed on each requirement feature to calculate the sensitivity coefficient of each gift size requirement feature. The sensitivity coefficient is the quantitative value of the impact of a small change in the size design parameter on the fit of the corresponding requirement feature while keeping other requirement features unchanged. The larger the sensitivity coefficient, the more sensitive the requirement feature is to changes in the size design parameter.

[0034] Specifically, a small perturbation value Δx is set for the dimensional design parameters. Δx is a preset proportion of the baseline value of the dimensional design parameters, used to simulate small changes in the dimensional parameters. Positive perturbations of Δx are applied sequentially to the dimensional design parameter dimensions corresponding to the displayed dimensional requirements, cost requirements, and transportation requirements, while keeping the dimensional design parameters corresponding to other requirement features unchanged. The change in single-dimensional fit of each requirement feature after the perturbation is calculated respectively. (Display size requirements) (Cost requirements) (Transportation demand), Change in single-dimensional fit = Single-dimensional fit after disturbance - Single-dimensional fit before disturbance; Sensitivity coefficients for each demand characteristic are calculated using the partial derivative method, with the formula: [Show size demand sensitivity coefficient] Cost demand sensitivity coefficient Transportation demand sensitivity coefficient .

[0035] For example, using the median value of each requirement as the baseline value for the size design parameters: display size baseline value 340mm, cost baseline value 14 yuan, and transportation size baseline value 340mm. Selecting 5% of the baseline value as a preset ratio, we obtain a small disturbance value Δx: display size Δx = 340mm × 5% = 17mm; cost Δx = 0.7 yuan; transportation size Δx = 17mm. Applying positive disturbances sequentially, keeping other parameters unchanged: the display size after disturbance is 357mm; the cost after disturbance is 14.7 yuan; the transportation size after disturbance is 357mm. The adaptability of each single dimension before disturbance is: display adaptability = |340-340| / 80 = 0.00; cost adaptability = |14-14| / 8 = 0.00; transportation adaptability = |340-340| / 40 = 0.00. After perturbation, the adaptability of each dimension is as follows: Display adaptability = |357-340| / 80 = 0.2125; Cost adaptability = |14.7-14| / 8 = 0.0875; Transportation adaptability = |357-340| / 40 = 0.425. Calculate the change in adaptability of each dimension: =0.2125; =0.0875; =0.425. Calculate the sensitivity coefficient: =0.2125 / 17=0.0125; =0.0875 / 0.7=0.125; =0.425 / 17=0.025.

[0036] Furthermore, the sensitivity coefficients of each demand feature are normalized to obtain the sensitivity weights of each gift size demand feature. The sensitivity weight refers to the normalized value of the sensitivity coefficient, reflecting the degree of sensitivity importance of each feature under the current demand. The normalization formula is: Sensitivity Weight = Sensitivity Coefficient of a Demand Feature / ( + + ).

[0037] For example, normalizing the sensitivity coefficient yields the sensitivity weights: + + =0.0125+0.125+0.025=0.1625; Display size sensitivity weight = 0.0125 / 0.1625≈0.0769; Cost sensitivity weight = 0.125 / 0.1625≈0.7692; Transportation sensitivity weight = 0.025 / 0.1625≈0.1539.

[0038] Finally, the historical average weight and the sensitivity weight are weighted and summed to obtain the weight coefficients for each gift size requirement characteristic. The weighted sum of the historical average weight and the sensitivity weight yields the final weight coefficients for display size requirements, cost requirements, and transportation requirements. During the weighted summation, the historical average weight and the sensitivity weight each account for 50%, i.e., the final weight = 0.5 × historical average weight + 0.5 × sensitivity weight.

[0039] For example, the known historical average weights are: display size demand weight 0.5, cost demand weight 0.3333, and transportation demand weight 0.1667; the sensitivity weights are: display size sensitivity weight 0.0769, cost sensitivity weight 0.7692, and transportation sensitivity weight 0.1539. The final weight for display size = 0.5 × 0.5 + 0.5 × 0.0769 = 0.28845, the final weight for cost = 0.5 × 0.3333 + 0.5 × 0.7692 = 0.55125, and the final weight for transportation = 0.5 × 0.1667 + 0.5 × 0.1539 = 0.1603.

[0040] In this embodiment of the invention, historical average weights are calculated using historical scheme data, and sensitivity weights are obtained by combining them with sensitivity analysis. The two are then weighted to obtain an objective weight coefficient, eliminating the subjectivity of manually setting weights. This allows for the precise quantification of the importance of the three types of requirements: display, cost, and transportation. It provides a scientific, reasonable, and practical weight basis for constructing a collaborative optimization objective function for gift sizes.

[0041] S300: Based on the gift size requirement feature set and the weight coefficient, construct a gift size collaborative optimization objective function, with size design parameters as independent variables and maximizing the comprehensive matching degree as the optimization objective, while setting cost constraint threshold and material size matching threshold as optimization constraints.

[0042] In this embodiment of the invention, a collaborative optimization objective function for gift size is constructed based on the gift size requirement feature set and the weighting coefficients. Size design parameters are used as independent variables, and maximizing the overall matching degree is the optimization objective. Cost constraint thresholds and material size matching thresholds are set as optimization constraints. Gift size design is a multi-objective optimization problem, with interrelationships among the three dimensions of display effect, cost control, and transportation adaptability: pursuing a better display effect may lead to larger sizes, thereby increasing material costs and reducing transportation loading rates; while excessively compressing sizes may reduce costs, it may affect the aesthetics of the display or the stability of transportation stacking. Therefore, it is necessary to construct a collaborative optimization objective function that weights and fuses the matching degrees of the three dimensions according to the weighting coefficients to form a unified overall matching degree index. Simultaneously, two hard constraints—cost and material adaptability—are set to ensure the feasibility of the design scheme in actual production. This objective function will serve as the evaluation basis for subsequent optimization solutions, guiding the search algorithm to iterate towards the direction of optimal overall performance.

[0043] Step S300 in the method provided in this embodiment of the invention includes: The quantitative values ​​of the length, width, and height of the gift are used as size design parameters and as independent variables of the objective function for optimization. Calculate the matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements in one dimension; Based on the weight coefficients of the size requirements of each gift, the single-dimensional matching degree is weighted and summed to obtain the objective function for collaborative optimization of gift size; The cost constraint threshold is set as the maximum cost quantification value corresponding to the gift size design, and the material size matching threshold is the minimum fit quantification value between the gift size and the standard size of the processed material, as optimization constraints.

[0044] First, the quantified values ​​of the gift's length, width, and height are used as size design parameters, serving as independent variables in the objective function for optimization. Size design parameters refer to the specific quantified values ​​of the gift packaging's length, width, and height. These are adjustable and optimizable parameters in the subsequent optimization process, directly determining the final design scheme of the gift's dimensions. The independent variables are the quantified values ​​of the gift packaging's length, width, and height. The optimal solution, maximizing the overall matching degree, is then found by adjusting these parameters. For example, the size design parameters might be: length = 350mm, width = 290mm, height = 150mm.

[0045] Secondly, calculate the matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements in one dimension.

[0046] This includes calculating the single-dimensional matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements, including: Calculate the deviation between the size design parameter and the median value of the display size requirement, divide the deviation by the range of the display size requirement to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional adaptation value corresponding to the size design parameter, and filter the minimum value among the single-dimensional adaptation values ​​as the single-dimensional matching degree of the display size requirement. Calculate the deviation between the cost corresponding to the size design parameter and the median value of the cost requirement. Divide the deviation by the range of the cost requirement to obtain the deviation percentage corresponding to the size design parameter. Subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter. Filter the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the cost requirement. Calculate the deviation between the size design parameter and the median value of the transportation size requirement, divide the deviation by the range of the transportation size requirement to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter, and select the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the transportation requirement.

[0047] First, calculate the deviation between the size design parameter and the median value of the display size requirement. Divide the deviation by the range of the display size requirement to obtain the deviation percentage corresponding to the size design parameter. Subtract the deviation percentage from 1 to obtain the single-dimensional adaptation value corresponding to the size design parameter. Filter the minimum value among the single-dimensional adaptation values ​​as the single-dimensional matching degree of the display size requirement.

[0048] Among them, the display size deviation is the difference between the median value of the size design parameters and the display size requirement. The display size deviation ratio is the ratio of the display size deviation to the range of the display size requirement, used to quantify the degree of deviation between the size design parameters and the median value of the display requirement. The single-dimensional adaptation value corresponding to the size design parameters = 1 - display size deviation ratio; the closer the value is to 1, the better the size design parameters fit the display size requirement. The single-dimensional matching degree of the display size is the maximum value selected from the single-dimensional adaptation values ​​of the length, width, and height dimensions, used to characterize the overall adaptation level of the size design parameters to the display requirements.

[0049] Specifically, first calculate the deviation between the size design parameters and the median value of the display size requirement, divide the deviation by the range of the display size requirement to obtain the deviation percentage of the corresponding dimension; then subtract the deviation percentage from 1 to obtain the single-dimensional adaptation value of the corresponding dimension; finally, select the minimum value from the single-dimensional adaptation values ​​of the length, width and height dimensions, and the minimum value is the single-dimensional matching degree of the display size requirement.

[0050] For example, the display size requirements are: length 300mm-380mm, midpoint 340mm, range 80mm; width 240mm-320mm, midpoint 280mm, range 80mm; height 100mm-180mm, midpoint 140mm, range 80mm; size design parameters: length 350mm, width 290mm, height 150mm. Length deviation = 350-340=10mm, length deviation percentage = 10 / 80 = 0.125, length single-dimensional adaptation value = 1-0.125 = 0.875; similarly, we calculate: width single-dimensional adaptation value = 0.875, height single-dimensional adaptation value = 0.875; display size single-dimensional matching degree = min(0.875, 0.875, 0.875) = 0.875.

[0051] Next, calculate the deviation between the cost corresponding to the size design parameter and the median value of the cost requirement. Divide the deviation by the range of the cost requirement to obtain the deviation percentage corresponding to the size design parameter. Subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter. Filter the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the cost requirement.

[0052] Here, cost deviation is the difference between the actual cost of a single mooncake gift box corresponding to the size design parameters and the median cost requirement. Cost deviation percentage is the ratio of cost deviation to the range of cost requirement intervals, used to quantify the degree of deviation between the actual cost and the median cost requirement. Cost single-dimensional fit value = 1 - cost deviation percentage; the closer the value is to 1, the closer the actual cost is to the cost requirement. Cost single-dimensional matching degree refers to the minimum value selected from the cost single-dimensional fit values ​​corresponding to the length, width, and height dimensions, used to characterize the overall fit level between the cost corresponding to the size design parameters and the cost requirement.

[0053] Specifically, first calculate the deviation between the actual cost of a single mooncake gift box and the median cost requirement corresponding to the size design parameters. Divide this deviation by the range of the cost requirement interval to obtain the deviation percentage of the corresponding dimension. Then subtract this deviation percentage from 1 to obtain the single-dimensional fit value of the corresponding dimension. Finally, select the minimum value from the single-dimensional fit values ​​of the three dimensions. This minimum value is the single-dimensional matching degree of the cost requirement.

[0054] For example, the cost requirement is 10-18 yuan, with a median of 14 yuan and a range of 8 yuan. The actual cost of a single mooncake gift box corresponding to the size design parameters is: 15 yuan for length, 14.8 yuan for width, and 15.2 yuan for height. The length cost deviation is 15-14=1 yuan, the length cost percentage is 1 / 8=0.125, and the length adaptation value is 1-0.125=0.875. Similarly, the width adaptation value is 0.9, and the height adaptation value is 1-0.15=0.85. The cost single-dimensional matching degree is min(0.875,0.9,0.85)=0.85.

[0055] Next, calculate the deviation between the size design parameter and the median value of the transportation size requirement, divide the deviation by the range of the transportation size requirement interval to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter, and select the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the transportation requirement.

[0056] Among them, the transportation size deviation is the difference between the dimensional design parameters and the median value of the transportation size requirements. The transportation size deviation ratio is the ratio of the transportation size deviation to the range of the transportation size requirements, used to quantify the degree of deviation between the dimensional design parameters and the median value of transportation requirements. The single-dimensional adaptation value of transportation = 1 - the transportation size deviation ratio; the closer the value is to 1, the better the dimensional design parameters fit the transportation size requirements. The single-dimensional matching degree of transportation size is the minimum value selected from the single-dimensional adaptation values ​​of the length, width, and height dimensions, used to characterize the overall adaptation level of the dimensional design parameters to the transportation requirements.

[0057] Specifically, first calculate the deviation between the size design parameters and the median value of the transportation size requirements, divide the deviation by the range of the transportation size requirements to obtain the deviation percentage of the corresponding dimension; then subtract the deviation percentage from 1 to obtain the single-dimensional fit value of the corresponding dimension; finally, select the minimum value from the single-dimensional fit values ​​of the three dimensions, and the minimum value is the single-dimensional matching degree of the transportation size requirements.

[0058] For example, the transportation size requirements are: length 320mm-360mm, midpoint 340mm, range 40mm; width 260mm-300mm, midpoint 280mm, range 40mm; height 120mm-160mm, midpoint 140mm, range 40mm; size design parameters: length 350mm, width 290mm, height 150mm. Length deviation = 350-340=10mm, length deviation percentage = 10 / 40 = 0.25, length single-dimensional adaptation value = 1-0.25 = 0.75; similarly, we can obtain: width single-dimensional adaptation value = 0.75, height single-dimensional adaptation value = 0.75; transportation size single-dimensional matching degree = min(0.75, 0.75, 0.75) = 0.75.

[0059] Next, based on the weighting coefficients of each gift size requirement feature, the single-dimensional matching degree is weighted and summed to obtain the gift size collaborative optimization objective function. The single-dimensional matching degrees of size requirement, cost requirement, and transportation requirement are multiplied one by one by the final weighting coefficients of the corresponding requirement features and then summed to obtain the gift size collaborative optimization objective function. Maximizing this function value is taken as the size optimization objective.

[0060] For example, the objective function for collaborative optimization of gift size is: Overall matching degree F = Display size requirement weight × Display size single-dimensional matching degree + Cost requirement weight × Cost single-dimensional matching degree + Transportation requirement weight × Transportation size single-dimensional matching degree = 0.28845 × 0.875 + 0.55125 × 0.85 + 0.1603 × 0.75 = 0.8412.

[0061] Finally, a cost constraint threshold is set as the maximum quantified cost corresponding to the gift size design, and a material size matching threshold is set as the minimum quantified fit between the gift size and the standard size of the processed material, serving as optimization constraints. The cost constraint threshold is set as the maximum allowable cost corresponding to the gift size design, and the material size matching threshold is set as the minimum fit between the gift size and the standard size of the processed material. These two thresholds are used as constraints for size optimization; only size design parameters that satisfy these constraints are considered valid solutions.

[0062] For example, the cost constraint threshold is set at a maximum cost requirement of 18 yuan, meaning the cost of a single gift is ≤18 yuan; the material size matching threshold is set at a minimum matching value of 0.8, meaning the matching value between the gift size and the standard material size is ≥0.8.

[0063] In this embodiment of the invention, the requirements of display, cost, and transportation are quantified into calculable single-dimensional matching degrees. The method of taking the minimum value is used to ensure that the length, width, and height parameters within each dimension meet the requirements, avoiding the problem of excellent performance in one dimension but serious deviation in other dimensions. The single-dimensional matching degrees are weighted and summed based on the weight coefficients calculated by S200 to form a comprehensive matching degree index, realizing the effective transformation of multi-objective optimization problem into single-objective optimization problem. The weight coefficients reflect the differences in importance of each requirement feature, so that the optimization results can accurately reflect user preferences. Cost constraint thresholds and material size matching thresholds are set as hard constraints to ensure that all candidate design schemes not only perform well in requirement matching, but also meet the feasibility requirements of actual production costs and production processes.

[0064] S400: Obtain the gift size design range, iterate through and solve the collaborative optimization objective function, obtain multiple candidate size design schemes that satisfy all constraints, and the comprehensive matching degree corresponding to each candidate scheme.

[0065] In this embodiment of the invention, a gift size design range is obtained, and the collaborative optimization objective function is solved iteratively to obtain multiple candidate size design schemes that satisfy all constraints, along with the comprehensive matching degree corresponding to each candidate scheme. After constructing the collaborative optimization objective function and constraints, a scientific solution method is needed to select candidate schemes that satisfy all constraints and have a high comprehensive matching degree from a reasonable size design range. Therefore, this step employs the particle swarm optimization method, combined with similarity deduplication and step size optimization strategies, to efficiently traverse and solve the objective function, obtaining multiple high-quality candidate size design schemes that satisfy the constraints, providing a comprehensive and reliable basis for subsequent selection of the optimal scheme.

[0066] Step S400 in the method provided in this embodiment of the invention includes: Based on the cost constraint threshold and material size matching threshold, the gift size design range is obtained; Based on the gift size design range, multiple size design parameters are randomly generated, with each size design parameter being treated as a particle. The number of size design parameters generated is obtained based on the range of the gift size design range. Initialize the particle's position and velocity. The particle position is the quantized value of the size design parameter, and the particle velocity is the adjustment step size of the size design parameter. The comprehensive matching degree calculated by the collaborative optimization objective function is used as the fitness value of the particle. The initial comprehensive matching degree value of each particle is calculated, and particles that meet the cost constraint threshold and material size matching threshold are selected as effective particles. The particle swarm is iteratively optimized until a preset number of iterations is reached; Get the gift size designs corresponding to multiple valid particles with a comprehensive matching degree greater than the comprehensive matching threshold, and use them as candidate size design schemes.

[0067] First, based on the cost constraint threshold and material size matching threshold, the gift size design range is obtained. The gift size design range is a reasonable range of values ​​for the size design parameters determined by combining the cost constraint threshold, the material size matching threshold, and the display and transportation size requirement range in S100. It is the basis for subsequent particle generation and traversal solution, and all size design parameters must be within this range.

[0068] Specifically, the following three conditions are considered to determine the respective size design ranges for length, width, and height: the display size requirement range in S100; the transportation size requirement range in S100; and the size range corresponding to the material size matching threshold. The intersection of these three factors is taken as the final gift size design range, ensuring that the size parameters within the range meet the basic requirements and constraints.

[0069] For example, the constraints and requirement ranges are as follows: Display size range: length 300mm-380mm, width 240mm-320mm, height 100mm-180mm; Transportation size range: length 320mm-360mm, width 260mm-300mm, height 120mm-160mm; Material size matching threshold 0.8, corresponding to standard cardboard 400mm×300mm, the suitable size range is length ≤360mm, width ≤280mm, height ≤160mm; Taking the intersection of the three, we get the gift size design range: length: 320mm-360mm; width: 260mm-280mm; height: 120mm-160mm.

[0070] Secondly, based on the gift size design range, multiple size design parameters are randomly generated, with each size design parameter treated as a particle. The number of generated size design parameters is determined by the range of the gift size design range. Each randomly generated size design parameter is analogous to a particle, and each particle corresponds to a potential gift size design scheme. The number of particles is determined by the range of the size design range, ensuring reasonable values ​​are covered within the range.

[0071] Specifically, the range of the dimension design intervals for each dimension is calculated; the number of particles is positively correlated with the range, the larger the range, the more particles there are, ensuring the comprehensiveness of the traversal. For example, the number of particles increases by 5 for every 20mm increase in the range; based on the dimension design interval, a corresponding number of dimension design parameter combinations are randomly generated, and each combination is treated as a particle.

[0072] The example shows the design range for each dimension: length 360-320=40mm, width 280-260=20mm, height 160-120=40mm. The number of particles is determined according to the set rules: a length range of 40mm corresponds to 10 particles, a width range of 20mm corresponds to 5 particles, and a height range of 40mm corresponds to 10 particles. A total of 10 particles are generated. Some of the particles are as follows: Particle 1: length 330mm, width 270mm, height 130mm; Particle 2: length 345mm, width 265mm, height 145mm; Particle 3: length 350mm, width 275mm, height 150mm.

[0073] Next, the particle position and velocity are initialized. The particle position is the quantized value of the size design parameter, and the particle velocity is the adjustment step size of the size design parameter. The particle position, i.e., the size design parameter corresponding to that particle, directly represents a potential gift size design scheme. The particle velocity, as the adjustment step size of the size design parameter, is used to adjust the particle position in subsequent iterations, i.e., to adjust the size parameter. The larger the step size, the greater the adjustment range of the size parameter, and the faster the iteration speed.

[0074] Specifically, particle position: the size design parameters corresponding to each particle are directly used as the initial position; particle velocity: based on the size design range setting, 5% of the range of each dimension is taken as the initial velocity to ensure that the adjustment range is reasonable and avoids that too large a range will lead to unstable iteration or too small a range will lead to low iteration efficiency.

[0075] For example, taking particle 3 as an example: Initial particle position: (350, 275, 150); Initial velocity in each dimension: Length: 40mm × 5% = 2mm, that is, it can be adjusted by ±2mm in each iteration; Width: 20mm × 5% = 1mm; Height: 40mm × 5% = 2mm; Other particles are initialized with positions and velocities according to the same rules.

[0076] Furthermore, the overall matching degree calculated by the collaborative optimization objective function is used as the fitness value of the particles. An initial overall matching degree value is calculated for each particle, and particles that satisfy the cost constraint threshold and the material size matching threshold are selected as valid particles. The fitness value refers to the overall matching degree calculated by the collaborative optimization objective function as the particle's fitness value; a higher fitness value indicates a better overall fit for the particle. Valid particles are those that satisfy both the cost constraint threshold and the material size matching threshold; invalid particles are discarded and do not participate in subsequent iterations.

[0077] Specifically, for each particle, calculate its matching degree in display, cost, and transportation dimensions using the method described above, and then calculate the overall matching degree, i.e., the fitness value, by combining the weighting coefficients; check whether each particle meets the constraints: single-unit cost ≤ 18, material fit value ≥ 0.8; filter out particles that meet both constraints simultaneously as valid particles.

[0078] For example, taking particle 3 as an example: Calculate the single-dimensional matching degree: display matching degree = 0.875, cost matching degree = 0.85, transportation matching degree = 0.75; Calculate the fitness value: F = 0.28845 × 0.875 + 0.55125 × 0.85 + 0.1603 × 0.75 ≈ 0.8412; Check the constraints: cost: the cost per unit for this size is 15.5 ≤ 18; material compatibility value: the compatibility value between this size and standard cardboard 400mm × 300mm is 0.89 ≥ 0.8, which satisfies the material constraint; therefore, particle 3 is a valid particle. Similarly, calculate whether the remaining particles are valid particles.

[0079] Subsequently, the particle swarm is iteratively optimized until a preset number of iterations is reached.

[0080] The process of iteratively optimizing the particle swarm until a certain number of iterations is reached includes: Record the historical best fitness value and corresponding position of each effective particle as the individual best, record the best fitness value and corresponding position of all effective particles in the particle swarm as the global best, update the position and velocity of effective particles based on the individual best and the global best, iteratively calculate the collaborative optimization objective function value of the particles as the comprehensive matching degree, and synchronously iteratively calculate the matching degree of each single dimension corresponding to each particle. During the iteration process, high-matching particle subsets are selected based on the single-dimensional matching degree of display size requirements, the single-dimensional matching degree of cost requirements, and the single-dimensional matching degree of transportation requirements. The similarity between any two valid particles is calculated within and between each subset. The similarity is the weighted sum of the dimensionless deviation value of the particle position parameter and the dimensionless deviation value of the fitness value. Particle combinations with a similarity greater than a preset similarity threshold are selected, wherein the particle combination includes two particles; Discard particles with low overall matching degree in the combination, retain particles with high overall matching degree, and obtain the optimization adjustment step size of the retained particles.

[0081] First, the historical best fitness value and corresponding position of each effective particle are recorded as the individual optimum. The best fitness value and corresponding position of all effective particles in the particle swarm are recorded as the global optimum. Based on the individual optimum and the global optimum, the position and velocity of the effective particles are updated. The collaborative optimization objective function value of the particles is calculated iteratively as the comprehensive matching degree. The matching degree of each single dimension corresponding to each particle is calculated simultaneously iteratively.

[0082] Here, individual optimality refers to the highest fitness value and corresponding particle position reached by each effective particle during iteration. Global optimality refers to the highest fitness value and corresponding particle position that appears during iteration of all effective particles, representing the optimal solution among all current schemes. Iterative computation refers to the process of continuously adjusting particle velocity and position, repeatedly calculating fitness values, and gradually approaching the global optimal solution.

[0083] Specifically, initialize individual optimality: use the initial fitness value of each effective particle as its initial individual optimal fitness value, and the initial position as its individual optimal position; initialize global optimality: select the maximum value from the initial fitness values ​​of all effective particles as the initial global optimal fitness value, and the corresponding particle position as the global optimal position; based on individual optimality and global optimality, update the velocity and position of each effective particle, using the following formulas: new velocity = initial velocity × preset coefficient 1 + individual optimal deviation × preset coefficient 2 + global optimal deviation × preset coefficient 3; new position = current position + new velocity; for particles with updated positions, recalculate their single-dimensional matching degree and comprehensive matching degree, and update individual optimality and global optimality; repeat the above steps until the number of iterations reaches the preset value.

[0084] For example, effective particles 1, 3, and 5 have initial fitness values ​​of 0.82, 0.8412, and 0.83, respectively. Initialization of individual and global optima: Individual optima: Particle 1 (0.82, 330, 270, 130), Particle 3 (0.8412, 350, 275, 150), Particle 5 (0.83, 355, 280, 155); Global optima: Particle 3; First iteration update (preset coefficient 1 = 0.5, preset coefficient 2 = 0.3, preset coefficient 3 = 0.2): Particle 3's initial velocity is (2, 1, 2), individual optima deviation = 0, global optima deviation = 0; New velocity = 2 × 0.5 + 0 × 0.3 + 0 × 0.2 = 1 mm; 1 × 0.5 + 0 + 0 = 0.5 mm; 2 × 0.5 + 0 + 0 = 1 mm; New position = 350 + 1 = 351 mm, 275 + 0.5 ≈ 276 mm, 150 + 1 = 151 mm; Recalculate the fitness value of particle 3's new position: display matching degree = 0.86, cost matching degree = 0.84, transportation matching degree = 0.74, overall matching degree ≈ 0.835, which is lower than the initial individual optimum, so the individual optimum is not updated; Repeat the iteration 10 times, and after the final update, the global optimum is still the initial position of particle 3, and the individual optima of particles 1 and 5 are updated to 0.83 and 0.835 respectively.

[0085] Secondly, during the iteration process, high-matching particle subsets are selected based on the single-dimensional matching degree of display size requirements, cost requirements, and transportation requirements. The similarity between any two valid particles is calculated within and between each subset. This similarity is the weighted sum of the dimensionless deviation of the particle position parameter and the dimensionless deviation of the fitness value. The high-matching particle subsets are defined as valid particles with a single-dimensional matching degree ≥ a preset threshold, selected based on the single-dimensional matching degree of display, cost, and transportation. These are categorized into three subsets: a high-matching subset for display, a high-matching subset for cost, and a high-matching subset for transportation. Particle similarity measures the degree of similarity between two valid particles and is the weighted sum of the dimensionless deviation of the particle position parameter and the dimensionless deviation of the fitness value. A larger value indicates higher similarity. The dimensionless deviation value is calculated as the absolute value of the difference between the two particle parameters divided by the corresponding dimensional range, ensuring that deviations across different dimensions and units are comparable.

[0086] Specifically, a single-dimensional high-match threshold is set, and three subsets are selected respectively. For any two valid particles within and between subsets, the following are calculated: the dimensionless deviation value of the position parameter, and the average of the three values ​​is taken; the dimensionless deviation value of the fitness value; similarity = 0.6 × average dimensionless deviation value of position + 0.4 × dimensionless deviation value of fitness. The specific weights can be adjusted according to actual needs.

[0087] For example, for effective particles 3 and 5: High-match subsets are selected: the two particles exhibit matching degrees (0.875, 0.86), cost matching degrees (0.85, 0.84), and transportation matching degrees (0.75, 0.74). Particles with display and cost matching degrees ≥ 0.8 belong to the high-match subset for display and cost. The dimensionless positional deviation is calculated as follows: Length: |350-355| / 40=0.125; Width: |275-280| / 20=0.25; Height: |150-155 | / 40=0.125; Average position dimensionless deviation = (0.125+0.25+0.125) / 3≈0.167; Calculate the dimensionless deviation value of fitness: The minimum fitness value of all effective particles is 0.82, fitness range = 0.8412-0.82=0.0212; Dimensionless deviation value of fitness = |0.8412-0.835| / 0.0212≈0.292; Similarity = 0.6×0.167+0.4×0.292≈0.217.

[0088] Next, particle combinations with a similarity greater than a preset similarity threshold are selected. Each particle combination consists of two particles. The preset similarity threshold is a critical value used to determine whether two particles are too similar; it ranges from 0 to 1, for example, 0.2, and can be adjusted as needed. Two particles with a similarity greater than this threshold are considered too similar and form a particle combination. A particle combination refers to a combination of two valid particles with a similarity greater than the preset threshold, used for subsequent deduplication.

[0089] For example, with a preset similarity threshold of 0.2: the similarity between particle 3 and particle 5 is approximately 0.217 > 0.2, forming particle combination 1: (particle 3, particle 5); the similarity between particle 1 and particle 6 is approximately 0.205 > 0.2, forming particle combination 2: (particle 1, particle 6); the similarity between particle 3 and particle 1 is approximately 0.18 < 0.2, and no combination is formed.

[0090] Furthermore, particles with low overall matching scores are discarded from the combinations, while particles with high overall matching scores are retained, and the optimization adjustment step size for the retained particles is obtained. Particle combination deduplication refers to discarding particles with low overall matching scores and retaining particles with high overall matching scores for each highly similar particle combination, avoiding redundant similar schemes and improving the diversity and quality of candidate schemes. For each particle combination, the overall matching scores of the two particles are compared, and particles with lower overall matching scores are directly discarded, retaining only particles with higher overall matching scores. The retained particles continue to participate in subsequent processes.

[0091] For example, particle combination 1 (particle 3, particle 5): particle 3 has a higher overall matching degree, so particle 5 is discarded and particle 3 is retained; particle combination 2 (particle 1, particle 6): particle 1 has a higher overall matching degree, so particle 6 is discarded and particle 1 is retained; after deduplication, the effective particles are: particle 3, particle 1, and other effective particles that do not form a high similarity combination.

[0092] The optimization adjustment step size for obtaining the retained particles includes: The adjustment coefficient is obtained by combining the ratio of the overall matching degree of discarded particles to the overall matching degree of retained particles with the proportion of single-dimensional similarity between discarded and retained particles. The optimized adjustment step size is obtained by multiplying the adjustment coefficient by the adjustment step size of the retained particles.

[0093] First, an adjustment coefficient is obtained based on the ratio of the overall matching degree of discarded particles to that of retained particles, combined with the proportion of single-dimensional similarity between discarded and retained particles. This adjustment coefficient is used to optimize the particle adjustment step size. Calculated based on the ratio of the overall matching degree of discarded and retained particles and the proportion of single-dimensional similarity, it allows for more targeted adjustment of the retained particle step size, improving the efficiency of subsequent iterations. The proportion of single-dimensional similarity refers to the ratio of the similarity values ​​of discarded and retained particles in the three single dimensions of display, cost, and transportation to their overall similarity, reflecting the impact of each single dimension on particle similarity.

[0094] Specifically, calculate the overall matching ratio of discarded particles to retained particles: Ratio = Overall matching ratio of discarded particles / Overall matching ratio of retained particles; calculate the similarity between two particles in three single dimensions: display, cost, and transportation; calculate the proportion of similarity in a single dimension: Proportion of similarity in a single dimension = Similarity in that single dimension / Sum of similarity in the three single dimensions; Adjustment coefficient = Overall matching ratio × (Display similarity proportion × 0.3 + Cost similarity proportion × 0.5 + Transportation similarity proportion × 0.2).

[0095] For example, taking particle combination 1 as an example: the overall matching ratio = 0.835 / 0.8412≈0.993; calculating single-dimensional similarity: display single-dimensional similarity ≈0.115; cost single-dimensional similarity ≈0.200; transportation single-dimensional similarity ≈0.115; single-dimensional similarity ratio: display similarity ratio = 0.115 / (0.115+0.200+0.115)≈0.288; cost similarity ratio ≈0.465; transportation similarity ratio ≈0.247; adjustment coefficient = 0.993×(0.288×0.3+0.465×0.5+0.247×0.2)≈0.365.

[0096] Secondly, the optimized adjustment step size is obtained by multiplying the adjustment coefficient by the adjustment step size of the retained particle. The optimized adjustment step size is the new adjustment step size obtained by multiplying the adjustment coefficient by the initial adjustment step size of the retained particle. This allows the adjustment range of the retained particle in subsequent iterations to better match the actual needs and avoid ineffective adjustments. Optimized adjustment step size = adjustment coefficient × initial adjustment step size of the retained particle.

[0097] For example, taking retained particle 3 as an example, the optimization adjustment step size is as follows: Initial step size: length 2mm, width 1mm, height 2mm; Optimization adjustment step size = 0.365 × initial step size, that is, length ≈ 0.73mm, width ≈ 0.365mm, height ≈ 0.73mm.

[0098] Finally, gift size designs corresponding to multiple valid particles with a comprehensive matching degree greater than the comprehensive matching threshold are obtained as candidate size design schemes. The comprehensive matching threshold is a critical value used to screen high-quality candidate schemes, set as the average comprehensive matching degree of all valid particles, for example, 0.83. Valid particles with a comprehensive matching degree greater than this threshold are used as candidate size design schemes.

[0099] Specifically, a comprehensive matching threshold is set, and all retained particles with a comprehensive matching degree greater than the comprehensive matching threshold are selected. Their corresponding size design parameters are the candidate size design schemes, and the comprehensive matching degree of each scheme is recorded.

[0100] For example, the comprehensive matching threshold is: the average comprehensive matching degree of all valid particles is approximately 0.83; candidate schemes are selected as follows: Particle 3: comprehensive matching degree 0.8412 > 0.83, corresponding size (350mm, 275mm, 150mm), as candidate scheme 1; Particle 1: comprehensive matching degree 0.83 ≥ 0.83, corresponding size (330mm, 270mm, 130mm), as candidate scheme 2; Particle 4: comprehensive matching degree 0.825 < 0.83, not included in the candidate schemes; for example, three candidate size design schemes are finally obtained, and the comprehensive matching degree and single-dimensional matching degree of each scheme are recorded simultaneously for subsequent optimal scheme selection.

[0101] In this embodiment of the invention, a reasonable size design range is determined by using particle swarm optimization (PSO) combined with constraints. Particles are generated and iteratively optimized. Simultaneously, redundant schemes are eliminated through similarity deduplication, and iterative efficiency is improved by adjusting the step size. Finally, multiple candidate size design schemes that meet all constraints and have a high overall matching degree are selected. This solves the problems of low efficiency and omission of optimal solutions in traditional traversal methods, while ensuring that candidate schemes are high-quality, diverse, and feasible. It provides a comprehensive and reliable basis for subsequent selection of the optimal gift size design scheme, and standardizes and quantifies the optimization process, improving the efficiency and rationality of gift size design.

[0102] S500: Select the candidate size design scheme with the highest overall matching degree and single-dimensional matching degree to obtain the optimal size design scheme set for the gift.

[0103] In this embodiment of the invention, the candidate size design schemes with the highest overall matching degree and single-dimensional matching degree are selected to obtain the optimal size design scheme set for gifts. The candidate schemes differ in terms of single-dimensional matching degree and overall matching degree in terms of display, cost, and transportation. When enterprises actually purchase gifts, they may have different scenario requirements, and selecting only a single overall optimal scheme cannot meet the diverse needs of these scenarios. Therefore, this step filters the optimal schemes for each single dimension and the optimal scheme for overall balance to construct the optimal size design scheme set, providing enterprises with the best choice for multiple scenarios and ensuring that the scheme set is both targeted and diverse.

[0104] Step S500 in the method provided in this embodiment of the invention includes: For all candidate size design schemes, extract the matching degree of each scheme in terms of display dimension, cost dimension, transportation dimension, and overall matching degree. Based on the multi-dimensional parameter set of the candidate solutions, the three single-dimensional optimal solutions with the highest matching degree in display, cost, and transportation were selected and added to the gift optimal size design solution set. The candidate size design scheme with the highest overall matching degree is selected as the optimal overall balance scheme and added to the set of optimal gift size design schemes.

[0105] First, for all candidate size design schemes, extract the single-dimensional matching degree for display, cost, transportation, and overall matching degree for each scheme. The multi-dimensional matching degree parameter refers to the single-dimensional matching degree for display size, cost requirement, and transportation size for each candidate size design scheme, and it serves as the basis for selecting the optimal scheme. Iterate through all candidate size design schemes selected by S400, and extract the single-dimensional matching degree for display, cost, and transportation for each scheme.

[0106] For example, candidate solution 1 (350, 275, 150) shows a matching degree of 0.875, a cost matching degree of 0.85, and a transportation matching degree of 0.75; candidate solution 2 (330, 27, 130) shows a matching degree of 0.85, a cost matching degree of 0.88, and a transportation matching degree of 0.78; and candidate solution 3 (340, 272, 140) shows a matching degree of 0.89, a cost matching degree of 0.84, and a transportation matching degree of 0.76.

[0107] Secondly, based on the multi-dimensional parameter set of the candidate solutions, the three single-dimensional optimal solutions with the highest matching degree in display, cost, and transportation are selected and added to the set of optimal gift size design solutions. The single-dimensional optimal solution refers to the candidate solution with the highest matching degree in one of the three dimensions of display, cost, and transportation, corresponding to the three categories of optimal display, optimal cost, and optimal transportation solutions, respectively.

[0108] Specifically, the matching degree of each candidate solution is compared in terms of presentation, cost, and transportation. The candidate solution with the highest matching degree is selected as the optimal solution for the corresponding dimension. If multiple candidate solutions have the same and the highest matching degree in a certain dimension, they are all included in the optimal solution for the corresponding dimension.

[0109] For example, comparing the multi-dimensional matching degree of the three candidate solutions: the highest matching degree is candidate solution 3, so candidate solution 3 is the best solution in terms of display; the highest matching degree is candidate solution 2, so candidate solution 2 is the best solution in terms of cost; the highest matching degree is candidate solution 2, so candidate solution 2 is the best solution in terms of transportation.

[0110] Finally, the candidate size design scheme with the highest overall matching degree is selected as the optimal balanced solution and added to the gift optimal size design scheme set. The optimal balanced solution refers to the candidate scheme with the highest overall matching degree, taking into account the adaptation needs of display, cost, and transportation, representing the best choice in a multi-dimensional balance. The gift optimal size design scheme set is a collection consisting of the three single-dimensional optimal solutions and the optimal balanced solution, providing enterprises with diverse optimal choices.

[0111] Specifically, the overall matching degree of all candidate solutions is compared, and the candidate solution with the highest overall matching degree is selected as the overall balanced optimal solution. The three single-dimensional optimal solutions selected above are integrated with the overall balanced optimal solution, and duplicate solutions are removed to finally form the set of optimal size design solutions for gifts.

[0112] For example, candidate solution 1 has the highest overall matching degree, therefore candidate solution 1 is the optimal solution in terms of overall balance. After integrating the optimal solution in one dimension and the optimal solution in terms of overall balance, and removing duplicate solutions, we obtain the set of optimal size design solutions for the gift, which includes 3 solutions: candidate solution 1, candidate solution 2, and candidate solution 3.

[0113] In this embodiment of the invention, by extracting multi-dimensional parameters of candidate solutions and filtering the optimal solutions for each single dimension and the optimal solutions for overall balance, a set of optimal size design solutions that is both targeted and diverse is constructed. The set of solutions includes both the optimal choices for each single dimension, which can meet the personalized needs of enterprises in different scenarios, and the optimal solutions for overall balance, which can serve as general optimal choices; thus providing enterprises with a comprehensive and reliable basis for finalizing the gift size design solution.

[0114] Through the specific implementation methods described above, the embodiments of the present invention achieve the following technical effects: This invention provides a method and system for optimizing gift design schemes by integrating multimodal user needs perception. It acquires and quantifies multimodal needs related to display, cost, and transportation, providing a realistic basis for subsequent optimization. By combining historical data and sensitivity analysis to calculate objective weights, it eliminates the subjectivity of manual weighting, ensuring accurate quantification of the importance of each need dimension. Based on these weights, it constructs a collaborative optimization objective function and sets constraints, clarifying the optimization direction and boundaries to avoid optimization results exceeding actual production and budget limitations. It employs particle swarm optimization combined with similarity deduplication and step size optimization strategies to efficiently traverse and solve the objective function, selecting high-quality candidate schemes that meet all constraints. It filters single-dimensional optimal and comprehensive balanced optimal schemes, constructing a diverse set of optimal schemes that considers both the personalized needs and multi-dimensional balanced needs of enterprises in different scenarios. This invention effectively improves the rationality, efficiency, and adaptability of gift size design, reduces design costs, and ensures that the final design scheme meets the needs of multimodal users.

[0115] Example 2, as Figure 2 As shown, this invention provides a gift design optimization system that integrates multimodal user demand perception. The system includes: The demand feature acquisition module 11 is used to acquire a gift size demand feature set, wherein the gift size demand feature set includes display size requirements, cost requirements and transportation requirements; The weighting coefficient calculation module 12 is used to calculate the weighting coefficient of each gift size requirement characteristic by combining sensitivity analysis with historical gift design schemes. The objective function construction module 13 is used to construct a collaborative optimization objective function for gift size based on the gift size requirement feature set and the weight coefficients, with size design parameters as independent variables and maximizing the comprehensive matching degree as the optimization objective, while setting cost constraint threshold and material size matching threshold as optimization constraints. The candidate solution acquisition module 14 is used to acquire the gift size design range, iterate through and solve the collaborative optimization objective function, and acquire multiple candidate size design schemes that satisfy all constraints, as well as the comprehensive matching degree corresponding to each candidate scheme. The optimal solution selection module 15 is used to select the candidate size design scheme with the highest comprehensive matching degree and single-dimensional matching degree to obtain the optimal size design scheme set for gifts.

[0116] In one embodiment, the demand feature acquisition module 11 is further configured to: Collect data on user gift usage scenarios and display space requirements to determine display size needs; Obtain the user's gift budget, and combine it with the unit price of gift materials and processing fee rates to obtain cost requirements; Quantitatively analyze historical gift procurement transportation and loading adaptation data to obtain transportation requirements; The display size requirements, cost requirements, and transportation requirements are integrated to form the gift size requirement feature set, wherein the display size requirements, cost requirements, and transportation requirements include the gift size requirement range.

[0117] In one embodiment, the weighting coefficient calculation module 12 is further configured to: Obtain historical gift design schemes and extract display size adaptation data, cost adaptation data, and transportation adaptation data from each scheme. Among them, the display size adaptation data is the quantified value of the deviation between the historical scheme size and the display size requirement, the cost adaptation data is the quantified value of the deviation between the historical scheme actual cost and the cost requirement, and the transportation adaptation data is the quantified value of the deviation between the historical scheme size and the transportation requirement. Based on the adaptation data of various dimensions of historical gift design schemes, the historical adaptation contribution of each gift size requirement characteristic is calculated and normalized to obtain the historical average weight. Based on the gift size demand feature set, sensitivity analysis is performed on each demand feature, and the sensitivity coefficient of each gift size demand feature is calculated. The sensitivity coefficients of each demand feature are normalized to obtain the sensitivity weights of each gift size demand feature. The weighted sum of the historical average weight and the sensitivity weight is used to obtain the weight coefficients of each gift size requirement feature.

[0118] In one embodiment, the objective function construction module 13 is further configured to: The quantitative values ​​of the length, width, and height of the gift are used as size design parameters and as independent variables of the objective function for optimization. Calculate the matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements in one dimension; Based on the weight coefficients of the size requirements of each gift, the single-dimensional matching degree is weighted and summed to obtain the objective function for collaborative optimization of gift size; The cost constraint threshold is set as the maximum cost quantification value corresponding to the gift size design, and the material size matching threshold is the minimum fit quantification value between the gift size and the standard size of the processed material, as optimization constraints.

[0119] This includes calculating the single-dimensional matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements, including: Calculate the deviation between the size design parameter and the median value of the display size requirement, divide the deviation by the range of the display size requirement to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional adaptation value corresponding to the size design parameter, and filter the minimum value among the single-dimensional adaptation values ​​as the single-dimensional matching degree of the display size requirement. Calculate the deviation between the cost corresponding to the size design parameter and the median value of the cost requirement. Divide the deviation by the range of the cost requirement to obtain the deviation percentage corresponding to the size design parameter. Subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter. Filter the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the cost requirement. Calculate the deviation between the size design parameter and the median value of the transportation size requirement, divide the deviation by the range of the transportation size requirement to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter, and select the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the transportation requirement.

[0120] In one embodiment, the candidate solution acquisition module 14 is further configured to: Based on the cost constraint threshold and material size matching threshold, the gift size design range is obtained; Based on the gift size design range, multiple size design parameters are randomly generated, with each size design parameter being treated as a particle. The number of size design parameters generated is obtained based on the range of the gift size design range. Initialize the particle's position and velocity. The particle position is the quantized value of the size design parameter, and the particle velocity is the adjustment step size of the size design parameter. The comprehensive matching degree calculated by the collaborative optimization objective function is used as the fitness value of the particle. The initial comprehensive matching degree value of each particle is calculated, and particles that meet the cost constraint threshold and material size matching threshold are selected as effective particles. The particle swarm is iteratively optimized until a preset number of iterations is reached; Get the gift size designs corresponding to multiple valid particles with a comprehensive matching degree greater than the comprehensive matching threshold, and use them as candidate size design schemes.

[0121] The process of iteratively optimizing the particle swarm until a certain number of iterations is reached includes: Record the historical best fitness value and corresponding position of each effective particle as the individual best, record the best fitness value and corresponding position of all effective particles in the particle swarm as the global best, update the position and velocity of effective particles based on the individual best and the global best, iteratively calculate the collaborative optimization objective function value of the particles as the comprehensive matching degree, and synchronously iteratively calculate the matching degree of each single dimension corresponding to each particle. During the iteration process, high-matching particle subsets are selected based on the single-dimensional matching degree of display size requirements, the single-dimensional matching degree of cost requirements, and the single-dimensional matching degree of transportation requirements. The similarity between any two valid particles is calculated within and between each subset. The similarity is the weighted sum of the dimensionless deviation value of the particle position parameter and the dimensionless deviation value of the fitness value. Particle combinations with a similarity greater than a preset similarity threshold are selected, wherein the particle combination includes two particles; Discard particles with low overall matching degree in the combination, retain particles with high overall matching degree, and obtain the optimization adjustment step size of the retained particles.

[0122] The optimization adjustment step size for obtaining the retained particles includes: The adjustment coefficient is obtained by combining the ratio of the overall matching degree of discarded particles to the overall matching degree of retained particles with the proportion of single-dimensional similarity between discarded and retained particles. The optimized adjustment step size is obtained by multiplying the adjustment coefficient by the adjustment step size of the retained particles.

[0123] In one embodiment, the optimal solution selection module 15 is further configured to: For all candidate size design schemes, extract the matching degree of each scheme in terms of display dimension, cost dimension, transportation dimension, and overall matching degree. Based on the multi-dimensional parameter set of the candidate solutions, the three single-dimensional optimal solutions with the highest matching degree in display, cost, and transportation were selected and added to the gift optimal size design solution set. The candidate size design scheme with the highest overall matching degree is selected as the optimal overall balance scheme and added to the set of optimal gift size design schemes.

[0124] It should be noted that 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 optimizing gift design schemes by integrating multimodal user demand perception, characterized in that, include: Obtain a gift size requirement feature set, wherein the gift size requirement feature set includes display size requirements, cost requirements, and transportation requirements; By combining sensitivity analysis with historical gift design schemes, the weighting coefficients of the size requirements for each gift were calculated. Based on the gift size requirement feature set and the weight coefficient, a collaborative optimization objective function for gift size is constructed, with size design parameters as independent variables and maximizing the comprehensive matching degree as the optimization objective. At the same time, cost constraint threshold and material size matching threshold are set as optimization constraints. Obtain the gift size design range, iterate through and solve the collaborative optimization objective function to obtain multiple candidate size design schemes that satisfy all constraints, and the comprehensive matching degree corresponding to each candidate scheme; Select the candidate size design scheme with the highest overall matching degree and single-dimensional matching degree to obtain the optimal size design scheme set for the gift.

2. The gift design optimization method based on multimodal user demand perception as described in claim 1, characterized in that, Obtain a gift size requirement feature set, wherein the gift size requirement feature set includes display size requirements, cost requirements, and transportation requirements, including: Collect data on user gift usage scenarios and display space requirements to determine display size needs; Obtain the user's gift budget, and combine it with the unit price of gift materials and processing fee rates to obtain cost requirements; Quantitatively analyze historical gift procurement transportation and loading adaptation data to obtain transportation requirements; The display size requirements, cost requirements, and transportation requirements are integrated to form the gift size requirement feature set, wherein the display size requirements, cost requirements, and transportation requirements include the gift size requirement range.

3. The gift design optimization method based on multimodal user demand perception as described in claim 1, characterized in that, By combining sensitivity analysis with historical gift design schemes, the weighting coefficients of the size requirements for each gift were calculated, including: Obtain historical gift design schemes and extract display size adaptation data, cost adaptation data, and transportation adaptation data from each scheme. Among them, the display size adaptation data is the quantified value of the deviation between the historical scheme size and the display size requirement, the cost adaptation data is the quantified value of the deviation between the historical scheme actual cost and the cost requirement, and the transportation adaptation data is the quantified value of the deviation between the historical scheme size and the transportation requirement. Based on the adaptation data of various dimensions of historical gift design schemes, the historical adaptation contribution of each gift size requirement characteristic is calculated and normalized to obtain the historical average weight. Based on the gift size demand feature set, sensitivity analysis is performed on each demand feature, and the sensitivity coefficient of each gift size demand feature is calculated. The sensitivity coefficients of each demand feature are normalized to obtain the sensitivity weights of each gift size demand feature. The weighted sum of the historical average weight and the sensitivity weight is used to obtain the weight coefficients of each gift size requirement feature.

4. The gift design optimization method based on multimodal user demand perception as described in claim 1, characterized in that, Based on the gift size requirement feature set and the weighting coefficients, a collaborative optimization objective function for gift size is constructed, using size design parameters as independent variables and maximizing the overall matching degree as the optimization objective. Simultaneously, cost constraint thresholds and material size matching thresholds are set as optimization constraints, including: The quantitative values ​​of the length, width, and height of the gift are used as size design parameters and as independent variables of the objective function for optimization. Calculate the matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements in one dimension; Based on the weight coefficients of the size requirements of each gift, the single-dimensional matching degree is weighted and summed to obtain the objective function for collaborative optimization of gift size; The cost constraint threshold is set as the maximum cost quantification value corresponding to the gift size design, and the material size matching threshold is the minimum fit quantification value between the gift size and the standard size of the processed material, as optimization constraints.

5. The gift design optimization method based on multimodal user demand perception as described in claim 4, characterized in that, Calculate the single-dimensional matching degree between the size design parameters and the display size requirements, cost requirements, and transportation requirements, including: Calculate the deviation between the size design parameter and the median value of the display size requirement, divide the deviation by the range of the display size requirement to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional adaptation value corresponding to the size design parameter, and filter the minimum value among the single-dimensional adaptation values ​​as the single-dimensional matching degree of the display size requirement. Calculate the deviation between the cost corresponding to the size design parameter and the median value of the cost requirement. Divide the deviation by the range of the cost requirement to obtain the deviation percentage corresponding to the size design parameter. Subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter. Filter the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the cost requirement. Calculate the deviation between the size design parameter and the median value of the transportation size requirement, divide the deviation by the range of the transportation size requirement to obtain the deviation percentage corresponding to the size design parameter, subtract the deviation percentage from 1 to obtain the single-dimensional fit value corresponding to the size design parameter, and select the minimum value among the single-dimensional fit values ​​as the single-dimensional matching degree of the transportation requirement.

6. The gift design optimization method based on multimodal user demand perception as described in claim 1, characterized in that, Obtain the gift size design range, iterate through and solve the collaborative optimization objective function to obtain multiple candidate size design schemes that satisfy all constraints, and the comprehensive matching degree corresponding to each candidate scheme, including: Based on the cost constraint threshold and material size matching threshold, the gift size design range is obtained; Based on the gift size design range, multiple size design parameters are randomly generated, with each size design parameter being treated as a particle. The number of size design parameters generated is obtained based on the range of the gift size design range. Initialize the particle's position and velocity. The particle position is the quantized value of the size design parameter, and the particle velocity is the adjustment step size of the size design parameter. The comprehensive matching degree calculated by the collaborative optimization objective function is used as the fitness value of the particle. The initial comprehensive matching degree value of each particle is calculated, and particles that meet the cost constraint threshold and material size matching threshold are selected as effective particles. The particle swarm is iteratively optimized until a preset number of iterations is reached; Get the gift size designs corresponding to multiple valid particles with a comprehensive matching degree greater than the comprehensive matching threshold, and use them as candidate size design schemes.

7. The gift design optimization method based on multimodal user demand perception as described in claim 6, characterized in that, The particle swarm is iteratively optimized until the required number of iterations is reached, including: Record the historical best fitness value and corresponding position of each effective particle as the individual best, record the best fitness value and corresponding position of all effective particles in the particle swarm as the global best, update the position and velocity of effective particles based on the individual best and the global best, iteratively calculate the collaborative optimization objective function value of the particles as the comprehensive matching degree, and synchronously iteratively calculate the matching degree of each single dimension corresponding to each particle. During the iteration process, high-matching particle subsets are selected based on the single-dimensional matching degree of display size requirements, the single-dimensional matching degree of cost requirements, and the single-dimensional matching degree of transportation requirements. The similarity between any two valid particles is calculated within and between each subset. The similarity is the weighted sum of the dimensionless deviation value of the particle position parameter and the dimensionless deviation value of the fitness value. Particle combinations with a similarity greater than a preset similarity threshold are selected, wherein the particle combination includes two particles; Discard particles with low overall matching degree in the combination, retain particles with high overall matching degree, and obtain the optimization adjustment step size of the retained particles.

8. The gift design optimization method based on multimodal user demand perception as described in claim 6, characterized in that, Obtain the optimization adjustment step size for retaining particles, including: The adjustment coefficient is obtained by combining the ratio of the overall matching degree of discarded particles to the overall matching degree of retained particles with the proportion of single-dimensional similarity between discarded and retained particles. The optimized adjustment step size is obtained by multiplying the adjustment coefficient by the adjustment step size of the retained particles.

9. The gift design optimization method based on multimodal user demand perception as described in claim 1, characterized in that, The candidate size design schemes with the highest overall matching degree and single-dimensional matching degree are selected to obtain the optimal size design scheme set for gifts, including: For all candidate size design schemes, extract the matching degree of each scheme in terms of display dimension, cost dimension, transportation dimension, and overall matching degree. Based on the multi-dimensional parameter set of the candidate solutions, the three single-dimensional optimal solutions with the highest matching degree in display, cost, and transportation were selected and added to the gift optimal size design solution set. The candidate size design scheme with the highest overall matching degree is selected as the optimal overall balance scheme and added to the set of optimal gift size design schemes.

10. A gift design optimization system integrating multimodal user demand perception, characterized in that, The method for optimizing gift design schemes that integrates multimodal user demand perception as described in any one of claims 1-9 includes: The demand feature acquisition module is used to acquire a gift size demand feature set, wherein the gift size demand feature set includes display size requirements, cost requirements, and transportation requirements. The weighting coefficient calculation module is used to calculate the weighting coefficients of the size requirements of each gift by combining sensitivity analysis with historical gift design schemes. The objective function construction module is used to construct a collaborative optimization objective function for gift size based on the gift size requirement feature set and the weight coefficients. The size design parameters are used as independent variables, and the optimization objective is to maximize the comprehensive matching degree. At the same time, cost constraint threshold and material size matching threshold are set as optimization constraints. The candidate solution acquisition module is used to obtain the gift size design range, iterate through and solve the collaborative optimization objective function, obtain multiple candidate size design schemes that satisfy all constraints, and the comprehensive matching degree corresponding to each candidate scheme; The optimal solution selection module is used to select the candidate size design scheme with the highest comprehensive matching degree and single-dimensional matching degree to obtain the optimal size design scheme set for gifts.

Images