A packaging design scheme generation method fusing domain knowledge base

Abstract

The application relates to the technical field of computer-aided design, in particular to a packaging design scheme generation method fusing a field knowledge base, which comprises the following steps: determining reference indexes of each reference surface of each historical packaging design three-dimensional model of a target object, combining the number of trend connecting lines corresponding to each contact point to obtain the reference degree of each contact point, and screening to obtain a deviation point; obtaining an enhanced weight of the deviation point based on the distance difference of the deviation point and other deviation points in a preset adjacent range of the deviation point to a corresponding standard surface of a standard model; obtaining a packaging enhancement buffer thickness of the deviation point from the enhanced weight, so as to generate a packaging design scheme of the target object. With the aid of computer-aided design, the packaging design scheme of the target object is obtained, the transportation safety of the target object is improved, the cost increase caused by blindly increasing the global buffer thickness is avoided, and a fine balance between safety and cost control is achieved.

Description

Technical Field

[0001] This invention relates to the field of computer-aided design technology, specifically to a method for generating packaging design schemes that integrates a domain knowledge base. Background Technology

[0002] During the transportation of goods, the cushioning thickness of the packaging is a crucial parameter for protecting the goods from damage during transit. For vulnerable areas of the goods, increased cushioning thickness is necessary to ensure safe transport, while also considering packaging costs. Current packaging design methods involve storing multiple historical 3D models of the goods' packaging in a knowledge base. Personnel manually select the model they deem most suitable to determine the packaging for the goods. However, existing methods only generate the packaging for the goods as a whole. After packaging, certain points on the goods may not be compatible with the packaging, resulting in inappropriate cushioning thickness at some points and increasing the risk of damage during transport. Summary of the Invention

[0003] The technical problem that this invention aims to solve is that the packaging design of the target item is not fully compatible with the shape of the target item, which easily leads to an unsuitable cushioning thickness at some points on the upper part of the target item.

[0004] The purpose of this invention is to provide a method for generating packaging design schemes that integrates a domain knowledge base. The specific technical solution adopted is as follows: This invention provides a method for generating packaging design schemes that integrate domain knowledge bases, including: Determine several comparison points on each reference surface of each historical packaging design 3D model in the knowledge base. Based on the distance and projection direction between the comparison points of each reference surface and the target item 3D model, obtain the reference index of each reference surface. The reference degree of each contact point is obtained based on the number of trend lines corresponding to each contact point and the reference index; the contact point is the position point corresponding to the comparison position point on the three-dimensional model of the target item, and the trend line is the line connecting the comparison position point and the corresponding contact point. Deviation points are obtained from each contact point using the reference value; The enhancement weight of the deviation point is obtained based on the distance difference between the deviation point and other deviation points within its preset neighborhood to the corresponding standard surface of the standard model. The packaging reinforcement buffer thickness at the deviation point is obtained from the reinforcement weight; the packaging reinforcement buffer thickness is used to indicate the packaging design scheme for generating the target item.

[0005] In an exemplary embodiment, the distance between the comparison point and the 3D model of the target object is the length of the trend line connecting the comparison points; the projection direction is the direction of the trend line connecting the comparison points. The process of obtaining the reference indicators includes: Obtain the length of the trend line connecting each comparison point on the candidate reference surface, and the degree of directional difference between the trend lines connecting the comparison points on the candidate reference surface; the candidate reference surface is any reference surface on any historical packaging design 3D model; Reference indices for candidate reference surfaces are obtained from the length and directional differences of the trend lines; the reference indices are inversely correlated with both the length and directional differences of the trend lines.

[0006] In an exemplary embodiment, the reference index for obtaining the candidate reference surface from the length and directional difference of the trend line includes: Based on the length of the trend line connecting the candidate comparison points and the overall degree of directional difference between the candidate comparison point and other comparison points on the candidate reference surface, the difference component of the candidate comparison point is obtained; the candidate comparison point is any comparison point on the candidate reference surface. The difference components of each comparison point on the candidate reference surface are fused to obtain the total difference component of the candidate reference surface; A reference index for the candidate reference surface is obtained based on the total difference component of the candidate reference surface; the reference index is inversely correlated with the total difference component.

[0007] In an exemplary embodiment, the process of obtaining the reference degree includes: Determine the number of trend lines connecting the contact points and the reference index of the reference surface where the contact points are located; The reference degree of the contact point is obtained based on the number of trend lines connecting the contact point and the reference index of the reference surface where the contact point is located; the reference degree is positively correlated with both the number of trend lines connecting the contact point and the reference index of the reference surface where the contact point is located.

[0008] In an exemplary embodiment, the process of obtaining the deviation point includes: taking the contact point whose reference degree is less than or equal to a preset reference degree threshold as the deviation point.

[0009] In an exemplary embodiment, the standard model is the smallest bounding cuboid of the three-dimensional model of the target object, and the standard faces are the various surfaces of the smallest bounding cuboid.

[0010] In an exemplary embodiment, the distance difference between a candidate deviation point and each other deviation point within its preset neighborhood to the corresponding standard surface is specifically: the difference between the distance from each other deviation point within its preset neighborhood to the corresponding standard surface and the distance from the candidate deviation point to the corresponding standard surface; the candidate deviation point is any deviation point. The process of obtaining the enhancement weights includes: Based on the distance difference between the candidate deviation point and each other deviation point within its preset neighborhood to the corresponding standard surface, and the projection distance on the corresponding standard surface, the attenuation coefficient of the candidate deviation point and each other deviation point within its preset neighborhood is obtained; the attenuation coefficient is positively correlated with the distance difference and negatively correlated with the projection distance. The enhancement weights of candidate deviation points are obtained based on the relationship between the attenuation coefficient and the preset standard attenuation coefficient.

[0011] In an exemplary embodiment, obtaining the enhancement weight of the candidate deviation point based on the relationship between the attenuation coefficient and the preset standard attenuation coefficient includes: Calculate the coefficient difference between the attenuation coefficient and the preset standard attenuation coefficient, and calculate the average value of the coefficient difference. Obtain the enhancement weight of the candidate deviation point based on the average value. The enhancement weight is positively correlated with the average value.

[0012] In one exemplary embodiment, the process of obtaining the packaging reinforcement cushioning thickness includes: The enhancement coefficient of the deviation point is obtained based on the enhancement weight; The packaging reinforcement buffer thickness is obtained by multiplying the reinforcement coefficient by the preset basic buffer thickness.

[0013] In an exemplary embodiment, after obtaining the packaging reinforcement buffer thickness at the deviation point from the enhancement weights, the packaging design scheme generation method of the fused domain knowledge base further includes: The transportation gain coefficient is obtained from the transportation time of the target item; the transportation gain coefficient is positively correlated with the transportation time. The final packaging cushioning thickness is obtained based on the transport gain coefficient and the packaging reinforcement cushioning thickness.

[0014] This invention offers the following advantages: It determines the reference indicators for each reference surface of the 3D model of each historical packaging design, providing a rich set of comparative benchmarks for subsequent anomaly (deviation) detection; trend lines connect points (comparison points) on the 3D model of historical packaging designs with points (contact points) on the target item, and the number of these points directly reflects the frequency with which the contact point was considered in the historical design. Combined with the reference indicators, the reference degree of the contact point is obtained, used to determine whether the normal distance relationship between the contact point and the 3D model of the historical packaging design is abnormal. The reference degree is a composite evaluation index that can automatically filter out contact points that have historically received less attention and deviate significantly from the normal buffer distance, achieving intelligent screening to automatically lock high-risk locations. This completes the focus from the overall model to key local locations, enabling design resources to be targeted; the enhancement weight not only reflects the magnitude of the deviation at a single point but also the prevalence and severity of the deviation in the local area. A higher enhancement weight means that the local area where the point is located is more distorted or more fragile, requiring more significant buffer compensation. This provides a precise calculation basis for how much to strengthen the buffer thickness; thus, based on the enhancement weight, the packaging reinforcement buffer thickness at the deviation point is obtained. This invention constructs a complete automated process from historical data mining → intelligent detection of local anomalies → quantitative assessment of deviation degree → precise output of design parameters. It changes the traditional packaging design mode that relies on overall model matching and human experience. It can proactively discover local buffer weaknesses that may be overlooked by humans. Based on historical data and local geometric relationships, it quantitatively determines the reinforcement thickness, reduces design subjectivity, and generates customized, locally optimized packaging solutions for target items. While improving transportation safety, it avoids the cost increase caused by blindly increasing the global buffer thickness. It achieves a fine balance between ensuring safety (strengthening vulnerable points) and controlling costs (not wasting materials in non-critical areas). Attached Figure Description

[0015] Figure 1 This is a flowchart of a method for generating a packaging design scheme that integrates a domain knowledge base, provided in one embodiment of the present invention. Detailed Implementation

[0016] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. All data and information collected in this application have been obtained with full consent.

[0018] In the packaging and transportation of target items, the cushioning thickness design is a crucial parameter for protecting the item from damage during transit. For more vulnerable parts of the target item, the cushioning thickness needs to be increased to ensure transportation safety. This embodiment provides a packaging design scheme generation method integrating a domain knowledge base to generate packaging design schemes for target items that need to be transported, ensuring the safety of the target item during transportation and effectively controlling the packaging cost. The target item is determined by the actual application scenario; for example, an engine block.

[0019] This embodiment requires retrieving information from a knowledge base. It should be understood that a knowledge base is a database used for systematically storing and managing information; it can include various knowledge related to packaging design, such as design principles, material information, market demands, and user experience. The application of knowledge base technology enables relevant personnel to quickly retrieve useful information, thereby improving design efficiency. Packaging design that integrates a domain knowledge base can generate design solutions using this integrated knowledge base. It can automatically generate multiple design options based on a large number of variables and constraints, helping users find the best solution.

[0020] To design and generate packaging design schemes based on multi-source knowledge and achieve packaging design scheme generation by integrating a domain knowledge base, this embodiment requires obtaining at least two historical 3D models of packaging designs for the target item. These historical 3D models can be 3D models of packaging designs used by the target item (or items of the same type as the target item) within a historical period. In this embodiment, the historical 3D models are stored in a knowledge base, and at least two historical 3D models are retrieved from the knowledge base. Simultaneously, a 3D model of the target item is obtained based on the target item.

[0021] like Figure 1 As shown in the figure, the method for generating a packaging design scheme that integrates a domain knowledge base provided in this embodiment includes the following steps: Step S1: Determine several comparison points on each reference surface of each historical packaging design 3D model in the knowledge base. Based on the distance and projection direction between the comparison points of each reference surface and the target item 3D model, obtain the reference index of each reference surface. Step S2: Based on the number of trend lines corresponding to each contact point and the reference indicators, obtain the reference degree of each contact point; Step S3: Select deviation points from each contact point using the reference value; Step S4: Based on the distance difference between the deviation point and other deviation points within its preset neighborhood to the corresponding standard surface of the standard model, obtain the enhancement weight of the deviation point; Step S5: Obtain the packaging reinforcement buffer thickness at the deviation point from the reinforcement weight.

[0022] The following is a detailed explanation of each step.

[0023] Step S1: Determine several comparison points on each reference surface of each historical packaging design 3D model in the knowledge base. Based on the distance and projection direction between the comparison points of each reference surface and the target item 3D model, obtain the reference index of each reference surface.

[0024] For the target item, items of the same type, such as engine blocks from different vehicles, exhibit a high degree of similarity in their 3D models. Therefore, historical packaging design 3D models can serve as a reference. Thus, this embodiment can utilize the similarity between historical packaging design 3D models and the target item's 3D model to generate packaging design schemes.

[0025] For ease of explanation, any historical packaging design 3D model is designated as a candidate historical packaging design 3D model. Each surface of the candidate historical packaging design 3D model is defined as a reference surface. It should be understood that the determined surfaces are not necessarily planes; they may also be curved surfaces. When the surfaces are curved, this embodiment attempts to divide the reference surfaces into relatively simple shapes, i.e., regular surfaces that can be described by simple mathematical equations, such as quadratic surfaces, and other simple analytical surfaces (cylinders, cones, spheres, etc.).

[0026] At least one location point is selected from each reference surface of the candidate historical packaging design 3D model and defined as a comparison location point. The selection method of the comparison location point is determined by the actual situation. This embodiment is not limited to a specific selection method. In an exemplary embodiment, as many comparison location points as possible can be selected on each reference surface, and the comparison location points are evenly selected on the candidate historical packaging design 3D model. For example, each reference surface of the candidate historical packaging design 3D model is divided into multiple small areas of equal area, such as 0.1 square centimeters. It should be understood that the area of ​​the small area is very small, so the small area can be equivalent to a plane, and the shape of the small area is set as a square area. The center point of each small area is used as the comparison location point, thereby obtaining the comparison location points on each reference surface. In this embodiment, the candidate historical packaging design 3D model is aligned with the target item 3D model, for example, by aligning the center points of the two models and aligning the orientations of the two models.

[0027] For ease of explanation, any reference plane of the candidate historical packaging design 3D model is set as the candidate reference plane, and any comparison point on the candidate reference plane is set as the candidate comparison point.

[0028] Determine the distances between candidate comparison points and various points on the surface of the target object's 3D model. Find the minimum distance, and the point on the target object's 3D model corresponding to this minimum distance is defined as the corresponding point on the target object's 3D model and thus the contact point. The line connecting the candidate comparison point and its corresponding contact point is defined as the trend line connecting the candidate comparison points.

[0029] Determine the distance between the candidate comparison points and the target object's 3D model. Since the trend line connecting the candidate comparison points represents the connection between the candidate comparison points and the target object's 3D model, the distance between the candidate comparison points and the target object's 3D model is the length of the trend line connecting the candidate comparison points.

[0030] Determine the projection direction of the candidate comparison point onto the 3D model of the target object, where the projection direction is the direction of the trend line connecting the candidate comparison point.

[0031] The reference index of the candidate reference surface is obtained by considering the distances between each comparison point on the candidate reference surface and the 3D model of the target item, as well as the projection directions of each comparison point on the candidate reference surface onto the 3D model of the target item. The greater the distance between each comparison point and the 3D model of the target item, the greater the difference between them, and the weaker the reference value of the candidate reference surface in packaging design; that is, the lower the reference index of the candidate reference surface. Similarly, the greater the difference in projection directions between any two comparison points on the candidate reference surface, the greater the difference in planar trend between the candidate reference surface and the 3D model of the target item, and the weaker the reference value of the candidate reference surface in packaging design; that is, the lower the reference index of the candidate reference surface.

[0032] In an exemplary embodiment, the degree of directional difference between the trend lines connecting the comparison points on the candidate reference surface is obtained. The degree of directional difference is obtained by the angle between the trend lines connecting any two comparison points. Specifically, for a candidate comparison point, the angle between the trend line connecting the candidate comparison point and each other comparison point on the candidate reference surface is obtained (in this embodiment, the angle is specifically an acute angle, i.e., the smallest angle is selected from the two angles formed by the two trend lines). The value range of this angle is 0 to... The larger the value, the greater the difference in direction between the two trend lines; the smaller the value, the more consistent the trends of the two trend lines. In an exemplary embodiment, this embodiment divides the included angle by... The included angle is normalized, and the normalized result is used as the degree of directional difference. This yields the degree of directional difference of the trend lines connecting the candidate comparison point to each other comparison point on the candidate reference surface. The average value of the directional difference of these trend lines is calculated and used as the mean directional difference of the candidate comparison point. This mean directional difference of the candidate comparison point represents the overall directional difference between the candidate comparison point and all other comparison points on the candidate reference surface. In this embodiment, the mean directional difference of the candidate comparison point is used as the overall directional difference between the candidate comparison point and all other comparison points on the candidate reference surface. This results in the mean directional difference of all comparison points on the candidate reference surface.

[0033] It should be understood that if two trend lines do not intersect, there are two possibilities. The first possibility is that the two trend lines are parallel, in which case the included angle is 0. The second possibility is that the two trend lines are not in the same plane. In this case, one of the trend lines is translated so that the two trend lines are in the same plane, and then the included angle between them is determined.

[0034] Reference indices for the candidate reference surface are derived based on the length of the trend lines connecting the comparison points on the candidate reference surface, and the average directional difference between these points. These indices are inversely correlated with both the length of the trend lines and the average directional difference. Specifically: Based on the length of the trend line connecting the candidate comparison points and the average directional difference of the candidate comparison points, the difference components of the candidate comparison points are obtained. The longer the trend line, the larger the difference component; the larger the average directional difference, the larger the difference component. Then, the difference components of each comparison point on the candidate reference surface are merged (e.g., by calculating the average value) to obtain the total difference component of the candidate reference surface. Finally, based on the total difference component of the candidate reference surface, the reference index of the candidate reference surface is obtained. The larger the total difference component, the smaller the reference index; the two are inversely correlated. Based on the above logic, a specific calculation method for the reference index is given below: ; in, This represents the reference index of the k-th reference surface in the r-th historical packaging design 3D model. This represents the average directional difference at the i-th comparison point on the k-th reference plane of the r-th historical packaging design 3D model. This represents the length of the trend line connecting the i-th comparison point on the k-th reference plane of the r-th historical packaging design 3D model. This represents the number of comparison points on the k-th reference plane of the r-th historical packaging design 3D model. This represents the maximum value among the trend line lengths connecting the comparison points on each reference plane of the r-th historical packaging design 3D model. Used to achieve Normalization.

[0035] This represents the difference component of the i-th comparison point on the k-th reference surface of the r-th historical packaging design 3D model. The average value is used to fuse the average length of the trend line and the average value of the directional difference.

[0036] This represents the total difference component of the k-th reference surface of the r-th historical packaging design 3D model.

[0037] This yields the reference indicators for each reference surface of the 3D model of each historical packaging design. The larger the reference indicator, the closer the corresponding reference surface is to the shape of the target item's 3D model, and the greater the degree of reference.

[0038] Step S2: Based on the number of trend lines corresponding to each contact point and the reference indicators, obtain the reference degree of each contact point.

[0039] After obtaining the reference indicators of each reference surface of the 3D model of each historical packaging design, it is necessary to convert them to each contact point on the 3D model of the target item for analysis. By analyzing the number of trend lines corresponding to the contact points and the reference indicators of the reference surfaces corresponding to the contact points, the reference degree of the contact points can be obtained.

[0040] For each candidate historical packaging design 3D model, each comparison point corresponds to a contact point on the target item 3D model. It should be understood that since different historical packaging design 3D models may not be entirely identical, the contact points corresponding to different historical packaging design 3D models may also not be entirely identical.

[0041] For ease of explanation, any contact point on the 3D model of the target item is designated as a candidate contact point. Each candidate contact point corresponds to at least one comparison point on a reference surface of one of the historical packaging design 3D models. Furthermore, a candidate contact point may simultaneously correspond to a comparison point on a reference surface of multiple historical packaging design 3D models. Each comparison point has a trend line; therefore, a candidate contact point may lie on multiple trend lines simultaneously, meaning it may correspond to a comparison point on a reference surface of multiple historical packaging design 3D models. The number of trend lines on which a candidate contact point lies (with a minimum value of 1) is determined. A higher number of trend lines indicates a greater number of historical packaging design 3D models providing reference for the candidate contact point, resulting in a higher reference value for the candidate contact point; the two are positively correlated. It should be understood that the maximum number of trend lines on which a candidate contact point may lie is equal to the number of historical packaging design 3D models.

[0042] After obtaining the trend lines connecting the candidate contact points, the corresponding comparison point for each trend line is determined, thus obtaining the reference surface for each comparison point. These are defined as the reference surfaces for the candidate contact points, and reference indices are obtained for each reference surface. The higher the reference indices of the candidate contact points, the higher the reference accuracy of the candidate contact points; the two are positively correlated. The average value of the reference indices for each reference surface is calculated.

[0043] The reference degree of a candidate contact point is obtained by considering the number of trend lines connecting the candidate contact points and the average value of the reference indices of each reference surface on which the candidate contact point is located. Based on the above analysis, a specific calculation method for the reference degree is given below: ; in, This represents the reference degree of the t-th contact point. R represents the number of trend lines connecting the t-th contact point, and R represents the number of historical 3D models of the packaging design, i.e., the maximum number of trend lines connecting candidate contact points. Indicates to Normalization, This represents the average value of the reference parameters of each reference surface at the t-th contact point. This represents the normalization function. In this embodiment, the normalization method can be the maximum / minimum value normalization method. The normalization process here is as follows: obtain the values ​​of each contact point. The maximum and minimum values ​​in the dataset, or the maximum and minimum values ​​of the reference value in the historical dataset, are then used for the maximum and minimum value normalization method for the t-th contact point. Normalize.

[0044] by As a weight, a larger value indicates a greater number of historical packaging design 3D models that provide a reference for the t-th contact point.

[0045] Step S3: Select the deviation points from each contact point using the reference value.

[0046] Step S2 obtains the reference value of each contact point of the target item's 3D model. A higher reference value indicates a closer similarity between the historical packaging design 3D model and the contact point's shape, providing a more reliable reference. Therefore, the packaging cushioning thickness design of the target item at the corresponding contact point is less likely to require modification. Conversely, a lower reference value indicates a greater difference between the historical packaging design 3D model and the contact point's shape, making it less reliable. In this case, the packaging cushioning thickness design of the target item at the corresponding contact point needs adjustment, requiring further analysis and adjustment for contact points with significant differences in subsequent steps.

[0047] This embodiment presets a reference threshold, the value of which ranges from 0 to 1. The specific value is set according to the actual judgment needs. If a more stringent judgment logic is required, the reference threshold can be set to a larger value. As an example, this embodiment uses 0.7.

[0048] Compare the reference degree of each contact point with the reference degree threshold. Contact points with a reference degree less than or equal to the reference degree threshold are designated as deviation points, while contact points with a reference degree greater than the reference degree threshold are designated as similar points.

[0049] The next step involves further analysis of the deviation points. For similar points, the buffer thickness at the location of the similar point can be determined directly based on the 3D model of the historical packaging design. This generates a relatively standard and reliable basic structure for the packaging design at the similar point. As an example, the average buffer thickness at the similar point in each historical packaging design 3D model is calculated as the buffer thickness of the target item at that similar point, and thus the reinforced buffer thickness of the packaging at that similar point.

[0050] Step S4: Based on the distance difference between the deviation point and other deviation points within its preset neighborhood to the corresponding standard surface of the standard model, obtain the enhancement weight of the deviation point.

[0051] This embodiment determines the standard model of the target item. In order to facilitate data processing, in an exemplary embodiment, the standard model is the minimum bounding cuboid of the three-dimensional model of the target item. The minimum bounding cuboid has six surfaces, and each surface of the minimum bounding cuboid is defined as a standard face.

[0052] The buffer thickness at each deviation point is dynamically adjusted by measuring the morphological differences between the deviation point and the standard model. This step analyzes the morphological distribution of the deviation points. In the packaging of the target item, protruding or sharp structures are often more vulnerable, so these areas require greater buffer thickness to protect the structure of the target item from damage during transportation.

[0053] For any deviation point, obtain the distance from that deviation point to each standard plane of the standard model (i.e., draw perpendicular lines from the deviation point to each standard plane and obtain the length of the perpendicular line segments). The standard plane with the smallest distance is taken as the corresponding standard plane for that deviation point, thus obtaining the corresponding standard plane for each deviation point. Simultaneously, obtain the distance between each deviation point and its corresponding standard plane. The smaller the distance between a deviation point and its corresponding standard plane, the more prominent the location of the deviation point.

[0054] Several deviation points sharing the same corresponding standard plane are identified. For ease of explanation, any one of these deviation points is designated as a candidate deviation point. A preset proximity range for each candidate deviation point is determined on the 3D model of the target object. In an exemplary embodiment, a sphere with the candidate deviation point as its center and a preset radius is formed; this spherical range is considered the preset proximity range of the candidate deviation point. The preset radius is set according to actual needs; for example, 3 cm is used. The distance difference between the candidate deviation point and each other deviation point within its preset proximity range to the corresponding standard plane is determined. This distance difference is obtained by subtracting the distance from the candidate deviation point to the corresponding standard plane from the distance from each other deviation point within the preset proximity range. Therefore, when the distance from a candidate deviation point to its corresponding standard surface is less than the distance from other deviation points within its preset neighboring range to the corresponding standard surface, the distance difference is positive; when the distance from a candidate deviation point to its corresponding standard surface is greater than the distance from other deviation points within its preset neighboring range to the corresponding standard surface, the distance difference is negative; and when the distance from a candidate deviation point to its corresponding standard surface is equal to the distance from other deviation points within its preset neighboring range to the corresponding standard surface, the distance difference is 0. The larger the distance difference, the more prominent the candidate deviation point is compared to its surrounding deviation points, indicating a greater need for packaging cushioning thickness, i.e., a greater reinforcement weight.

[0055] It should be understood that if there are no other deviation points within the preset neighborhood of a candidate deviation point, it means that the candidate deviation point is relatively isolated and will not be further processed. If the candidate deviation point is likely to be noisy data, it will be treated as a nearby point.

[0056] Obtain the projection points (i.e., the intersections of the aforementioned vertical segments on the corresponding standard plane) of the candidate deviation point and each other deviation point within its preset neighborhood. Then, calculate the distance between these two projection points on the corresponding standard plane, which is used as the projection distance between the candidate deviation point and each other deviation point within its preset neighborhood on the corresponding standard plane. The smaller the projection distance, the more prominent the distance difference between the two deviation points, and the greater the packaging buffer thickness required for the candidate deviation point, i.e., the greater the enhancement weight.

[0057] Based on the distance difference between the candidate deviation point and each other deviation point within its preset neighborhood to the corresponding standard surface, and the projected distance on the standard surface, the attenuation coefficient of the candidate deviation point and each other deviation point within its preset neighborhood is obtained. The larger the attenuation coefficient, the greater the enhancement weight. As analyzed above, the attenuation coefficient is positively correlated with the distance difference and inversely correlated with the projected distance. In an exemplary embodiment, a specific calculation method for the attenuation coefficient is given below: ; in, This represents the attenuation coefficient of the j-th deviation point among the m-th deviation point within its preset neighborhood. This represents the distance from the m-th deviation point to the corresponding standard surface. This represents the distance from the j-th deviation point to the corresponding standard surface. This represents the difference in distance from the j-th deviation point to the m-th deviation point within its preset neighborhood to the corresponding standard surface. This represents the projected distance between the j-th deviation point and the m-th deviation point on the corresponding standard plane.

[0058] It should be understood that when the projection distance is too small, approaching 0, the denominator will be 0, causing the calculation result to overflow. To avoid this situation, this embodiment presets a lower limit value for the projection distance. If the projection distance is less than this lower limit value, the projection distance is set to this lower limit value.

[0059] The larger the attenuation coefficient between the j-th deviation point and the m-th deviation point, the greater the decrease in the deviation distance between the m-th and j-th deviation points, and the more pronounced the protruding shape of the j-th deviation point relative to the m-th deviation point. Similarly, the attenuation coefficients of the j-th deviation point and each other deviation point within its preset neighborhood are obtained, and the attenuation coefficients are used to represent the shape changes of the neighborhood of the j-th deviation point.

[0060] In this embodiment, for the j-th deviation point, a standard attenuation coefficient is preset. This standard attenuation coefficient serves as a benchmark. The enhancement weight of the j-th deviation point is obtained based on the relationship between the attenuation coefficients of the j-th deviation point and each other deviation point within its preset neighboring range, and this standard attenuation coefficient. The larger the attenuation coefficient of the j-th deviation point and each other deviation point within its preset neighboring range is compared to the standard attenuation coefficient, the greater the enhancement weight of the j-th deviation point. In an exemplary embodiment, the average attenuation coefficient of the j-th deviation point and each other deviation point within its preset neighboring range is calculated, and this average value is used as the standard attenuation coefficient of the j-th deviation point.

[0061] Calculate the difference between the attenuation coefficient of the j-th deviation point and the preset standard attenuation coefficient of each other deviation point within its preset neighborhood. The larger the difference, the more prominent the j-th deviation point is within its preset neighborhood. Since positive enhancement of the deviation point is required, the difference is set to 0 when it is less than 0. Then, calculate the average of the difference values ​​and obtain the enhancement weight of the j-th deviation point based on this average value; the enhancement weight is positively correlated with this average value. Based on the above logical analysis, the following is a specific calculation method for the enhancement weight of the j-th deviation point: ; in, This represents the enhancement weight at the j-th deviation point. This represents the standard attenuation coefficient at the j-th deviation point. This represents the number of other deviation points within a preset neighborhood of the j-th deviation point. express The function sets the value of the defined object to 0 when it is less than 0, to avoid negative enhancement when the attenuation coefficient is not significant. Similarly, the maximum and minimum value normalization method can be used to obtain the values ​​of each deviation point. The maximum value in the range is used here, where the minimum value in the maximum-minimum-value normalization method is set to 0. The maximum-minimum-value normalization method is then applied to each deviation point. Normalize.

[0062] Step S5: Obtain the packaging reinforcement buffer thickness at the deviation point from the reinforcement weight.

[0063] Step S4 yields the reinforcement weights for each deviation point. These reinforcement weights, representing the fragility of the local shape centered on the deviation point, reflect the local shape and are used to adjust the cushioning thickness. A larger reinforcement weight results in a greater increase in cushioning thickness, leading to a thicker package reinforcement cushioning.

[0064] In this embodiment, a basic buffer thickness is preset for the j-th deviation point, and this basic buffer thickness serves as the benchmark for enhancing the buffer thickness at the j-th deviation point. In an exemplary embodiment, the buffer thickness at the j-th deviation point in each historical packaging design 3D model is obtained, and then the average value of the buffer thickness at the j-th deviation point in each historical packaging design 3D model is calculated as the basic buffer thickness corresponding to the j-th deviation point.

[0065] The enhancement coefficient for the j-th deviation point is obtained based on the enhancement weight of the j-th deviation point. Then, the enhancement coefficient for the j-th deviation point is multiplied by the basic buffer thickness of the j-th deviation point to obtain the packaging enhancement buffer thickness of the j-th deviation point. The calculation formula is as follows: ; in, This represents the thickness of the packaging reinforcement cushioning at the j-th deviation point. This represents the enhancement coefficient at the j-th deviation point. This represents the basic buffer thickness corresponding to the j-th deviation point.

[0066] Therefore, this embodiment can be understood as follows: when the contact points are close to each other, the corresponding reinforcement weight is recorded as 0, and the packaging reinforcement buffer thickness of the close points is obtained by using the above calculation formula.

[0067] At this point, the packaging reinforcement and cushioning thickness at each contact point of the target item is obtained. This thickness is analyzed in conjunction with the historical 3D model of the packaging design and the shape of the target item, resulting in a more personalized and targeted cushioning thickness design. Based on the packaging reinforcement and cushioning thickness at each contact point of the target item, a packaging design scheme for the target item can be obtained. The core of the packaging design scheme is the numerical sequence composed of the packaging reinforcement and cushioning thickness at each contact point of the target item. As an example, this embodiment can also automatically generate a lightweight cushioning structure based on the specific shape of the target item using existing topology optimization algorithms, under the constraint of the packaging reinforcement and cushioning thickness at each contact point, thus obtaining a packaging design scheme for the target item.

[0068] It should be understood that this embodiment can also set a maximum allowable buffer thickness according to the shape of the target item and transportation conditions. If the packaging reinforcement buffer thickness obtained above is greater than the maximum allowable buffer thickness, the packaging reinforcement buffer thickness is set to the maximum allowable buffer thickness.

[0069] In an exemplary embodiment, after obtaining the packaging reinforcement cushioning thickness at each contact point as described above, this embodiment can further consider the impact of the target item's transportation time on the packaging reinforcement cushioning thickness. Based on the transportation requirements of the target item, the packaging reinforcement cushioning thickness is updated to achieve the fusion of data from multiple domains. The longer the transportation time of the target item, the greater the need for increased packaging reinforcement cushioning thickness to ensure transportation safety. The transportation time can be estimated from the transportation planning of the target item.

[0070] The transportation gain coefficient is obtained from the transportation time of the target item; the transportation gain coefficient is positively correlated with the transportation time. Then, based on the transportation gain coefficient and the packaging reinforcement cushioning thickness at each contact point, the final packaging cushioning thickness at each contact point is obtained. Based on the above logical analysis, the following is a specific calculation method for the final packaging cushioning thickness at each contact point: ; in, This represents the final packaging buffer thickness at the t-th contact point. This represents the thickness of the packaging reinforcement cushioning at the t-th contact point. Indicates the transportation time of the target item. This represents the average historical transportation time of the target item, which is calculated by taking the historical transportation time of multiple shipments of items of the same type as the target item and averaging these historical transportation times.

[0071] This indicates the relative increase rate of the target item's transportation time compared to the historical average transportation time, expressed through... The function has a limitation: if the transit time of the target item is lower than the historical average transit time, the packaging reinforcement and cushioning thickness will not be increased. The longer the transit time of the target item is compared to the historical average transit time, the more necessary it is to further increase the packaging reinforcement and cushioning thickness to ensure the packaging strength of the target item during long-term transit.

[0072] Similarly, if the final packaging buffer thickness obtained above is greater than the maximum allowable buffer thickness, then the final packaging buffer thickness is set to the maximum allowable buffer thickness.

[0073] It should be understood that if the impact of transportation time on the packaging cushioning thickness is not considered, the process for obtaining the final packaging cushioning thickness described above may not be necessary.

[0074] This embodiment also provides a packaging design scheme generation system that integrates a domain knowledge base, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above embodiment of the packaging design scheme generation method that integrates a domain knowledge base when the program instructions are executed.

[0075] In one exemplary embodiment, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the above-described embodiment of the method for generating packaging design schemes based on a fusion domain knowledge base.

[0076] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0077] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for generating packaging design schemes that integrates a domain knowledge base, characterized in that, include: Determine several comparison points on each reference surface of each historical packaging design 3D model in the knowledge base. Based on the distance and projection direction between the comparison points of each reference surface and the target item 3D model, obtain the reference index of each reference surface. The reference degree of each contact point is obtained based on the number of trend lines corresponding to each contact point and the reference index. The contact point is the position point on the target item's 3D model corresponding to the comparison position point, and the trend line is the line connecting the comparison position point and the corresponding contact point. Deviation points are obtained from each contact point using the reference value; The enhancement weight of the deviation point is obtained based on the distance difference between the deviation point and other deviation points within its preset neighborhood to the corresponding standard surface of the standard model; The packaging reinforcement buffer thickness at the deviation point is obtained from the reinforcement weight; the packaging reinforcement buffer thickness is used to indicate the packaging design scheme for generating the target item.

2. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, The distance between the comparison point and the 3D model of the target item is the length of the trend line connecting the comparison points; The projection direction is the direction of the trend line connecting the comparison position points; The process of obtaining the reference indicators includes: Obtain the length of the trend line connecting each comparison point on the candidate reference surface, and the degree of directional difference between the trend lines connecting the comparison points on the candidate reference surface; the candidate reference surface is any reference surface on any historical packaging design 3D model; Reference indices for candidate reference surfaces are obtained from the length and directional differences of the trend lines; the reference indices are inversely correlated with both the length and directional differences of the trend lines.

3. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 2, characterized in that, The reference indicators for obtaining candidate reference surfaces based on the length and direction difference of trend lines include: Based on the length of the trend line connecting the candidate comparison points and the overall degree of directional difference between the candidate comparison point and other comparison points on the candidate reference surface, the difference component of the candidate comparison point is obtained; the candidate comparison point is any comparison point on the candidate reference surface. The difference components of each comparison point on the candidate reference surface are fused to obtain the total difference component of the candidate reference surface; A reference index for the candidate reference surface is obtained based on the total difference component of the candidate reference surface; the reference index is inversely correlated with the total difference component.

4. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, The process of obtaining the reference level includes: Determine the number of trend lines connecting the contact points and the reference index of the reference surface where the contact points are located; The reference degree of the contact point is obtained based on the number of trend lines connecting the contact point and the reference index of the reference surface where the contact point is located; the reference degree is positively correlated with both the number of trend lines connecting the contact point and the reference index of the reference surface where the contact point is located.

5. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, The process of obtaining the deviation point includes: taking the contact point where the reference degree is less than or equal to the preset reference degree threshold as the deviation point.

6. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, The standard model is the smallest bounding cuboid of the three-dimensional model of the target object, and the standard faces are each surface of the smallest bounding cuboid.

7. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, The distance difference between a candidate deviation point and each other deviation point within its preset neighboring range to the corresponding standard surface is specifically: the difference between the distance from each other deviation point within the preset neighboring range of the candidate deviation point to the corresponding standard surface and the distance from the candidate deviation point to the corresponding standard surface; the candidate deviation point is any deviation point; The process of obtaining the enhancement weights includes: Based on the distance difference between the candidate deviation point and each other deviation point within its preset neighborhood to the corresponding standard surface, and the projection distance on the corresponding standard surface, the attenuation coefficient of the candidate deviation point and each other deviation point within its preset neighborhood is obtained; the attenuation coefficient is positively correlated with the distance difference and negatively correlated with the projection distance. The enhancement weights of candidate deviation points are obtained based on the relationship between the attenuation coefficient and the preset standard attenuation coefficient.

8. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 7, characterized in that, The step of obtaining the enhancement weights of candidate deviation points based on the relationship between the attenuation coefficient and the preset standard attenuation coefficient includes: Calculate the coefficient difference between the attenuation coefficient and the preset standard attenuation coefficient, and calculate the average value of the coefficient difference. Obtain the enhancement weight of the candidate deviation point based on the average value. The enhancement weight is positively correlated with the average value.

9. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, The process of obtaining the thickness of the packaging reinforcement cushioning includes: The enhancement coefficient of the deviation point is obtained based on the enhancement weight; The packaging reinforcement buffer thickness is obtained by multiplying the reinforcement coefficient by the preset basic buffer thickness.

10. The method for generating packaging design schemes by integrating a domain knowledge base as described in claim 1, characterized in that, After obtaining the packaging reinforcement buffer thickness at the deviation point from the enhanced weights, the packaging design scheme generation method of the fusion domain knowledge base further includes: The transportation gain coefficient is obtained from the transportation time of the target item; the transportation gain coefficient is positively correlated with the transportation time. The final packaging cushioning thickness is obtained based on the transport gain coefficient and the packaging reinforcement cushioning thickness.

Images