Cigarette quality judgment method and device and electronic equipment

Abstract

The invention provides a cigarette quality judgment method and device and electronic equipment, and the method comprises the steps: constructing a standard cigarette sample set which comprises a plurality of standard cigarette samples and standard physical indexes of each standard cigarette sample, the standard physical indexes comprise at least one of standard density, standard water content, standard draw resistance, standard ventilation rate, standard circumference, standard length, standard draw resistance, standard total air permeability and standard hardness; determining a standard index boundary of the standard cigarette sample set according to standard physical indexes of all standard cigarette samples in the standard cigarette sample set; obtaining physical indexes of the to-be-detected cigarette; and calculating the distance between the physical index of the to-be-detected cigarette and the standard index boundary so as to judge the quality of the to-be-detected cigarette. Therefore, through reasonable standard index boundary determination and quality judgment based on the standard index boundary determination, the combustion characteristics and the like of the cigarettes can be ensured to meet safety requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of cigarettes, and in particular to a method, device and electronic equipment for judging the quality of cigarettes. Background Art

[0002] Cigarette quality, circumference, draw resistance, hardness, length, total ventilation rate, etc. are the main physical indicators for evaluating product processing quality, and are also important factors affecting sensory and smoke quality. In the cigarette production process, in addition to raw material factors, physical indicators are the main control variables that directly affect cigarette quality.

[0003] It is of vital importance to set the qualified threshold of cigarette products according to the physical indicators of cigarettes, which is directly related to the quality, taste, safety and market competitiveness of cigarette products. Cigarette physical indicators such as density, water content, draw resistance, ventilation rate, etc. determine the characteristics of cigarette combustion and the law of smoke release. Setting a reasonable qualified threshold can ensure that each cigarette has similar physical properties during the production process, thereby ensuring the consistency of product quality and avoiding poor taste caused by quality differences.

[0004] Setting qualified thresholds based on physical indicators of cigarettes is a key link in ensuring the quality, safety and market competitiveness of cigarette products. This requires comprehensive consideration of scientific research, market demand and consumer preferences, and the formulation of reasonable standards and norms to promote the healthy development of the cigarette industry.

[0005] However, no relevant effective solution is provided in the prior art. Summary of the invention

[0006] In order to solve the above technical problems, the present application provides a method, device and electronic device for judging the quality of cigarettes, so as to at least solve or alleviate the above problems existing in the prior art.

[0007] A method for judging cigarette quality, comprising: Constructing a standard cigarette sample set, the standard cigarette sample set comprising a plurality of standard cigarette samples and a standard physical index of each standard cigarette sample, the standard physical index comprising: at least one of a standard density, a standard water content, a standard draw resistance, a standard ventilation rate, a standard circumference, a standard length, a standard draw resistance, a standard total air permeability and a standard hardness; Determining the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set; Obtain physical indicators of the cigarettes to be tested; The distance between the physical index of the cigarette to be inspected and the boundary of the standard index is calculated to determine the quality of the cigarette to be inspected.

[0008] Optionally, determining the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set includes: Vectorizing the standard physical index of each standard cigarette sample to form a standard index vector; Based on the standard indicator vectors corresponding to all standard cigarette samples, a standard vector set is formed; Based on the standard vector set, the minimum circumscribed shape description including the standard physical indicators of all standard cigarette samples is calculated, and the minimum circumscribed shape description is used as the standard indicator boundary.

[0009] Optionally, the step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Counting the maximum and minimum values ​​of the standard physical indicators of each dimension in the standard vector set to form a maximum value set and a minimum value set; Based on the extreme value set, a minimum circumscribed rectangle is constructed, and the vertex coordinates of the minimum circumscribed rectangle are limited by the elements in the minimum value set and the elements in the maximum value set according to the dimensions of the standard physical indicators.

[0010] Optionally, the step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Counting the centroids of standard physical indicators of all dimensions in the standard vector set; Calculate the distance from the standard physical indicator of each dimension to the center of mass to obtain a distance set; The largest distance is selected from the distance set as the radius of constructing a minimum circumscribed sphere to construct a minimum circumscribed sphere description, and the minimum circumscribed sphere description is used as the standard indicator boundary.

[0011] Optionally, the step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Calculating the distance between different standard indicator vectors in the standard vector set as an indicator similarity measure; For each dimension of the standard physical indicator, multiple standard indicator vectors whose indicator similarity measure is greater than a set similarity measure threshold are selected as its neighbors to form a neighbor set; Based on the neighbor set of each standard indicator vector, a series of line segments or polygons are constructed to form the standard indicator boundary.

[0012] Optionally, the calculating the distance between the physical index of the cigarette to be inspected and the boundary of the standard index to judge the quality of the cigarette to be inspected includes: The distance between each physical indicator of the cigarette to be inspected and the boundary of the standard indicator is calculated to judge the quality of the cigarette to be inspected based on the distance.

[0013] Optionally, the calculating the distance between each physical indicator of the cigarette to be inspected and the boundary of the standard indicator to judge the quality of the cigarette to be inspected based on the distance includes: Taking the minimum value of the distances from all physical indicators to the boundary of the standard indicator, and determining the boundary point on the boundary of the standard indicator corresponding to the minimum value; Calculating the distance between the boundary point and the centroid of the standard indicator boundary, and the distance between the physical indicator and the centroid of the standard indicator boundary when the minimum value is obtained, so as to determine whether the physical indicator is located within the standard indicator boundary when the minimum value is obtained; If it is, then output the minimum value; The point cluster distance threshold is set based on the homogeneity rate so that the minimum value is compared with the point cluster distance threshold. If it is lower than the point cluster distance threshold, the cigarette to be inspected is determined to be qualified; otherwise, the cigarette to be inspected is determined to be qualified.

[0014] Optionally, the method further comprises: Based on the homogeneity rate, multiple point cluster distance thresholds are set, so that the minimum value is compared with the multiple point cluster distance thresholds respectively to obtain the point cluster distance group analysis result; Based on the point cluster distance group analysis results, generating a receiver operating characteristic curve and a precision-recall curve; The accuracy of determining whether the cigarette to be inspected is qualified is determined based on the receiver operating characteristic curve and the precision-recall curve.

[0015] A device for judging cigarette quality, comprising: A first program unit is used to construct a standard cigarette sample set, wherein the standard cigarette sample set includes a plurality of standard cigarette samples and a standard physical index of each standard cigarette sample, wherein the standard physical index includes at least one of a standard density, a standard moisture content, a standard draw resistance, a standard ventilation rate, a standard circumference, a standard length, a standard draw resistance, a standard total air permeability and a standard hardness; A second program unit is used to determine the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set; The third program unit is used to obtain the physical indicators of the cigarette to be tested; The fourth program unit is used to calculate the distance between the physical index of the cigarette to be tested and the boundary of the standard index to determine the quality of the cigarette to be tested.

[0016] An electronic device comprises: a memory and a processor, wherein the memory stores a computer executable program, and the processor is used to execute the computer executable program to implement the method described in any one of the embodiments of the present application.

[0017] In the present invention, by constructing a standard cigarette sample set, a plurality of standard cigarette samples with clear standard physical indicators (covering standard density, standard water content, standard draw resistance, etc.) are collected. This is equivalent to establishing a reliable reference benchmark for subsequent judgments, solving the problem of lack of reference standards in the past when there was no relevant effective solution. In the past, due to the lack of such a clear and systematic set of standard samples, it was difficult to accurately measure whether cigarette products were qualified and what their quality was. This step makes the judgment have a specific basis object, and relies on the abstract concept of qualified threshold value on actual samples and their indicators. For example, cigarettes of different brands and batches that meet high quality requirements can be included as standard cigarette samples, and the standard density, water content and other indicators corresponding to each sample are clearly recorded to form a comprehensive reference system.

[0018] Determining the standard index boundary based on the standard physical indexes of all samples in the constructed standard cigarette sample set is a key quantitative process. It integrates the scattered sample indexes and obtains a boundary range that can define whether the sample is qualified or not through scientific and reasonable statistical or analytical methods.

[0019] After obtaining the physical indicators of the cigarettes to be tested, the distance between them and the boundary of the standard indicators is calculated to judge the quality. This method establishes a direct connection between the cigarettes to be tested and the standard samples, and uses intuitive data distance to measure their quality. Regardless of density, ventilation rate or other physical indicators, as long as the corresponding indicators of the cigarettes to be tested are within a reasonable range from the boundary of the standard indicators, their quality can be considered qualified, otherwise it is unqualified.

[0020] To this end, the above-mentioned judgment method can ensure that the physical properties of each cigarette are measured according to the established standard indicator boundaries during the production process. As long as this boundary range is met, it means that cigarettes produced in different batches and different time periods can remain similar in key physical indicators such as density and draw resistance, thereby ensuring the consistency of product quality. It avoids the uneven quality caused by the lack of an effective judgment method, so that the quality of the cigarettes purchased by consumers can be relatively stable in terms of physical performance-related quality, and there will be no obvious taste differences and other quality problems. Reasonable standard indicator boundary determination and quality judgment based on this will help ensure that the combustion characteristics of cigarettes meet safety requirements. For example, if physical indicators such as ventilation rate and water content do not meet the standards, it may lead to incomplete combustion, increased harmful components and other safety hazards. This method can screen out cigarettes that do not meet the standard physical indicator boundaries, so that they cannot enter the market, thereby improving the overall safety of cigarette products and protecting consumer health. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings constituting a part of this application are used to provide a further understanding of this application, so that other features, purposes and advantages of this application become more obvious. The schematic embodiment drawings and their descriptions of this application are used to explain this application and do not constitute an improper limitation on this application. In the drawings: Figure 1 This is a schematic flow chart of a method for determining cigarette quality according to an embodiment of the present application.

[0022] Figure 2-Figure 4 Shown are ROC and PROC examples, and G-measures index examples. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0024] Figure 1 This is a schematic diagram of a method for judging cigarette quality according to an embodiment of the present application. Figure 1 As shown, it includes: Constructing a standard cigarette sample set, the standard cigarette sample set comprising a plurality of standard cigarette samples and a standard physical index of each standard cigarette sample, the standard physical index comprising: at least one of a standard density, a standard water content, a standard draw resistance, a standard ventilation rate, a standard circumference, a standard length, a standard draw resistance, a standard total air permeability and a standard hardness; Determining the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set; Obtain physical indicators of the cigarettes to be tested; The distance between the physical index of the cigarette to be inspected and the boundary of the standard index is calculated to determine the quality of the cigarette to be inspected.

[0025] In the present invention, by constructing a standard cigarette sample set, a plurality of standard cigarette samples with clear standard physical indicators (covering standard density, standard water content, standard draw resistance, etc.) are collected. This is equivalent to establishing a reliable reference benchmark for subsequent judgments, solving the problem of lack of reference standards in the past when there was no relevant effective solution. In the past, due to the lack of such a clear and systematic set of standard samples, it was difficult to accurately measure whether cigarette products were qualified and what their quality was. This step makes the judgment have a specific basis object, and relies on the abstract concept of qualified threshold value on actual samples and their indicators. For example, cigarettes of different brands and batches that meet high quality requirements can be included as standard cigarette samples, and the standard density, water content and other indicators corresponding to each sample are clearly recorded to form a comprehensive reference system.

[0026] Determining the standard index boundary based on the standard physical indexes of all samples in the constructed standard cigarette sample set is a key quantitative process. It integrates the scattered sample indexes and obtains a boundary range that can define whether the sample is qualified or not through scientific and reasonable statistical or analytical methods.

[0027] After obtaining the physical indicators of the cigarettes to be tested, the distance between them and the boundary of the standard indicators is calculated to judge the quality. This method establishes a direct connection between the cigarettes to be tested and the standard samples, and uses intuitive data distance to measure their quality. Regardless of density, ventilation rate or other physical indicators, as long as the corresponding indicators of the cigarettes to be tested are within a reasonable range from the boundary of the standard indicators, their quality can be considered qualified, otherwise it is unqualified.

[0028] To this end, the above-mentioned judgment method can ensure that the physical properties of each cigarette are measured according to the established standard indicator boundaries during the production process. As long as this boundary range is met, it means that cigarettes produced in different batches and different time periods can remain similar in key physical indicators such as density and draw resistance, thereby ensuring the consistency of product quality. It avoids the uneven quality caused by the lack of an effective judgment method, so that the quality of the cigarettes purchased by consumers can be relatively stable in terms of physical performance-related quality, and there will be no obvious taste differences and other quality problems. Reasonable standard indicator boundary determination and quality judgment based on this will help ensure that the combustion characteristics of cigarettes meet safety requirements. For example, if physical indicators such as ventilation rate and water content do not meet the standards, it may lead to incomplete combustion, increased harmful components and other safety hazards. This method can screen out cigarettes that do not meet the standard physical indicator boundaries, so that they cannot enter the market, thereby improving the overall safety of cigarette products and protecting consumer health.

[0029] Optionally, determining the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set includes: Vectorizing the standard physical index of each standard cigarette sample to form a standard index vector; Based on the standard indicator vectors corresponding to all standard cigarette samples, a standard vector set is formed; Based on the standard vector set, the minimum circumscribed shape description including the standard physical indicators of all standard cigarette samples is calculated, and the minimum circumscribed shape description is used as the standard indicator boundary.

[0030] Optionally, the step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Counting the maximum and minimum values ​​of the standard physical indicators of each dimension in the standard vector set to form a maximum value set and a minimum value set; Based on the extreme value set, a minimum circumscribed rectangle is constructed, and the vertex coordinates of the minimum circumscribed rectangle are limited by the elements in the minimum value set and the elements in the maximum value set according to the dimensions of the standard physical indicators.

[0031] Optionally, in a specific application scenario, the standard vector set is recorded as: , where each It is a vector composed of all standard physical indicators of the jth standard cigarette sample, that is, the standard indicator vector. If there are d standard physical indicators, the dimension of the standard indicator vector is d, which can be generally recorded as: , n is the total number of standard cigarette samples, n≥j≥1.

[0032] When the maximum and minimum values ​​of the standard physical indicators of each dimension in the standard vector set are counted to form a maximum value set and a minimum value set, wherein, and are the minimum and maximum values ​​of the standard physical indicators of any standard cigarette sample in the i-th dimension (i = 1, 2, ..., d).

[0033] Based on the extreme value set, a minimum bounding rectangle is constructed. When the vertex coordinates of the minimum bounding rectangle are limited by the elements in the minimum value set and the elements in the maximum value set according to the dimensions of the standard physical index, the minimum bounding rectangle B is recorded as: .

[0034] To this end, the standard physical indicators of each standard cigarette sample are vectorized to form a standard indicator vector, so that the originally scattered and different types of physical indicators (such as density is a specific value, water content is a percentage, and other data with different forms and meanings) can be expressed in a unified vector form. In this way, whether it is the use of computer algorithms for data processing or the use of mathematical models for analysis, it becomes more convenient and efficient.

[0035] In addition, the quality of a cigarette is determined by multiple physical indicators. The vectorization process integrates these related but relatively independent physical indicators into a vector space, which can fully reflect the characteristics of cigarette samples in various physical dimensions. It avoids the one-sidedness that may occur when analyzing each physical indicator separately, but starts from a holistic perspective and comprehensively considers the mutual influence and synergy between various indicators, so that the standard indicator boundaries determined based on these vectors can more accurately reflect the comprehensive physical characteristics that high-quality cigarettes should have.

[0036] Furthermore, aggregating the standard indicator vectors corresponding to all standard cigarette samples into a standard vector set is equivalent to constructing a feature space specifically for standard cigarette samples. In this space, the distribution of different high-quality cigarette samples in various physical indicator dimensions can be clearly seen. This helps to grasp the overall range and combination of various physical indicators of cigarette products when they meet high-quality requirements from a macro perspective, and provides a comprehensive data basis for further exploring the appropriate boundaries of standard indicators. For example, by observing this standard vector set, the commonalities and differences in the matching of indicators such as density and draw resistance of different brands of high-quality cigarettes can be found, and then the boundary range that can cover the characteristics of these high-quality samples can be more reasonably determined.

[0037] Finally, the minimum enclosing rectangle is constructed based on the extreme value as the standard indicator boundary. By statistically analyzing the maximum and minimum values ​​of the standard physical indicators of each dimension in the standard vector set, the maximum value set and the minimum value set are formed, and then the minimum enclosing rectangle is constructed to determine the standard indicator boundary. This method is very intuitive and clear. For quality control personnel and related technical personnel in the production process, it is easy to understand and master its judgment logic. For example, in an actual cigarette production line, quality inspection personnel can easily compare the maximum and minimum value ranges corresponding to each physical indicator dimension (that is, the boundary of the minimum enclosing rectangle) to quickly determine whether the physical indicators of the cigarettes to be inspected exceed this boundary, thereby efficiently making a preliminary judgment on the product quality without complex calculations or professional data analysis background knowledge. Moreover, the construction of the minimum enclosing rectangle can cover the standard physical indicator conditions of all standard cigarette samples in a relatively loose but reasonable way. It ensures that the physical indicator combinations within this boundary range are consistent with the characteristics of high-quality cigarette samples. As long as the physical indicators of the cigarettes to be inspected fall within the range defined by this minimum enclosing rectangle, there is a high probability that they have similar quality performance. At the same time, it also avoids overly strict or unreasonable boundary setting that may lead to misjudgment of some cigarette products that should belong to the qualified high-quality category. On the basis of ensuring product quality consistency, it takes into account the effective inclusion of different high-quality sample characteristics, which helps to improve the accuracy and rationality of quality judgments.

[0038] Since the vertex coordinates of the minimum circumscribed rectangle are limited by the minimum and maximum elements corresponding to the dimensions of each standard physical indicator, this achieves the comprehensive consideration and boundary definition of multi-dimensional physical indicators. Cigarette quality is the result of the combined effect of multiple physical indicators. This boundary built on multiple dimensions can comprehensively measure whether the cigarette is qualified from the perspective of various physical properties. For example, it will not only focus on whether the draw resistance meets the standard and ignore other indicators such as water content, but will comprehensively consider the reasonable range of various physical indicators to fully judge the overall quality of the cigarette, ensuring the comprehensiveness and scientificity of the quality judgment.

[0039] Optionally, the step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Counting the centroids of standard physical indicators of all dimensions in the standard vector set; Calculate the distance from the standard physical indicator of each dimension to the center of mass to obtain a distance set; The largest distance is selected from the distance set as the radius of constructing a minimum circumscribed sphere to construct a minimum circumscribed sphere description, and the minimum circumscribed sphere description is used as the standard indicator boundary.

[0040] The standard indicator boundaries It is expressed by the minimum circumscribed sphere description, Represents the bounds of the minimum circumscribed sphere. yes Any one of . yes The centroid of is calculated as follows:

[0041] r is the radius described by the smallest circumscribed sphere, calculated using the formula:

[0042] in, yes The distance from the jth indicator vector to the centroid is calculated as follows:

[0043] To this end, by calculating the centroid of the standard physical indicators of all dimensions in the standard vector set, it is equivalent to finding a "central balance point" of these high-quality cigarette samples in the multi-dimensional physical indicator space. This centroid can comprehensively reflect the average characteristic performance of all standard cigarette samples in various physical indicator dimensions. For example, for the vector space composed of multiple physical indicators such as density, water content, and draw resistance, the position of the centroid represents the core position where these indicator values ​​tend to gather as a whole in the group of cigarettes that meet high-quality requirements. It helps to grasp the typical characteristics of the physical indicators of high-quality cigarettes from a macro perspective and provide an important reference benchmark for the subsequent definition of reasonable standard indicator boundaries.

[0044] Furthermore, using the centroid as a reference, it is easy to compare the differences between each standard cigarette sample and the overall average characteristics. In the process of quality control and quality research, it can be clearly seen in which physical indicator dimensions a specific sample deviates from the center position of the whole, and the degree of deviation. This is of great significance for in-depth understanding of the subtle differences between different high-quality cigarette samples, and further analysis of the impact of these differences on quality-related aspects such as cigarette taste and burning characteristics. It enables more comprehensive consideration of the diversity within the high-quality sample group and its correlation with the overall characteristics when determining the boundaries of standard indicators in the future.

[0045] In addition, the distance from the standard physical indicator of each dimension to the centroid is calculated to quantify the relationship between the originally relatively abstract sample and the overall characteristics. This distance value intuitively reflects the degree to which each physical indicator deviates from the average level in its corresponding dimension, and presents the discreteness of each sample characteristic in the entire high-quality cigarette group in the form of specific data. For example, the distance from the draw resistance indicator to the centroid of a certain cigarette sample is large, indicating that the sample has a significant difference in the average draw resistance of other high-quality samples in the dimension of draw resistance. This quantitative method helps to more accurately analyze the uniqueness of each sample and its impact on the overall sample set characteristics, and provides detailed data support for reasonable judgment of the qualified range of product quality.

[0046] Moreover, after obtaining the distance set from each dimensional indicator to the centroid, the characteristic distribution law of the entire standard cigarette sample set can be comprehensively analyzed from multiple dimensions. The distribution trend of different physical indicators in terms of distance from the centroid can be observed. For example, whether some indicators are generally close to the centroid means that these indicators are relatively more stable and concentrated in high-quality cigarettes; while other indicators have a larger range of distance from the centroid, indicating that they have a richer value situation in the high-quality sample group. Based on these laws, when determining the boundaries of standard indicators, we can more scientifically consider which indicators require relatively strict boundary limitations and which can be appropriately relaxed according to the actual distribution characteristics of each indicator, so that the final standard indicator boundaries are more in line with the actual quality characteristic distribution of high-quality cigarettes.

[0047] Finally, the minimum circumscribed sphere description is constructed with the maximum distance selected from the distance set as the radius of the minimum circumscribed sphere, and it is used as the standard indicator boundary. This method can ensure that the standard physical indicators of all standard cigarette samples are included in this boundary as much as possible. From the perspective of spatial geometry, the minimum circumscribed sphere includes the multi-dimensional physical indicator characteristics of high-quality cigarette samples in a relatively loose but reasonable form, avoiding the erroneous exclusion of some cigarette samples that are still high-quality despite having certain differences from the average characteristics in individual dimensions due to too narrow boundary settings, ensuring the effective inclusion of the overall characteristics of the high-quality sample group, and helping to improve the accuracy and comprehensiveness of quality judgment. The quality of cigarettes is determined by the synergistic effect of multiple physical indicators. Using the minimum circumscribed sphere to define the standard indicator boundary is based on a holistic, multi-dimensional spatial perspective. Unlike only considering the maximum and minimum value range of a single dimension, the boundary of the sphere comprehensively considers the mutual relationship and overall distribution of various physical indicators in space. When judging the quality of cigarettes, it is not to look at whether each physical indicator meets the standard in isolation, but to measure whether the combination of physical indicators falls within this reasonable sphere from a multi-dimensional collaborative perspective. This is more in line with the actual situation that cigarette quality is affected by a combination of multiple factors, and can thus more scientifically determine whether the product meets high quality requirements.

[0048] In practical applications, the minimum circumscribed sphere is used as the standard indicator boundary, and its judgment logic is relatively simple and clear. For quality inspectors, they only need to calculate the distance from the physical indicator vector of the cigarette to be inspected to the center of mass, and then compare it with the predetermined sphere radius (i.e., the maximum distance), and they can intuitively determine whether the cigarette is within the standard indicator boundary. The operation is simple and easy, and does not require complicated steps such as complex multi-dimensional interval comparison, which is conducive to improving the efficiency of quality inspection in the production process and ensuring the consistency and stability of cigarette product quality.

[0049] Optionally, the step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Calculating the distance between different standard indicator vectors in the standard vector set as an indicator similarity measure; For each dimension of the standard physical indicator, multiple standard indicator vectors whose indicator similarity measure is greater than a set similarity measure threshold are selected as its neighbors to form a neighbor set; Based on the neighbor set of each standard indicator vector, a series of line segments or polygons are constructed to form the standard indicator boundary.

[0050] Optionally, when calculating the distance between different standard indicator vectors in the standard vector set as an indicator similarity measure, based on , Represents the standard index vector The distance between them, k is the kth neighbor, that is, the kth standard indicator vector; For each dimension of the standard physical indicator, multiple standard indicator vectors whose indicator similarity measure is greater than the set similarity measure threshold are selected as its neighbors to form a neighbor set. The neighbor set is recorded as: , which means that for the i-th standard indicator vector, find the k nearest neighbors to it, where is the set of k nearest neighbors of the i-th standard indicator vector, k is the preset number of neighbors, and k is greater than or equal to 1.

[0051] Based on the neighbor set of each standard indicator vector, a series of line segments or polygons are constructed to form the standard indicator boundary B. express.

[0052] To this end, by calculating the distance between different standard indicator vectors in the standard vector set as a measure of indicator similarity, the similarity of each standard cigarette sample in terms of multi-dimensional physical indicators can be presented in an accurate and quantitative manner. Different combinations of physical indicators constitute different vectors, and the size of the distance between vectors intuitively reflects the differences in the comprehensive performance of the corresponding samples in multiple key physical indicators such as density, water content, and resistance to draw. For example, if the distance between two standard indicator vectors is close, it means that the two cigarette samples are closer in the values ​​of various physical indicators, and their product quality is more similar in terms of physical properties. This quantitative method provides a solid data foundation for the subsequent in-depth analysis of the relationship between samples, and helps to more accurately grasp the structural characteristics within the high-quality cigarette sample group.

[0053] Furthermore, after calculating and analyzing the distances between all standard indicator vectors, the distribution patterns of the entire set of standard cigarette samples in the multi-dimensional physical indicator space can be mined. It can be observed which areas have relatively more clustered samples, that is, the distances between vectors are generally smaller, indicating that the physical indicator combinations corresponding to these areas appear more frequently and are more typical in high-quality cigarettes; and which areas have more dispersed samples, reflecting that the corresponding physical indicator values ​​are somewhat diverse in high-quality cigarettes. Based on such patterns, the subsequent determination of the boundaries of standard indicators can better fit the actual sample distribution, scientifically and reasonably define the range of physical indicators that qualified products should have, and ensure that the boundary setting covers both common high-quality characteristics and certain reasonable differences.

[0054] In addition, for the standard physical indicators of each dimension, multiple standard indicator vectors with indicator similarity measures greater than the set similarity measure threshold are selected as their neighbors and a neighbor set is formed. This approach is equivalent to focusing on a group of samples with high similarity to each sample in the local dimension from the overall sample set. In the multi-dimensional cigarette physical indicator system, different dimensions have different effects on cigarette quality. By constructing a neighbor set, we can deeply analyze which samples have similar performances and the subtle differences between them in a specific physical indicator dimension. For example, for the dimension of draw resistance, the neighbor set can be used to find cigarette samples with similar draw resistance indicators, and then study their matching in other related physical indicators, which helps to more carefully grasp the relationship between the mutual influence and synergy between the physical indicators of each dimension, and provide a richer local perspective for comprehensive consideration of cigarette quality.

[0055] Finally, the standard indicator boundaries are determined based on the neighbor set, so that the determination of the boundaries is no longer a simple global unified consideration, but can be set in a targeted manner according to the specific similarity distribution of physical indicators in different dimensions. Physical indicators in different dimensions may have different reasonable fluctuation ranges and similarity requirements due to their own nature and the importance of their impact on cigarette quality. Through the neighbor set, the tightness of the boundary can be flexibly adjusted according to the actual sample similarity of each dimension to ensure that the indicator value range that meets the high quality requirements can be accurately captured in each dimension, avoiding the problem of insufficient consideration or excessive constraints on certain dimensional characteristics that may result from the use of a single fixed standard, so that the final standard indicator boundary is more in line with the complexity of cigarette quality being affected by multiple factors.

[0056] Furthermore, based on the neighbor set of each standard indicator vector, a series of line segments or polygons are constructed to form the standard indicator boundary. This method can closely fit the actual distribution pattern of standard cigarette samples in the multi-dimensional physical indicator space. Since the neighbor set itself reflects the local similarity between samples, by connecting the vectors corresponding to these samples with similar characteristics with line segments or polygons, the formed boundary can well surround the high-quality cigarette sample group and accurately outline their distribution range. Compared with some simple regular shape boundaries (such as rectangles, spheres, etc.), the boundaries constructed in this way can better adapt to the actual distribution characteristics of samples. Whether it is an area where samples are relatively concentrated or an area with a certain degree of dispersion, it can be more reasonably covered, effectively reducing the situation of misjudgment of product quality due to unreasonable boundary shape, and improving the accuracy of quality judgment.

[0057] The boundary formed by line segments or polygons is synthesized from multiple standard indicator vectors and their neighbor sets, reflecting the synergy and mutual relationship between multi-dimensional physical indicators. The quality of cigarettes is not determined by a single physical indicator alone, but is the result of the coordinated influence of multiple indicators. When judging the quality of the cigarettes to be inspected, the boundary constructed in this way will comprehensively consider the position of its physical indicator vector in the entire area defined by the line segments or polygons, that is, to measure whether it meets the high quality requirements from the perspective of multi-dimensional coordination. For example, all the physical indicators of a cigarette need to be satisfied at the same time within this complex boundary constructed by multi-dimensional line segments and polygons before it can be judged as a qualified product. This multi-dimensional collaborative boundary definition method is more in line with the internal logic of cigarette quality formation and ensures the scientificity and comprehensiveness of quality judgment.

[0058] Optionally, the calculating the distance between the physical index of the cigarette to be inspected and the boundary of the standard index to judge the quality of the cigarette to be inspected includes: The distance between each physical indicator of the cigarette to be inspected and the boundary of the standard indicator is calculated to judge the quality of the cigarette to be inspected based on the distance.

[0059] Optionally, the calculating the distance between each physical indicator of the cigarette to be inspected and the boundary of the standard indicator to judge the quality of the cigarette to be inspected based on the distance includes: Taking the minimum value of the distances from all physical indicators to the boundary of the standard indicator, and determining the boundary point on the boundary of the standard indicator corresponding to the minimum value; Calculating the distance between the boundary point and the centroid of the standard indicator boundary, and the distance between the physical indicator and the centroid of the standard indicator boundary when the minimum value is obtained, so as to determine whether the physical indicator is located within the standard indicator boundary when the minimum value is obtained; If it is, then output the minimum value; The point cluster distance threshold is set based on the homogeneity rate so that the minimum value is compared with the point cluster distance threshold. If it is lower than the point cluster distance threshold, the cigarette to be inspected is determined to be qualified; otherwise, the cigarette to be inspected is determined to be qualified.

[0060] Optionally, in a specific implementation, the minimum value of the distances from all physical indicators to the boundary of the standard indicator is recorded as: , Indicates that from all physical indicators Medium to standard indicator boundaries The minimum distance, is the standard indicator boundary To find all the points on the boundary, , calculate them to Distance , and then take the minimum value of these distances, which is the shortest distance , and the boundary point corresponding to the minimum distance is recorded as .

[0061] Calculate the distance between the boundary point and the centroid of the standard indicator boundary, and the distance between the physical indicator and the centroid of the standard indicator boundary when the minimum value is obtained, so as to determine whether the physical indicator is located within the standard indicator boundary when the minimum value is obtained. To the centroid Distance , as well as To the centroid Distance .

[0062] if , but Outside the boundary, output distance .

[0063] if , but Within the boundary, no distance is output , Right now .

[0064] Optionally, the method further comprises: setting a plurality of point cluster distance thresholds based on the homogeneity rate, so that the minimum value is respectively compared with the plurality of point cluster distance thresholds to obtain a point cluster distance group analysis result; Based on the point cluster distance group analysis results, generating a receiver operating characteristic curve and a precision-recall curve; The accuracy of determining whether the cigarette to be inspected is qualified is determined based on the receiver operating characteristic curve (denoted as ROC) and the precision-recall curve (denoted as PR-ROC).

[0065] When judging the accuracy of whether the cigarette to be tested is qualified according to the receiver operating characteristic curve (denoted as ROC) and the precision-recall curve (denoted as PR-ROC), the classification accuracy ROC curve and the PR-ROC curve are traversed from low to high different thresholds to form a comprehensive G-measure index to judge the accuracy of whether the cigarette to be tested is qualified.

[0066] For example, Figure 2-Figure 4 The following are examples of ROC and PROC, and examples of G-measures index. As can be seen from the figure, the optimal threshold of the point cluster distance is 0.176, and the accuracy of judging whether the cigarette sample is qualified or not by this threshold is about 80.95%. The AUC area under the ROC curve of this brand of cigarettes is 0.82, indicating that the threshold classification effect of this cigarette sample group is good.

[0067] To this end, by calculating the distance from each physical indicator of the cigarette to be tested to the boundary of the standard indicator, the degree of difference in various physical properties between the cigarette to be tested and the standard high-quality cigarettes can be comprehensively and meticulously measured from multiple dimensions. The quality of cigarettes is determined by the joint action of multiple physical indicators. For example, indicators such as density, draw resistance, and water content have a key impact on its combustion characteristics, taste, etc. By calculating the distance from each physical indicator to the boundary separately, it is possible to know exactly whether the cigarette to be tested is close to or deviates from the standard requirements at each specific indicator level, avoiding the details that may be missed due to only a single dimension or general comprehensive judgment, providing a rich and accurate data basis for the accurate evaluation of cigarette quality, and helping to more accurately grasp the actual quality of the cigarettes to be tested.

[0068] Furthermore, this method of calculating distances separately helps to quickly locate the root cause of the quality problems of the cigarettes to be tested. If the distance between the draw resistance index of a cigarette to be tested and the boundary of the standard index is large, while the distances of other physical indicators are relatively reasonable, then we can focus on whether there are deviations in the process, raw materials and other factors related to the draw resistance in the production process, so as to make targeted adjustments and improvements. For cigarette manufacturers, this can more efficiently detect quality risks, optimize production processes, and improve the accuracy and effectiveness of overall product quality control.

[0069] Take the minimum value of the distance from all physical indicators to the boundary of the standard indicator, and determine the corresponding boundary point on the boundary of the standard indicator when the minimum value is obtained. This operation is equivalent to screening out the key dimensions that are closest to or most deviated from the standard sample from the differences in many physical indicators. In multi-dimensional comparisons, the physical indicator corresponding to this minimum value is often the key to highlighting the difference between the quality characteristics of the cigarettes to be tested and the standard requirements. For example, if the minimum distance corresponds to the moisture content indicator, it means that the difference in moisture content is the most prominent in the current judgment, which helps quality inspectors to quickly focus on the key points and give priority to the inspection and analysis of the production links related to this key indicator, thereby improving the efficiency of quality judgment and problem solving.

[0070] Further calculations are made to calculate the distance between the boundary point and the centroid of the standard indicator boundary, as well as the distance between the physical indicator and the centroid of the standard indicator boundary when the minimum value is obtained, so as to determine whether the physical indicator is within the standard indicator boundary when the minimum value is obtained. This judgment method based on the relationship between the centroid and the boundary point considers the degree of fit between the cigarettes to be tested and the characteristics of the standard high-quality cigarette group in key difference dimensions from the perspective of spatial geometry and overall sample characteristic distribution. The centroid represents the overall center position of the standard sample. By comparing these two distances, it can be determined whether the key physical indicator is within or outside the reasonable area defined by the standard sample, providing a more rigorous and in-depth basis for subsequent accurate judgment of cigarette quality, and avoiding possible misjudgments caused by simple comparison of distance sizes.

[0071] The point cluster distance threshold is set based on the homogeneity rate, and the calculated minimum value is compared with the threshold to determine whether the cigarettes to be tested are qualified. This fully takes into account the requirement that cigarette products should maintain a certain degree of homogeneity in quality. In actual production and market circulation, consumers expect the cigarette products they buy to have relatively stable and consistent quality. By setting such a threshold, the degree of similarity between the cigarettes to be tested and the standard high-quality cigarette group is controlled as a whole, ensuring that only those cigarettes that are close enough to the standard samples in various physical indicators and meet the homogeneity requirements can be judged as qualified products, which is conducive to ensuring the stability and consistency of cigarette product quality and enhancing consumers' usage experience and trust in product brands.

[0072] The homogeneity rate can be flexibly adjusted according to factors such as different cigarette brand positioning, market segmentation, and quality control objectives, and the corresponding point cluster distance threshold is also flexible. For example, for high-end cigarette products with extremely strict quality requirements, the homogeneity rate requirements may be higher, and the corresponding point cluster distance threshold will be set smaller to screen out products with quality closer to the perfect standard; while for some cigarette products targeting the mass market and focusing more on cost-effectiveness, the homogeneity rate requirements can be appropriately relaxed and the threshold adjusted accordingly. This flexibility makes the determination method widely applicable to various cigarette production and quality control scenarios with different quality standards, enhancing its adaptability and versatility in industry applications.

[0073] By comparing with the point cluster distance threshold, the final judgment of whether the cigarette to be inspected is qualified or not provides an objective and quantitative judgment standard. It no longer relies on subjective experience judgment or vague qualitative description, but makes clear judgments based on clear distance threshold values, reducing the interference of human factors on the quality judgment results, making the judgment process of cigarette product quality more scientific, fair and accurate, which helps to standardize the quality inspection process of the entire cigarette industry and promote the industry to develop in a more standardized and normalized direction.

[0074] The embodiment of the present application also provides a device for judging cigarette quality, which includes: A first program unit is used to construct a standard cigarette sample set, wherein the standard cigarette sample set includes a plurality of standard cigarette samples and a standard physical index of each standard cigarette sample, wherein the standard physical index includes at least one of a standard density, a standard moisture content, a standard draw resistance, a standard ventilation rate, a standard circumference, a standard length, a standard draw resistance, a standard total air permeability and a standard hardness; A second program unit is used to determine the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set; The third program unit is used to obtain the physical indicators of the cigarette to be tested; The fourth program unit is used to calculate the distance between the physical index of the cigarette to be tested and the boundary of the standard index to determine the quality of the cigarette to be tested.

[0075] An embodiment of the present application further provides an electronic device, comprising: a memory and a processor, wherein the memory stores a computer executable program, and the processor is used to execute the computer executable program to implement any method described in the embodiments of the present application.

[0076] For an exemplary description of the above-mentioned apparatus and electronic device, please refer to the above-mentioned Figure 1 .

[0077] In summary, in the method for judging the quality of cigarettes provided in the present application, in the step of constructing a set of standard cigarette samples, it is pointed out that the standard physical indicators cover at least one of the multiple aspects such as standard density, standard water content, standard draw resistance, standard ventilation rate, standard circumference, standard length, standard draw resistance, standard total air permeability and standard hardness. From the very beginning, multiple physical indicators closely related to the quality of cigarettes were taken into consideration, and the standard cigarette samples were not defined by focusing on a single indicator. This laid the foundation for subsequent comprehensive judgments, embodied the description and definition of the characteristics that high-quality cigarettes should possess from a multi-dimensional perspective, and was in line with the concept of comprehensively judging the quality of cigarettes based on multiple physical indicators.

[0078] Furthermore, the process of determining the boundaries of standard indicators based on the standard physical indicators of all samples in the standard cigarette sample set is actually the process of integrating and analyzing multiple physical indicators. Whether it is by calculating the minimum circumscribed shape description after vectorization (such as the minimum circumscribed rectangle, the minimum circumscribed sphere, or constructing line segments or polygons based on neighbor sets, etc.), it is processing standard indicator vectors containing physical indicator information in multiple dimensions. These operations combine the values ​​of each sample on different physical indicators to determine a boundary range that can cover the characteristics of all high-quality cigarette samples. This boundary is based on the synergy of multiple indicators and is not limited to a single physical indicator. It essentially realizes the integrated use of multiple physical indicators to determine a multi-dimensional judgment boundary.

[0079] Finally, obtaining the physical indicators of the cigarettes to be tested is also a process of collecting multi-dimensional information, involving data from multiple physical aspects. When subsequently calculating the distance between the physical indicators of the cigarettes to be tested and the boundaries of the standard indicators, the cigarettes to be tested are also regarded as an object with multiple physical indicator dimensions to measure the differences between it and the boundaries defined by the standard samples in each dimension. Moreover, in the specific operation of calculating the distance to judge the quality (for example, taking the minimum value of the distance from all physical indicators to the boundary, and then combining it with multiple factors such as the distance from the center of mass to judge, and setting the threshold comparison through the homogeneity rate, etc.), all are from a holistic and multi-dimensional perspective, comprehensively considering the impact of each physical indicator, and the final distance value is the result obtained after integrating the overall performance of multiple physical indicators, which is used as the basis for judging the quality of cigarettes, emphasizing that a comprehensive judgment is made through a multi-dimensional vector composed of multiple physical indicators.

[0080] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for judging the quality of cigarettes, characterized in that: include: Constructing a standard cigarette sample set, the standard cigarette sample set comprising a plurality of standard cigarette samples and a standard physical index of each standard cigarette sample, the standard physical index comprising: at least one of a standard density, a standard water content, a standard draw resistance, a standard ventilation rate, a standard circumference, a standard length, a standard draw resistance, a standard total air permeability and a standard hardness; Determining the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set; Obtain physical indicators of the cigarettes to be tested; The distance between the physical index of the cigarette to be inspected and the boundary of the standard index is calculated to determine the quality of the cigarette to be inspected.

2. A method for judging cigarette quality according to claim 1, characterized in that: The step of determining the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set comprises: Vectorizing the standard physical index of each standard cigarette sample to form a standard index vector; Based on the standard indicator vectors corresponding to all standard cigarette samples, a standard vector set is formed; Based on the standard vector set, the minimum circumscribed shape description including the standard physical indicators of all standard cigarette samples is calculated, and the minimum circumscribed shape description is used as the standard indicator boundary.

3. A method for judging cigarette quality according to claim 2, characterized in that: The step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Counting the maximum and minimum values ​​of the standard physical indicators of each dimension in the standard vector set to form a maximum value set and a minimum value set; Based on the extreme value set, a minimum circumscribed rectangle is constructed, and the vertex coordinates of the minimum circumscribed rectangle are limited by the elements in the minimum value set and the elements in the maximum value set according to the dimensions of the standard physical indicators.

4. A method for judging cigarette quality according to claim 1, characterized in that: The step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Counting the centroids of standard physical indicators of all dimensions in the standard vector set; Calculate the distance from the standard physical indicator of each dimension to the center of mass to obtain a distance set; The largest distance is selected from the distance set as the radius of constructing a minimum circumscribed sphere to construct a minimum circumscribed sphere description, and the minimum circumscribed sphere description is used as the standard indicator boundary.

5. A method for judging cigarette quality according to claim 1, characterized in that: The step of calculating the minimum circumscribed shape description of the standard physical indicators of all standard cigarette samples based on the standard vector set, and using the minimum circumscribed shape description as the standard indicator boundary, comprises: Calculating the distance between different standard indicator vectors in the standard vector set as an indicator similarity measure; For each dimension of the standard physical indicator, multiple standard indicator vectors whose indicator similarity measure is greater than a set similarity measure threshold are selected as its neighbors to form a neighbor set; Based on the neighbor set of each standard indicator vector, a series of line segments or polygons are constructed to form the standard indicator boundary.

6. A method for judging cigarette quality according to claim 1, characterized in that: The calculating the distance between the physical index of the cigarette to be inspected and the boundary of the standard index to judge the quality of the cigarette to be inspected includes: The distance between each physical indicator of the cigarette to be inspected and the boundary of the standard indicator is calculated to judge the quality of the cigarette to be inspected based on the distance.

7. A method for judging cigarette quality according to claim 6, characterized in that: The calculating the distance between each physical indicator of the cigarette to be inspected and the boundary of the standard indicator to judge the quality of the cigarette to be inspected based on the distance includes: Taking the minimum value of the distances from all physical indicators to the boundary of the standard indicator, and determining the boundary point on the boundary of the standard indicator corresponding to the minimum value; Calculating the distance between the boundary point and the centroid of the standard indicator boundary, and the distance between the physical indicator and the centroid of the standard indicator boundary when the minimum value is obtained, so as to determine whether the physical indicator is located within the standard indicator boundary when the minimum value is obtained; If it is, then output the minimum value; The point cluster distance threshold is set based on the homogeneity rate so that the minimum value is compared with the point cluster distance threshold. If it is lower than the point cluster distance threshold, the cigarette to be inspected is determined to be qualified; otherwise, the cigarette to be inspected is determined to be qualified.

8. A method for judging cigarette quality according to claim 7, characterized in that: The method further comprises: Based on the homogeneity rate, multiple point cluster distance thresholds are set, so that the minimum value is compared with the multiple point cluster distance thresholds respectively to obtain the point cluster distance group analysis result; Based on the point cluster distance group analysis results, generating a receiver operating characteristic curve and a precision-recall curve; The accuracy of determining whether the cigarette to be inspected is qualified is determined based on the receiver operating characteristic curve and the precision-recall curve.

9. A device for judging cigarette quality, characterized in that: include: A first program unit is used to construct a standard cigarette sample set, wherein the standard cigarette sample set includes a plurality of standard cigarette samples and a standard physical index of each standard cigarette sample, wherein the standard physical index includes at least one of a standard density, a standard moisture content, a standard draw resistance, a standard ventilation rate, a standard circumference, a standard length, a standard draw resistance, a standard total air permeability and a standard hardness; A second program unit is used to determine the standard index boundary of the standard cigarette sample set according to the standard physical indexes of all the standard cigarette samples in the standard cigarette sample set; The third program unit is used to obtain the physical indicators of the cigarette to be tested; The fourth program unit is used to calculate the distance between the physical index of the cigarette to be tested and the boundary of the standard index to determine the quality of the cigarette to be tested.

10. An electronic device, characterized in that: include: A memory and a processor, wherein the memory stores a computer executable program, and the processor is used to execute the computer executable program to implement the method according to any one of claims 1 to 8.

Images