A method for evaluating silk making process quality

Abstract

The present invention relates to a silk-making process quality evaluation method, which includes: determining silk-making quality evaluation indicators, constructing a hierarchical model according to subordinate relationships for each evaluation indicator, and dividing the hierarchical model from top to bottom into: a target layer, an intermediate layer, and a single indicator layer; performing single indicator quality evaluation; performing comprehensive quality evaluation of intermediate indicators; performing comprehensive quality evaluation of target indicators, which includes: obtaining the comprehensive quality evaluation index of the target indicator according to the comprehensive quality evaluation index of each intermediate indicator of the corresponding intermediate layer, and the weight corresponding to each intermediate indicator. The evaluation method enriches the connotation of process quality evaluation, provides a methodological basis and a practical guide for scientifically and standardizedly carrying out silk-making process quality evaluation work, ensures the standardization, reliability, and comparability of the evaluation results, and is conducive to the comprehensive improvement of the stability, uniformity, coordination, and consistency of the silk-making process quality.

Description

Technical Field

[0001] The invention relates to the technical field of cigarette process evaluation, and in particular to a method for evaluating the quality of a tobacco making process. Background Art

[0002] In the production process of cigarettes, tobacco making is a very important link. It is also the link with the most processes and the longest process flow in the cigarette manufacturing process. Tobacco making is mainly responsible for producing qualified tobacco to provide for the next process. The quality of tobacco processing will not only affect the physical quality of the product, but also the intrinsic quality of the product. The control of tobacco making process is the quality guarantee for brand development. After years of process quality research and practice, the industry has made breakthroughs in key technologies such as raw material processing characteristics, group processing methods, and process quality evaluation methods. Improving and stabilizing the quality of cigarette tobacco making process is the focus of current process quality research and an important link in achieving the product quality and benefit goals of enterprises. With the integration and reorganization of cigarette industrial enterprises, multi-point processing of the same brand has become normalized. There are certain differences in the compliance and stability of cigarette rolling quality between different processing points, different machines, and different batches. This is objective and inevitable. The key to the problem is how to scientifically evaluate and accurately identify this difference, and on this basis, reduce this difference through parameter optimization. However, current domestic research mainly focuses on the study of factors affecting the quality of the silk-making process, and has not been linked to the conformity and stability of quality, nor has a systematic method for the conformity, stability and comprehensive evaluation of the quality of the silk-making process been formed. Summary of the invention

[0003] In order to solve or partially solve the problems existing in the related art, the present invention provides a method for evaluating the quality of a silk making process.

[0004] The present invention provides a method for evaluating the quality of a silk-making process, which comprises the following steps:

[0005] S01), determining silk quality evaluation indicators, and constructing a hierarchical structure model for each evaluation indicator according to a subordinate relationship, wherein the hierarchical structure model is divided from top to bottom into: a target layer, an intermediate layer, and a single indicator layer; the single indicator layer corresponds to a single process parameter or a single quality indicator;

[0006] S02), performing single indicator quality evaluation; the single indicator quality evaluation includes: single indicator quality conformity evaluation, single indicator quality stability evaluation, and single indicator quality comprehensive evaluation;

[0007] S03) Conduct a comprehensive quality evaluation of the intermediate indicators, including:

[0008] According to the single indicator quality comprehensive evaluation index of each single indicator of the corresponding single indicator layer and the weight corresponding to each single indicator, the quality comprehensive evaluation index of the intermediate indicator is obtained;

[0009] S04) Conduct a comprehensive quality evaluation of the target indicators, including:

[0010] The quality comprehensive evaluation index of the target indicator is obtained according to the quality comprehensive evaluation index of each intermediate indicator of the corresponding intermediate layer and the weight corresponding to each intermediate indicator.

[0011] Furthermore, the single indicator quality comprehensive evaluation in step S02) is characterized by a single indicator quality comprehensive evaluation index, and the larger the single indicator quality comprehensive evaluation index is, the better the quality is;

[0012] The single indicator quality comprehensive evaluation index is calculated according to the following formula:

[0013]

[0014] In the formula, C pm is the comprehensive evaluation index of single indicator quality; USL and LSL are the specification limits of single indicators, and the difference between them is the tolerance; μ is the overall mean of single indicator, σ 2 is the overall variance of a single indicator, and m is the target value of a single indicator.

[0015] Furthermore, in the step S03), the following formula is used to calculate the comprehensive quality evaluation index of the intermediate indicator;

[0016] Z M =w1×C pm1 +w2×C pm2 ...+w n ×C pmn

[0017] In the formula, Z M is the comprehensive quality evaluation index of the intermediate indicators, C pm1 , C pm2 ……C pmn is the comprehensive quality evaluation index of each single indicator corresponding to the intermediate indicator, w1, w2, ... w n is the corresponding weight of each single indicator.

[0018] Furthermore, in step S04), the following formula is used to calculate the comprehensive quality evaluation index of the intermediate indicator:

[0019] Z=w M1 ×Z M1 +w M2 ×Z M2 ...+wMn ×Z Mn

[0020] In the formula, Z is the comprehensive quality evaluation index of the target indicator, Z M1 , Z M2 ...Z Mn is the comprehensive quality evaluation index of each intermediate indicator corresponding to the target indicator, w M1 、w M2 ,……w Mn is the corresponding weight of each intermediate indicator.

[0021] Furthermore, the weights corresponding to the single indicators and / or intermediate indicators are determined as follows:

[0022] a) Constructing the judgment matrix

[0023] Compare the evaluation indicators of the same layer in pairs, quantify their relative importance using the nine-level scaling method, and construct the judgment matrix A;

[0024] b) Calculate the index weight coefficient

[0025] Calculate the geometric mean of all elements in each row of the judgment matrix to obtain the unnormalized weight of the corresponding indicator in this row;

[0026] The weight coefficient is normalized, and the unnormalized weight of each indicator is divided by the sum of the unnormalized weights of all indicators to obtain the weight coefficient vector w;

[0027] c) Check the consistency of the judgment matrix.

[0028] Furthermore, the step a) is specifically as follows:

[0029] The five judgments of equally important, slightly important, relatively important, strongly important and extremely important are used to characterize the relative importance of indicators and quantify them according to the following rules: i With indicator A j When equally important ij =1, when A i Ratio A j Slightly important ij =3, when A i Ratio A j Stronger important time ij =5, when A i Ratio A j Strongly important ij =7, when A i Ratio A j Extremely important ij =9, if the relative importance is between the two, use 2, 4, 6, 8 for the score; and a ij=1 / a ji , and construct an n×n judgment matrix A = (a ij ); n is the number of lower-level indicators belonging to the same upper-level indicator.

[0030] Furthermore, the step c) is specifically as follows:

[0031] c1) Calculate the maximum eigenvalue λ of the judgment matrix:

[0032] The judgment matrix A is multiplied by the weight coefficient vector w on the right to obtain the vector Aw; then, the ratio of the corresponding elements of the vector Aw and the weight coefficient vector w is calculated respectively, and the sum of the ratios is the maximum eigenvalue λ;

[0033] c2) Check the consistency of the judgment matrix:

[0034] Calculate the consistency index CI, which is equal to the difference between the maximum eigenvalue λ and the index number n divided by the index number n minus 1;

[0035] According to the number of indicators n, the average random consistency index RI value is obtained;

[0036] Calculate the random consistency ratio CR of the judgment matrix. The random consistency ratio CR is equal to the ratio of the consistency index CI to the average random consistency index RI. Check whether the consistency of the judgment matrix meets the requirements based on the random consistency ratio CR.

[0037] Furthermore, the single indicator quality conformity evaluation in step S02) adopts t test; the single indicator quality stability evaluation adopts χ 2 test.

[0038] Furthermore, the t-test adopts a two-tailed test or a one-tailed test.

[0039] Furthermore, the x 2 The test was one-tailed.

[0040] For the existing silk-making process quality evaluation method, the technical solution provided by the present invention can include the following beneficial effects:

[0041] Analyzing and evaluating the quality of the cigarette silk-making process is conducive to improving quality control capabilities and improving the control accuracy of the quality of in-process and finished products. The present invention constructs a hierarchical model for silk-making process quality evaluation, provides single-indicator quality compliance, stability and comprehensive evaluation methods, as well as intermediate and target indicator quality comprehensive evaluation methods, enriches the connotation of process quality evaluation, and provides a methodological basis and practical operation guide for scientific and standardized silk-making process quality evaluation, ensuring the standardization, reliability and comparability of the evaluation results, which is conducive to the overall improvement of silk-making process quality stability, uniformity, coordination and consistency, and creates conditions for deepening the "transformation from control indicators to control parameters, from result control to process control, and from manual control experience decision-making to automatic control scientific decision-making", so as to better achieve the company's product quality goals and benefit goals.

[0042] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The above and other objects, features and advantages of the present invention will become more apparent through a more detailed description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, wherein like reference numerals generally represent like components throughout the exemplary embodiments of the present invention.

[0044] Figure 1 It is a flowchart of a silk-making process quality evaluation method shown in an embodiment of the present invention;

[0045] Figure 2 It is a hierarchical model for comprehensive evaluation of quality of a certain process in an embodiment of the present invention;

[0046] Figure 3 It is a schematic diagram of the operation flow of using Minitab statistical software to evaluate the quality compliance of the moisture content rate at the blade outlet in the loosening and rehumidification process in an embodiment of the present invention;

[0047] Figure 4 This is a schematic diagram of the operation flow of using Minitab statistical software to evaluate the quality stability of the moisture content rate at the blade outlet in the loosening and rehumidification process in an embodiment of the present invention;

[0048] Figure 5 The figure is a schematic diagram of the operation flow of using Minitab statistical software to comprehensively evaluate the quality of the moisture content rate at the blade outlet in the loosening and rehumidification process in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the" and "the" used in the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0051] It should be understood that although the terms "first", "second", "third", etc. may be used to describe various information in the present invention, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present invention, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0052] The present invention provides a method for evaluating the quality of a silk-making process. Figure 1 The present embodiment provides a method for evaluating the quality of the silk making process, comprising the following steps:

[0053] S01), determining silk quality evaluation indicators, and constructing a hierarchical structure model for each evaluation indicator according to a subordinate relationship, wherein the hierarchical structure model is divided from top to bottom into: a target layer, an intermediate layer, and a single indicator layer; the single indicator layer corresponds to a single process parameter or a single quality indicator;

[0054] S02), performing single indicator quality evaluation; the single indicator quality evaluation includes: single indicator quality conformity evaluation, single indicator quality stability evaluation, and single indicator quality comprehensive evaluation;

[0055] S03) Conduct a comprehensive quality evaluation of the intermediate indicators, including:

[0056] According to the single indicator quality comprehensive evaluation index of each single indicator of the corresponding single indicator layer and the weight corresponding to each single indicator, the quality comprehensive evaluation index of the intermediate indicator is obtained;

[0057] S04) Conduct a comprehensive quality evaluation of the target indicators, including:

[0058] The quality comprehensive evaluation index of the target indicator is obtained according to the quality comprehensive evaluation index of each intermediate indicator of the corresponding intermediate layer and the weight corresponding to each intermediate indicator.

[0059] The silk-making process quality evaluation method provided by the present invention uses mathematical statistics methods to organize product process quality data in a hierarchical manner, and scientifically and normatively carries out silk-making process quality analysis and evaluation, which is beneficial to improving the process quality control capability and improving the control accuracy of the quality of work-in-progress and finished products.

[0060] In order to improve the accuracy and reliability of the evaluation results, the analysis and evaluation data used in the embodiment of the present invention is constructed according to the following collection rules:

[0061] 1) Data source

[0062] Process data is detected by offline detection instruments and online detection instruments. The detection data can be obtained through automatic collection, manual intervention or manual collection. There are detailed calibration and management requirements for measuring instruments and relevant detection operating procedures. According to the characteristics of important process quality indicators and key process parameters of cigarettes, determine the appropriate data collection method. Perform variation source analysis on important parameters and indicators of key processes, and determine the appropriate time interval and sample size for data collection, and standardize the collection frequency, data batch attribution, and the definition of batch valid data and steady-state data.

[0063] 2) Basic principles of automatic data collection

[0064] The data collection system can determine the appropriate collection start time point according to the collection rules; reasonably determine the minimum sample size for each batch of data; the data collection frequency should be able to take into account the sample size and system capabilities and reflect the actual production and processing process.

[0065] Determine the principles for eliminating automatically collected data. Abnormal data caused by the following reasons can be included in the elimination range: abnormal data due to inherent performance of the equipment; abnormal data due to material replacement or abnormality; abnormal data due to abnormality of the automatic collection system itself; abnormal data caused by changing production batches or changing brands; abnormal data due to other reasons.

[0066] 3) Basic principles of manual intervention in automatic data collection

[0067] Automatic data collection with manual intervention should be used in situations where there is no automatic data collection; the working parameters of the testing instruments should be standardized; automatic data collection with manual intervention should have a complete electronic control system, data recording, and backup system; manual data collection should be used in situations where there is no automatic data collection and when there is manual intervention in automatic data collection, and complete records should be kept.

[0068] 4) Inspection standards and operating requirements for testing instruments and online testing instruments

[0069] The measurement management department shall be responsible for the management of testing instruments and online testing instruments, and keep them in good condition (or effective); testing instruments and online testing instruments shall have complete calibration procedures and operating procedures.

[0070] 5) Measurement cycle and measurement conditions

[0071] Determine the appropriate measurement cycle based on the characteristics and importance of the parameters and indicators; the statistical sample of each key parameter and indicator should maintain at least 60 data points.

[0072] The above step S01 is to construct a hierarchical structure model based on the hierarchical analysis method for each evaluation index according to the subordinate relationship. The key to the comprehensive evaluation of multiple indicators is to determine the relative importance of each quality index, that is, the weight. The methods for determining the weight of evaluation indicators can be divided into two types: subjective weighting method and objective weighting method. The subjective weighting method is characterized by concentrating the experience and opinions of experts to determine the weight of each indicator; the objective weighting method is to determine the weight based on the statistical properties of the evaluation indicators, and the size of the weight depends on the distribution of the original data of the evaluation indicators. Commonly used subjective weighting methods include: hierarchical analysis method, Delphi method, quantitative scoring method, etc.; commonly used objective weighting methods include: deviation weight method, correlation analysis method, principal component analysis method, etc. Among them, the hierarchical analysis method is a subjective weighting method that combines qualitative and quantitative methods. It was proposed by American operations researcher Sadie in the 1970s and is a commonly used indicator weighting method. For the determination of the weights of process parameters and quality indicators of each process, YC / Z 602-2014 "Guidelines for Data Collection and Processing of Cigarette Silk Process" and YC / T 295-2009 "Guidelines for Capability Evaluation of Cigarette Manufacturing Process" give the relative importance of some process parameters and indicators, which can be used as a reference for evaluating the weights of process parameters and indicators. On this basis, the present invention can use the hierarchical analysis method to determine the weights of process parameters and indicators of each process. Therefore, this embodiment first constructs a hierarchical structure model based on the hierarchical analysis method according to the subordinate relationship of each evaluation indicator. On the basis of in-depth analysis of the evaluation object and the evaluation purpose, each evaluation indicator is decomposed into several levels from top to bottom according to different attributes, and the related indicators of the same layer are subordinate to the indicators of the upper layer. The hierarchical structure model is divided from top to bottom into: target layer, middle layer and single indicator layer; among them, the highest layer is the target layer, which usually has only one indicator, and the middle layer can have one or more layers. The single indicator layer corresponds to a single process parameter or a single quality indicator. The establishment of a hierarchical structure model provides a scientific framework for the subsequent calculation of indicator weights and the quantification of comprehensive quality evaluation.

[0073] See also Figure 2, the establishment of the hierarchical model is explained below by taking the comprehensive evaluation of a process quality as an example. The comprehensive evaluation of a process quality is the highest level, that is, the target level; the comprehensive evaluation of process quality can be evaluated from two aspects, namely, process parameter A and quality index B, which is the middle level. The lowest level A1, A2 and A3 are subordinate to process parameter A, and B1, B2 and B3 are subordinate to quality index B. Further, taking the comprehensive evaluation of the loosening and moisture regain process quality as an example, the comprehensive evaluation of the loosening and moisture regain quality is the highest level, that is, the target level; the comprehensive evaluation of process quality can be evaluated from two aspects, namely, process parameter A and quality index B, which is the middle level. The lowest level material flow, hot air velocity, hot air temperature, drum speed, steam injection amount and water addition amount are subordinate to process parameter A, and the moisture content at the blade outlet, outlet temperature, loosening effect, etc. are subordinate to quality index B.

[0074] Preferably, in the above hierarchical model:

[0075] The evaluation index of the target layer is selected from one or more of the vacuum rehydration process, loose rehydration process, leaf feeding process, hot air moistening process, shredding process, shredded process and aromatizing process; the evaluation index of the intermediate layer is the process parameter and / or quality index corresponding to the above process index. The single index layer is a single process parameter corresponding to the above process parameter, and / or a single quality index corresponding to the above quality index.

[0076] or,

[0077] The evaluation index of the target layer is the comprehensive quality of cigarette shreds; the middle layer is two layers, and the evaluation index of the first middle layer from top to bottom is selected from one or more of the vacuum rehumidification process, loose rehumidification process, leaf feeding process, hot air moisturizing process, shredding process, shredding drying process and flavoring process; the evaluation index of the second middle layer is the process parameters and / or quality indicators corresponding to the above process indicators. The single index layer is a single process parameter corresponding to the above process parameter, and / or a single quality index corresponding to the above quality index.

[0078] The above step S02 is to evaluate the single indicator from three dimensions: single indicator quality conformity evaluation, single indicator quality stability evaluation and single indicator quality comprehensive evaluation, where:

[0079]

About the construction of single indicator quality conformity evaluation method

[0080] The quality conformity evaluation of a single indicator (a single quality indicator or a single process parameter) in the silk making process is also called consistency evaluation. Its essence is the conformity evaluation of a single process parameter or a single quality indicator with the design value of the standard (or specification). That is, it is to evaluate whether there is a statistically significant difference between the overall mean (distribution center) of the process parameter or quality indicator and the design value of the corresponding standard (or specification): if there is a statistically significant difference between the two, it indicates that the process parameter or quality indicator does not meet the requirements of the standard; if there is no statistically significant difference between the two, it indicates that the process parameter or quality indicator meets the requirements of the standard.

[0081] In statistics, the conformity evaluation of a single process parameter or quality indicator with a standard (or specification) design value is essentially a statistical single population mean test. Since the population standard deviation of a process parameter or quality indicator cannot be accurately determined under normal circumstances, and in actual evaluation applications, it may be a large sample or a small sample. Therefore, this embodiment preferably uses a single population mean t-test method to evaluate the quality conformity of a single indicator.

[0082] t-test statistic: in, is the sample mean, s is the sample standard deviation, and n is the number of samples.

[0083] The t-test can be a one-tailed test or a two-tailed test. The two-tailed test is used to determine the conformity of the process parameters or quality indicators with the standard (or specification) design values; the one-tailed test is used to determine whether the overall mean (distribution center) of the process parameters or quality indicators is greater than or less than the standard (or specification) design value.

[0084] Significance level: Generally, α=0.05, that is, the confidence level is 95%.

[0085] Statistical analysis tools can use Minitab statistical software or Excel statistical functions.

[0086] The single indicator quality compliance evaluation is described below in conjunction with a specific embodiment:

[0087] Example 1 - Quality compliance evaluation of the moisture content rate at the blade outlet during the loosening and rehumidification process

[0088] The moisture content of the loose rehydration (module A) leaves of a certain leaf group of the "Lotus" brand on May 5, 2020 was extracted, with a total of 10 minutes of data, as shown in Table 1. Among them, the process quality requirement for the moisture content of the loose rehydration leaves of module A is: 18.8% ± 1.0%. The conformity of the moisture content of the loose rehydration (module A) leaves of this batch with the process quality standard was evaluated.

[0089] Table 1 Loose rehydration (module A) blade outlet moisture content

[0090]

[0091]

[0092] (1) Evaluation method

[0093] Conformity evaluation based on Minitab statistical software

[0094] First, enter the moisture content data in Table 1 into the worksheet of Minitab statistical software, such as Figure 3 As shown in a.

[0095] In Minitab statistical software, select "Statistics" → "Basic Statistics" → "Single Sample t" (b in Figure 3).

[0096] Click "Single Sample t" to get Figure 3 c. Enter "water content" in "sample column" (which means testing the compliance of water content), select "perform hypothesis test (T)", and enter "18.8" in the hypothesized mean (H).

[0097] Click "Options" to get Figure 3 In the confidence level, enter "95" (which means testing the compliance of the moisture content at a 5% significance level). In the alternative (A), select "Not equal" (which means a two-tailed test).

[0098] If you want to perform a one-tailed test, you can select "Less than" or "Greater than" in "Alternative (A)". If you select "Less than", you will test whether the sample mean is less than the standard design value; if you select "Greater than", you will test whether the sample mean is greater than the standard design value.

[0099] Click "OK" in each dialog box. Finally, you will get the analysis results.

[0100] (2) Analysis of evaluation results

[0101] First, you can copy and paste the analysis results into a Word document as shown below:

[0102] Single sample T: moisture content

[0103] Test of mu=18.8 and ≠18.8

[0104]

[0105] Interpret the results:

[0106] For H0: mu=18.8 (ie, the null hypothesis H0: population mean μ=18.8), the calculated result of the test statistic T is -0.39.

[0107] The P value of this test or the probability of obtaining a closer extreme value of the test statistic when the null hypothesis is true is 0.701, which is called the achieved significance level or P value. Therefore, if the acceptable significance level (α level, usually 0.05) is less than the P value or 0.701, H0 cannot be rejected. That is to say, there is no significant difference between the outlet moisture content of the loose rehydrated (A module) blades of this batch and the design value of the process quality standard, that is, the compliance (consistency) of the two is good.

[0108] The 95% confidence interval of the overall mean μ is (18.7721, 18.8189). The standard design value of 18.8 is within the interval, which also shows that there is no significant difference between the outlet moisture content of the loose rehydrated (A module) blades of this batch and the design value of the process quality standard.

[0109]

Construction of single index quality stability evaluation method

[0110] The quality stability evaluation of a single indicator (a single quality indicator or a single process parameter) in the silk making process is to evaluate whether the degree of fluctuation of the process parameter or quality indicator meets the requirements of the standard (or specification), that is, whether there is a significant difference between its variance and the variance controlled by the standard (or specification): if it exceeds the variance controlled by the standard (or specification), it indicates that the stability of the rolling quality indicator is poor (the fluctuation is too large); if it is less than or equal to the variance controlled by the standard (or specification), it indicates that the stability of the rolling quality indicator is good (the fluctuation meets the quality requirements).

[0111] In statistics, the stability evaluation of process parameters or quality indicators is actually a single population variance test, that is, χ 2 Because the population variance of process parameters or quality indicators cannot be accurately determined under normal circumstances, in this implementation, the population variance σ 2 is the sample variance s 2 To estimate.

[0112] According to sampling distribution theory, the test statistic is: Obey x 2 (n-1).

[0113] Given a significance level α, we have: H0:

[0114] H1:

[0115] The test rules are: Note: For the stability evaluation of process parameters or quality indicators, this embodiment can only use a single-tail test, because the quality generally requires that the fluctuation should not exceed a certain degree.

[0116] Significance level: Generally, α=0.05, that is, the confidence level is 95%.

[0117] Statistical analysis tools can use Minitab statistical software or Excel statistical functions.

[0118] The single indicator quality compliance evaluation is described below in conjunction with a specific embodiment:

[0119] Example 2 - Evaluation of the quality stability of the moisture content at the blade outlet during the loosening and rehumidification process

[0120] The moisture content of the loose rehydration (module A) leaves of a certain leaf group of the "Lotus" brand on May 5, 2020 was extracted, with a total of 10 minutes of data, as shown in Table 1. Among them, the process quality requirement for the moisture content of the loose rehydration leaves of module A is: 18.8% ± 1.0%. The stability of the moisture content of the loose rehydration (module A) leaves of this batch was evaluated.

[0121] Due to the process quality requirements for the moisture content at the outlet of the loose rehumidified blades of module A, there are no specific requirements for the standard deviation. However, we can calculate the maximum allowable standard deviation of the corresponding indicator as 0.33% based on the quality rule of 3 / 1000 (3 sigma level), that is, tolerance = 3σ. At this time, the qualified rate of the product process quality can be guaranteed to reach 99.73% and above.

[0122] (1) Evaluation method implementation

[0123] Stability evaluation based on Minitab statistical software

[0124] First, enter the raw data in Table 1 into the worksheet of Minitab statistical software. In Minitab statistical software, select "Statistics" → "Basic Statistics" → "One Variance".

[0125] Click "Single Variance" to get Figure 4 a. In the "Sample Column", enter "Water Content". Select to perform hypothesis test. In the "Hypothesized Standard Deviation", enter "0.33".

[0126] Click "Options" to get Figure 4 b. Under Alternative, select Less Than. Click OK.

[0127] (2) Analysis of evaluation results

[0128] First, you can copy and paste the above analysis results into a Word document as shown below:

[0129] One-way variance test and confidence interval: moisture content

[0130] method

[0131] Null hypothesis Sigma = 0.33

[0132] Alternative hypothesis Sigma < 0.33

[0133] The chi-square method is applicable only to normal distributions.

[0134] Bonett's method works for any continuous distribution.

[0135] Statistics

[0136]

[0137] 95% one-sided confidence interval

[0138]

[0139] test

[0140]

[0141] Interpret the results:

[0142] The moisture content data of the materials in the silk-making process can usually be regarded as coming from a normally distributed population, so the standard method, the chi-square test, is used as a reference. The P value of the one-sided hypothesis test is close to 0.000. This value is low enough to reject the null hypothesis and infer that the standard deviation of the moisture content of the loose rehydrated (A module) blade outlet of this batch is less than 0.33%.

[0143] The estimate of the population variance (or standard deviation) can be made more accurate by examining the 95% upper confidence limit, which provides a value that the population variance (or standard deviation) may be lower than (variance is 0.01143%, standard deviation is 0.1069%). From this analysis, it should be inferred that the variance (less than 0.01143%) and standard deviation (less than 0.1069%) of the moisture content at the blade outlet are small enough to fully meet the process standard requirements, that is, the stability is very good.

[0144]

Construction of a comprehensive evaluation method for single indicator quality

[0145] The comprehensive quality evaluation of a single index (single quality index or single process parameter) in the silk-making process is essentially an overall evaluation of the process parameters or quality indicators based on the conformity evaluation and stability evaluation. In other words, the evaluation should consider not only the degree of closeness (i.e. conformity) of the process parameters or quality indicators to the design values ​​of the standards (or specifications), but also the impact of the fluctuation (i.e. stability) of the characteristic values ​​of the process parameters or quality indicators on the product quality.

[0146] In many cases, the process parameter or quality indicator X has a target value m, and the further X deviates from the target value m, the greater the quality loss, and the closer X is to the target value m, the higher the product quality. Japanese quality expert Dr. Genichi Taguchi suggested using the quadratic loss function L = (Xm) 2 To measure this loss, since the process parameter or quality index X is a random variable, the loss function L is also a random variable, and random variables are not easy to compare, so the average loss E(L) is used to measure its quality loss. If the mean of X is μ and the variance is σ 2 , then the average loss is:

[0147] E(L)=E(Xm) 2 =E[(X-μ)+(μ-m)] 2 =E(X-μ) 2 +(μ-m) 2 =σ 2 +(μ-m) 2

[0148] This shows that the average quality loss consists of two parts: one is the overall variance σ of the process parameter or quality indicator X 2 , which can characterize the stability of process parameters or quality indicators; the second is the square of the deviation of the overall mean of process parameters or quality indicators from the target value (μ-m) 2 , which can characterize the conformity of process parameters or quality indicators. Therefore, the average quality loss is a comprehensive characterization of the stability and conformity of process parameters or quality indicators. In this case, to improve quality is to reduce the variance σ 2 , and reduce the deviation |μ-m| of the mean μ to the target value m. From this perspective, it is obviously inappropriate to use 6σ to measure process capability, and we should use targeted process capability instead, and its measurement formula is:

[0149]

[0150] At this point, the process capability based on quality loss can fully characterize the two aspects of compliance and stability of process parameters or quality indicators.

[0151] When the process parameter or quality index X has a target value m, the calculation of the process capability index should consider the process capability based on quality loss. At this time, the process capability index based on the target value m is the single indicator quality comprehensive evaluation index. Therefore, preferably, the single indicator quality comprehensive evaluation in step S02) is characterized by a single indicator quality comprehensive evaluation index. The larger the single indicator quality comprehensive evaluation index, the better the quality.

[0152] The single indicator quality comprehensive evaluation index is calculated according to the following formula:

[0153]

[0154] In the formula, C pm is the comprehensive evaluation index of single indicator quality; USL and LSL are the specification limits of single indicators, and the difference between them is the tolerance; μ is the overall mean of single indicator, σ 2 is the overall variance of a single indicator, and m is the target value of a single indicator.

[0155] Statistical analysis tools can use Minitab statistical software or Excel statistical functions.

[0156] The single indicator quality compliance evaluation is described below in conjunction with a specific embodiment:

[0157] Example 3 - Comprehensive evaluation of the quality of the moisture content at the blade outlet during the loosening and rehumidification process

[0158] The moisture content of the loose rehydrated leaves (module A) of a certain production batch of the "Lotus" brand leaf group on May 5, 2020 was extracted, with a total of 10 minutes of data, as shown in Table 1. Among them, the process quality requirement for the moisture content of the loose rehydrated leaves of module A is: 18.8% ± 1.0%. Try to evaluate the comprehensive quality level of the moisture content of the loose rehydrated leaves (module A) of this batch.

[0159] (1) Implementation of evaluation method

[0160] Comprehensive evaluation based on Minitab statistical software

[0161] First, enter the raw test data in Table 1 into the worksheet of Minitab statistical software. In Minitab statistical software, select "Statistics" → "Quality Tools" → "Capability Analysis" → "Normal".

[0162] Click "Normal" to get Figure 5 a. Select Single Column, enter Moisture Content, select Subgroup Size, enter 1, enter 17.8 for Lower Spec, and 19.8 for Upper Spec.

[0163] Click Estimate to get Figure 5 b. Check "Use unbiasing constant to calculate overall standard deviation", click "OK", and the analysis results are as follows: Figure 5 Middle c.

[0164] (2) Analysis of evaluation results

[0165] According to the above analysis results ( Figure 5 From c) we can know: C pm =3.65

[0166] It can also be calculated directly by the formula:

[0167] From the above compliance and stability evaluation results, we can know that:

[0168] Mean moisture content = 18.7955

[0169] Moisture content variance = 0.00820

[0170] Design value of moisture content = 18.8%, tolerance = 1%

[0171]

[0172] After the single indicator is evaluated for quality conformity, quality stability and comprehensive quality, the intermediate indicator can be evaluated for comprehensive quality, i.e., step S03), which is to obtain the comprehensive quality evaluation index of the intermediate indicator according to the comprehensive quality evaluation index of each single indicator of the corresponding single indicator layer and the weight corresponding to each single indicator. Preferably, in this step, the comprehensive quality evaluation index of the intermediate indicator is calculated using the following formula:

[0173] Z M =w1×C pm1 +w2×C pm2 ...+w n ×C pmn

[0174] In the formula, Z M is the comprehensive quality evaluation index of the intermediate indicators, C pm1 , C pm2 ……C pmn is the comprehensive quality evaluation index of each single indicator corresponding to the intermediate indicator, w1, w2, ... w n is the corresponding weight of each single indicator.

[0175] After the intermediate indicators are comprehensively evaluated, the target indicators can be comprehensively evaluated, i.e., step S04), which is to obtain the comprehensive quality evaluation index of the target indicator according to the comprehensive quality evaluation index of each intermediate indicator of the corresponding intermediate layer and the weight corresponding to each intermediate indicator. Preferably, in this step, the comprehensive quality evaluation index of the intermediate indicator is calculated using the following formula;.

[0176] Z=w M1 ×Z M1 +w M2 ×Z M2 ...+w Mn ×Z Mn

[0177] In the formula, Z is the comprehensive quality evaluation index of the target indicator, Z M1 , Z M2 ...Z Mn is the comprehensive quality evaluation index of each intermediate indicator corresponding to the target indicator, w M1 、w M2 ,……w Mn is the corresponding weight of each intermediate indicator.

[0178] The weights corresponding to the above single indicators and / or intermediate indicators are preferably determined in the following manner based on the establishment of the aforementioned hierarchical structure model:

[0179] a) Constructing the judgment matrix

[0180] Compare the evaluation indicators of the same layer in pairs, quantify their relative importance using the nine-level scaling method, and construct the judgment matrix A;

[0181] b) Calculate the index weight coefficient

[0182] Calculate the geometric mean of all elements in each row of the judgment matrix to obtain the unnormalized weight of the corresponding indicator in this row;

[0183] The weight coefficient is normalized, and the unnormalized weight of each indicator is divided by the sum of the unnormalized weights of all indicators to obtain the weight coefficient vector w;

[0184] c) Check the consistency of the judgment matrix.

[0185] The above step a) usually starts from the second level of the hierarchical model, compares the relevant indicators of each level, and quantifies their relative importance using the 9-scale method to construct a judgment matrix until the lowest level. Specifically, it is:

[0186] The five judgments of equally important, slightly important, relatively important, strongly important and extremely important are used to characterize the relative importance of indicators and quantify them according to the following rules: i With indicator A j When equally important ij =1, when A i Ratio A j Slightly important ij =3, when A i Ratio A j Stronger important time ij =5, when A i Ratio A j Strongly important ij =7, when A i Ratio A j Extremely important ij =9, if the relative importance is between the two, then use 2, 4, 6, 8 as the score; and aij = 1 / a ji (Use the reciprocal of 1 to 9 to indicate the opposite situation, such as A i With A j The comparison result of a ij , and A j With A i The comparison result of a ji and construct an n×n judgment matrix A=(a ij ); n is the number of lower-level indicators belonging to the same upper-level indicator.

[0187] The following formula is the result of comparing the three sub-indicators A1, A2 and A3 of aroma characteristic A using the 9-scale method. Since A1 is slightly more important than A2, 12 Equal to 3; A1 is stronger and more important than A3, a 13 Equals 5.

[0188]

[0189] The calculation of the index weight coefficient in the above step b) includes the following two steps:

[0190] Calculate the geometric mean of all elements in each row of the judgment matrix to obtain the unnormalized weight of the indicator corresponding to the row; the so-called geometric mean refers to the nth root of the product of n score values. For example: according to the judgment matrix, the unnormalized weight of indicator A1 is the cube root of the three scores of 1, 3, and 5 in the row where indicator A1 is located, and the result is equal to 2.4662; similarly, the unnormalized weights of A2 and A3 are 1 and 0.4055 respectively;

[0191] The weight coefficients are normalized, and the unnormalized weights of each indicator are divided by the sum of the unnormalized weights of all indicators to obtain the weight coefficient vector w; for this example, the sum of the unnormalized weights of the three indicators is 2.4662+1+0.4055=3.8717. Therefore, the normalized weight of indicator A1 is 2.4662÷3.8717=0.6370, the normalized weight of indicator A2 is 1÷3.8717=0.2583, and the normalized weight of indicator A3 is 0.4055÷3.8717=0.1047. Based on this, the weight coefficient vector w can be obtained, which is also the eigenvector w corresponding to the maximum eigenvalue λ.

[0192]

[0193] The above step c) is preferably:

[0194] c1) Calculate the maximum eigenvalue λ of the judgment matrix:

[0195] Multiply the judgment matrix A by the weight coefficient vector w to obtain the vector Aw. Then, calculate the ratio of the corresponding elements of the vector Aw and the weight coefficient vector w respectively, and the sum of the ratios is the maximum eigenvalue λ. For the above example, the maximum eigenvalue λ of the judgment matrix A is 3.0385.

[0196]

[0197]

[0198] c2) Check the consistency of the judgment matrix:

[0199] Calculate the consistency index CI, which is equal to the difference between the maximum eigenvalue λ and the index number n divided by the index number n minus 1; for the above example, CI = (3.0385-3) ÷ (3-1) = 0.0193;

[0200] According to the number of indicators n, the average random consistency index RI value is obtained; for this example, there are 3 indicators in total. By looking up the table, it can be found that the average random consistency index RI value is 0.58;

[0201]

[0202]

[0203] Calculate the random consistency ratio CR of the judgment matrix. The random consistency ratio CR is equal to the ratio of the consistency index CI to the average random consistency index RI. Check whether the consistency of the judgment matrix meets the requirements based on the random consistency ratio CR. For the above example, CR = 0.0193 ÷ 0.58 = 0.0332; Conclusion: Since CR = 0.0332 < 0.1, the judgment matrix has satisfactory consistency, indicating that the weight distribution is reasonable.

[0204] After determining the indicator weights, the quality comprehensive evaluation index of the intermediate indicator and the quality comprehensive evaluation index of the target indicator can be calculated according to the above method to characterize the quality comprehensive performance of the intermediate index and the target index. Figure 2 Take the hierarchical model shown as an example:

[0205] Assume that the weights of process parameters A1, A2, and A3 are w1, w2, and w3 respectively, and their corresponding comprehensive evaluation indexes (single index) are C pm1 , C pm2 and C pm3 , then the comprehensive quality evaluation index Z of process parameter A is A for:

[0206] Z A =w1*C pm1 +w2*C pm2 +w3*C pm3

[0207] Similarly, the weights of quality indicators B1, B2, and B3 are w 1' 、w 2' and w 3' , and their corresponding comprehensive evaluation indexes (single index) are C pm1' , C pm2' and C pm3' , then the comprehensive quality evaluation index Z of process parameter B is B for:

[0208] Z B =w 1' *C pm1' +w 2' *C pm2' +w 3' *C pm3'

[0209] Assuming that the weights of process parameter A and quality index B are a1 and a2 respectively, the comprehensive evaluation index Z of the process quality is:

[0210] Z=a1*Z A +a2*Z B

[0211] It can be seen from the above content that, compared with the prior art, the silk making process quality evaluation method provided by the embodiment of the present invention has the following advantages:

[0212] Analyzing and evaluating the quality of the cigarette silk-making process is conducive to improving quality control capabilities and improving the control accuracy of the quality of in-process and finished products. The present invention constructs a hierarchical model for silk-making process quality evaluation, provides single-indicator quality compliance, stability and comprehensive evaluation methods, as well as intermediate and target indicator quality comprehensive evaluation methods, enriches the connotation of process quality evaluation, and provides a methodological basis and practical operation guide for scientific and standardized silk-making process quality evaluation, ensuring the standardization, reliability and comparability of the evaluation results, which is conducive to the overall improvement of silk-making process quality stability, uniformity, coordination and consistency, and creates conditions for deepening the "transformation from control indicators to control parameters, from result control to process control, and from manual control experience decision-making to automatic control scientific decision-making", so as to better achieve the company's product quality goals and benefit goals.

[0213] The embodiments of the present invention have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for evaluating the quality of a silk-making process, characterized in that: The following steps are involved: S01), determining silk quality evaluation indicators, and constructing a hierarchical structure model for each evaluation indicator according to a subordinate relationship, wherein the hierarchical structure model is divided from top to bottom into: a target layer, an intermediate layer, and a single indicator layer; the single indicator layer corresponds to a single process parameter or a single quality indicator; S02), conduct single indicator quality evaluation; The single indicator quality evaluation includes: single indicator quality conformity evaluation, single indicator quality stability evaluation, and single indicator quality comprehensive evaluation; The single indicator quality comprehensive evaluation is characterized by a single indicator quality comprehensive evaluation index. The larger the single indicator quality comprehensive evaluation index, the better the quality. The single indicator quality comprehensive evaluation index is calculated according to the following formula: In the formula, C pm is the comprehensive evaluation index of single indicator quality; USL and LSL are the specification limits of single indicators, and the difference between the two is the tolerance; μ is the overall mean of single indicators, σ 2 is the overall variance of a single indicator, and m is the target value of a single indicator; S03) Conduct a comprehensive quality evaluation of the intermediate indicators, including: According to the single indicator quality comprehensive evaluation index of each single indicator of the corresponding single indicator layer and the weight corresponding to each single indicator, the quality comprehensive evaluation index of the intermediate indicator is obtained; S04) Conduct a comprehensive quality evaluation of the target indicators, including: The quality comprehensive evaluation index of the target indicator is obtained according to the quality comprehensive evaluation index of each intermediate indicator of the corresponding intermediate layer and the weight corresponding to each intermediate indicator.

2. The silk making process quality evaluation method according to claim 1, characterized in that: In the step S03), the following formula is used to calculate the comprehensive quality evaluation index of the intermediate indicator; WITH M =w1×C pm1 +w2×C pm2 ……+in n ×C pmn In the formula, Z M is the comprehensive quality evaluation index of the intermediate indicators, C pm1 , C pm2 ……C pmn is the comprehensive quality evaluation index of each single indicator corresponding to the intermediate indicator, w1, w2, ... w n is the corresponding weight of each single indicator.

3. The silk making process quality evaluation method according to claim 2, characterized in that: In the step S04), the following formula is used to calculate the comprehensive quality evaluation index of the intermediate indicator: From M1 ×Z M1 +in M2 ×Z M2 ……+in Mn ×Z Mn In the formula, Z is the comprehensive quality evaluation index of the target indicator, Z M1 , Z M2 ...Z Mn is the comprehensive quality evaluation index of each intermediate indicator corresponding to the target indicator, w M1 、w M2 ,……w Mn is the corresponding weight of each intermediate indicator.

4. The silk making process quality evaluation method according to claim 3, characterized in that: The weights corresponding to the single indicators and / or intermediate indicators are determined as follows: a) Constructing the judgment matrix The evaluation indicators of the same layer are compared pairwise, and their relative importance is quantified using the nine-level scaling method to construct the judgment matrix A; b) Calculate the index weight coefficient Calculate the geometric mean of all elements in each row of the judgment matrix to obtain the unnormalized weight of the corresponding indicator in this row; The weight coefficient is normalized, and the unnormalized weight of each indicator is divided by the sum of the unnormalized weights of all indicators to obtain the weight coefficient vector w; c) Check the consistency of the judgment matrix.

5. The silk making process quality evaluation method according to claim 4, characterized in that: The step a) is specifically as follows: The five judgments of equally important, slightly important, relatively important, strongly important and extremely important are used to characterize the relative importance of indicators and quantify them according to the following rules: i With indicator A j When equally important ij =1, when A i Ratio A j Slightly important ij =3, when A i Ratio A j Stronger important time ij =5, when A i Ratio A j Strongly important ij =7, when A i Ratio A j Extremely important ij =9, if the relative importance is between the two, use 2, 4, 6, 8 for the score; and a ij =1 / a ji , and construct an n×n judgment matrix A=(a ij ); n is the number of lower-level indicators belonging to the same upper-level indicator.

6. The silk-making process quality evaluation method according to claim 5, characterized in that: The step c) is specifically as follows: c1) Calculate the maximum eigenvalue λ of the judgment matrix: The judgment matrix A is multiplied by the weight coefficient vector w on the right to obtain the vector Aw; then, the ratio of the corresponding elements of the vector Aw and the weight coefficient vector w is calculated respectively, and the sum of the ratios is the maximum eigenvalue λ; c2) Check the consistency of the judgment matrix: Calculate the consistency index CI, which is equal to the difference between the maximum eigenvalue λ and the index number n divided by the index number n minus 1; According to the number of indicators n, the average random consistency index RI value is obtained; Calculate the random consistency ratio CR of the judgment matrix. The random consistency ratio CR is equal to the ratio of the consistency index CI to the average random consistency index RI. Check whether the consistency of the judgment matrix meets the requirements based on the random consistency ratio CR.

7. The method for evaluating silk making process quality according to any one of claims 1 to 6, characterized in that: The single index quality conformity evaluation in step S02) adopts t test; the single index quality stability evaluation adopts χ 2 test.

8. The silk-making process quality evaluation method according to claim 7, characterized in that: The t-test adopts a two-tailed test or a one-tailed test.

9. The silk-making process quality evaluation method according to claim 7, characterized in that: The x 2 The test was one-tailed.

Images