Kiln process parameter tracing and early warning method for medicinal glass bottle defects

Abstract

The invention discloses a kiln process parameter traceability and early warning method for medicinal glass bottle defects, and relates to the technical field of production and manufacturing process optimizing.The method comprises the steps that S1, key process flows of glass bottle production in a kiln are disassembled, and process parameter characteristics of all the process flows are determined; s2, dividing defect categories of the medicinal glass bottles, and constructing a corresponding set of the defect categories and corresponding process parameter characteristics; s3, sorting the influence degrees of the process parameter characteristics relative to the corresponding defect categories to obtain main process parameter characteristics causing causes of each defect category, and determining a data qualification range; s4, identifying and listing technological parameter characteristics exceeding a qualified range based on operation data collected by a sensor in the kiln in real time; and S5, according to the monitored abnormal process parameter characteristics, rapidly tracing the possible glass bottle defect types, determining an instant processing method for each defect type, constructing a method processing set, and prompting the corresponding processing method through a pop-up window.

Description

Technical Field

[0001] This invention relates to the field of manufacturing process optimization technology, specifically to a method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles. Background Technology

[0002] Pharmaceutical glass bottles are a key material for drug packaging, and their quality directly affects the safety and efficacy of medicines. However, various defects may occur in pharmaceutical glass bottles during the production process. These defects not only affect the appearance but may also negatively impact the preservation and use of medicines. Summary of the Invention

[0003] To address the aforementioned technical problems and provide a method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles, this technical solution resolves the aforementioned issues.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for tracing and providing early warning of kiln process parameters for defects in pharmaceutical glass bottles, including the following steps: S1. Based on the current production process of pharmaceutical glass bottles in the kiln, the key process flow of glass bottle production in the kiln is broken down, and the process parameter characteristics of each process flow are clarified. S2. Based on historical production data, classify the defect categories of pharmaceutical glass bottles, analyze the process parameter characteristics that lead to the corresponding defect categories in glass bottle production, and construct a set of corresponding defect categories and process parameter characteristics. S3. Based on the set of corresponding defect categories and process parameter features, as well as historical production data, the influence of process parameter features on the corresponding defect categories is ranked to obtain the main process parameter features that cause each defect category, and the data acceptable range of each process parameter feature is determined. S4. Based on the actual operating data of various process parameters of glass production collected in real time by sensors inside the kiln, compare them with the data qualified range of the determined process parameter characteristics, identify and list the process parameter characteristics that exceed the qualified range. S5. Based on the detected abnormal process parameter characteristics, and combined with the corresponding set of defect categories and process parameter characteristics, quickly trace the possible types of glass bottle defects, determine the immediate handling methods for each defect category, construct a set of handling methods, and prompt the corresponding immediate handling methods through pop-up windows. Preferably, step S1 specifically includes: Based on the current production process of pharmaceutical glass bottles in kilns, the key process flow of glass bottle production in kilns is broken down and divided into melting stage, clarification stage, homogenization stage, and annealing and cooling stage, and the key process parameter characteristics of each of the above process flows are identified. The process parameters in the melting stage include melting temperature, melting pressure, molten glass viscosity, molten gas content, bubble line, and melting time; the process parameters in the refining stage include refining temperature, refining pressure, refining gas content, refining agent dosage, refining time, and refining glass viscosity; the process parameters in the homogenization stage include homogenization temperature, homogenizing gas content, homogenization pressure, homogenization time, and homogenizing glass viscosity in each area of ​​the furnace; and the process parameters in the annealing and cooling stage include annealing and cooling temperature, annealing temperature gradient, and annealing and cooling time. Preferably, step S2 specifically includes: Based on historical production data from the kiln, the physical and chemical defects of pharmaceutical glass bottles were recorded. All characteristic data were standardized and normalized to establish a set of defective characteristics for each pharmaceutical glass bottle. Assumptions were made regarding the defect characteristics of each pharmaceutical glass bottle. Given a d-dimensional feature vector, the initial preset defect feature set has k clusters, and each cluster has a unique cluster center. Calculate the Euclidean distance between the defect feature vector of each medicine glass bottle and the corresponding cluster center. Establish the objective function The cluster centers of the target k clusters are obtained iteratively through the K-Means clustering algorithm. Based on the defect feature set, all defective pharmaceutical glass bottles are classified and the defect categories of pharmaceutical glass bottles are divided. Furthermore, step S2 also includes: Based on the categorized defects of pharmaceutical glass bottles, the causes of these defects are preliminarily listed. Several kiln process experts were invited to score the causes of defects in various categories of glass bottles, and the scoring results were assigned different weights based on the qualifications and experience of the kiln process experts, with the sum of the weights being 1. The expert's score for each cause is multiplied by the weight assigned by the expert. The weighted comprehensive score is calculated for each cause of defects in pharmaceutical glass bottles. The n causes with the highest scores are selected into the initial cause matching pool. The n process parameter features that lead to the corresponding defect category in glass bottle production are comprehensively correlated to construct a set of defect categories and corresponding process parameter features. Preferably, step S3 specifically includes: Based on the set of corresponding process parameter features and historical production data, the influence of process parameter features on the corresponding defect categories is ranked to obtain the main process parameter features that cause each defect category. Grey relational analysis is used to assess the influence of process parameter features on their corresponding defect categories. First, the local correlation degree between the q-th process parameter feature and the p-th defect category is calculated using the correlation coefficient formula: ; in This represents the sample data of the k-th pharmaceutical glass bottle, representing the q-th type of process parameter characteristics. For the sample data of the k-th pharmaceutical glass bottle of the p-th defect category, The resolution coefficient is typically set to 0.5. The correlation between the q-th process parameter feature and the p-th defect category is calculated as follows: ,in The weight of the k-th sample data of the drug glass bottle is 1. The calculated correlation results are sorted to obtain the overall influence of each process parameter feature on each defect category. The greater the correlation, the more significant the influence of the corresponding process parameter feature on the defect category. The main process parameter features that cause each defect category are obtained. Furthermore, step S3 also includes: The weights of the k-th sample data of the medicine glass bottle were determined using the analytic hierarchy process. First, kiln process experts were asked to compare the sample data of medicine glass bottles in pairs according to the importance criteria and score them. The target layer is the importance analysis of the sample data of medicine glass bottles, the criteria layer includes the timeliness, completeness, representativeness, accuracy and consistency of the sample data, and the scheme layer is the sample data of each medicine glass bottle. For each importance criterion and sample data, a judgment matrix is ​​constructed, and the weight vector for the importance analysis of each importance criterion relative to the sample data of pharmaceutical glass bottles is calculated. The sample data of each drug glass bottle relative to the t-th criterion in the criterion layer The weight vector is Finally, the sample data of the kth medicine glass bottle in the scheme layer was obtained through calculation. The total weight is ; Furthermore, step S3 also includes: Based on historical production data, that is, based on the maximum and minimum values ​​of various process parameter characteristics of qualified pharmaceutical glass bottles that have been produced, the data qualification range of various process parameter characteristics is determined. Preferably, step S4 specifically includes: Based on the actual operating data of various process parameters of glass production collected in real time by sensors inside the kiln, the data is compared with the data qualified range of the determined process parameter characteristics. The process parameter characteristics that exceed the qualified range are identified and listed. The deviation value and deviation rate between the monitored value and the data qualified range are calculated. The absolute value is taken and sorted from high to low to obtain the abnormal process parameter characteristics and their order of processing. Preferably, a set of processing methods is constructed by determining an immediate processing method for each defect category; Based on the monitored abnormal process parameter characteristics, combined with the corresponding set of defect categories and process parameter characteristics, the possible types of glass bottle defects can be quickly traced, and corresponding handling methods can be output through pop-up windows to warn staff to make adjustments. Furthermore, according to the order in which abnormal process parameter features are to be processed, pop-up windows sequentially warn of the corresponding immediate processing methods. After adjustment, until the actual operating data of the corresponding process parameter feature collected in real time by the internal sensors are within the qualified range of the corresponding process parameter feature, the pop-up window closes, and the next abnormal process parameter feature to be processed and its corresponding processing method are sequentially displayed according to the order in which abnormal process parameter features are to be processed, until the actual operating data of all abnormal process parameter features are within the qualified range.

[0005] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The kiln production process is broken down into four core stages: melting, clarification, homogenization, and annealing and cooling. The key process parameters of each stage (such as temperature, pressure, viscosity, time, etc.) are clearly defined to form a comprehensive parameter characteristic system, avoid traceability blind spots caused by parameter omissions, and lay a data foundation for subsequent analysis. 2. The K-Means clustering algorithm is used to standardize and normalize the defect features before clustering. Combined with expert experience for weighted verification, it breaks through the subjective limitations of traditional manual classification, realizes the objective classification of defect categories and preliminary matching of causes, and provides a classification basis for accurate source tracing. 3. By matching defect categories with process parameters, rapid tracing of abnormal parameters to potential defects is achieved, and a standardized set of handling methods is constructed. Immediate handling solutions are precisely pushed out in the form of pop-up windows, reducing the reliance of staff on experience and shortening the time for handling abnormalities. Based on the comparison of real-time data collected by sensors with preset acceptable ranges, deviation values ​​and deviation rates are calculated and sorted to help staff quickly focus on high-priority abnormal parameters, avoid blind adjustments, and improve the efficiency of abnormality handling. Attached Figure Description

[0006] Figure 1 Flowchart of methods for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles; Figure 2 A flowchart for constructing a set of defect categories and corresponding process parameter characteristics; Figure 3 To obtain the characteristic flow chart of the main process parameters that lead to the causes of each defect category; Figure 4 Flowchart for determining the weights of sample data for pharmaceutical glass bottles; Figure 5 This is a flowchart of the early warning and handling method. Detailed Implementation

[0007] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0008] Reference Figure 1 As shown, the method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles includes the following steps: S1. Based on the current production process of pharmaceutical glass bottles in the kiln, the key process flow of glass bottle production in the kiln is broken down, and the process parameter characteristics of each process flow are clarified. S2. Based on historical production data, classify the defect categories of pharmaceutical glass bottles, analyze the process parameter characteristics that lead to the corresponding defect categories in glass bottle production, and construct a set of corresponding defect categories and process parameter characteristics. S3. Based on the set of corresponding defect categories and process parameter features, as well as historical production data, the influence of process parameter features on the corresponding defect categories is ranked to obtain the main process parameter features that cause each defect category, and the data acceptable range of each process parameter feature is determined. S4. Based on the actual operating data of various process parameters of glass production collected in real time by sensors inside the kiln, compare them with the data qualified range of the determined process parameter characteristics, identify and list the process parameter characteristics that exceed the qualified range. S5. Based on the detected abnormal process parameter characteristics, and combined with the corresponding set of defect categories and process parameter characteristics, quickly trace the possible types of glass bottle defects, determine the immediate handling methods for each defect category, construct a set of handling methods, and prompt the corresponding immediate handling methods through pop-up windows.

[0009] Reference Figure 2 As shown, step S1 specifically includes: Based on the current production process of pharmaceutical glass bottles in kilns, the key process flow of glass bottle production in kilns is broken down and divided into melting stage, clarification stage, homogenization stage, and annealing and cooling stage, and the key process parameter characteristics of each of the above process flows are identified. The process parameters in the melting stage include melting temperature, melting pressure, molten glass viscosity, molten gas content, bubble boundary, and melting time; the process parameters in the refining stage include refining temperature, refining pressure, refining gas content, refining agent addition amount, refining time, and refining glass viscosity; the process parameters in the homogenization stage include homogenization temperature, homogenizing gas content, homogenization pressure, homogenization time, and homogenizing glass viscosity in each area of ​​the furnace; and the process parameters in the annealing and cooling stage include annealing and cooling temperature, annealing temperature gradient, and annealing and cooling time.

[0010] Preferably, step S2 specifically includes: recording the physical and chemical defect characteristics of pharmaceutical glass bottles based on historical production data of the kiln; standardizing and normalizing all characteristic data; establishing a defect characteristic set for defective pharmaceutical glass bottles; and assuming defect characteristics for each pharmaceutical glass bottle. Given a d-dimensional feature vector, the initial preset defect feature set has k clusters, and each cluster has a unique cluster center. Calculate the Euclidean distance between the defect feature vector of each medicine glass bottle and the corresponding cluster center. Establish the objective function The cluster centers of the target k clusters are obtained through iterative K-Means clustering algorithm. Based on the defect feature set, all defective pharmaceutical glass bottles are classified and the defect categories of pharmaceutical glass bottles are divided.

[0011] Furthermore, step S2 also includes: Based on the categorized defects of pharmaceutical glass bottles, the causes of these defects are preliminarily listed. Several kiln process experts were invited to score the causes of defects in various categories of glass bottles, and the scoring results were assigned different weights based on the qualifications and experience of the kiln process experts, with the sum of the weights being 1. The expert's score for each cause is multiplied by the weight assigned by the expert. The weighted comprehensive score is calculated for each cause of defects in pharmaceutical glass bottles. The n causes with the highest scores are selected into the initial cause matching pool. The n process parameter features that lead to the corresponding defect category in glass bottle production are comprehensively correlated to construct a set of corresponding defect categories and process parameter features.

[0012] Reference Figure 3 As shown, step S3 specifically includes: based on the set of corresponding process parameter features and historical production data, sorting the degree of influence of process parameter features relative to the corresponding defect categories to obtain the main process parameter features that cause each defect category; Grey relational analysis is used to assess the influence of process parameter features on their corresponding defect categories. First, the local correlation degree between the q-th process parameter feature and the p-th defect category is calculated using the correlation coefficient formula: ; in This represents the sample data of the k-th pharmaceutical glass bottle, representing the q-th type of process parameter characteristics. For the sample data of the k-th pharmaceutical glass bottle of the p-th defect category, The resolution coefficient is typically set to 0.5. The correlation between the q-th process parameter feature and the p-th defect category is calculated as follows: ,in The weight of the k-th sample of pharmaceutical glass bottle is 1. The calculated correlation results are sorted to obtain the overall influence of each process parameter feature on each defect category. The greater the correlation, the more significant the influence of the corresponding process parameter feature on the defect category. The main process parameter features that cause each defect category are obtained.

[0013] Reference Figure 4 As shown, step S3 further includes: The weights of the k-th sample data of the medicine glass bottle were determined using the analytic hierarchy process. First, kiln process experts were asked to compare the sample data of medicine glass bottles in pairs according to the importance criteria and score them. The target layer is the importance analysis of the sample data of medicine glass bottles, the criteria layer includes the timeliness, completeness, representativeness, accuracy and consistency of the sample data, and the scheme layer is the sample data of each medicine glass bottle. For each importance criterion and sample data, a judgment matrix is ​​constructed, and the weight vector for the importance analysis of each importance criterion relative to the sample data of pharmaceutical glass bottles is calculated. The sample data of each drug glass bottle relative to the t-th criterion in the criterion layer The weight vector is Finally, the sample data of the kth medicine glass bottle in the scheme layer was obtained through calculation. The total weight is .

[0014] Furthermore, step S3 also includes: based on historical production data, that is, based on the maximum and minimum values ​​of the various process parameter characteristics of the qualified pharmaceutical glass bottles that have been produced, determining the data qualification range of each process parameter characteristic.

[0015] Reference Figure 5 As shown, step S4 specifically includes: comparing the actual operating data of various process parameters of glass production collected in real time by sensors inside the kiln with the data qualified range of the determined process parameter characteristics, identifying and listing process parameter characteristics that exceed the qualified range, calculating the deviation value and deviation rate between the monitored value and the data qualified range, taking the absolute value and sorting them from high to low to obtain abnormal process parameter characteristics and their order of processing.

[0016] Preferably, step S5 specifically includes: For each defect category, determine the method for immediate handling and construct a set of handling methods; Based on the monitored abnormal process parameter characteristics, combined with the corresponding set of defect categories and process parameter characteristics, the possible types of glass bottle defects can be quickly traced, and corresponding handling methods can be output through pop-up windows to warn staff to make adjustments.

[0017] Furthermore, step S5 also includes: According to the order in which abnormal process parameter features are to be processed, the corresponding immediate processing methods are warned sequentially through pop-up windows. After adjustment, the pop-up windows are closed until the actual operating data of the corresponding process parameter features collected in real time by the internal sensors are within the qualified range of the corresponding process parameter features. Then, the pop-up windows are closed, and the next abnormal process parameter feature to be processed and its corresponding processing method are displayed sequentially according to the order in which abnormal process parameter features are to be processed, until the actual operating data of all abnormal process parameter features are within the qualified range.

[0018] The process of using this invention is as follows: Based on the current production process of pharmaceutical glass bottles in the kiln, the key technological processes of glass bottle production in the kiln are broken down to clarify the process parameter characteristics of each process; combined with historical production data, pharmaceutical glass bottle defect categories are classified, and the process parameter characteristics that lead to the corresponding defect categories in glass bottle production are analyzed to construct a set of corresponding defect categories and corresponding process parameter characteristics; based on the set of corresponding defect categories and process parameter characteristics and historical production data, the degree of influence of process parameter characteristics relative to the corresponding defect categories is ranked to obtain the main process parameter characteristics that cause each defect category, and the data acceptable range of each process parameter characteristic is determined; combined with the actual operating data of various process parameter characteristics of glass production collected in real time by sensors inside the kiln, the data acceptable range of the determined process parameter characteristics is compared to identify and list the process parameter characteristics that exceed the acceptable range; for the abnormal process parameter characteristics detected, combined with the set of corresponding defect categories and process parameter characteristics, the possible glass bottle defect types are quickly traced, and an immediate handling method is determined for each defect category, constructing a method handling set, and prompting the corresponding immediate handling method through a pop-up window.

[0019] In summary, the advantages of this invention are as follows: It breaks down the kiln production process into four core stages: melting, clarification, homogenization, and annealing / cooling. Key process parameters for each stage (such as temperature, pressure, viscosity, and time) are clearly defined, forming a comprehensive parameter feature system. This avoids blind spots in traceability caused by parameter omissions and lays a data foundation for subsequent analysis. The K-Means clustering algorithm is used to standardize and normalize defect features before clustering. Combined with expert experience for weighted verification, this overcomes the subjective limitations of traditional manual classification, achieving objective classification of defect categories and preliminary matching of causes, providing a classification basis for accurate traceability. Through the corresponding set of defect categories and process parameters, rapid traceability from abnormal parameters to potential defects is achieved. A standardized set of processing methods is constructed, and immediate processing solutions are precisely pushed out in a pop-up window, reducing staff reliance on experience and shortening anomaly handling time. Based on real-time sensor data collection and comparison with preset acceptable ranges, deviation values ​​and deviation rates are calculated and sorted, helping staff quickly focus on high-priority abnormal parameters, avoiding blind adjustments, and improving anomaly handling efficiency.

[0020] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles, characterized in that, Includes the following steps: S1. Based on the current production process of pharmaceutical glass bottles in the kiln, the key process flow of glass bottle production in the kiln is broken down, and the process parameter characteristics of each process flow are clarified. S2. Based on historical production data, classify the defect categories of pharmaceutical glass bottles, analyze the process parameter characteristics that lead to the corresponding defect categories in glass bottle production, and construct a set of corresponding defect categories and process parameter characteristics. S3. Based on the set of corresponding defect categories and process parameter features, as well as historical production data, the influence of process parameter features on the corresponding defect categories is ranked to obtain the main process parameter features that cause each defect category, and the data acceptable range of each process parameter feature is determined. S4. Based on the actual operating data of various process parameters of glass production collected in real time by sensors inside the kiln, compare them with the data qualified range of the determined process parameter characteristics, identify and list the process parameter characteristics that exceed the qualified range. S5. Based on the detected abnormal process parameter characteristics, and combined with the corresponding set of defect categories and process parameter characteristics, quickly trace the possible types of glass bottle defects, determine the immediate handling methods for each defect category, construct a set of handling methods, and prompt the corresponding immediate handling methods through pop-up windows.

2. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 1, characterized in that, Step S1 specifically includes: Based on the current production process of pharmaceutical glass bottles in kilns, the key process flow of glass bottle production in kilns is broken down and divided into melting stage, clarification stage, homogenization stage, and annealing and cooling stage, and the key process parameter characteristics of each of the above process flows are identified. The process parameters in the melting stage include melting temperature, melting pressure, molten glass viscosity, molten gas content, bubble boundary, and melting time; the process parameters in the refining stage include refining temperature, refining pressure, refining gas content, refining agent addition amount, refining time, and refining glass viscosity; the process parameters in the homogenization stage include homogenization temperature, homogenizing gas content, homogenization pressure, homogenization time, and homogenizing glass viscosity in each area of ​​the furnace; and the process parameters in the annealing and cooling stage include annealing and cooling temperature, annealing temperature gradient, and annealing and cooling time.

3. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 2, characterized in that, Step S2 specifically includes: Based on historical production data from the kiln, the physical and chemical defects of pharmaceutical glass bottles were recorded. All characteristic data were standardized and normalized to establish a set of defective characteristics for each pharmaceutical glass bottle. Assumptions were made regarding the defect characteristics of each pharmaceutical glass bottle. Given a d-dimensional feature vector, the initial preset defect feature set has k clusters, and each cluster has a unique cluster center. Calculate the Euclidean distance between the defect feature vector of each medicine glass bottle and the corresponding cluster center. Establish the objective function The cluster centers of the target k clusters are obtained through iterative K-Means clustering algorithm. Based on the defect feature set, all defective pharmaceutical glass bottles are classified and the defect categories of pharmaceutical glass bottles are divided.

4. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 3, characterized in that, Step S2 further includes: Based on the categorized defects of pharmaceutical glass bottles, the causes of these defects are preliminarily listed. Several kiln process experts were invited to score the causes of defects in various categories of glass bottles, and the scoring results were assigned different weights based on the qualifications and experience of the kiln process experts, with the sum of the weights being 1. The expert's score for each cause is multiplied by the weight assigned by the expert. The weighted comprehensive score is calculated for each cause of defects in pharmaceutical glass bottles. The n causes with the highest scores are selected into the initial cause matching pool. The n process parameter features that lead to the corresponding defect category in glass bottle production are comprehensively correlated to construct a set of corresponding defect categories and process parameter features.

5. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 4, characterized in that, Step S3 specifically includes: Based on the set of corresponding process parameter features and historical production data, the influence of process parameter features on the corresponding defect categories is ranked to obtain the main process parameter features that cause each defect category. Grey relational analysis is used to assess the influence of process parameter features on their corresponding defect categories. First, the local correlation degree between the q-th process parameter feature and the p-th defect category is calculated using the correlation coefficient formula: ； in This represents the sample data of the k-th pharmaceutical glass bottle, representing the q-th type of process parameter characteristics. For the sample data of the k-th pharmaceutical glass bottle of the p-th defect category, The resolution coefficient is typically set to 0.

5. The correlation between the q-th process parameter feature and the p-th defect category is calculated as follows: ,in The weight of the k-th sample of pharmaceutical glass bottle is 1. The calculated correlation results are sorted to obtain the overall influence of each process parameter feature on each defect category. The greater the correlation, the more significant the influence of the corresponding process parameter feature on the defect category. The main process parameter features that cause each defect category are obtained.

6. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 5, characterized in that, Step S3 further includes: The weights of the k-th sample data of the medicine glass bottle were determined using the analytic hierarchy process. First, kiln process experts were asked to compare the sample data of medicine glass bottles in pairs according to the importance criteria and score them. The target layer is the importance analysis of the sample data of medicine glass bottles, the criteria layer includes the timeliness, completeness, representativeness, accuracy and consistency of the sample data, and the scheme layer is the sample data of each medicine glass bottle. For each importance criterion and sample data, a judgment matrix is ​​constructed, and the weight vector for the importance analysis of each importance criterion relative to the sample data of pharmaceutical glass bottles is calculated. The sample data of each drug glass bottle relative to the t-th criterion in the criterion layer The weight vector is Finally, the sample data of the kth medicine glass bottle in the scheme layer was obtained through calculation. The total weight is .

7. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 6, characterized in that, Step S3 further includes: Based on historical production data, specifically the maximum and minimum values ​​of various process parameter characteristics of the produced qualified pharmaceutical glass bottles, the acceptable range of data for each process parameter characteristic is determined.

8. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 7, characterized in that, Step S4 specifically includes: Based on the actual operating data of various process parameters of glass production collected in real time by sensors inside the kiln, the data is compared with the data qualified range of the determined process parameter characteristics. The process parameter characteristics that exceed the qualified range are identified and listed. The deviation value and deviation rate between the monitored value and the data qualified range are calculated. The absolute value is taken and sorted from high to low to obtain the abnormal process parameter characteristics and their order of processing.

9. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 8, characterized in that, Step S5 specifically includes: For each defect category, determine the method for immediate handling and construct a set of handling methods; Based on the monitored abnormal process parameter characteristics, combined with the corresponding set of defect categories and process parameter characteristics, the possible types of glass bottle defects can be quickly traced, and corresponding handling methods can be output through pop-up windows to warn staff to make adjustments.

10. The method for tracing and early warning of kiln process parameters for defects in pharmaceutical glass bottles according to claim 9, characterized in that, Step S5 further includes: According to the order in which abnormal process parameter features are to be processed, the corresponding immediate processing methods are warned sequentially through pop-up windows. After adjustment, the pop-up windows are closed until the actual operating data of the corresponding process parameter features collected in real time by the internal sensors are within the qualified range of the corresponding process parameter features. Then, the pop-up windows are closed, and the next abnormal process parameter feature to be processed and its corresponding processing method are displayed sequentially according to the order in which abnormal process parameter features are to be processed, until the actual operating data of all abnormal process parameter features are within the qualified range.

Images