A method and system for evaluating and predicting building material production quality based on data feedback

Abstract

The present invention discloses a method and system for evaluating and predicting the quality of building materials production based on data feedback, which belongs to the field of building materials production. The method includes: determining the production links of building materials according to the types of building materials, and obtaining a production parameter data group; screening out building materials with similar working conditions; constructing a performance defect evaluation model based on production parameters of different production links, training the performance defect evaluation model, and obtaining an optimal performance defect evaluation model; after each batch of building materials is produced, the corresponding optimal performance defect evaluation model is retrieved according to the types of performance defects that may be generated, and the production parameter data group is input into the optimal performance defect evaluation model to predict performance defect data for building materials quality evaluation and prediction. The system includes a quality supervision feedback module, a production data acquisition module, and a data analysis module. It is adaptable to the use scenarios of building materials under multiple working conditions, and provides reliable data support and research methods for subsequent building materials performance optimization and production process improvement.

Description

Technical Field

[0001] The present invention relates to the field of building material production, and in particular to a method and system for evaluating and predicting building material production quality based on data feedback. Background Art

[0002] In the process of building materials production, product quality is affected by many factors, including raw material characteristics, production process parameters, production equipment status and environmental conditions. Traditional quality control methods often rely on sampling inspection and post-analysis, which makes it difficult to grasp the quality changes in the production process in real time and comprehensively, and cannot effectively predict future product quality trends, timely improve and correct production processes, resulting in an inability to adapt to real market demand. However, due to the variety of building materials and the ever-changing use conditions, the same type of building materials will also have a variety of production processes. In order to adapt to the various use conditions of building materials, it is often not accurate to evaluate the production quality of building materials from a single data perspective. Therefore, it is necessary to combine the defect feedback during the use of building materials and the uniformity of the use conditions to accurately evaluate and predict the production quality of building materials. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method and system for evaluating and predicting the quality of building materials production based on data feedback, which discusses the quality of building materials production under similar working conditions by monitoring the performance of building materials during use.

[0004] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0005] A method for evaluating and predicting the quality of building materials production based on data feedback is provided, which comprises:

[0006] Step S1: Determine the production links of building materials according to the types of building materials, and obtain the key production parameters that affect the quality of building materials in each production link to obtain a production parameter data set:

[0007] ;

[0008] in, m is the number of the production link, For the m The production parameter number of each production link, For the m The first stage of production Production parameters;

[0009] Step S2: using the performance defect data collected by the quality supervision feedback module during the use of building materials, and collecting the use condition data of the building materials, from the building materials produced in the same batch, screen out building materials with similar use conditions;

[0010] Step S3: construct a performance defect assessment model based on production parameters of different production links, use the historical performance defect data fed back during the use of building materials and the production parameters of the production links of the corresponding building materials to train the performance defect assessment model, and optimize the performance defect assessment model based on the new performance defect data to obtain the optimal performance defect assessment model;

[0011] Step S4: After each batch of building materials is produced, the corresponding optimal performance defect assessment model is retrieved according to the types of performance defects that may occur, and the production parameter data group is input into the optimal performance defect assessment model to predict the performance defect data for the evaluation and prediction of the building materials production quality.

[0012] Furthermore, in step S2, the method for screening building materials with similar working conditions is:

[0013] Step S21: construct a text database of building material usage conditions, split the text or words in the text database into individual fonts, and establish an embedded data mapping for each font;

[0014] ;

[0015] in, u Number the font. N is the number of strokes of the font, is the semantic classification label of the font. Fonts with the same semantics have the same semantic classification label. is the embedded data mapping function, For font u Embedded data;

[0016] Step S22: each font obtains corresponding embedded data through embedded data mapping to obtain an initial embedded data set of the font in the text database;

[0017] Step S23: determining the difference between the embedded data according to the number of strokes of the font and the semantic classification label, constructing an embedded data difference function, and correcting the same embedded data in the initial embedded data set;

[0018] Step S24: After the same embedded data in the initial embedded data set is corrected using the embedded data difference function, a standard embedded data set is obtained;

[0019] Step S25: According to the text combination of the building material usage condition type, retrieve the font embedding data combination corresponding to the text combination of the usage condition type from the standard embedding data set. , x The number of fonts for the condition type text combination. Embedding data for fonts of condition type text combinations;

[0020] Step S26: Calculate the similarity coefficient of the building material usage conditions, use the similarity coefficient to evaluate the similarity of the building material usage conditions, and screen building materials with similar usage conditions.

[0021] Furthermore, the embedding data difference function is:

[0022] ;

[0023] in, and are the two font numbers with the same embedded data in the initial embedding dataset, Fonts and fonts The embedded data Fonts and fonts The number of strokes, Fonts and fonts Semantic classification labels.

[0024] Further, step S26 includes:

[0025] Step S261: Calculate the similarity coefficient of building material usage conditions;

[0026] ;

[0027] in, are semantically similar terms, is the parameter similarity term, i The number of the font embedding data. They are the first characters corresponding to the usage condition type combination of two building materials. i font embedding data, They are the operating parameters of two building materials respectively;

[0028] Step S262: When When the parameter similarity As the similarity coefficient and the similarity threshold set For comparison, if , then the use conditions of the two building materials are judged to be similar, otherwise, the use conditions of the two building materials are judged to be dissimilar; when When the working conditions of the two building materials are different, it is directly determined that they are not similar.

[0029] Further, step S3 includes:

[0030] Step S31: Extract performance defect data from building materials with similar working conditions, and construct a performance defect data set for building materials with similar working conditions , Mis the type of performance defect data, For the M performance defect data;

[0031] Step S32: Constructing a performance defect evaluation model based on production parameters of different production links ;

[0032] ;

[0033] in, For the m The first stage of production Production parameters Corresponding regression coefficients, performance defect evaluation model According to the number of production links, m a performance defect assessment sub-model;

[0034] Step S33: Generate a training data set using the historical performance defect data fed back during the use of building materials and the production parameters of the corresponding building materials production process to evaluate the performance defect model. Perform training, output the optimal regression coefficient, and obtain the trained performance defect evaluation model ;

[0035] Step S34: The quality supervision feedback module collects new performance defect data , and carry the usage data of the building materials, use the usage data to perform steps S21-S27, and calculate new performance defect data The similarity coefficient of the corresponding operating conditions is used to retrieve the trained performance defect assessment model. ;

[0036] Step S35: New performance defect data The corresponding production parameter data set of building materials is input into the trained performance defect assessment model In the example, the sub-model output is evaluated according to each performance defect m Performance defect prediction data ;

[0037] Step S36: Calculate the average value of performance defect prediction data ;

[0038] ;

[0039] in, is the number of the performance defect prediction data, For the Performance defect prediction data;

[0040] Step S37: Calculate the average value With new performance defect data The difference between them is calculated and the performance defect evaluation model after training is completed The coefficient of precision ;

[0041] ;

[0042] Step S38: Setting the threshold of the accuracy coefficient ;

[0043] like , then the performance defect evaluation model currently trained is If the accuracy meets the requirements, the model is used as the current optimal performance defect assessment model, and the process returns to step S34, where the quality supervision feedback module continues to collect new performance defect data.

[0044] like , then the performance defect evaluation model currently trained is If the accuracy does not meet the requirement, return to step S33 and write the new performance defect data As historical performance defect data, update the training data set and retrain the performance defect assessment model , output the new trained performance defect assessment model , as the current optimal performance defect assessment model.

[0045] A building material production quality assessment and prediction system is provided, which executes the above-mentioned building material production quality assessment and prediction method based on data feedback, and comprises:

[0046] Quality supervision and feedback module, including performance data monitoring module, data feedback module and evaluation module;

[0047] The performance data monitoring module includes several performance sensors installed on the building materials, which are used to collect performance data during the use of the building materials;

[0048] The data feedback module is used to input the building material performance data fed back by the staff;

[0049] The evaluation module is used to process and filter the performance data and output performance defect data;

[0050] The production data collection module is used to collect key production parameters that affect the quality of building materials in each production link and obtain a production parameter data group;

[0051] The data analysis module combines the production parameter data set with the performance defect data to process and analyze the data, and evaluate and predict the production quality of building materials.

[0052] The beneficial effects of the present invention are as follows: the present invention screens performance defects in the process of using building materials by monitoring performance changes during the use of building materials, and is used to reversely study defects in the production process of building materials, which can assist in optimizing the production process and process of building materials under different use conditions. At the same time, the present invention discusses the relationship between performance defects and production parameters of building materials under the same working conditions, constructs an evaluation model, ensures the evaluation and prediction accuracy of quality defects in building material production, adapts to building material use scenarios under multiple working conditions, reduces the impact of other factors on the production quality evaluation process, and provides reliable data support and research methods for subsequent building material performance optimization and production process improvement. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The flowchart of the building materials production quality assessment and prediction method based on data feedback. DETAILED DESCRIPTION

[0054] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0055] like Figure 1 As shown, a building material production quality assessment and prediction method based on data feedback includes:

[0056] Step S1: Determine the production links of building materials according to the types of building materials, and obtain the key production parameters that affect the quality of building materials in each production link to obtain a production parameter data set:

[0057] ;

[0058] in, m is the number of the production link, For the m The production parameter number of each production link, For the m The first stage of production There are many kinds of building materials, and the production process of different building materials will be different. The production process includes mixing, stirring, molding, heat treatment, etc. The production parameters include mixing and stirring time, stirring speed, molding pressure, processing temperature, etc.

[0059] Step S2: Using the performance defect data collected by the quality supervision feedback module during the use of building materials, and collecting the usage condition data of building materials, from the same batch of building materials, screen out building materials with similar usage conditions; the method for screening building materials with similar usage conditions is: the performance defect data of building materials include bending deformation, insufficient strength, abnormal hardness, cracking, dimensional deviation, abnormal service life, etc.

[0060] Step S21: construct a text database of building material usage conditions, split the text or words in the text database into individual fonts, and establish an embedded data mapping for each font; the usage condition type includes phrases and condition parameters corresponding to the condition semantics;

[0061] ;

[0062] in, u Number the font. N is the number of strokes of the font, is the semantic classification label of the font. Fonts with the same semantics have the same semantic classification label. is the embedded data mapping function, For font u Embedded data;

[0063] Step S22: each font obtains corresponding embedded data through embedded data mapping to obtain an initial embedded data set of the font in the text database;

[0064] Step S23: Determine the difference between the embedded data according to the number of strokes of the font and the semantic classification label, construct an embedded data difference function, and correct the same embedded data in the initial embedded data set; the embedded data difference function is:

[0065] ;

[0066] in, and are the two font numbers with the same embedded data in the initial embedding dataset, Fonts and fonts The embedded data Fonts and fonts The number of strokes, Fonts and fonts Semantic classification labels.

[0067] The embedded data of the operating condition type may be substantially the same or accidentally the same. The present invention introduces an embedded data difference function to correct the accidentally identical embedded data, ensuring that one operating condition type corresponds to only one embedded data.

[0068] Step S24: After the same embedded data in the initial embedded data set is corrected using the embedded data difference function, a standard embedded data set is obtained;

[0069] Step S25: According to the text combination of the building material usage condition type, retrieve the font embedding data combination corresponding to the text combination of the usage condition type from the standard embedding data set. , x The number of fonts for the condition type text combination. Embedding data for fonts of condition type text combinations;

[0070] Step S26: Calculate the similarity coefficient of the building material usage conditions, use the similarity coefficient to evaluate the similarity of the building material usage conditions, and select building materials with similar usage conditions. Step S26 specifically includes:

[0071] Step S261: Calculate the similarity coefficient of building material usage conditions;

[0072] ;

[0073] in, are semantically similar terms, is the parameter similarity term, i The number of the font embedding data. They are the first characters corresponding to the usage condition type combination of two building materials. i font embedding data, They are the operating parameters of two building materials respectively;

[0074] Step S262: When When the parameter similarity As the similarity coefficient and the similarity threshold set For comparison, if , then the use conditions of the two building materials are judged to be similar, otherwise, the use conditions of the two building materials are judged to be dissimilar; when When , it is directly determined that the usage conditions of the two building materials are not similar. Similarity threshold Indicates the maximum allowable value of the similarity coefficient for judging the similarity of operating conditions.

[0075] Step S3: Construct a performance defect assessment model based on production parameters of different production links, use the historical performance defect data fed back during the use of building materials and the production parameters of the corresponding building materials production links to train the performance defect assessment model, and optimize the performance defect assessment model based on the new performance defect data to obtain the optimal performance defect assessment model. Step S3 specifically includes:

[0076] Step S31: Extract performance defect data from building materials with similar working conditions, and construct a performance defect data set for building materials with similar working conditions , M is the type of performance defect data, For the M performance defect data;

[0077] Step S32: Constructing a performance defect evaluation model based on production parameters of different production links ;

[0078] ;

[0079] in, For the m The first stage of production Production parameters Corresponding regression coefficients, performance defect evaluation model According to the number of production links, m a performance defect assessment sub-model;

[0080] Step S33: Generate a training data set using the historical performance defect data fed back during the use of building materials and the production parameters of the corresponding building materials production process to evaluate the performance defect model. Perform training, output the optimal regression coefficient, and obtain the trained performance defect evaluation model ;

[0081] Step S34: The quality supervision feedback module collects new performance defect data , and carry the usage data of the building materials, use the usage data to perform steps S21-S27, and calculate new performance defect data The similarity coefficient of the corresponding operating conditions is used to retrieve the trained performance defect assessment model. ;

[0082] Step S35: New performance defect data The corresponding production parameter data set of building materials is input into the trained performance defect assessment model In the example, the sub-model output is evaluated according to each performance defect m Performance defect prediction data ;

[0083] Step S36: Calculate the average value of performance defect prediction data ;

[0084] ;

[0085] in, is the number of the performance defect prediction data, For the Performance defect prediction data;

[0086] Step S37: Calculate the average value With new performance defect data The difference between them is calculated and the performance defect evaluation model after training is completed The coefficient of precision ;

[0087] ;

[0088] Step S38: Setting the threshold of the accuracy coefficient ;Threshold Indicates the maximum value allowed for the accuracy coefficient that satisfies the model accuracy requirements.

[0089] like , then the performance defect evaluation model currently trained is If the accuracy meets the requirements, the model is used as the current optimal performance defect assessment model, and the process returns to step S34, where the quality supervision feedback module continues to collect new performance defect data.

[0090] like , then the performance defect evaluation model currently trained is If the accuracy does not meet the requirement, return to step S33 and write the new performance defect data As historical performance defect data, update the training data set and retrain the performance defect assessment model , output the new trained performance defect assessment model , as the current optimal performance defect assessment model.

[0091] Step S4: After each batch of building materials is produced, the corresponding optimal performance defect assessment model is retrieved according to the type of performance defects that may occur, and the production parameter data group is input into the optimal performance defect assessment model to predict the performance defect data for the assessment and prediction of the quality of building materials production. In general, the larger the performance defect data calculated according to the optimal performance defect assessment model, the worse the performance of the produced building materials under the working conditions of drinking, and the worse the quality of the building materials production.

[0092] A building material production quality assessment and prediction system, which executes the above-mentioned building material production quality assessment and prediction method based on data feedback, comprises:

[0093] Quality supervision and feedback module, including performance data monitoring module, data feedback module and evaluation module;

[0094] The performance data monitoring module includes several performance sensors installed on the building materials, which are used to collect performance data during the use of the building materials;

[0095] The data feedback module is used to input the building material performance data fed back by the staff;

[0096] The evaluation module is used to process and filter the performance data and output performance defect data;

[0097] The production data collection module is used to collect key production parameters that affect the quality of building materials in each production link and obtain a production parameter data group;

[0098] The data analysis module combines the production parameter data set with the performance defect data to process and analyze the data, and evaluate and predict the production quality of building materials.

[0099] The present invention screens performance defects in the process of using building materials by monitoring performance changes during the use of building materials, and is used to reversely study defects in the production process of building materials, which can assist in optimizing the production process and process of building materials under different use conditions. At the same time, the present invention discusses the relationship between performance defects and production parameters of building materials under the same working conditions, constructs an evaluation model, ensures the evaluation and prediction accuracy of quality defects in building material production, adapts to building material use scenarios under multiple working conditions, reduces the impact of other factors on the production quality evaluation process, and provides reliable data support and research methods for subsequent building material performance optimization and production process improvement.

Claims

1. A method for evaluating and predicting building material production quality based on data feedback, characterized in that: include: Step S1: Determine the production links of building materials according to the types of building materials, and obtain the key production parameters that affect the quality of building materials in each production link to obtain a production parameter data set: ； in, m is the number of the production link, For the m The production parameter number of each production link, For the m The first stage of production Production parameters; Step S2: using the performance defect data collected by the quality supervision feedback module during the use of building materials, and collecting the use condition data of the building materials, from the building materials produced in the same batch, screen out building materials with similar use conditions; Step S3: construct a performance defect assessment model based on production parameters of different production links, use the historical performance defect data fed back during the use of building materials and the production parameters of the production links of the corresponding building materials to train the performance defect assessment model, and optimize the performance defect assessment model based on the new performance defect data to obtain the optimal performance defect assessment model; Step S4: After each batch of building materials is produced, the corresponding optimal performance defect assessment model is retrieved according to the types of performance defects that may occur, and the production parameter data group is input into the optimal performance defect assessment model to predict the performance defect data for the evaluation and prediction of the building materials production quality.

2. The method for evaluating and predicting building material production quality based on data feedback according to claim 1 is characterized in that: The method for screening building materials with similar working conditions in step S2 is: Step S21: construct a text database of building material usage conditions, split the text or words in the text database into individual fonts, and establish an embedded data mapping for each font; ； in, u Number the font. N is the number of strokes of the font, is the semantic classification label of the font. Fonts with the same semantics have the same semantic classification label. is the embedded data mapping function, For font u Embedded data; Step S22: each font obtains corresponding embedded data through embedded data mapping to obtain an initial embedded data set of the font in the text database; Step S23: determining the difference between the embedded data according to the number of strokes of the font and the semantic classification label, constructing an embedded data difference function, and correcting the same embedded data in the initial embedded data set; Step S24: After the same embedded data in the initial embedded data set is corrected using the embedded data difference function, a standard embedded data set is obtained; Step S25: According to the text combination of the building material usage condition type, retrieve the font embedding data combination corresponding to the text combination of the usage condition type from the standard embedding data set. , x The number of fonts for the condition type text combination. Embedding data for fonts of condition type text combinations; Step S26: Calculate the similarity coefficient of the building material usage conditions, use the similarity coefficient to evaluate the similarity of the building material usage conditions, and screen building materials with similar usage conditions.

3. The method for evaluating and predicting building material production quality based on data feedback according to claim 2 is characterized in that: The embedded data difference function is: ； in, and are the two font numbers with the same embedded data in the initial embedding dataset, Fonts and fonts The embedded data Fonts and fonts The number of strokes, Fonts and fonts Semantic classification labels.

4. The method for evaluating and predicting building material production quality based on data feedback according to claim 3 is characterized in that: The step S26 comprises: Step S261: Calculate the similarity coefficient of building material usage conditions; ； in, are semantically similar terms, is the parameter similarity term, i The number of the font embedding data. They are the first characters corresponding to the usage condition type combination of two building materials. i font embedding data, They are the operating parameters of two building materials respectively; Step S262: When When the parameter similarity As the similarity coefficient and the similarity threshold set For comparison, if , then the use conditions of the two building materials are judged to be similar, otherwise, the use conditions of the two building materials are judged to be dissimilar; when When the working conditions of the two building materials are different, it is directly determined that they are not similar.

5. The method for evaluating and predicting building material production quality based on data feedback according to claim 4 is characterized in that: The step S3 comprises: Step S31: Extract performance defect data from building materials with similar working conditions, and construct a performance defect data set for building materials with similar working conditions , M is the type of performance defect data, For the M performance defect data; Step S32: Constructing a performance defect evaluation model based on production parameters of different production links ; ； in, For the m The first stage of production Production parameters Corresponding regression coefficients, performance defect evaluation model According to the number of production links, m a performance defect assessment sub-model; Step S33: Generate a training data set using the historical performance defect data fed back during the use of building materials and the production parameters of the corresponding building materials production process to evaluate the performance defect model. Perform training, output the optimal regression coefficient, and obtain the trained performance defect evaluation model ; Step S34: The quality supervision feedback module collects new performance defect data , and carry the usage data of the building materials, use the usage data to perform steps S21-S26, and calculate new performance defect data The similarity coefficient of the corresponding operating conditions is used to retrieve the trained performance defect assessment model. ; Step S35: New performance defect data The corresponding production parameter data set of building materials is input into the trained performance defect assessment model In the example, the sub-model output is evaluated according to each performance defect m Performance defect prediction data ; Step S36: Calculate the average value of performance defect prediction data ; ； in, is the number of the performance defect prediction data, For the Performance defect prediction data; Step S37: Calculate the average value With new performance defect data The difference between them is calculated and the performance defect evaluation model after training is completed The coefficient of precision ; ； Step S38: Setting the threshold of the accuracy coefficient ; like , then the performance defect evaluation model currently trained is If the accuracy meets the requirements, the model is used as the current optimal performance defect assessment model, and the process returns to step S34, where the quality supervision feedback module continues to collect new performance defect data. like , then the performance defect evaluation model currently trained is If the accuracy does not meet the requirement, return to step S33 and write the new performance defect data As historical performance defect data, update the training data set and retrain the performance defect assessment model , output the new trained performance defect assessment model , as the current optimal performance defect assessment model.

6. A building material production quality assessment and prediction system, executing the building material production quality assessment and prediction method based on data feedback according to any one of claims 1 to 5, characterized in that: include: Quality supervision and feedback module, including performance data monitoring module, data feedback module and evaluation module; The performance data monitoring module includes a number of performance sensors installed on the building materials, which are used to collect performance data during the use of the building materials; The data feedback module is used to input the building material performance data fed back by the staff; The evaluation module is used to process and screen the performance data and output performance defect data; The production data collection module is used to collect key production parameters that affect the quality of building materials in each production link and obtain a production parameter data group; The data analysis module combines the production parameter data set with the performance defect data to process and analyze the data, and evaluate and predict the production quality of building materials.

Images