A prediction model validation method for architectural ceramics production

Abstract

The present invention relates to the field of model verification technology, and in particular to a prediction model verification method for building ceramic production, comprising the following steps: step S1: offline verification of a prediction model to be verified to obtain an offline verification result; step S2: judging whether the offline verification result meets the offline production requirements, if yes, proceeding to step S3, if not, adjusting the prediction model, and repeating step S1; step S3: online verification of the prediction model that has passed the offline verification to obtain an online verification result; step S4: judging whether the online verification result meets the online production requirements, if yes, completing the verification, if not, adjusting the prediction model, and repeating step S1. The present invention can verify the feasibility and accuracy of the prediction model for building ceramic production, while ensuring the effectiveness of the prediction model in the verification process and reducing the trial and error cost.

Description

Technical Field

[0001] The invention relates to the technical field of model verification, and in particular to a prediction model verification method for building ceramic production. Background Art

[0002] With the continuous improvement of my country's national economic level and the continuous progress of social productivity, more and more ceramic furniture is used in home decoration. When people choose from many different brands and manufacturers, they are paying more and more attention to the variety and quality of ceramic furniture. Ceramic tiles are a must-have product for product decoration, and their quality varies. How to accurately detect the quality of ceramic tiles has become an important issue of concern in construction projects.

[0003] At present, in order to realize the evaluation of ceramic tile quality and to better improve the quality analysis of ceramic tiles, the big data prediction model of building ceramics has come into being. It mainly collects raw material data, process data and equipment data of building ceramic production, analyzes the production status and predicts the adjustment of production parameters through big data modeling, and realizes the early estimation of ceramic quality indicators. However, this prediction model relies on the rigid generation of existing data and is not suitable for actual production. A method is needed to verify its feasibility and accuracy. Summary of the invention

[0004] The purpose of the present invention is to propose a prediction model verification method for building ceramic production, which can verify the feasibility and accuracy of the prediction model for building ceramic production, while ensuring the effectiveness of the prediction model during the verification process and reducing the trial and error cost.

[0005] To achieve this object, the present invention adopts the following technical solutions:

[0006] A prediction model validation method for building ceramic production, comprising the following steps:

[0007] Step S1: Perform offline verification on the prediction model to be verified to obtain an offline verification result;

[0008] Step S2: Determine whether the offline verification result meets the offline production requirements. If yes, proceed to step S3. If no, adjust the prediction model and repeat step S1.

[0009] Step S3: Performing online verification on the prediction model that has passed offline verification to obtain an online verification result;

[0010] Step S4: Determine whether the online verification result meets the online production requirements. If so, complete the verification. If not, adjust the prediction model and repeat step S1.

[0011] Preferably, in step S1, performing offline verification on the prediction model to be verified to obtain an offline verification result comprises the following steps:

[0012] Step S11: inputting N actual production historical data into the prediction model to be verified, obtaining N offline prediction production parameters, and drawing the N prediction production parameters into an offline prediction production curve;

[0013] Step S12: Plotting actual production history parameters corresponding to N actual production history data into an actual production history curve;

[0014] Step S13: Draw a predicted production upper limit curve and a predicted production lower limit curve according to the upper and lower limits of the prediction standard of the prediction model, and define the predicted production upper limit curve and the predicted production lower limit curve; wherein the area between the predicted production upper limit curve and the predicted production lower limit curve is defined as a prediction area;

[0015] Step S14: Draw a process standard upper limit curve and a process standard lower limit curve according to the upper and lower limits of the production process standard, wherein the area between the process standard upper limit curve and the process standard lower limit curve is defined as a qualified area;

[0016] Step S15: Draw the offline predicted production curve, the actual production history curve, the predicted production upper limit curve, the predicted production lower limit curve, the process standard upper limit curve and the process standard lower limit curve into a curve graph to obtain the offline verification result.

[0017] Preferably, in step S2, determining whether the offline verification result meets the offline production requirements comprises the following steps:

[0018] Step S21: determine whether the actual production history curve is within the prediction area;

[0019] Step S22: Determine whether the predicted production upper limit curve and the predicted production lower limit curve are within the qualified area.

[0020] Preferably, in step S21, the determining whether the actual production history curve is within the prediction area comprises the following steps:

[0021] Step S211: Calculate the predicted proportion of actual production history curves that are not within the predicted area;

[0022] Step S212: Determine whether the prediction ratio is less than the prediction error determination point. If the prediction ratio is less than the prediction error determination point, execute step S22; if the prediction ratio is greater than or equal to the prediction error determination point, output information that the prediction model is incorrect, modify the prediction model, and repeat step S1.

[0023] Preferably, in step S22, the step of determining whether the predicted production upper limit curve and the predicted production lower limit curve are within a qualified region comprises the following steps:

[0024] Step S221: Calculate the qualified ratio of those not in the qualified area;

[0025] Step S222: Determine whether the qualified proportion is less than the qualified error determination point. If the qualified proportion is less than the qualified error determination point, output the information that the prediction model passes the offline verification and execute step S3; if the qualified proportion is greater than or equal to the qualified error determination point, output the information that the prediction model is wrong and repeat step S1.

[0026] Preferably, in step S3, the prediction model that has passed offline verification is verified online to obtain an online verification result, which includes the following steps:

[0027] Step S31: inputting N real-time production data into the prediction model that has been verified offline to obtain N online prediction production parameters;

[0028] Step S32: Plotting N online predicted production parameters into an online production curve to obtain an online verification result.

[0029] Preferably, in step S4, determining whether the online verification result meets the online production requirements comprises the following steps:

[0030] Step S41: Send the online verification result to the expert module and receive the expert judgment information;

[0031] Step S42: If the expert judges that the information is in compliance, production is carried out according to the online predicted production parameters, the actual parameters during production are recorded, and steps S43-S44 are executed;

[0032] If the expert judges that the information is not in compliance, the prediction model is adjusted and step S1 is repeated;

[0033] Step S43: sending the actual real parameters and the online predicted production parameters to the expert module to obtain difference analysis information;

[0034] Step S44: According to the difference analysis results, the prediction model is optimized, and step S3 is repeated until the difference between the actual real parameters and the online predicted production parameters is within the set allowable range, and the prediction model is judged to be effective.

[0035] Preferably, the difference analysis information includes offline test difference analysis information and online test difference analysis information;

[0036] The offline test difference analysis information includes the parameter error rate between the actual real parameter and the online predicted production parameter, and the parameter error rate is (online predicted production parameter-actual real parameter) / actual real parameter*100%;

[0037] The online test difference analysis information includes production efficiency difference results and quality difference results;

[0038] The production efficiency difference result is the production efficiency difference result between the actual production efficiency obtained by the actual real parameters and the predicted production efficiency obtained by the online predicted production parameters, and the production efficiency difference result is (predicted production efficiency-actual production efficiency) / actual production efficiency*100%;

[0039] The quality difference result is the quality difference result between the actual quality obtained by the actual real parameters and the predicted quality obtained by the online predicted production parameters, and the quality difference result is (predicted quality-actual quality) / actual quality*100%.

[0040] One of the above technical solutions has the following beneficial effects: first, after offline verification to ensure that it meets the offline production requirements, the prediction model is verified online based on the established online production requirements; through multiple adjustments and judging whether the prediction data of the prediction model meets the process production requirements and predicted production requirements designed in the production requirements, not only the verification operability and rationality of the prediction model are met, but also the prediction data will gradually meet the production requirements, so as to better evaluate and analyze the quality of ceramic tiles actually produced. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic flow chart of a prediction model verification method for building ceramic production according to the present invention;

[0042] Figure 2 It is a schematic diagram of the principle of the prediction model verification method for building ceramic production of the present invention;

[0043] Figure 3 A schematic diagram of offline verification results of Embodiment 1 of the prediction model verification method for building ceramic production of the present invention;

[0044] Figure 4 A schematic diagram of offline verification results of Example 2 of the prediction model verification method for building ceramic production of the present invention;

[0045] Figure 5 A schematic diagram of online verification results of Example 3 of the prediction model verification method for building ceramic production of the present invention;

[0046] Figure 6A schematic diagram of online verification results of Example 4 of the prediction model verification method for building ceramic production of the present invention; DETAILED DESCRIPTION

[0047] The technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0048] like Figure 1 As shown, a prediction model verification method for building ceramic production includes the following steps:

[0049] Step S1: Perform offline verification on the prediction model to be verified to obtain an offline verification result;

[0050] Step S2: Determine whether the offline verification result meets the offline production requirements. If yes, proceed to step S3. If no, adjust the prediction model and repeat step S1.

[0051] Step S3: Performing online verification on the prediction model that has passed offline verification to obtain an online verification result;

[0052] Step S4: Determine whether the online verification result meets the online production requirements. If so, complete the verification. If not, adjust the prediction model and repeat step S1.

[0053] As for how to verify the feasibility of the prediction model used for building ceramic production, and at the same time to ensure the effectiveness of the model during the verification process and reduce the cost of trial and error, the prediction model verification method in this embodiment is to first verify offline that it meets the offline production requirements, and then verify the prediction model online based on the established online production requirements; through multiple adjustments and judging whether the prediction data of the prediction model meets the process production requirements and predicted production requirements designed in the production requirements, not only the verification operability and rationality of the prediction model are met, but also the prediction data will gradually meet the production requirements, so as to better evaluate and analyze the quality of ceramic tiles actually produced.

[0054] To further illustrate, in step S1, the prediction model to be verified is offline verified to obtain an offline verification result, including the following steps:

[0055] Step S11: inputting N actual production historical data into the prediction model to be verified, obtaining N offline prediction production parameters, and drawing the N prediction production parameters into an offline prediction production curve;

[0056] Step S12: Plotting actual production history parameters corresponding to N actual production history data into an actual production history curve;

[0057] Step S13: Draw a predicted production upper limit curve and a predicted production lower limit curve according to the upper and lower limits of the prediction standard of the prediction model, and define the predicted production upper limit curve and the predicted production lower limit curve; wherein the area between the predicted production upper limit curve and the predicted production lower limit curve is defined as a prediction area;

[0058] Step S14: Draw a process standard upper limit curve and a process standard lower limit curve according to the upper and lower limits of the production process standard, wherein the area between the process standard upper limit curve and the process standard lower limit curve is defined as a qualified area;

[0059] Step S15: Draw the offline predicted production curve, the actual production history curve, the predicted production upper limit curve, the predicted production lower limit curve, the process standard upper limit curve and the process standard lower limit curve into a curve graph to obtain the offline verification result.

[0060] like Figure 2 As shown, this embodiment plots the offline verification result on a curve chart through multiple parameter data, which can intuitively feedback the changes in statistical data.

[0061] To further illustrate, in step S2, the determination of whether the offline verification result meets the offline production requirements includes the following steps:

[0062] Step S21: determine whether the actual production history curve is within the prediction area;

[0063] Step S22: Determine whether the predicted production upper limit curve and the predicted production lower limit curve are within the qualified area.

[0064] This embodiment improves the verification quality of offline verification through continuous judgment during offline verification.

[0065] To further illustrate, in step S21, the step of determining whether the actual production history curve is within the prediction region includes the following steps:

[0066] Step S211: Calculate the predicted proportion of actual production history curves that are not within the predicted area;

[0067] Step S212: Determine whether the prediction ratio is less than the prediction error determination point. If the prediction ratio is less than the prediction error determination point, execute step S22; if the prediction ratio is greater than or equal to the prediction error determination point, output information that the prediction model is incorrect, modify the prediction model, and repeat step S1.

[0068] To further illustrate, in step S22, the step of determining whether the predicted production upper limit curve and the predicted production lower limit curve are within the qualified region includes the following steps:

[0069] Step S221: Calculate the qualified ratio of those not in the qualified area;

[0070] Step S222: Determine whether the qualified proportion is less than the qualified error determination point. If the qualified proportion is less than the qualified error determination point, output the information that the prediction model passes the offline verification and execute step S3; if the qualified proportion is greater than or equal to the qualified error determination point, output the information that the prediction model is wrong and repeat step S1.

[0071] This embodiment combines offline predicted production parameters with production process standards and actual production history parameters for comparative analysis to further improve the verification quality of offline verification. The process includes plotting the N offline predicted production parameters output by the prediction model into predicted production upper limit curves and predicted production lower limit curves, and judging that all of them fall within the process standard upper limit curve and process standard lower limit curve required by the production process standard, and there is no statistically significant difference compared with the actual production history curve of the actual production history parameters, that is, 95% of the points of the actual production history parameters fall within the prediction parameter range output by the model, indicating that the model is usable and can learn from the workers' experience, so that the offline verification results meet the actual production needs.

[0072] To further illustrate, in step S3, the prediction model that has passed offline verification is verified online to obtain an online verification result, which includes the following steps:

[0073] Step S31: inputting N real-time production data into the prediction model that has been verified offline to obtain N online prediction production parameters;

[0074] Step S32: Plotting N online predicted production parameters into an online production curve to obtain an online verification result.

[0075] To further illustrate, in step S4, determining whether the online verification result meets the online production requirements includes the following steps:

[0076] Step S41: Send the online verification result to the expert module and receive the expert judgment information;

[0077] Step S42: If the expert judges that the information is in compliance, production is carried out according to the online predicted production parameters, the actual parameters during production are recorded, and steps S43-S44 are executed;

[0078] If the expert judges that the information is not in compliance, the prediction model is adjusted and step S1 is repeated;

[0079] Step S43: sending the actual real parameters and the online predicted production parameters to the expert module to obtain difference analysis information;

[0080] Step S44: According to the difference analysis results, the prediction model is optimized, and step S3 is repeated until the difference between the actual real parameters and the online predicted production parameters is within the set allowable range, and the prediction model is judged to be effective.

[0081] Since the use of model prediction parameters may lead to unstable production or even production quality accidents, it is necessary to conduct multiple offline verifications before online verification to see whether they meet production requirements and ensure the identity of production conditions. Secondly, the online prediction production parameters predicted by the prediction model need to be discussed and verified with process experts and firing experts. If they meet the production process, they are adjusted and produced according to the online prediction production parameters. At the same time, big data model experts, process experts, firing experts and quality control experts track online to prevent production accidents. At the same time, the changes in production are recorded at all times, especially the final product quality status, to identify product quality defects that are not related to the kiln, in order to judge the accuracy of the online prediction production parameters predicted by the prediction model.

[0082] Preferably, the difference analysis information includes offline test difference analysis information and online test difference analysis information;

[0083] The offline test difference analysis information includes the parameter error rate between the actual real parameter and the online predicted production parameter, and the parameter error rate is (online predicted production parameter-actual real parameter) / actual real parameter*100%;

[0084] The online test difference analysis information includes production efficiency difference results and quality difference results;

[0085] The production efficiency difference result is the production efficiency difference result between the actual production efficiency obtained by the actual real parameters and the predicted production efficiency obtained by the online predicted production parameters, and the production efficiency difference result is (predicted production efficiency-actual production efficiency) / actual production efficiency*100%;

[0086] The quality difference result is the quality difference result between the actual quality obtained by the actual real parameters and the predicted quality obtained by the online predicted production parameters, and the quality difference result is (predicted quality-actual quality) / actual quality*100%.

[0087] To further illustrate the feasibility and rationality of the verification of the present invention, four examples are provided below. Example 1: Offline verification of ball milling prediction model

[0088] Table 1 Offline verification results of ball mill prediction model for predicting mud fineness

[0089]

[0090]

[0091] Summary: The offline verification pass rate of the ball mill model predicting mud quality is 90%, and its normal production product index is 90%. Therefore, the offline prediction production parameters of the prediction model meet the offline production requirements, indicating that the ball mill prediction model meets the production requirements.

[0092] Example 2: Offline verification of kiln prediction model

[0093] Table 2 Offline verification results of the kiln prediction model predicted temperature

[0094]

[0095]

[0096]

[0097] Summary: The temperatures predicted by the kiln model are all within the set range of the process standard temperature, and the actual production trends are basically consistent, indicating that the kiln model prediction has met the production process requirements.

[0098] Example 3: In-service verification of ball milling prediction model

[0099] Table 3 Online verification results of ball mill prediction model for predicting slurry fineness

[0100]

[0101] Summary: Through the above results, it is found that the predicted mud fineness is basically consistent with the actual mud fineness, the maximum absolute value of the error rate is 4%, and the error rate is ≤5%, indicating that the model effect meets the requirements. Therefore, it is shown that this ball mill fineness prediction model meets the production process requirements and has a high prediction accuracy.

[0102] Example 4: Online verification of kiln prediction model

[0103] Table 4 Online verification results of the production quality rate of kiln model prediction parameters

[0104]

[0105] Summary: According to the daily error value of the excellent rate during normal production and verification, it was found that the error value was within the range of ±5, indicating that the kiln prediction model was available.

[0106] The technical principle of the present invention is described above in conjunction with specific embodiments. These descriptions are only for explaining the principle of the present invention and cannot be interpreted as limiting the scope of protection of the present invention in any way. Based on the explanations herein, those skilled in the art can associate other specific embodiments of the present invention without creative work, and these equivalent variations or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A prediction model verification method for building ceramic production, characterized in that: The following steps are involved: Step S1: Perform offline verification on the prediction model to be verified to obtain an offline verification result; Step S2: Determine whether the offline verification result meets the offline production requirements. If yes, proceed to step S3. If no, adjust the prediction model and repeat step S1. Step S3: Performing online verification on the prediction model that has passed offline verification to obtain an online verification result; Step S4: determine whether the online verification result meets the online production requirements. If so, complete the verification. If not, adjust the prediction model and repeat step S1. In step S1, the prediction model to be verified is offline verified to obtain an offline verification result, including the following steps: Step S11: inputting N actual production historical data into the prediction model to be verified, obtaining N offline prediction production parameters, and drawing the N prediction production parameters into an offline prediction production curve; Step S12: Plotting actual production history parameters corresponding to N actual production history data into an actual production history curve; Step S13: Draw a predicted production upper limit curve and a predicted production lower limit curve according to the upper and lower limits of the prediction standard of the prediction model, and define the predicted production upper limit curve and the predicted production lower limit curve; wherein the area between the predicted production upper limit curve and the predicted production lower limit curve is defined as a prediction area; Step S14: Draw a process standard upper limit curve and a process standard lower limit curve according to the upper and lower limits of the production process standard, wherein the area between the process standard upper limit curve and the process standard lower limit curve is defined as a qualified area; Step S15: Draw the offline predicted production curve, the actual production history curve, the predicted production upper limit curve, the predicted production lower limit curve, the process standard upper limit curve and the process standard lower limit curve into a curve graph to obtain an offline verification result; In step S2, the determination of whether the offline verification result meets the offline production requirements includes the following steps: Step S21: determine whether the actual production history curve is within the prediction area; Step S22: judging whether the predicted production upper limit curve and the predicted production lower limit curve are within the qualified area; In step S21, the determination of whether the actual production history curve is within the prediction area includes the following steps: Step S211: Calculate the predicted proportion of actual production history curves that are not within the predicted area; Step S212: determine whether the prediction ratio is less than the prediction error determination point. If the prediction ratio is less than the prediction error determination point, execute step S22; if the prediction ratio is greater than or equal to the prediction error determination point, output information indicating that the prediction model is wrong, modify the prediction model, and repeat step S1; In step S22, the step of judging whether the predicted production upper limit curve and the predicted production lower limit curve are within the qualified region comprises the following steps: Step S221: Calculate the qualified ratio of those not in the qualified area; Step S222: Determine whether the qualified proportion is less than the qualified error determination point. If the qualified proportion is less than the qualified error determination point, output the information that the prediction model passes the offline verification and execute step S3; if the qualified proportion is greater than or equal to the qualified error determination point, output the information that the prediction model is wrong and repeat step S1.

2. A prediction model verification method for building ceramic production according to claim 1, characterized in that: In step S3, the prediction model that has passed offline verification is verified online to obtain an online verification result, including the following steps: Step S31: inputting N real-time production data into the prediction model that has been verified offline to obtain N online prediction production parameters; Step S32: Plotting N online predicted production parameters into an online production curve to obtain an online verification result.

3. A prediction model verification method for building ceramic production according to claim 2, characterized in that: In step S4, the determination of whether the online verification result meets the online production requirements comprises the following steps: Step S41: Send the online verification result to the expert module and receive the expert judgment information; Step S42: If the expert judges that the information is in compliance, production is carried out according to the online predicted production parameters, the actual parameters during production are recorded, and steps S43-S44 are executed; If the expert judges that the information is not in compliance, the prediction model is adjusted and step S1 is repeated; Step S43: sending the actual real parameters and the online predicted production parameters to the expert module to obtain difference analysis information; Step S44: Optimize the prediction model based on the difference analysis information, and repeat step S3 until the difference between the actual real parameters and the online predicted production parameters is within the set allowable range, then the prediction model is judged to be valid and the verification is completed.

4. A prediction model verification method for building ceramic production according to claim 3, characterized in that: The difference analysis information includes offline test difference analysis information and online test difference analysis information; The offline test difference analysis information includes the parameter error rate between the actual real parameter and the online predicted production parameter, and the parameter error rate is (online predicted production parameter-actual real parameter) / actual real parameter*100%; The online test difference analysis information includes production efficiency difference results and quality difference results; The production efficiency difference result is the production efficiency difference result between the actual production efficiency obtained by the actual real parameters and the predicted production efficiency obtained by the online predicted production parameters, and the production efficiency difference result is (predicted production efficiency-actual production efficiency) / actual production efficiency*100%; The quality difference result is the quality difference result between the actual quality obtained by the actual real parameters and the predicted quality obtained by the online predicted production parameters, and the quality difference result is (predicted quality-actual quality) / actual quality*100%.

Images