Heat setting process parameter recommendation method and device, computer equipment and storage medium

Abstract

The invention discloses a heat setting process parameter recommendation method and device, computer equipment and a storage medium. The method comprises the steps of obtaining to-be-predicted production order data; preprocessing the to-be-predicted production order data to obtain a preprocessing result; inputting the preprocessing result into a prediction model to perform process parameter prediction so as to obtain a prediction result; wherein the prediction model is obtained by preprocessing production order data, labeling the data as a sample set and training a multi-weight main model formed by stacking a plurality of base models with weights and meta-models, and the weights of the base models are determined by the prediction performance of the base models; a loss function in the training process is added into the quality score of each piece of preprocessed and labeled data; and outputting the prediction result. By implementing the method disclosed by the invention, the effect of enhancing stacking by multiple models can be exerted, the dependence on artificial experience is reduced, the accuracy of process parameters and the one-time success rate of production are improved, and the product repair and production cost is reduced.

Description

Technical Field

[0001] The present invention relates to the field of machine learning, and more specifically to a method, device, computer equipment and storage medium for recommending heat setting process parameters. Background Art

[0002] Heat setting is a key step in textile dyeing and finishing production. The process involves heating the fabric at a certain size at a high temperature for a certain period of time and then cooling it at an appropriate speed. The purpose of heat setting is to ensure that the fabric can meet the predetermined requirements such as flatness, width, weight and moisture content during the heating process.

[0003] At present, the selection and adjustment of heat setting process mainly rely on long-term experience accumulation and lack of precise process design, resulting in low accuracy and poor adaptability of process design. Such practices easily lead to instability in shrinkage, weight and width of the same batch of products, and the setting effect is not ideal. Due to reliance on experience, high scrap rate and production cost often occur, especially when customer demand or production conditions change, traditional methods are difficult to meet new needs. Even if the process parameters are determined, it is difficult to accurately predict the quality indicators after setting. Usually, product quality can only be judged by manual sampling inspection, which increases uncertainty. In the end, regardless of whether the setting quality is qualified, the fabric must go through a complete heat setting process. This process not only wastes time and resources, but also increases manpower and material costs. In general, traditional methods cannot ensure a stable and efficient production process. In addition, there are currently some parameters that are predicted and recommended by models, but the process parameters of heat setting include target requirements for machine cloth sealing, target requirements for machine weight, target requirements for machine WPI, machine CPI, target temperature, target machine speed, target width reference, target overfeed, first target width deviation, and second target width deviation. When a single model is used for stacking prediction, the setting of the base model weight is often ignored, resulting in the influence of some base models being over-magnified and the role of other models being weakened. In this way, the advantages of each base model cannot be fully utilized, which affects the prediction accuracy and generalization ability of the entire model.

[0004] Therefore, it is necessary to design a new method to give full play to the role of multi-model enhanced stacking, reduce the reliance on manual experience, improve the accuracy of process parameters and the first-time success rate of production, and reduce product rework and production costs. Summary of the invention

[0005] The purpose of the present invention is to overcome the defects of the prior art and provide a method, device, computer equipment and storage medium for recommending heat setting process parameters.

[0006] To achieve the above object, the present invention adopts the following technical solution: a method for recommending heat setting process parameters, comprising:

[0007] Obtain production order data to be predicted;

[0008] Preprocessing the production order data to be predicted to obtain a preprocessing result;

[0009] The preprocessing result is input into the prediction model to predict the process parameters to obtain the prediction result; wherein the prediction model is obtained by preprocessing the production order data and labeling it as a sample set, and training a multi-weight main model formed by stacking several weighted base models and meta models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data;

[0010] The prediction result is outputted.

[0011] The further technical solution is: the prediction model is obtained by preprocessing and labeling the production order data as a sample set, training a multi-weighted main model formed by stacking several weighted base models and meta models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data, including:

[0012] Acquire production order data, and preprocess and label the production order data to obtain a sample set, wherein the sample set includes feature data and label data;

[0013] Construct a multi-weighted master model;

[0014] Calculating a data quality score of the feature data;

[0015] Input the characteristic data into each base model to predict process parameters, and calculate the mean square error of each base model according to the prediction result and the corresponding label data;

[0016] Calculate the weight of the corresponding base model according to the mean square error of each base model;

[0017] Multiplying the weight of the base model by the result predicted by the base model to obtain a weighted prediction result;

[0018] Extracting corresponding process parameters from the weighted prediction results to form a matrix;

[0019] Determine the final prediction result according to the matrix combined with the linear regression model;

[0020] Using mean error as a loss function, and adding a data quality score of the feature data into the loss function;

[0021] The prediction model is obtained by minimizing the loss function, adjusting the parameters of the meta-model, and repeating the training until the meta-model converges.

[0022] A further technical solution is: the construction of a multi-weighted main model includes:

[0023] Several models whose prediction performance meets the requirements are selected as base models, and a linear regression model is selected as a meta-model. The base models and the meta-model are stacked to form a multi-weight main model.

[0024] A further technical solution is: the step of calculating the data quality score of the feature data includes:

[0025] use Calculate the data quality score of the feature data, where s represents the quantity quality score of each feature data; S2 is the target deviation, which represents the degree of deviation between the actual test result and the target value of the parameter, and takes a value between (0, 1]; E M E is the relative error ratio of the cloth seal; G is the relative error ratio of gram weight; M is the target deviation critical value when the difference between the actual test result and the target value is equal to the error value; actual is the actual value of the cloth seal; M target is the sealing target value; t M is the error range of the cloth seal; G actual G is the actual value of gram weight; target is the target weight value; t G is the error range of gram weight; the calculated data quality score of each characteristic data is combined to form a data quality score set.

[0026] A further technical solution is: the weight of the corresponding base model is calculated according to the mean square error of each base model, including:

[0027] use Calculate the weight of the corresponding base model, where W i is the weight of the i-th base model; MSE i is the mean square error of the i-base models, m represents the number of base models; n is the amount of training data; h is the number of process parameters; y is the number of training data; ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the feature data in the i-th row; s i is the data quality score of the i-th feature data.

[0028] A further technical solution is: determining the final prediction result according to the matrix combined with the linear regression model includes:

[0029] Determining a coefficient matrix of the linear regression model;

[0030] The coefficient matrix is ​​split and transposed row by row to obtain a new coefficient matrix;

[0031] Performing a point multiplication operation on the matrix and each of the new coefficient matrices to obtain a process parameter prediction value for each feature data;

[0032] All the process parameter prediction values ​​are combined to form a final prediction result.

[0033] Its further technical solution is: the base model includes a GBR model, an RFR model and a KNN model.

[0034] The present invention also provides a device for recommending heat setting process parameters, comprising:

[0035] An acquisition unit, used for acquiring production order data to be predicted;

[0036] A preprocessing unit, used for preprocessing the production order data to be predicted to obtain a preprocessing result;

[0037] A prediction unit, used for inputting the preprocessing result into a prediction model to predict the process parameters, so as to obtain a prediction result; wherein the prediction model is obtained by preprocessing and labeling the production order data as a sample set, and training a multi-weighted main model formed by stacking a number of weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data;

[0038] An output unit is used to output the prediction result.

[0039] The present invention further provides a computer device, which includes a memory and a processor. The memory stores a computer program, and the processor implements the above method when executing the computer program.

[0040] The present invention also provides a storage medium, wherein the storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0041] Compared with the prior art, the present invention has the following beneficial effects: the present invention performs automatic parameter prediction through a trained multi-weight main model, specifically obtains production order data to be predicted, and performs preprocessing to eliminate noise and normalize the data format; labels the preprocessed data, constructs a standardized sample set, and ensures the consistency of data quality; trains through a multi-weight main model, which is composed of a stack of multiple base models and meta-models, and the weights of the base models are adjusted according to their prediction performance; during the training process, a loss function with a quality score is used to strengthen the impact of high-quality data on the model, thereby improving the prediction accuracy; the advantages of each base model are integrated through the stacking model, the prediction results of the process parameters are further optimized, and the deviation that may be caused by a single model is reduced; the role of multi-model enhanced stacking is brought into play, the reliance on manual experience is reduced, the accuracy of process parameters and the one-time success rate of production are improved, and product rework and production costs are reduced.

[0042] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0044] Figure 1 A schematic diagram of an application scenario of the method for recommending heat setting process parameters provided by an embodiment of the present invention;

[0045] Figure 2 A schematic diagram of a process flow of a method for recommending heat setting process parameters provided by an embodiment of the present invention;

[0046] Figure 3 A schematic diagram of a sub-process of a method for recommending heat setting process parameters provided by an embodiment of the present invention;

[0047] Figure 4 A schematic diagram of a sub-process of a method for recommending heat setting process parameters provided by an embodiment of the present invention;

[0048] Figure 5 A schematic diagram of a flow chart of a training process of a multi-weighted main model provided by an embodiment of the present invention;

[0049] Figure 6 A schematic diagram of an interface for inputting basic batch information and target information provided by an embodiment of the present invention;

[0050] Figure 7 Schematic diagram of an output page provided by an embodiment of the present invention Figure 1 ;

[0051] Figure 8 Schematic diagram of an output page provided by an embodiment of the present invention Figure 2 ;

[0052] Fig. 9 A schematic block diagram of a device for recommending heat setting process parameters provided by an embodiment of the present invention;

[0053] Fig.10 A schematic block diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0056] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0057] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0058] See also Figure 1 and Figure 2 , Figure 1 A schematic diagram of an application scenario of the method for recommending heat setting process parameters provided in an embodiment of the present invention. Figure 2A schematic flow chart of a method for recommending heat setting process parameters provided in an embodiment of the present invention. The method for recommending heat setting process parameters is applied to a server. The server interacts with a terminal for data exchange, wherein the server may be a controller of a fabric heat setting device, etc., and optimizes the selection of process parameters through advanced machine learning techniques to meet specific production goals. This method not only improves production efficiency, but also ensures the consistency and stability of product quality. The method of this embodiment applies the Stacking algorithm technology to the recommendation of multiple process parameters in the textile dyeing and finishing industry, and improves the overall prediction performance by integrating multiple base models. The data quality score calculation based on the label error and target deviation feedback of the initial model is combined, the quality of the training data is evaluated by the deviation value between the setting test result and the target value, and these quality scores are converted into weights and applied to model training, thereby improving the robustness and accuracy of the model. A base model weight allocation mechanism is introduced into the stacking model, and the influence weight of each base model is dynamically adjusted according to the loss value feedback of a single base model to optimize the performance of the stacking technology.

[0059] Figure 2 FIG. 1 is a flow chart of a method for recommending heat setting process parameters provided by an embodiment of the present invention. Figure 2 As shown, the method includes the following steps S110 to S140.

[0060] S110. Obtain production order data to be predicted.

[0061] In this embodiment, the production order data to be predicted includes data from the order data management system, specifically including cylinder number, process route, material composition, cloth seal, gram weight, density, temperature and other data, which can be obtained from Figure 6 The interface shown is filled in manually, by device scanning, or by automatic crawling.

[0062] S120: preprocessing the production order data to be predicted to obtain a preprocessing result.

[0063] In this embodiment, the preprocessing result refers to the preprocessed production order data. After receiving the production order data, preprocessing is required to improve the accuracy and efficiency of the prediction.

[0064] Specifically, in production order data, missing data is often encountered, which may be due to omissions or errors in data collection. Missing data will affect the accuracy of model training and subsequent analysis, so it must be effectively processed. In the method of this embodiment, for the missing pre-cooking and post-cooking data, first try to match them in the data source of the development office by the cylinder number. If the missing data still cannot be completed, choose to delete these incomplete data. This method helps to avoid the adverse effects of missing data on the model, ensure the integrity and accuracy of the data, and thus improve the reliability of subsequent analysis and model prediction.

[0065] There may be some duplicate records in the production order data. If these duplicate data are not processed, they may lead to deviations in the analysis results. In order to eliminate the impact of duplicate data on the model, the repeatability analysis of feature data is used. Specifically, by comparing the feature data of each production order, when it is found that there is exactly the same data, the record with the most recent retention time is selected according to the order time, and other duplicate records are deleted. This step can ensure that the model will not cause errors due to duplicate data, thereby improving the accuracy of the analysis and the validity of the data.

[0066] In the actual production process, due to manual input or other operational problems, there may be some outliers in the production order data. These outliers will seriously interfere with the learning and prediction of the model, so it is necessary to detect and process the outliers in the data. Specifically, the following formula is used to determine the outliers; K={k1,k2,…,k n}; where n is the amount of training data, and k is the ratio of the target specification to the pre-cooking specification; b i Utilize the seal for the i-th data object; g i b′ is the target weight of the i-th data; i g′ is the seal before cooking the i-th data; i is the weight of the i-th data before cooking.

[0067] The method of this embodiment combines data features and uses the interquartile range (IQR) method to detect outliers. The specific operation is to calculate the 25th percentile (q25) and 75th percentile (q75) of the data, and then determine the upper and lower limits of the outliers based on the IQR (i.e., q75-q25). Any data outside this range will be regarded as an outlier and deleted. This step can ensure that each data in the data set is within a reasonable range, avoid unnecessary interference of outliers to the model, and thus improve the accuracy and prediction effect of the model.

[0068] Specifically, the first quantile is also called the quarter quantile or the 25% quantile, and the third quantile is also called the quarter quantile or the 75% quantile. Let q 25is the first quantile, which means that 25% of the data points in data K are less than q 25 ,q 75 is the third quantile, which means that 75% of the data points in data K are greater than q 75 , then IQR=q 75 -q 25 ;

[0069] Thus, we can obtain the upper bound of the threshold and the lower bound of the threshold: lower = q 25 -1.5*IQR; upper=q 75 +1.5*IQR;

[0070] The data K is then filtered by the threshold [lower, upper], and any values ​​that are not within the range are considered outliers and deleted.

[0071] By processing missing data, duplicate data, and outliers, we ensure that input data is more complete, accurate, and consistent, thus providing a reliable foundation for subsequent analysis and model training. Data processing can eliminate noise and bias, ensuring that the model can better identify and learn the true relationship between data and improve the reliability of predictions. By deleting incomplete or abnormal records, we avoid the negative impact of erroneous data on analysis results, helping to obtain more accurate insights into production data.

[0072] These processing steps ensure the quality of the dataset, ultimately optimizing the decision support and predictive capabilities of process parameters during production.

[0073] S130, inputting the preprocessing result into a prediction model to perform process parameter prediction to obtain a prediction result.

[0074] In this embodiment, the prediction result refers to the predicted values ​​corresponding to parameters such as required cloth sealing, required grammage, required WPI, required CPI, temperature, machine speed, width reference, overfeed, first width deviation and second width deviation.

[0075] Among them, WPI and CPI respectively represent the coil density index of the fabric. CPI (Courses Per Inch) refers to the number of horizontal rows of coils in 1 inch of the fabric, that is, the lateral density of the fabric, and WPI (Wales Per Inch) refers to the number of vertical rows of coils in 1 inch of the fabric, that is, the longitudinal density of the fabric.

[0076] The prediction model is obtained by preprocessing and labeling the production order data as a sample set, and training a multi-weight main model formed by stacking several weighted base models and meta-models. The weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data.

[0077] In one embodiment, see Figure 3 and Figure 5 The above-mentioned prediction model is obtained by preprocessing and labeling the production order data as a sample set, and training a multi-weight main model formed by stacking several weighted base models and meta-models. The weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data, including steps S131 to S1310.

[0078] S131. Obtain production order data, and preprocess and label the production order data to obtain a sample set, wherein the sample set includes feature data and label data.

[0079] In this embodiment, the characteristic data refers to the data such as cylinder number, process route, material composition, cloth seal, gram weight, density, temperature, etc., and the label data refers to the actual values ​​corresponding to the parameters such as the required cloth seal, required gram weight, required WPI, required CPI, temperature, machine speed, width reference, upper overfeed, first width deviation and second width deviation, etc., which are called

[0080] The preprocessing in this step is consistent with the above-mentioned step S120 and will not be described in detail here.

[0081] S132. Construct a multi-weight main model.

[0082] In this embodiment, the multi-weight main model refers to a stacking model that incorporates weights.

[0083] Specifically, several models whose prediction performance meets the requirements are selected as base models, and a linear regression model is selected as a meta-model, and the base model and the meta-model are stacked to form a multi-weight main model.

[0084] The training process of the multi-weighted main model is actually stacked ensemble learning, which is a method of fusing the prediction results of multiple base models through a meta-model to improve the overall prediction performance, especially when dealing with complex regression problems. The structure of the stacked model usually consists of two main parts: the base model of the first layer and the meta-model of the second layer. For the selection of the base model, first, the original data set is divided into multiple subsets, and K-fold cross-validation is usually used to effectively reduce the overfitting risk of the base model. Each base model (for example, KNN, SVR, GBR, RFR, LR, DTR, etc.) is trained on each subset and generates prediction results. The prediction results of these base models are independent and may differ to some extent because each base model has different assumptions and preferences when processing data. Selecting base models with strong predictive power can capture different patterns in the data. The diversity of base models can make the integration effect better. For example, models such as KNN, SVR, GBR, RFR, DTR, etc. can be used as base models because they have different advantages under different data characteristics. For the meta-model, the prediction results of all base models in the first layer are combined with the original features to form a new data set as the input of the meta-model. The meta-model is usually a relatively simple model (such as linear regression LR), which is responsible for further integrating and adjusting the output of the first layer base model to generate the final prediction result; the meta-model generally selects a model with a relatively simple structure and high computational efficiency, such as linear regression (LR). A simple meta-model can effectively avoid overfitting because it does not have too many degrees of freedom to overfit the output of the first layer.

[0085] In practical applications, different base models are usually selected by evaluating their regression effects. Common evaluation indicators include regression coefficient R2 and mean square error (MSE). According to these evaluation indicators, the performance of each base model is ranked, and the best performing model is selected as the base model of the stacked model.

[0086] For example, the following base models were evaluated:

[0087] KNN (k nearest neighbor regression); SVR (support vector machine regression); GBR (gradient boosting regression); RFR (random forest regression); LR (linear regression); DTR (decision tree regression);

[0088] By calculating the R 2 and MSE, and sort them from small to large according to the size of MSE. Assume that the top three models with the best performance are selected, which are: GBR, RFR and KNN. These models will be used as the base models of the first layer.

[0089] Next, a model with a simple structure, such as LR (linear regression), is selected as the meta-model of the second layer. The output of the base model of the first layer is fused through it to finally obtain the prediction result of the stacked model.

[0090] Stacked ensemble learning improves the overall prediction performance by passing the prediction results of multiple base models to a meta-model to fuse their predictions. The key steps include:

[0091] In the first layer, select base models that perform well and can capture different characteristics of the data.

[0092] A simple meta-model (such as linear regression) is used for fusion in the second layer to avoid overfitting.

[0093] By sorting the base models based on evaluation indicators such as MSE, selecting the optimal base model, and combining it with a simple meta-model to improve the stability and accuracy of the entire model, this method can effectively improve the prediction accuracy of regression problems, especially for complex practical problems.

[0094] S133: Calculate the data quality score of the feature data.

[0095] In this embodiment, the Calculate the data quality score of the feature data, where s represents the quantity quality score of each feature data; S2 is the target deviation, which represents the degree of deviation between the actual test result and the target value of the parameter, and takes a value between (0, 1]; E M E is the relative error ratio of the cloth seal; G is the relative error ratio of gram weight; M is the target deviation critical value when the difference between the actual test result and the target value is equal to the error value; actual is the actual value of the cloth seal; M target is the sealing target value; t M is the error range of the cloth seal; G actual G is the actual value of gram weight; target is the target weight value; t G is the error range of gram weight; the data quality scores of each characteristic data calculated are combined to form a data quality score set. S = [s1, s2, ...s n ]; where n is the amount of training data, s n is the data quality score of the nth data.

[0096] Step S133 is a quality scoring mechanism based on experimental data, which aims to evaluate the deviation of the test results of parameters such as cloth seal (M) and grammage (G) during the production process from the target value, and then assign a quality score to each piece of data. This method combines the difference in error range, degree of deviation and target value, and can provide high-quality data support for the model while retaining data diversity.

[0097] Specifically, the test results of the two parameters of cloth cover and gram weight are analyzed to calculate the quality score of each data. The key indicator of the evaluation is the deviation between the test result and the target value. The evaluation formula takes into account the following aspects:

[0098] Target Deviation (S): This indicator reflects the degree of deviation between the actual test results and the target value, and the score range is (0, 1). The closer the score is to 1, the higher the quality of the data is and the closer it is to the target value.

[0099] Error range (E M and E G ): The relative error between the actual and target values ​​of cloth seal and gram weight is used to calculate the error ratio (E M and E G If these two error ratios are within the allowable range (i.e. E M ≤1 and E G ≤1), the standard formula is used to calculate the target deviation; if the error ratio exceeds the allowable range (i.e. E M >1 or E G >1), the target deviation decreases rapidly.

[0100] The quality score of each data is calculated based on the error ratio of the cloth cover and gram weight. The calculation of the quality score takes into account whether the test result meets the expected value and how to quickly reduce the quality score when the error exceeds the range. For test results outside the error range, although the data is not directly deleted, a lower quality score will be given to reflect its poor quality.

[0101] The quality scores of the training data form a data set that represents the quality of all the training data. It can reflect the quality level of the overall data and help screen out high-quality sample data for subsequent analysis or model training.

[0102] Even if some data test results exceed the error range, they will not be deleted directly, but retained by lowering the quality score. This helps maintain the diversity of the data set while ensuring that the quality issues of unqualified data can be accurately reflected in subsequent analysis. By calculating the target deviation, the difference between the data and the target value can be quantified, and the quality score can be quickly adjusted according to the actual error. This precise evaluation method makes the quality score of each data more objective and scientific. For cases where the test results exceed the allowable error range, the quality score will drop rapidly. This design can effectively prevent the overall quality assessment deviation caused by individual abnormal data. The closer the test result is to the target value, the higher the quality score, indicating that the data has a higher reference value for actual production and model training, so that the quality of the final training set is higher and the prediction effect is better. By using actual production historical data as a training set, this method can more realistically reflect various possible situations in the production process and enhance the adaptability and generalization ability of the model.

[0103] In summary, this quality scoring mechanism not only ensures data diversity but also ensures that the data used in the model training process has high quality through refined deviation processing, error range control and dynamic adjustment of quality scores, thereby improving the overall prediction accuracy and optimization effect of the production process.

[0104] S134, inputting the characteristic data into each base model to predict process parameters, and calculating the mean square error of each base model according to the prediction result and the corresponding label data.

[0105] S135. Calculate the weight of the corresponding base model according to the mean square error of each base model.

[0106] In this embodiment, the Calculate the weight of the corresponding base model, where W i is the weight of the i-th base model; MSE i is the mean square error of the i-base models, m represents the number of base models; n is the amount of training data; h is the number of process parameters; y is the number of training data; ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the feature data in the i-th row; s i is the data quality score of the i-th feature data.

[0107] In the traditional stacking method, the prediction results of multiple base models are usually synthesized by simple averaging. However, this embodiment adds a base model weighting mechanism, namely WeightStacking, based on the traditional stacking algorithm. Through this method, each base model is given a different weight according to its prediction performance and reliability, thereby achieving more accurate and stable results.

[0108] Specifically, the performance of each base model is evaluated by calculating the mean square error (MSE) of each base model, and this error value is converted into the weight of the model. The smaller the mean square error (MSE), the more accurate the model prediction is, and the larger the corresponding weight is; conversely, the larger the error is, the smaller the weight is. During the calculation process, the weights of all base models are normalized so that their sum is 1. In this way, Weight Stacking can effectively improve the accuracy and stability of the overall prediction, giving more reliable base models higher weights.

[0109] Weight Stacking introduces a weighting mechanism for base models, so that the contribution of each model is proportional to its performance. Models with stronger performance automatically receive higher weights, thereby optimizing the accuracy of the final prediction results. Through the weighting mechanism, models with larger errors are assigned lower weights, so they have less impact on the final prediction results, effectively reducing the interference of model errors on the results. By considering the mean square error of each model, Weight Stacking can improve the stability of the combined prediction results. In the face of data noise or instability of certain base models, other more stable models will compensate with higher weights to avoid fluctuations in the overall prediction results. As an evaluation criterion, the mean square error can directly reflect the prediction performance of the base model. Through this criterion, Weight Stacking can automatically select models with excellent performance and stability, thereby enhancing the reliability of the overall model combination. When faced with situations with complex data structures or multiple process parameters, Weight Stacking can effectively assign appropriate weights to different base models, thereby better adapting to diverse data features and avoiding the risk that a single model may not cover all data patterns. As the training data and errors change, Weight Stacking can dynamically adjust the weight distribution of each model, so that the model combination can be continuously optimized and adapted to the new data environment. This flexibility improves the long-term applicability of the algorithm.

[0110] In summary, Weight Stacking not only optimizes the prediction accuracy but also improves the stability and reliability of the model through a weighting mechanism based on model errors. It is an effective improvement on the traditional stacking method, especially suitable for the combination of multiple models, and can show stronger adaptability on complex data sets.

[0111] S136: Multiply the weight of the base model by the result predicted by the base model to obtain a weighted prediction result.

[0112] In this embodiment, GBR, RFR, and KNN are used as base models, and LR is used as a meta-model for stacking. In the production order, the characteristic data fields include target practical cloth seal, target gram weight, D number, proportion of each component, cloth seal before cooking, gram weight before cooking, WPI before cooking, CPI before cooking, cloth seal after cooking, gram weight after cooking, WPI after cooking, and CPI after cooking. The parameters that need to be recommended include required cloth seal, required gram weight, required WPI, required CPI, temperature, machine speed, width reference, upper overfeed and other parameters.

[0113] Assuming that the model input data is an n×d matrix, the feature data is represented as X: Among them, n is the amount of training data, d is the number of feature fields; x ij ,i=1,2,…,n、j=1,2,…,d represents the jth feature field of the i-th row of data. Passing X to the base model, the prediction result is Y: Where n is the amount of training data; h is the number of process parameters; y ij ,i=1,2,…,n,j=1,2,…,h, represents the predicted value of the jth process parameter of the i-th row of data.

[0114] Each base model will generate its own prediction results, and the prediction results generated by the m base models are integrated to form Y ′ : Where m represents the number of base models; Y i Represents the predicted data of the i-th base model; y ij ,i=1,2,…,n、j=1,2,…,h represents the jth predicted value of the i-th row of data of the first base model. j=h+1,h+2,…,h*2 represents the jth predicted value of the i-th row of data of the second base model, and so on.

[0115] The weight of each base model is calculated in this way. A weighted mechanism is introduced on the basis of traditional stacking, which can assign different weights to each basic model, making the adjustment of model combination more flexible.

[0116] Substitute the weights into the new data set. In the subsequent meta-model training and prediction process, the weights of each basic model will play a role. The data after substituting the weights is: Where m represents the number of base models; W m is the weight of the mth base model; Y mis the process parameter matrix predicted by the mth base model; h is the number of process parameters; n is the amount of training data; y n(h*m) It is the predicted value of the hth parameter of the nth data under the mth base model.

[0117] S137. Extract corresponding process parameters from the weighted prediction results to form a matrix.

[0118] In this embodiment, the value of YY is extracted from the process parameters of each base model according to the process parameters, and the result is: A = [A1, A2, ..., A h ]; Where m represents the number of base models; W m is the weight of the mth base model; h is the number of process parameters; n is the amount of training data; y n(h*m) It is the predicted value of the hth parameter of the nth data under the mth base model.

[0119] S138. Determine the final prediction result according to the matrix combined with the linear regression model.

[0120] In this embodiment, the linear regression model performs estimation analysis through the linear relationship between the independent variable and the dependent variable, and quantifies the influence of the independent variable on the dependent variable through the coefficient.

[0121] In one embodiment, see Figure 4 , the above-mentioned step S138 may include S1381~S1384.

[0122] S1381, determining the coefficient matrix of the linear regression model;

[0123] S1382, splitting the coefficient matrix row by row and performing a transposition operation to obtain a new coefficient matrix;

[0124] S1383, performing a point multiplication operation on the matrix and each of the new coefficient matrices to obtain a process parameter prediction value for each feature data;

[0125] S1384. Combine all the process parameter prediction values ​​to form a final prediction result.

[0126] Specifically, each independent variable corresponds to a coefficient, which is multiplied by the independent variable in the model and then summed. The corresponding meta-model coefficient matrix of the present invention is P: Among them, h is the number of process parameters; m is the number of base models; p hm is the coefficient of the h-th process under the m-th base model.

[0127] Split P by rows and transpose it to get a new coefficient matrix: Among them, h is the number of process parameters; m is the number of base models; p hm is the coefficient of the h-th process under the m-th base model.

[0128] Finally, matrix multiplication is performed, A h The dimension of is (n,m), The dimension is (m, 1), and the dot product is used to obtain the hth process parameter of each data, which has a dimension of (n, 1). After combining, the final prediction result Z is formed, and the shape of Z is (n, h). The calculation formula is as follows: z nh =W1y n1 p h1 +W2y n(h*2) p h2 +…+W m y n(h*m) p hm ; Among them, z ij , i=1,2,…,n、j=1,2,…,h represents the jth predicted value of the i-th row of data; m represents the number of base models; h is the number of process parameters; n is the amount of training data; W m is the weight of the mth base model; y n(h*m) is the predicted value of the hth parameter of the nth data under the mth base model; p hm is the coefficient of the h-th process under the m-th base model.

[0129] Compared with the traditional stacking method that directly combines each base model for prediction, this embodiment introduces a feature weighting mechanism to weight the prediction results of different base models, which makes it possible to more flexibly adjust the contribution of each model, thereby optimizing the prediction performance of the model in a specific scenario. Weight stacking gives higher weights to models with better performance, which can more effectively utilize the strengths of each model, thereby improving the prediction accuracy of the integrated model. Lower weights are given to models with poorer performance, thereby reducing their impact on the final prediction results and reducing the risk of the integrated model.

[0130] S139. Use the mean error as a loss function, and add the data quality score of the feature data into the loss function.

[0131] In this embodiment, during model training, the optimization direction of the model is controlled by calculating the loss value, the selected loss function is the mean square error, and the data quality score is applied in the calculation. Where J(θ) is the objective function, which is a function of the metamodel parameter θ; n is the amount of training data; h is the number of process parameters; y ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the i-th row of data; s i is the quality score of the i-th sample.

[0132] S1310, adjusting the parameters of the meta-model by minimizing the loss function, and repeating the training until the meta-model converges to obtain a prediction model.

[0133] In this embodiment, the loss function is calculated by the data quality score s i The influence of each sample is adjusted. Samples with higher quality scores contribute more to the loss function, thus affecting the model training process.

[0134] During the model training process, the goal is to minimize the loss function and optimize the prediction results by continuously adjusting the parameters θ of the meta-model. The training process is repeated until the loss function of the model converges, and finally a multi-weighted master model is obtained.

[0135] The introduction of a weighted mechanism for data quality scores allows model training to pay more attention to the impact of high-quality samples. Samples with high quality scores contribute more to the loss, which helps the model learn and fit these important samples more accurately, thereby improving the overall prediction accuracy. In practical applications, data quality often varies. Through data quality scores, the model can dynamically adjust the degree of attention to data of different qualities, reduce the interference of low-quality data, and improve the learning ability of high-quality data. By minimizing the loss function with quality score weighting, the training process can avoid the negative impact of abnormal samples or low-quality data on model training. This improves the stability of model training, especially when facing a noisy data set. In the traditional training process, the loss function simply calculates the difference between the predicted value and the true value without considering the quality of the sample. By introducing the quality score, the loss function can automatically adjust the weight during the calculation process, so that the model can pay more attention to high-quality samples during the optimization process, thereby effectively improving the optimization efficiency of the model. Due to the weighted influence of high-quality samples, the model can identify important features faster and avoid being disturbed by noise or low-quality data, thereby accelerating the convergence process. Compared with models without quality score weighting, such a weighted mechanism can make training more efficient. By introducing data quality scores, the model can better adapt to diverse data sets and avoid overfitting low-quality data, thereby improving its ability to generalize to new data. This approach has significant advantages, especially when data quality varies. During model training, the quality score can be flexibly adjusted according to different data sources or application scenarios. This flexibility enables the model to perform well in different environments, especially when facing real-time data, and to dynamically adjust the training strategy.

[0136] By introducing the data quality score into the loss function and dynamically adjusting the weights during model training, the multi-weighted main model optimization method proposed in this invention can effectively improve the accuracy, stability and efficiency of model training. This method can not only reduce the negative impact of low-quality data, but also accelerate the training convergence speed and enhance the generalization ability of the model. It is particularly suitable for application scenarios with uneven data quality or high noise.

[0137] In addition, for each final prediction result output by the entire multi-weight main model, the mean square error corresponding to the label data is calculated, and the mean square error is mapped to obtain a mapping result.

[0138] In this embodiment, the prediction matrix of the process parameters is obtained, and the predicted value and the actual value are calculated to obtain the mean square error of each training data: L = [l1, l2, ..., l i ,…,l n ]; Where L is the set of average errors between the predicted values ​​and actual values ​​of the label data of the training data; n is the amount of training data; h is the number of process parameters; y is the average error ... ij is the predicted value of the jth process parameter of the i-th data; is the true value of the jth process parameter of the ith data.

[0139] L is mapped to between 1 and 0 through an exponential function. The smaller the mean square error l is, that is, the smaller the deviation is, the higher the score is, with the highest being 1. in, is the set of mean square errors after mapping; n is the amount of training data; l i is the average error between the predicted value and the actual value of the label data of the i-th data; e is a natural constant.

[0140] The data quality score is adjusted according to the mapping calculation, the parameters of the multi-weight main model are adjusted using the adjusted data quality score, and the training is repeated until the multi-weight main model converges to obtain the prediction model.

[0141] In this embodiment, during the training process of the multi-weighted main model, different weights are formulated according to the influence of the target deviation and the model feedback error, and a new data quality score is obtained through fusion calculation: S ′ =[s1 ′ ,s2 ′ ,…,s i ′ ,…,s n ′ ]; Among them, S ′ is the set of adjusted data quality scores; n is the amount of training data; w1 is the weight of target deviation; si is the target deviation of the i-th data; w2 is the weight of the model feedback error; The average error between the predicted value and the actual value of the label data after mapping the i-th data.

[0142] After obtaining a new data quality score, the training process needs to be re-executed, that is, steps S134 to S1310.

[0143] S140: Output the prediction result.

[0144] In this embodiment, the prediction results include the results of the two outputs of the multi-weight main model, specifically including the predicted values ​​of the required cloth seal, required weight, required WPI, required CPI, temperature, machine speed, upper overfeed, width reference, first width deviation and second width deviation, and are finally presented on the specified interface, such as Figures 7 and 8 shown.

[0145] The above-mentioned heat setting process parameter recommendation method performs automatic parameter prediction through the trained multi-weight main model, specifically obtains the production order data to be predicted, and performs preprocessing to eliminate noise and normalize the data format; labels the preprocessed data, constructs a standardized sample set, and ensures the consistency of data quality; trains through the multi-weight main model, which is composed of multiple base models and meta-models stacked, and the weights of the base models are adjusted according to their prediction performance; during the training process, a loss function with a quality score is used to strengthen the impact of high-quality data on the model, thereby improving the prediction accuracy; through the stacking model, the advantages of each base model are combined to further optimize the prediction results of the process parameters and reduce the deviation that may be caused by a single model; the role of multi-model enhanced stacking is brought into play, the reliance on manual experience is reduced, the accuracy of process parameters and the first-time success rate of production are improved, and product rework and production costs are reduced.

[0146] Fig. 9 FIG. 3 is a schematic block diagram of a heat setting process parameter recommendation device 300 provided in an embodiment of the present invention. Fig. 9 As shown, corresponding to the above heat setting process parameter recommendation method, the present invention also provides a heat setting process parameter recommendation device 300. The heat setting process parameter recommendation device 300 includes a unit for executing the above heat setting process parameter recommendation method, and the device can be configured in a server. Fig. 9 The heat setting process parameter recommendation device 300 includes an acquisition unit 301 , a preprocessing unit 302 , a prediction unit 303 and an output unit 304 .

[0147] An acquisition unit 301 is used to acquire production order data to be predicted; a preprocessing unit 302 is used to preprocess the production order data to be predicted to obtain a preprocessing result; a prediction unit 303 is used to input the preprocessing result into a prediction model to predict process parameters to obtain a prediction result; wherein the prediction model is obtained by preprocessing and labeling the production order data as a sample set, and training a multi-weight main model formed by stacking several weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data; an output unit 304 is used to output the prediction result.

[0148] In one embodiment, the dyeing and finishing industry process parameter recommendation system further includes a model training unit, and the model training unit includes:

[0149] An acquisition subunit, used for acquiring production order data, and preprocessing and labeling the production order data to obtain a sample set, wherein the sample set includes feature data and label data;

[0150] Constructing subunits for constructing multi-weighted main models;

[0151] The score calculation subunit is used to calculate the data quality score of the feature data; specifically,

[0152] Calculate the data quality score of the feature data, where s represents the quantity quality score of each feature data; S2 is the target deviation, which represents the degree of deviation between the actual test result and the target value of the parameter, and takes a value between (0, 1]; E M E is the relative error ratio of the cloth seal; G is the relative error ratio of gram weight; M is the target deviation critical value when the difference between the actual test result and the target value is equal to the error value; actual is the actual value of the cloth seal; M target is the sealing target value; t M is the error range of the cloth seal; G actual G is the actual value of gram weight; target is the target weight value; t G is the error range of gram weight; the calculated data quality score of each characteristic data is combined to form a data quality score set.

[0153] A base model prediction subunit, used for inputting the feature data into each base model to perform process parameter prediction, and calculating the mean square error of each base model according to the prediction result and the corresponding label data;

[0154] The weight calculation subunit is used to calculate the weight of the corresponding base model according to the mean square error of each base model; specifically, Calculate the weight of the corresponding base model, where W i is the weight of the i-th base model; MSE i is the mean square error of the i-base models, m represents the number of base models; n is the amount of training data; h is the number of process parameters; y is the number of training data; ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the feature data in the i-th row; s i is the data quality score of the i-th feature data.

[0155] A multiplication subunit, used for multiplying the weight of the base model with the result predicted by the base model to obtain a weighted prediction result;

[0156] An extraction subunit, used for extracting corresponding process parameters from the weighted prediction results to form a matrix;

[0157] A regression processing subunit, used to determine the final prediction result according to the matrix in combination with the linear regression model;

[0158] A loss function construction subunit, used for using the mean error as the loss function and adding the data quality score of the feature data into the loss function;

[0159] The adjustment subunit is used to adjust the parameters of the meta-model by minimizing the loss function, and repeat the training until the meta-model converges to obtain a multi-weight main model.

[0160] In one embodiment, the construction subunit is used to select several models whose prediction performance meets the requirements as base models, and select a linear regression model as a meta-model, and stack the base model and the meta-model to form a multi-weight main model.

[0161] In one embodiment, the regression processing subunit includes:

[0162] A coefficient matrix determination module, used to determine the coefficient matrix of the linear regression model;

[0163] A matrix processing module, used for performing row-wise splitting and transposition operations on the coefficient matrix to obtain a new coefficient matrix;

[0164] A dot product module, used for performing a dot product operation on the matrix and each of the new coefficient matrices to obtain a process parameter prediction value of each feature data;

[0165] The combination module is used to combine all the process parameter prediction values ​​to form a final prediction result.

[0166] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned heat setting process parameter recommendation device 300 and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0167] The heat setting process parameter recommendation device 300 can be implemented in the form of a computer program. Fig.10 Runs on the computer device shown.

[0168] See also Fig.10 , Fig.10 5 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device 500 may be a server, wherein the server may be an independent server or a server cluster composed of multiple servers.

[0169] See also Fig.10 The computer device 500 includes a processor 502 , a memory and a network interface 505 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0170] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, and when the program instructions are executed, the processor 502 can execute a method for recommending heat setting process parameters.

[0171] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0172] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a method for recommending heat setting process parameters.

[0173] The network interface 505 is used to communicate with other devices over the network. Fig.10 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0174] The processor 502 is used to run the computer program 5032 stored in the memory to implement the following steps:

[0175] Acquire production order data to be predicted; preprocess the production order data to be predicted to obtain preprocessing results; input the preprocessing results into a prediction model to predict process parameters to obtain prediction results; wherein the prediction model is obtained by preprocessing and labeling the production order data as a sample set, training a multi-weight main model formed by stacking several weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data; output the prediction result.

[0176] In one embodiment, the processor 502 implements the prediction model by preprocessing and labeling the production order data as a sample set, training a multi-weighted main model formed by stacking a plurality of weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process is added to the quality score step of each preprocessed and labeled data, and specifically implements the following steps:

[0177] Acquire production order data, and preprocess and label the production order data to obtain a sample set, wherein the sample set includes feature data and label data; construct a multi-weighted main model; calculate the data quality score of the feature data; input the feature data into each base model to predict process parameters, and calculate the mean square error of each base model based on the predicted result and the corresponding label data; calculate the weight of the corresponding base model based on the mean square error of each base model; multiply the weight of the base model by the result predicted by the base model to obtain a weighted prediction result; extract the corresponding process parameters from the weighted prediction result to form a matrix; determine the final prediction result based on the matrix combined with the linear regression model; use the mean error as the loss function, and add the data quality score of the feature data to the loss function; adjust the parameters of the meta-model by minimizing the loss function, and repeat the training until the meta-model converges to obtain a prediction model.

[0178] In one embodiment, when implementing the step of building a multi-weighted main model, the processor 502 specifically implements the following steps:

[0179] Several models whose prediction performance meets the requirements are selected as base models, and a linear regression model is selected as a meta-model. The base models and the meta-model are stacked to form a multi-weight main model.

[0180] In one embodiment, when the processor 502 implements the step of calculating the data quality score of the feature data, the processor 502 specifically implements the following steps:

[0181] use Calculate the data quality score of the feature data, where s represents the quantity quality score of each feature data; S2 is the target deviation, which represents the degree of deviation between the actual test result and the target value of the parameter, and takes a value between (0, 1]; E M E is the relative error ratio of the cloth seal; G is the relative error ratio of gram weight; M is the target deviation critical value when the difference between the actual test result and the target value is equal to the error value; actual is the actual value of the cloth seal; M target is the sealing target value; t M is the error range of the cloth seal; G actual G is the actual value of gram weight; target is the target weight value; t G is the error range of gram weight; the calculated data quality score of each characteristic data is combined to form a data quality score set.

[0182] In one embodiment, when the processor 502 implements the step of calculating the weight of the corresponding base model according to the mean square error of each base model, it specifically implements the following steps:

[0183] use Calculate the weight of the corresponding base model, where W i is the weight of the i-th base model; MSE i is the mean square error of the i-base models, m represents the number of base models; n is the amount of training data; h is the number of process parameters; y is the number of training data; ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the feature data in the i-th row; s i is the data quality score of the i-th feature data.

[0184] In one embodiment, when the processor 502 implements the step of determining the final prediction result according to the matrix combined with the linear regression model, the processor 502 specifically implements the following steps:

[0185] Determine the coefficient matrix of the linear regression model; split the coefficient matrix by rows and transpose it to obtain a new coefficient matrix; perform a dot multiplication operation on the matrix and each of the new coefficient matrices to obtain a process parameter prediction value for each feature data; combine all the process parameter prediction values ​​to form a final prediction result.

[0186] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0187] It can be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0188] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein when the computer program is executed by a processor, the processor executes the following steps:

[0189] Acquire production order data to be predicted; preprocess the production order data to be predicted to obtain preprocessing results; input the preprocessing results into a prediction model to predict process parameters to obtain prediction results; wherein the prediction model is obtained by preprocessing and labeling the production order data as a sample set, training a multi-weight main model formed by stacking several weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data; output the prediction result.

[0190] In one embodiment, the processor implements the prediction model by preprocessing and labeling the production order data as a sample set, training a multi-weighted main model formed by stacking a plurality of weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score step of each preprocessed and labeled data, and specifically implements the following steps:

[0191] Acquire production order data, and preprocess and label the production order data to obtain a sample set, wherein the sample set includes feature data and label data;

[0192] Construct a multi-weighted main model; calculate the data quality score of the feature data; input the feature data into each of the base models to predict the process parameters, and calculate the mean square error of each base model based on the predicted results and the corresponding label data; calculate the weight of the corresponding base model based on the mean square error of each base model; multiply the weight of the base model by the result predicted by the base model to obtain a weighted prediction result; extract the corresponding process parameters from the weighted prediction result to form a matrix; determine the final prediction result based on the matrix combined with the linear regression model; use the mean error as the loss function, and add the data quality score of the feature data to the loss function; adjust the parameters of the meta-model by minimizing the loss function, and repeat the training until the meta-model converges to obtain a prediction model.

[0193] In one embodiment, when the processor executes the computer program to implement the step of building a multi-weighted main model, the processor specifically implements the following steps:

[0194] Several models whose prediction performance meets the requirements are selected as base models, and a linear regression model is selected as a meta-model. The base models and the meta-model are stacked to form a multi-weight main model.

[0195] In one embodiment, when the processor executes the computer program to implement the step of calculating the data quality score of the feature data, the processor specifically implements the following steps:

[0196] use Calculate the data quality score of the feature data, where s represents the quantity quality score of each feature data; S2 is the target deviation, which represents the degree of deviation between the actual test result and the target value of the parameter, and takes a value between (0, 1]; E M E is the relative error ratio of the cloth seal; G is the relative error ratio of gram weight; M is the target deviation critical value when the difference between the actual test result and the target value is equal to the error value; actual is the actual value of the cloth seal; M target is the sealing target value; t M is the error range of the cloth seal; G actual G is the actual value of gram weight; target is the target weight value; t G is the error range of gram weight; the calculated data quality score of each characteristic data is combined to form a data quality score set.

[0197] In one embodiment, when the processor executes the computer program to implement the step of calculating the weight of the corresponding base model according to the mean square error of each base model, the processor specifically implements the following steps:

[0198] use Calculate the weight of the corresponding base model, where W i is the weight of the i-th base model; MSE i is the mean square error of the i-base models, m represents the number of base models; n is the amount of training data; h is the number of process parameters; y is the number of training data; ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the feature data in the i-th row; s i is the data quality score of the i-th feature data.

[0199] In one embodiment, when the processor executes the computer program to implement the step of determining the final prediction result according to the matrix combined with the linear regression model, the processor specifically implements the following steps:

[0200] Determine the coefficient matrix of the linear regression model; split the coefficient matrix by rows and transpose it to obtain a new coefficient matrix; perform a dot multiplication operation on the matrix and each of the new coefficient matrices to obtain a process parameter prediction value for each feature data; combine all the process parameter prediction values ​​to form a final prediction result.

[0201] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0202] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0203] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0204] The steps in the method of the embodiment of the present invention can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0205] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0206] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for recommending heat setting process parameters, characterized in that: include: Obtain production order data to be predicted; Preprocessing the production order data to be predicted to obtain a preprocessing result; The preprocessing result is input into the prediction model to predict the process parameters to obtain the prediction result; wherein the prediction model is obtained by preprocessing the production order data and labeling it as a sample set, and training a multi-weight main model formed by stacking several weighted base models and meta models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data; The prediction result is output.

2. The method for recommending heat setting process parameters according to claim 1, characterized in that: The prediction model is obtained by preprocessing and labeling the production order data as a sample set, and training a multi-weighted main model formed by stacking several weighted base models and meta models. The weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data, including: Acquire production order data, and preprocess and label the production order data to obtain a sample set, wherein the sample set includes feature data and label data; Construct a multi-weighted master model; Calculating a data quality score of the feature data; Input the characteristic data into each base model to predict process parameters, and calculate the mean square error of each base model according to the prediction result and the corresponding label data; Calculate the weight of the corresponding base model according to the mean square error of each base model; Multiplying the weight of the base model by the result predicted by the base model to obtain a weighted prediction result; Extracting corresponding process parameters from the weighted prediction results to form a matrix; Determine the final prediction result according to the matrix combined with the linear regression model; Using mean error as a loss function, and adding a data quality score of the feature data into the loss function; The prediction model is obtained by minimizing the loss function, adjusting the parameters of the meta-model, and repeating the training until the meta-model converges.

3. The method for recommending heat setting process parameters according to claim 2, characterized in that: The construction of the multi-weight main model comprises: Several models whose prediction performance meets the requirements are selected as base models, and a linear regression model is selected as a meta-model. The base models and the meta-model are stacked to form a multi-weight main model.

4. The method for recommending heat setting process parameters according to claim 2, characterized in that: The calculating the data quality score of the feature data comprises: use Calculate the data quality score of the feature data, where s represents the quantity quality score of each feature data; S2 is the target deviation, which represents the degree of deviation between the actual test result and the target value of the parameter, and takes a value between (0, 1]; E M E is the relative error ratio of the cloth seal; G is the relative error ratio of gram weight; M is the target deviation critical value when the difference between the actual test result and the target value is equal to the error value; actual is the actual value of the cloth seal; M target is the sealing target value; t M is the error range of the cloth seal; G actual G is the actual value of gram weight; target is the target weight value; t G is the error range of gram weight; the calculated data quality score of each characteristic data is combined to form a data quality score set.

5. The method for recommending heat setting process parameters according to claim 2, characterized in that: The step of calculating the weight of the corresponding base model according to the mean square error of each base model comprises: use Calculate the weight of the corresponding base model, where W i is the weight of the i-th base model; MSE i is the mean square error of the i-base models, m represents the number of base models; n is the amount of training data; h is the number of process parameters; y is the number of training data; ij is the predicted value of the jth process parameter of the i-th row of data; is the true value of the jth process parameter of the feature data in the i-th row; s i is the data quality score of the i-th feature data.

6. The method for recommending heat setting process parameters according to claim 2, characterized in that: Determining the final prediction result according to the matrix in combination with the linear regression model includes: Determining a coefficient matrix of the linear regression model; The coefficient matrix is ​​split and transposed row by row to obtain a new coefficient matrix; Performing a point multiplication operation on the matrix and each of the new coefficient matrices to obtain a process parameter prediction value for each feature data; All the process parameter prediction values ​​are combined to form a final prediction result.

7. The method for recommending heat setting process parameters according to claim 3, characterized in that: The base models include a GBR model, an RFR model and a KNN model.

8. Heat setting process parameter recommendation device, characterized in that: include: An acquisition unit, used for acquiring production order data to be predicted; A preprocessing unit, used for preprocessing the production order data to be predicted to obtain a preprocessing result; A prediction unit, used for inputting the preprocessing result into a prediction model to predict the process parameters, so as to obtain a prediction result; wherein the prediction model is obtained by preprocessing and labeling the production order data as a sample set, and training a multi-weighted main model formed by stacking a number of weighted base models and meta-models, the weight of the base model is determined by the prediction performance of the base model, and the loss function in the training process adds the quality score of each preprocessed and labeled data; An output unit is used to output the prediction result.

9. A computer device, characterized in that: The computer device comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images