Method and system for predicting productivity of equipment in sock weaving workshop

Abstract

The invention relates to the technical field of footwear products, in particular to a method and a system for predicting productivity of equipment in a sock weaving workshop. The method comprises the following steps: firstly, acquiring historical production environment data, historical productivity data and current production environment data of a sock weaving workshop; thirdly, a historical production environment data set meeting a threshold value judgment condition is obtained through calculation by means of a sliding window and the production environment similarity, and mean value calculation is carried out on historical productivity data corresponding to the historical production environment data set to obtain first predicted productivity data; then, determining a training strategy of a sock productivity prediction model, and inputting the current meteorological data into the sock productivity prediction model to obtain second predicted productivity data; evaluating the first predicted productivity data and the second predicted productivity data to obtain a sock productivity predicted value; and finally, correcting the sock productivity predicted value by using a sock weaving workshop equipment productivity correction formula to obtain a final sock productivity predicted value.

Description

Technical Field

[0001] The present invention relates to the technical field of footwear and sock products, and particularly to a method and system for predicting the production capacity of equipment in a sock weaving workshop. Background Art

[0002] Sock weaving is a production process of weaving fiber materials into sock products. The sock weaving process includes: raw material processing, textile operation, dyeing and printing, post-treatment and other links. At the same time, specific equipment is configured for each link to perform corresponding link operations. Therefore, a large number of equipment need to be introduced into the sock weaving workshop to meet the needs of the entire production line.

[0003] The production capacity prediction of equipment in a sock weaving workshop is essential for sock weaving manufacturers, because the production capacity prediction of equipment can optimize the allocation of production resources, improve production efficiency and reduce overhead costs. At present, the production capacity prediction of equipment in a sock weaving workshop requires manual recording of the output information of each equipment to infer the production capacity situation, which is time-consuming and laborious. At the same time, unexpected situations such as equipment failures and power outages will also cause deviations in the production capacity prediction results of equipment, thus reducing the efficiency and accuracy of the production capacity prediction of equipment in a sock weaving workshop.

[0004] Therefore, a method and system for predicting the production capacity of equipment in a sock weaving workshop are proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for predicting the production capacity of equipment in a sock weaving workshop. First, obtain the historical production environment data, historical production capacity data and current production environment data of the sock weaving workshop; then, use a sliding window and production environment similarity calculation to obtain a set of historical production environment data that meets the threshold judgment conditions, and calculate the mean value of the historical production capacity data corresponding to the set of historical production environment data to obtain the first predicted production capacity data; then, determine the training strategy of the sock production capacity prediction model, input the current meteorological data into the sock production capacity prediction model to obtain the second predicted production capacity data; evaluate the first predicted production capacity data and the second predicted production capacity data to obtain the sock production capacity prediction value; finally, use the sock weaving workshop equipment production capacity correction formula to correct the sock production capacity prediction value to obtain the final sock production capacity prediction value.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for predicting the production capacity of equipment in a sock weaving workshop, comprising:

[0008] Obtain the historical production environment data, historical production capacity data and current production environment data of the sock weaving workshop;

[0009] Construct a current production environment matrix and a historical production environment matrix according to the current production environment data and the historical production environment data respectively;

[0010] Use the window sliding method to extract the historical production environment matrix to obtain a set of historical production environment data, calculate the production environment similarity with the current production environment matrix, and combine the production environment similarity and the historical production capacity data to obtain the first predicted production capacity data;

[0011] Construct a sock production capacity prediction model, specifically train the sock production capacity prediction model according to the first predicted production capacity data, the historical production environment data and the historical production capacity data to obtain the final sock production capacity prediction model; input the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data; perform weighted summation on the first predicted production capacity data and the second predicted production capacity data to obtain the sock production capacity prediction value;

[0012] Use the sock weaving workshop equipment production capacity correction formula to correct the sock production capacity prediction value to obtain the final sock production capacity prediction value.

[0013] Further, the historical production environment data includes: historical production temperature data, historical production humidity data and historical production equipment data; the current production environment data includes: current production temperature data, current production humidity data and current production equipment data.

[0014] Further, the specific implementation process of using the window sliding method to extract the historical production environment matrix to obtain a set of historical production environment data, calculate the production environment similarity with the current production environment matrix, and combine the production environment similarity and the historical production capacity data to obtain the first predicted production capacity data includes:

[0015] Obtain the current production environment matrix and the historical production environment matrix;

[0016] Use a window of size N to slide through the historical production environment matrix, each time extracting a set of sliding data as a set of historical production environment data, and calculate the production environment similarity with the current production environment matrix to obtain the production environment similarity of the set of historical production environment data;

[0017] If there is a set of historical production environment data whose production environment similarity is greater than the similarity threshold, then sort the set of historical production environment data by time, record the proportion of the set that meets the judgment condition, and calculate the average value of the historical production capacity data corresponding to the first N sets of historical production environment data to obtain the first predicted production capacity data; otherwise, the first predicted production capacity data is 0.

[0018] Further, a sock production capacity prediction model is constructed, and the sock production capacity prediction model is specifically trained according to the first predicted production capacity data, the historical production environment data, and the historical production capacity data to obtain a final sock production capacity prediction model; the current production environment data is input into the final sock production capacity prediction model to obtain second predicted production capacity data; the specific implementation process of weighted summation of the first predicted production capacity data and the second predicted production capacity data to obtain a sock production capacity prediction value includes:

[0019] Construct a sock production capacity prediction model;

[0020] Further, historical production environment data, current production environment data, and first predicted production capacity data are obtained;

[0021] Further, window sliding and production environment similarity calculation are performed on the historical production environment data and the current production environment data to obtain a set of historical production environment data that meets the similarity judgment conditions;

[0022] Further, according to the proportion of the first predicted production capacity data and the set of historical production environment data, a model training strategy is determined; wherein, the model training strategy includes: a first training strategy, a second training strategy, and a third training strategy;

[0023] Further, the first training strategy includes: if the first predicted production capacity data is 0, the preprocessed historical production environment data is input into the sock production capacity prediction model for training to obtain a final sock production capacity prediction model;

[0024] Further, the second training strategy includes: if the first predicted production capacity data is not 0 and the proportion of the set of historical production environment data is higher than a preset threshold, the preprocessed set of historical production environment data is input into the sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0025] Further, the third training strategy includes: if the first predicted production capacity data is not 0 and the proportion of the set of historical production environment data is not higher than the preset threshold, the preprocessed historical production environment data is input into the sock production capacity prediction model for training to obtain a pre-trained sock production capacity prediction model; then, the preprocessed set of historical production environment data is input into the pre-trained sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0026] Further, the current production environment data is input into the final sock production capacity prediction model to obtain second predicted production capacity data;

[0027] Further, a comprehensive evaluation is performed on the first predicted production capacity data and the second predicted production capacity data to obtain a predicted sock production capacity value.

[0028] Further, the predicted sock production capacity value is corrected by using the equipment production capacity correction formula for the sock weaving workshop, and the calculation formula for the final predicted sock production capacity value is:

[0029]

[0030] Among them, PCF fin is the final predicted sock production capacity value, PCF is the predicted sock production capacity value, R lh is the aging coefficient of the equipment in the sock weaving workshop, Δqe is the error between the actual defective rate and the defective rate threshold, α is the production temperature coefficient, T is the actual production temperature, T ref is the reference production temperature, β is the production humidity coefficient, H is the actual production humidity, H ref is the reference production humidity.

[0031] A system for predicting the production capacity of equipment in a sock weaving workshop includes: a system control module, a data acquisition module, a data processing module, an equipment production capacity prediction module, a production capacity prediction correction module, and a data output module;

[0032] Among them, the system control module is used to control the start, pause, and stop of the system;

[0033] The data acquisition module is used to obtain the production environment data and equipment production capacity data of the sock weaving workshop;

[0034] The data processing module is used to preprocess the production environment data and the equipment production capacity data;

[0035] The equipment production capacity prediction module is used to predict and evaluate the preprocessed data to obtain a predicted sock production capacity value; among them, the equipment production capacity prediction module includes: a first production capacity prediction unit and a second production capacity prediction unit;

[0036] The production capacity prediction correction module is used to correct the predicted sock production capacity value according to the equipment production capacity correction formula for the sock weaving workshop;

[0037] The data output module is used to output and display the corrected final predicted sock production capacity value.

[0038] Further, the specific implementation process of the first production capacity prediction unit using the window sliding method to extract the historical production environment matrix, obtain the historical production environment data set, calculate the production environment similarity with the current production environment matrix, and combine the production environment similarity and historical production capacity data to obtain the first predicted production capacity data includes:

[0039] Obtain the current production environment matrix and the historical production environment matrix;

[0040] Use a window of size N to slide through the historical production environment matrix, extract a set of sliding data as the historical production environment data set each time, and calculate the production environment similarity with the current production environment matrix to obtain the production environment similarity of the historical production environment data set;

[0041] If there is a historical production environment data set with a production environment similarity greater than the similarity threshold, sort the historical production environment data set by time, record the proportion of the set that meets the judgment conditions, and calculate the average value of the historical production capacity data corresponding to the first N historical production environment data sets to obtain the first predicted production capacity data; otherwise, the first predicted production capacity data is 0.

[0042] Further, the calculation formula of the production environment similarity is:

[0043]

[0044] Among them, PES represents the production environment similarity; pee() represents the production environment error function; R P represents the current production environment matrix; R h represents the historical production environment data set; N represents the window size; M represents the number of types of production environment data; represents the j-th data in the i-th group of data of the current production environment matrix; represents the j-th data in the i-th group of data of the historical production environment data set.

[0045] Further, the second production capacity prediction unit specifically trains the sock production capacity prediction model according to the first predicted production capacity data, historical production environment data, and historical production capacity data to obtain the final sock production capacity prediction model; the specific implementation process of inputting the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data includes:

[0046] Construct a sock production capacity prediction model;

[0047] Further, obtain historical production environment data, current production environment data, and the first predicted production capacity data;

[0048] Further, perform window sliding and production environment similarity calculation on the historical production environment data and the current production environment data to obtain a historical production environment data set that meets the similarity judgment conditions;

[0049] Further, a model training strategy is determined according to the ratio of the first predicted production capacity data and the historical production environment data set; wherein, the model training strategy includes: a first training strategy, a second training strategy, and a third training strategy;

[0050] Further, the first training strategy includes: if the first predicted production capacity data is 0, then the preprocessed historical production environment data is input into the sock production capacity prediction model for training to obtain a final sock production capacity prediction model;

[0051] Further, the second training strategy includes: if the first predicted production capacity data is not 0 and the ratio of the historical production environment data set is higher than a preset threshold, then the preprocessed historical production environment data set is input into the sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0052] Further, the third training strategy includes: if the first predicted production capacity data is not 0 and the ratio of the historical production environment data set is not higher than the preset threshold, then the preprocessed historical production environment data is input into the sock production capacity prediction model for training to obtain a pre-trained sock production capacity prediction model; then, the preprocessed historical production environment data set is input into the pre-trained sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0053] Further, the current production environment data is input into the final sock production capacity prediction model to obtain second predicted production capacity data.

[0054] Further, the production capacity prediction correction module corrects the sock production capacity prediction value by using the sock weaving workshop equipment production capacity correction formula, and the calculation formula for obtaining the final sock production capacity prediction value is:

[0055]

[0056] wherein, PCF fin is the final sock production capacity prediction value, PCF is the sock production capacity prediction value, R lh is the aging coefficient of the sock weaving workshop equipment, Δqe is the error between the actual sock defective rate and the defective rate threshold, α is the production temperature coefficient, T is the actual production temperature, T ref is the reference production temperature, β is the production humidity coefficient, H is the actual production humidity, H ref is the reference production humidity.

[0057] Compared with the prior art, the beneficial effects of the present invention are:

[0058] 1. The present invention proposes a production capacity prediction function based on window sliding and similarity comparison to obtain the first predicted production capacity data. This function first performs window sliding on historical production environment data, and calculates the production environment similarity between each historical production environment data set and the current production environment matrix. Then, it determines whether there is historical production environment data similar to the current production environment data through threshold judgment, and calculates the average value of the corresponding historical production capacity data to obtain the first predicted production capacity data. This function uses the similarity between historical data and current data to initially predict the production capacity of the sock weaving workshop equipment, which can effectively improve the efficiency and accuracy of the production capacity prediction of the sock weaving workshop equipment.

[0059] 2. The present invention proposes a sock production capacity prediction function based on a model training strategy to obtain the second predicted production capacity data. This function inputs the current production environment data into the sock production capacity prediction model for processing, and outputs the second predicted production capacity data. Among them, the training process of the sock production capacity prediction model is determined according to the proportion of the first predicted production capacity data and the historical production environment data set that meets the production environment similarity judgment conditions. This function combines different model training strategies and model predictions to effectively improve the efficiency and accuracy of the production capacity prediction of the sock weaving workshop equipment.

[0060] 3. The present invention proposes a sock weaving workshop equipment production capacity correction function to correct the sock production capacity prediction value according to the actual situation. This function uses the sock weaving workshop equipment production capacity correction formula to modify the sock production capacity prediction value to obtain the final sock production capacity prediction value. Among them, the sock weaving workshop equipment production capacity correction formula is set according to the aging degree of the equipment, the sock defect rate, the production temperature, and the production humidity. This function uses the sock weaving workshop equipment production capacity correction formula to ensure that the final prediction value is more in line with the actual situation, which can effectively improve the accuracy of the production capacity prediction of the sock weaving workshop equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic flow chart of a method for predicting the production capacity of sock weaving workshop equipment of the present invention;

[0062] Figure 2 It is a schematic flow chart for obtaining the first predicted production capacity data of the present invention;

[0063] Figure 3 It is a schematic structural diagram of the sock production capacity prediction model of the present invention;

[0064] Figure 4 It is a schematic structural diagram of a system for predicting the production capacity of sock weaving workshop equipment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] Sock weaving is a production process that weaves fiber materials into sock products. The sock weaving process includes: raw material processing, textile operations, dyeing and printing, post-processing and other links. At the same time, each link is equipped with specific equipment to perform the corresponding link operations. Therefore, the sock weaving workshop needs to introduce a large number of equipment to meet the needs of the entire production line.

[0067] Capacity forecasting of equipment in hosiery weaving workshops is essential for hosiery weaving manufacturers because it can optimize production resource allocation, improve production efficiency, and reduce overhead costs. At present, the capacity forecasting of equipment in hosiery weaving workshops requires manual recording of the output information of each device to infer the capacity situation, which is time-consuming and labor-intensive. At the same time, unexpected situations such as equipment failure and power outages will also cause significant deviations in the equipment capacity forecast results, thereby reducing the efficiency and accuracy of the equipment capacity forecasting in the hosiery weaving workshop.

[0068] Embodiment 1

[0069] In the embodiment of the present application, the specific implementation process will be implemented by a method for predicting the production capacity of equipment in a hosiery weaving workshop of the present invention, see Figure 1 The process of the method proposed in the present invention is described in the following; the method for predicting the equipment capacity of a hosiery weaving workshop comprises:

[0070] S10. Obtain the historical production environment data, historical production capacity data and current production environment data of the hosiery weaving workshop;

[0071] S20. Construct a current production environment matrix and a historical production environment matrix, perform window sliding processing on the historical production environment matrix, and calculate the production environment similarity with the current production environment matrix to obtain the first predicted production capacity data;

[0072] S30. Targetedly train the socks production capacity prediction model according to the first predicted production capacity data, the historical production environment data and the historical production capacity data, and input the current production environment data into the trained socks production capacity prediction model to obtain second predicted production capacity data;

[0073] S40. Comprehensively evaluate the first predicted capacity data and the second predicted capacity data to obtain a predicted value of hosiery capacity;

[0074] S50. Use the equipment production capacity correction formula of the sock weaving workshop to correct the predicted sock production capacity value to obtain the final predicted sock production capacity value.

[0075] Furthermore, the specific implementation process of a method for predicting the equipment production capacity of a sock weaving workshop is as follows:

[0076] Obtain the historical production environment data, historical production capacity data, and current production environment data of the sock weaving workshop;

[0077] Construct a current production environment matrix and a historical production environment matrix based on the current production environment data and the historical production environment data respectively;

[0078] Use the window sliding method to extract the historical production environment matrix to obtain a historical production environment data set, calculate the production environment similarity with the current production environment matrix, and combine the production environment similarity and the historical production capacity data to obtain the first predicted production capacity data;

[0079] Construct a sock production capacity prediction model, conduct targeted training on the sock production capacity prediction model according to the first predicted production capacity data, the historical production environment data, and the historical production capacity data to obtain the final sock production capacity prediction model; input the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data; perform weighted summation on the first predicted production capacity data and the second predicted production capacity data to obtain the sock production capacity prediction value;

[0080] Use the equipment production capacity correction formula of the sock weaving workshop to correct the predicted sock production capacity value to obtain the final predicted sock production capacity value.

[0081] In this embodiment, a method for predicting the equipment production capacity of a sock weaving workshop is proposed; first, obtain the historical production environment data, historical production capacity data, and current production environment data of the sock weaving workshop; then, use the sliding window and production environment similarity calculation to obtain a historical production environment data set that meets the threshold judgment conditions, calculate the average value of the historical production capacity data corresponding to the historical production environment data set to obtain the first predicted production capacity data; then, determine the training strategy of the sock production capacity prediction model, input the current meteorological data into the sock production capacity prediction model to obtain the second predicted production capacity data; evaluate the first predicted production capacity data and the second predicted production capacity data to obtain the sock production capacity prediction value; finally, use the equipment production capacity correction formula of the sock weaving workshop to correct the predicted sock production capacity value to obtain the final predicted sock production capacity value; this method can effectively improve the efficiency and accuracy of predicting the equipment production capacity of the sock weaving workshop.

[0082] For specific illustration, it is elaborated in combination with the following embodiments as follows:

[0083] Obtain the historical production environment data, historical production capacity data, and current production environment data of the sock weaving workshop;

[0084] In this embodiment, the historical production environment data and historical production capacity data are extracted from the historical database; among them, the historical production capacity data is calculated based on the historical equipment quantity, historical equipment production efficiency, and historical equipment production time; the current production environment data is a set of production record data that is continuous in time.

[0085] Further, the historical production environment data includes: historical production temperature data, historical production humidity data, and historical production equipment data; the current production environment data includes: current production temperature data, current production humidity data, and current production equipment data.

[0086] In this embodiment, the production temperature data and production humidity data are collected through environmental sensors; the production equipment data is obtained by calling the equipment control terminal; the production equipment data includes equipment production data and equipment operation status data; the historical production environment data and current production environment data can provide a solid data foundation for subsequent model prediction and prediction value correction, thereby effectively improving the efficiency and accuracy of equipment production capacity prediction in the sock weaving workshop.

[0087] Further, use the window sliding method to extract the historical production environment matrix to obtain a historical production environment data set, and calculate the production environment similarity with the current production environment matrix. Combining the production environment similarity and the historical production capacity data, the acquisition process schematic of the first predicted production capacity data can be referred to Figure 2 , including:

[0088] S110. Obtain the current production environment data and historical production environment data, and construct the current production environment matrix and historical production environment matrix;

[0089] S120. In the historical production environment matrix, slide a window of size N, and each time extract a set of sliding data as the historical production environment data set. The window slides one time step each time until the historical production environment matrix is completely traversed to obtain a group of historical production environment data sets;

[0090] S130. Calculate the production environment similarity between each historical production environment data set in the historical production environment data set group and the current production environment matrix to obtain the production environment similarity of each historical production environment data set;

[0091] S140. Judge the threshold of the production environment similarity for each of the historical production environment data sets. If there is a situation where the production environment similarity is greater than the similarity threshold, sort the data sets that meet the judgment conditions by time, and calculate the average value of the historical production capacity data corresponding to the first N data sets to obtain the first predicted production capacity data; otherwise, set the first predicted production capacity data to 0.

[0092] In this embodiment, the time step for each window sliding can be 6 hours, 12 hours, 24 hours, 36 hours, etc. The time step is flexibly set according to the quantity of historical production environment data and is not unique.

[0093] Among them, the calculation formula for the production environment similarity is:

[0094]

[0095] Among them, PES represents the production environment similarity; pee() represents the production environment error function; R P represents the current production environment matrix; R h represents the historical production environment data set; N represents the window size; M represents the number of types of production environment data; represents the j-th data in the i-th group of data of the current production environment matrix; represents the j-th data in the i-th group of data of the historical production environment data set.

[0096] In this embodiment, the number N of historical production environment data sets selected is set to 5; the number M of types of production environment data is set to 4; of course, these values can be adjusted according to the actual situation.

[0097] To facilitate the description of the production capacity prediction function based on window sliding and similarity comparison proposed by the present invention, this embodiment selects three links of textile, dyeing, and drying for testing, extracts 500 groups of historical production environment data respectively, and simultaneously obtains the normal current production environment data of each link. The window size and the number of selected data sets are both set to 10, and the similarity threshold is 0.9; then, combining the process methods of S110 - S140, the first predicted production capacity data of each link is obtained, and the test results of the first predicted production capacity data are shown in Table 1:

[0098] Table 1. Test results of the first predicted production capacity data

[0099]

[0100] In this embodiment, a production capacity prediction function based on window sliding and similarity comparison is proposed to obtain the first predicted production capacity data. This function first performs window sliding on the historical production environment data, and calculates the production environment similarity between each historical production environment data set and the current production environment matrix. Then, it determines whether there is historical production environment data similar to the current production environment data through a threshold, and calculates the average value of the corresponding historical production capacity data to obtain the first predicted production capacity data. This function uses the similarity between historical data and current data to initially predict the production capacity of the sock weaving workshop equipment, which can effectively improve the efficiency and accuracy of the production capacity prediction of the sock weaving workshop equipment.

[0101] Further, a sock production capacity prediction model is constructed, and the sock production capacity prediction model is specifically trained according to the first predicted production capacity data, the historical production environment data, and the historical production capacity data to obtain the final sock production capacity prediction model. Inputting the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data. The specific implementation process of weighted summation of the first predicted production capacity data and the second predicted production capacity data to obtain the sock production capacity prediction value includes:

[0102] Construct a sock production capacity prediction model;

[0103] Further, obtain the historical production environment data, the current production environment data, and the first predicted production capacity data;

[0104] Further, perform window sliding and production environment similarity calculation on the historical production environment data and the current production environment data to obtain a set of historical production environment data that meets the similarity judgment conditions;

[0105] Further, determine the model training strategy according to the proportion of the first predicted production capacity data and the historical production environment data set. The model training strategy includes: the first training strategy, the second training strategy, and the third training strategy;

[0106] Further, the first training strategy includes: if the first predicted production capacity data is 0, input the preprocessed historical production environment data into the sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0107] Further, the second training strategy includes: if the first predicted production capacity data is not 0 and the proportion of the historical production environment data set is higher than the preset threshold, input the preprocessed historical production environment data set into the sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0108] Further, the third training strategy includes: if the first predicted production capacity data is not 0 and the proportion of the historical production environment data set is not higher than the preset threshold, input the preprocessed historical production environment data into the sock production capacity prediction model for training to obtain a pre-trained sock production capacity prediction model; then, input the preprocessed historical production environment data set into the pre-trained sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0109] In this embodiment, the preset threshold is set to 0.65, and the preset threshold will vary with the amount of historical production environment data and can be flexibly adjusted according to the actual data volume.

[0110] In this embodiment, the preprocessing process includes data cleaning, deduplication, and normalization; data cleaning and data deduplication are used to exclude missing data, error data, and duplicate data; data normalization is used to improve the training speed of the prediction model; the preprocessed data will be divided into a training set and a test set in a ratio of 8:2.

[0111] Further, input the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data;

[0112] In this embodiment, the structure of the sock production capacity prediction model can refer to Figure 3 , including: an input layer, a feature extraction layer, 3 GRU layers, 3 LSTM layers, a feature fusion layer, and an output layer; the input layer is used to convert the current production environment data from the data space to the feature space for subsequent feature processing; the feature extraction layer is used to extract deep features to enhance the feature expression ability; the GRU layer and the LSTM layer are respectively in different branches, each used to obtain local dependence features and global dependence features; the feature fusion layer is used to fuse the local dependence features and the global dependence features; the output layer is used to convert the features into the second predicted production capacity data and output it.

[0113] In this embodiment, a sock production capacity prediction function based on a model training strategy is proposed to obtain the second predicted production capacity data; this function inputs the current production environment data into the sock production capacity prediction model for processing and outputs the second predicted production capacity data; among them, the training process of the sock production capacity prediction model is determined according to the proportion of the first predicted production capacity data and the historical production environment data set that meets the production environment similarity judgment condition; this function can effectively improve the efficiency and accuracy of equipment production capacity prediction in the sock weaving workshop by combining different model training strategies and model predictions.

[0114] Further, comprehensively evaluate the first predicted production capacity data and the second predicted production capacity data to obtain a predicted sock production capacity value; wherein, the calculation formula for the predicted sock production capacity value is:

[0115] P o = η * P f + (1 - η) * P s ;

[0116]

[0117] wherein, P o represents the predicted sock production capacity value; η represents the adaptive weight factor; ω represents the first predicted weight factor, and this value is set to 0.5; P f represents the first predicted production capacity data; P s represents the second predicted production capacity data.

[0118] Further, use the equipment production capacity correction formula for the sock weaving workshop to correct the predicted sock production capacity value, and the calculation formula for the final predicted sock production capacity value is:

[0119]

[0120] α + β = 1.0;

[0121] wherein, PCF fin is the final predicted sock production capacity value, PCF is the predicted sock production capacity value, R lh is the aging coefficient of the equipment in the sock weaving workshop, Δqe is the error between the actual defective rate and the defective rate threshold, α is the production temperature coefficient, T is the actual production temperature, T ref is the reference production temperature, β is the production humidity coefficient, H is the actual production humidity, H ref is the reference production humidity.

[0122] In this embodiment, the aging coefficient is expressed as the ratio of the number of equipment failures and anomalies during the production cycle to the total number of equipment productions; wherein, the production cycle can be a quarter, half a year, one year, etc.; the actual defective rate is expressed as the ratio of the number of defective socks in actual production to the total actual production volume.

[0123] In this embodiment, both the production temperature coefficient α and the production humidity coefficient β are set to 0.5; the selection of the coefficients can be flexibly adjusted according to the actual situation and is not unique.

[0124] To facilitate the description of the function for correcting the equipment production capacity in the sock weaving workshop proposed by the present invention, in this embodiment, three processes of textile, dyeing, and drying are selected for correction testing, and 1000 sets of historical production environment data are extracted respectively. At the same time, the normal current production environment data of each process are obtained. The window size and the number of selected data sets are both 20, and the similarity threshold is 0.9. Then, by combining the process methods of S10 - S40, the sock production capacity prediction values of each process are obtained. Next, the reference production temperatures and reference production humidities of the three processes are set to (24°C, 50%), (70°C, 60%), and (80°C, 40%) respectively. At the same time, the aging coefficients of the equipment in each process are calculated to be 0.05, 0.10, and 0.11 respectively, and the defective rate errors are 0.05, 0.09, and 0.07 respectively. According to the sock weaving workshop equipment production capacity correction formula, the final sock production capacity prediction value of each process is obtained. The test results of the final sock production capacity prediction value are shown in Table 2:

[0125] Table 2. Test Results of the Final Sock Production Capacity Prediction Value

[0126]

[0127] In this embodiment, a function for correcting the equipment production capacity in the sock weaving workshop is proposed to correct the sock production capacity prediction value according to the actual situation. This function uses the sock weaving workshop equipment production capacity correction formula to modify the sock production capacity prediction value to obtain the final sock production capacity prediction value. Among them, the sock weaving workshop equipment production capacity correction formula is set according to the aging degree of the equipment, the defective rate of socks, the production temperature, and the production humidity. This function uses the sock weaving workshop equipment production capacity correction formula to ensure that the final prediction value is more in line with the actual situation, which can effectively improve the accuracy of sock production capacity prediction in the sock weaving workshop.

[0128] The present invention collects the current production environment data, the historical production environment data and historical production capacity data of the past year in Workshop A for sock weaving. Among them, the current production environment data are multiple sets of data collected under good production environment conditions. By performing data cleaning, duplicate removal, and normalization on the historical production environment data and the historical production capacity data, preprocessed data are obtained. The preprocessed data are divided into a training set and a test set according to a ratio of 8:2.

[0129] To verify the actual effect of a method for predicting the equipment production capacity in a hosiery weaving workshop proposed by the present invention, multiple groups of comparative experiments were designed. Among them, Method 1 applied a method for predicting the equipment production capacity in a hosiery weaving workshop proposed by the present invention, which included a first production capacity prediction based on a sliding window and production environment similarity and a second production capacity prediction based on a hosiery production capacity prediction model; Method 2 only used the second production capacity prediction method based on the hosiery production capacity prediction model for production capacity prediction, omitting the first production capacity prediction method; Method 3 used a traditional empirical prediction method to predict the production capacity.

[0130] The first two methods first used the training set to train their respective prediction models, and then input the current production environment data into their respective trained prediction models to obtain the predicted production capacity; while Method 3 did not require training the model and could obtain the predicted production capacity based on historical production environment data or current production environment data, and calculated the ratio of the predicted production capacity within a reasonable range respectively.

[0131] Table 3. Comparison of the effects of different prediction schemes

[0132]

[0133] As shown in Table 3, the comprehensive method (Method 1) combining the first production capacity prediction based on a sliding window and production environment similarity and the second production capacity prediction based on a hosiery production capacity prediction model performed the best, indicating that the method proposed by the present invention is the most effective in predicting the equipment production capacity in a hosiery weaving workshop.

[0134] Embodiment 2

[0135] As an implementation manner of the present invention, referring to Figure 4 , a system for predicting the equipment production capacity in a hosiery weaving workshop includes: a system control module, a data acquisition module, a data processing module, an equipment production capacity prediction module, a production capacity prediction correction module, and a data output module;

[0136] Among them, the system control module is used to control the start, pause, and stop of the system;

[0137] The data acquisition module is used to obtain the production environment data and equipment production capacity data of the hosiery weaving workshop;

[0138] The data processing module is used to preprocess the production environment data and the equipment production capacity data;

[0139] The equipment production capacity prediction module is used to predict and evaluate the preprocessed data to obtain the hosiery production capacity prediction value; among them, the equipment production capacity prediction module includes: a first production capacity prediction unit and a second production capacity prediction unit;

[0140] Further, the specific implementation process of the first production capacity prediction unit using the window sliding method to extract the historical production environment matrix, obtain the historical production environment data set, calculate the production environment similarity with the current production environment matrix, and combine the production environment similarity and historical production capacity data to obtain the first predicted production capacity data includes:

[0141] Obtain the current production environment matrix and the historical production environment matrix;

[0142] Use a window of size N to slide through the historical production environment matrix, extract a set of sliding data as the historical production environment data set each time, and calculate the production environment similarity with the current production environment matrix to obtain the production environment similarity of the historical production environment data set;

[0143] If there is a historical production environment data set with the production environment similarity greater than the similarity threshold, sort the historical production environment data set by time, record the proportion of the set that meets the judgment condition, and calculate the average value of the historical production capacity data corresponding to the first N historical production environment data sets to obtain the first predicted production capacity data; otherwise, the first predicted production capacity data is 0.

[0144] Further, the calculation formula of the production environment similarity is:

[0145]

[0146] where PES represents the production environment similarity; pee() represents the production environment error function; R P represents the current production environment matrix; R h represents the historical production environment data set; N represents the window size; M represents the number of types of production environment data; represents the j-th data in the i-th group of data of the current production environment matrix; represents the j-th data in the i-th group of data of the historical production environment data set.

[0147] Further, the second production capacity prediction unit specifically trains the sock production capacity prediction model according to the first predicted production capacity data, historical production environment data, and historical production capacity data to obtain the final sock production capacity prediction model; the specific implementation process of inputting the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data includes:

[0148] Construct a sock production capacity prediction model;

[0149] Further, obtain historical production environment data, current production environment data, and the first predicted production capacity data;

[0150] Further, window sliding and production environment similarity calculation are performed on the historical production environment data and the current production environment data to obtain a set of historical production environment data that meets the similarity judgment conditions;

[0151] Further, according to the proportion of the first predicted production capacity data and the set of historical production environment data, a model training strategy is determined; wherein, the model training strategy includes: a first training strategy, a second training strategy, and a third training strategy;

[0152] Further, the first training strategy includes: if the first predicted production capacity data is 0, then the preprocessed historical production environment data is input into the sock production capacity prediction model for training to obtain a final sock production capacity prediction model;

[0153] Further, the second training strategy includes: if the first predicted production capacity data is not 0 and the proportion of the set of historical production environment data is higher than a preset threshold, then the preprocessed set of historical production environment data is input into the sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0154] Further, the third training strategy includes: if the first predicted production capacity data is not 0 and the proportion of the set of historical production environment data is not higher than the preset threshold, then the preprocessed historical production environment data is input into the sock production capacity prediction model for training to obtain a pre-trained sock production capacity prediction model; then, the preprocessed set of historical production environment data is input into the pre-trained sock production capacity prediction model for training to obtain the final sock production capacity prediction model;

[0155] To further verify the accuracy of the present invention in predicting production capacity under different data sets, multiple sets of comparative experiments are designed; the present invention collects the current production environment data and the historical production environment data and historical production capacity data of the B sock weaving workshop in the past two years; wherein, the current production environment data is collected under good production environment conditions; by screening the historical production environment data to meet the condition requirements of three model training strategies, three historical production environment data sets with the same quantity are obtained, denoted as the first data set, the second data set, and the third data set; then, the sock production capacity prediction model proposed by the present invention and the traditional convolutional network model are respectively trained using the three historical production environment data sets, and then the current production environment data is input into their respective trained models to obtain the second predicted production capacity data; wherein, the present invention adopts different training strategies for different data sets, while the traditional convolutional network model is directly trained; finally, the second predicted production capacity data output by each is compared.

[0156] Table 4. Comparison of prediction effects of different data sets

[0157]

[0158] As shown in Table 4, compared with the traditional convolutional network model, the neural network model proposed by the present invention has the highest accuracy in predicting production capacity under different training sets. Even in the first data set containing the most abnormal production equipment data, it still has relatively high prediction accuracy, indicating that the neural network and training strategy proposed by the present invention can effectively improve the accuracy and reliability of equipment production capacity prediction in the sock weaving workshop.

[0159] Further, input the current production environment data into the final sock production capacity prediction model to obtain the second predicted production capacity data; wherein, the calculation formula for the sock production capacity prediction value is as follows:

[0160] P o = η * P f +(1 - η) * P s ;

[0161]

[0162] wherein, P o represents the sock production capacity prediction value; η represents the adaptive weight factor; ω represents the first prediction weight factor, and this value is set to 0.5; P f represents the first predicted production capacity data; P s represents the second predicted production capacity data.

[0163] The production capacity prediction correction module is used to correct the sock production capacity prediction value according to the equipment production capacity correction formula in the sock weaving workshop;

[0164] Further, the production capacity prediction correction module corrects the sock production capacity prediction value by using the equipment production capacity correction formula in the sock weaving workshop, and the calculation formula for obtaining the final sock production capacity prediction value is as follows:

[0165]

[0166] wherein, PCF fin is the final sock production capacity prediction value, PCF is the sock production capacity prediction value, R lh is the aging coefficient of the equipment in the sock weaving workshop, Δqe is the error between the actual sock defective rate and the defective rate threshold, α is the production temperature coefficient, T is the actual production temperature, T ref is the reference production temperature, β is the production humidity coefficient, H is the actual production humidity, H ref is the reference production humidity.

[0167] The data output module is used to output and display the corrected final sock production capacity prediction value.

[0168] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting equipment capacity in a hosiery weaving workshop, characterized in that: include: Obtain historical production environment data, historical production capacity data, and current production environment data of the hosiery weaving workshop; Constructing a current production environment matrix and a historical production environment matrix according to the current production environment data and the historical production environment data respectively; The historical production environment matrix is ​​extracted by using a window sliding method to obtain a historical production environment data set, and the production environment similarity is calculated with the current production environment matrix, and the first predicted production capacity data is obtained by combining the production environment similarity and the historical production capacity data; Constructing a socks production capacity prediction model, and performing targeted training on the socks production capacity prediction model according to the first predicted production capacity data, the historical production environment data, and the historical production capacity data to obtain a final socks production capacity prediction model; inputting the current production environment data into the final socks production capacity prediction model to obtain second predicted production capacity data; performing weighted summation on the first predicted production capacity data and the second predicted production capacity data to obtain a socks production capacity prediction value; The socks production capacity forecast value is corrected using the socks weaving workshop equipment capacity correction formula to obtain the final socks production capacity forecast value.

2. The method for predicting equipment capacity in a hosiery weaving workshop according to claim 1, characterized in that: The historical production environment data includes: historical production temperature data, historical production humidity data and historical production equipment data; the current production environment data includes: current production temperature data, current production humidity data and current production equipment data.

3. The method for predicting equipment capacity in a hosiery weaving workshop according to claim 1, characterized in that: The specific implementation process of extracting the historical production environment matrix by using the window sliding method to obtain a historical production environment data set, calculating the production environment similarity with the current production environment matrix, and combining the production environment similarity with the historical production capacity data to obtain the first predicted production capacity data includes: Get the current production environment matrix and the historical production environment matrix; The historical production environment matrix is ​​traversed by sliding using a window of size N, and a set of sliding data is extracted each time as a historical production environment data set, and the production environment similarity is calculated with the current production environment matrix to obtain the production environment similarity of the historical production environment data set; If there exists a historical production environment data set whose production environment similarity is greater than the similarity threshold, the historical production environment data set is sorted by time, the proportion of the set that meets the judgment conditions is recorded, and the historical production capacity data corresponding to the first N historical production environment data sets are averaged to obtain the first predicted production capacity data; otherwise, the first predicted production capacity data is 0.

4. The method for predicting equipment capacity in a hosiery weaving workshop according to claim 1, characterized in that: Constructing a socks production capacity prediction model, and performing targeted training on the socks production capacity prediction model according to the first predicted production capacity data, the historical production environment data, and the historical production capacity data to obtain a final socks production capacity prediction model; inputting the current production environment data into the final socks production capacity prediction model to obtain second predicted production capacity data; The specific implementation process of performing weighted summation on the first predicted capacity data and the second predicted capacity data to obtain the predicted value of the socks capacity includes: Construct a socks production capacity forecasting model; Obtain historical production environment data, current production environment data and first predicted production capacity data; Performing window sliding and production environment similarity calculation on the historical production environment data and the current production environment data to obtain a set of historical production environment data that meets similarity judgment conditions; Determine a model training strategy according to the proportion of the first predicted capacity data and the historical production environment data set; wherein the model training strategy includes: a first training strategy, a second training strategy and a third training strategy; The first training strategy includes: if the first predicted capacity data is 0, inputting the pre-processed historical production environment data into the socks capacity prediction model for training to obtain a final socks capacity prediction model; The second training strategy includes: if the first predicted capacity data is not 0 and the proportion of the historical production environment data set is higher than a preset threshold, inputting the pre-processed historical production environment data set into the socks capacity prediction model for training to obtain the final socks capacity prediction model; The third training strategy includes: if the first predicted capacity data is not 0 and the proportion of the historical production environment data set is not higher than the preset threshold, then inputting the pre-processed historical production environment data into the socks capacity prediction model for training to obtain a pre-trained socks capacity prediction model; then, inputting the pre-processed historical production environment data set into the pre-trained socks capacity prediction model for training to obtain the final socks capacity prediction model; Inputting the current production environment data into the final socks production capacity prediction model to obtain second predicted production capacity data; A comprehensive evaluation is performed on the first predicted capacity data and the second predicted capacity data to obtain a predicted value of hosiery capacity.

5. The method for predicting equipment capacity in a hosiery weaving workshop according to claim 1, characterized in that: The socks production capacity forecast value is corrected using the socks weaving workshop equipment capacity correction formula, and the calculation formula for the final socks production capacity forecast value is obtained as follows: Among them, PCF fin is the final socks production capacity forecast value, PCF is the socks production capacity forecast value, R lh is the aging coefficient of the equipment in the hosiery weaving workshop, Δqe is the error between the actual defective rate and the defective rate threshold, α is the production temperature coefficient, T is the actual production temperature, T ref is the reference production temperature, β is the production humidity coefficient, H is the actual production humidity, H ref For reference production humidity.

6. A system for predicting equipment capacity in a hosiery weaving workshop, characterized in that: include: System control module, data acquisition module, data processing module, equipment capacity prediction module, capacity prediction correction module and data output module; Wherein, the system control module is used to control the start, pause and stop of the system; The data acquisition module is used to obtain the production environment data and equipment capacity data of the hosiery weaving workshop; The data processing module is used to pre-process the production environment data and the equipment capacity data; The equipment capacity prediction module is used to predict and evaluate the preprocessed data to obtain a predicted value of the socks capacity; wherein the equipment capacity prediction module includes: a first capacity prediction unit and a second capacity prediction unit; The capacity prediction correction module is used to correct the socks capacity prediction value according to the equipment capacity correction formula of the socks weaving workshop; The data output module is used to output and display the corrected final socks production capacity forecast value.

7. The system for predicting equipment capacity in a hosiery weaving workshop according to claim 6, characterized in that: The first capacity prediction unit extracts the historical production environment matrix using the window sliding method to obtain a historical production environment data set, and calculates the production environment similarity with the current production environment matrix, and combines the production environment similarity and the historical capacity data to obtain the first predicted capacity data. The specific implementation process includes: Get the current production environment matrix and the historical production environment matrix; The historical production environment matrix is ​​traversed by sliding using a window of size N, and a set of sliding data is extracted each time as a historical production environment data set, and the production environment similarity is calculated with the current production environment matrix to obtain the production environment similarity of the historical production environment data set; If there exists a historical production environment data set whose production environment similarity is greater than the similarity threshold, the historical production environment data set is sorted by time, the proportion of the set that meets the judgment conditions is recorded, and the historical production capacity data corresponding to the first N historical production environment data sets are averaged to obtain the first predicted production capacity data; otherwise, the first predicted production capacity data is 0.

8. The system for predicting equipment capacity in a hosiery weaving workshop according to claim 7, characterized in that: The calculation formula of the production environment similarity is: Wherein, PES represents the similarity of the production environment; pee() represents the production environment error function; R P Represented as the current production environment matrix; R h It is represented as a set of historical production environment data; N is the window size; M is the number of types of production environment data; Represented as the jth data in the i-th group of data of the current production environment matrix; Represented as the j-th data in the i-th group of data in the historical production environment data set.

9. The system for predicting equipment capacity in a hosiery weaving workshop according to claim 6, characterized in that: The second capacity prediction unit performs targeted training on the socks capacity prediction model according to the first predicted capacity data, the historical production environment data and the historical capacity data to obtain a final socks capacity prediction model; the specific implementation process of inputting the current production environment data into the final socks capacity prediction model to obtain the second predicted capacity data includes: Construct a socks production capacity forecasting model; Obtain historical production environment data, current production environment data and first predicted production capacity data; Performing window sliding and production environment similarity calculation on the historical production environment data and the current production environment data to obtain a set of historical production environment data that meets similarity judgment conditions; Determine a model training strategy according to the proportion of the first predicted capacity data and the historical production environment data set; wherein the model training strategy includes: a first training strategy, a second training strategy and a third training strategy; The first training strategy includes: if the first predicted capacity data is 0, inputting the pre-processed historical production environment data into the socks capacity prediction model for training to obtain a final socks capacity prediction model; The second training strategy includes: if the first predicted capacity data is not 0 and the proportion of the historical production environment data set is higher than a preset threshold, inputting the pre-processed historical production environment data set into the socks capacity prediction model for training to obtain the final socks capacity prediction model; The third training strategy includes: if the first predicted capacity data is not 0 and the proportion of the historical production environment data set is not higher than the preset threshold, then inputting the pre-processed historical production environment data into the socks capacity prediction model for training to obtain a pre-trained socks capacity prediction model; then, inputting the pre-processed historical production environment data set into the pre-trained socks capacity prediction model for training to obtain the final socks capacity prediction model; The current production environment data is input into the final socks production capacity prediction model to obtain second predicted production capacity data.

10. The system for predicting equipment capacity in a hosiery weaving workshop according to claim 6, characterized in that: The capacity prediction correction module uses the capacity correction formula of the hosiery weaving workshop equipment to correct the hosiery capacity prediction value, and the calculation formula for obtaining the final hosiery capacity prediction value is: Among them, PCF fin is the final socks production capacity forecast value, PCF is the socks production capacity forecast value, R lh is the aging coefficient of the equipment in the hosiery weaving workshop, Δqe is the error between the actual hosiery defective rate and the defective rate threshold, α is the production temperature coefficient, T is the actual production temperature, T ref is the reference production temperature, β is the production humidity coefficient, H is the actual production humidity, H ref For reference production humidity.

Images