Data identification method and device, electronic equipment and computer readable storage medium

Abstract

This application discloses a data identification method, comprising: collecting data from a target information system according to preset data types to obtain sample data corresponding to each preset data type; training an initial neural network model using the sample data to obtain a data identification model; processing the operational data of the target information system using the data identification model to obtain identification results, wherein the operational data includes operational data corresponding to each preset data type; and determining the key influencing factors of the target information system among all preset data types based on the identification results. Applying the technical solution provided in this application allows for more efficient data identification of the target information system, enabling the rapid identification of its key influencing factors. This application also discloses a data identification device, electronic device, and computer-readable storage medium, which also possess the aforementioned technical effects.

Description

Technical Field

[0001] This application relates to the field of information and data processing technology, and in particular to a data identification method, a data identification device, an electronic device, and a computer-readable storage medium. Background Technology

[0002] Currently, in the production practice of automotive component electronics manufacturing, DOE (Design of Experiment, a statistical method for optimizing and improving product process parameters and production processes) is commonly used to analyze various factors in the production process to identify key influencing factors and make improvements. However, with the increasing digitalization in manufacturing, the amount of data in the wave soldering manufacturing process is also increasing, such as preheating temperature, wave soldering temperature, and solder wave height. Tens of thousands of data points are generated every moment. If DOE technology continues to be used, a large number of experiments and data collections will be required, which may lead to a waste of time and costs, and consequently, inefficiency in the identification process.

[0003] Therefore, how to conduct more efficient data identification of target information systems in order to quickly find their key influencing factors is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide a data identification method that can more efficiently identify data in a target information system in order to quickly find its key influencing factors; another purpose of this application is to provide a data identification device, electronic device, and computer-readable storage medium, all of which have the above-mentioned beneficial effects.

[0005] Firstly, this application provides a data identification method, including:

[0006] Data is collected from the target information system according to the preset data types to obtain sample data corresponding to each preset data type.

[0007] The initial neural network model is trained using the sample data to obtain a data recognition model;

[0008] The data recognition model is used to process the operational data of the target information system to obtain recognition results. The operational data includes operational data corresponding to each of the preset data types.

[0009] Based on the identification results, the key influencing factors of the target information system are determined among all the preset data types.

[0010] Optionally, training the initial neural network model using the sample data to obtain the data recognition model includes:

[0011] The sample data is divided according to a preset ratio to obtain training samples, validation samples, and test samples.

[0012] The initial neural network model is iteratively trained using the training samples and the validation samples to obtain an initial data recognition model;

[0013] The initial data recognition model is evaluated using the test samples to obtain evaluation results;

[0014] When the evaluation result is that the evaluation fails, return to the step of iteratively training the initial neural network model using the training samples and the validation samples to obtain the initial data recognition model;

[0015] When the evaluation result is "evaluation passed", the initial data recognition model is determined to be the data recognition model.

[0016] Optionally, before training the initial neural network model using the sample data to obtain the data recognition model, the method further includes:

[0017] The sample data is preprocessed; the preprocessing includes one or more of the following: data cleaning, outlier handling, data standardization, and missing value imputation.

[0018] Optionally, determining the key influencing factors of the target information system based on the identification results among all the preset data types includes:

[0019] The influence weight of each preset data type is determined based on the identification results;

[0020] The preset data type corresponding to the largest influence weight is taken as the key influence factor of the target information system.

[0021] Optionally, the data identification method further includes:

[0022] The operational data corresponding to the key influencing factors are filtered to obtain abnormal operational data;

[0023] Calculate the failure rate of the target information system based on the amount of abnormal operation data;

[0024] Output the key influencing factors, the abnormal operation data, and the defect rate.

[0025] Optionally, the data identification method further includes:

[0026] The target information system is monitored based on the key influencing factors to obtain monitoring data;

[0027] When the monitoring data does not exceed the first threshold range, return to the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data;

[0028] When the monitoring data exceeds the first threshold range but does not exceed the second threshold range, the log information of the target information system is obtained.

[0029] When the log information does not contain any abnormal information, return to the step of monitoring the target information system based on the key influencing factors and obtaining monitoring data;

[0030] When the log information contains the abnormal information, an alarm message is output;

[0031] When the monitoring data exceeds the second threshold range, the monitoring data is sent to the target terminal so that the target terminal user can remotely debug the target information system through the target terminal.

[0032] Optionally, before the step of returning to the step of monitoring the target information system based on the key influencing factors and obtaining monitoring data when the monitoring data does not exceed the first threshold range, the method further includes:

[0033] Perform a normality check on the monitoring data and obtain the check result;

[0034] When the verification result does not meet the preset requirements, return to the step of monitoring the target information system based on the key influencing factors and obtaining monitoring data.

[0035] Secondly, this application also discloses a data identification device, comprising:

[0036] The data acquisition module is used to acquire data from the target information system according to preset data types and obtain sample data corresponding to each preset data type.

[0037] The training module is used to train the initial neural network model using the sample data to obtain a data recognition model;

[0038] The processing module is used to process the operating data of the target information system using the data recognition model to obtain recognition results. The operating data includes operating data corresponding to each of the preset data types.

[0039] The determination module is used to determine the key influencing factors of the target information system among all the preset data types based on the identification results.

[0040] Thirdly, this application also discloses an electronic device, comprising:

[0041] Memory, used to store computer programs;

[0042] A processor for executing the computer program to implement any of the data recognition methods described above.

[0043] Fourthly, this application also discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the data recognition methods described above.

[0044] This application provides a data identification method, comprising: collecting data from a target information system according to preset data types to obtain sample data corresponding to each preset data type; training an initial neural network model using each sample data to obtain a data identification model; processing the operational data of the target information system using the data identification model to obtain an identification result, wherein the operational data includes operational data corresponding to each preset data type; and determining the key influencing factors of the target information system among all preset data types based on the identification result.

[0045] Applying the technical solution provided in this application, firstly, for the target information system, data is collected according to preset data types to obtain corresponding sample data. Each preset data type represents a potential influencing factor within the target information system. Then, a neural network model is trained using the sample data corresponding to each preset data type to obtain a data recognition model. Therefore, for the target information system, the trained data recognition model can be directly used to identify and process the operational data of each preset data type, obtaining recognition results. Based on these results, the key influencing factors of the target information system are determined among all preset data types. Clearly, compared to DOE technology, this technical solution eliminates the need for extensive experiments, effectively simplifying the operational process and enabling more efficient data recognition of the target information system to quickly identify its key influencing factors, thus possessing better practicality and application prospects.

[0046] The data identification device, electronic device, and computer-readable storage medium provided in this application also have the above-mentioned technical effects, and will not be described in detail here. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the prior art and the embodiments of this application, the accompanying drawings used in the description of the prior art and the embodiments of this application will be briefly introduced below. Of course, the accompanying drawings described below with respect to the embodiments of this application are only a part of the embodiments in this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort, and such other drawings also fall within the protection scope of this application.

[0048] Figure 1 A flowchart illustrating a data identification method provided in this application;

[0049] Figure 2 A flowchart illustrating another data identification method provided in this application;

[0050] Figure 3 A schematic diagram of the structure of a data identification device provided in this application;

[0051] Figure 4 This is a schematic diagram of the structure of an electronic device provided in this application. Detailed Implementation

[0052] The core of this application is to provide a data identification method that can more efficiently identify data in a target information system to quickly find its key influencing factors; another core aspect of this application is to provide a data identification device, electronic device, and computer-readable storage medium, all of which have the aforementioned beneficial effects.

[0053] To provide a clearer and more complete description of the technical solutions in the embodiments of this application, the technical solutions in the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0054] This application provides a data identification method.

[0055] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a data identification method provided in this application, which may include the following steps S101 to S104.

[0056] S101: Collect data from the target information system according to the preset data types to obtain sample data corresponding to each preset data type.

[0057] First, it should be noted that the data identification method provided in this application aims to achieve data identification, specifically, to identify data in a target information system in order to determine the key influencing factors of the target information system. These key influencing factors refer to data types that have a significant impact on the actual operation of the target information system. In one implementation, the aforementioned target information system can specifically be a MES system (a common management information system in the electronics manufacturing industry, used to monitor and optimize real-time production data during the manufacturing process).

[0058] Furthermore, this step aims to achieve data acquisition to obtain sample data from the target information system, which will be used for subsequent model training. During sample data acquisition, data can be collected from the target information system according to preset data types to obtain sample data corresponding to each preset data type. The preset data types are the actual data types existing in the target information system and represent potential influencing factors. The data identification method provided in this application identifies the data types that have a key impact on the actual operation of the target information system—the key influencing factors—among all preset data types. It is understood that the number and types of preset data types can be set by technical personnel according to actual conditions, and this application does not limit this. For example, for an MES system, preset data types can include preheating zone temperature, wave soldering zone temperature, and solder wave height.

[0059] S102: Train the initial neural network model using the sample data to obtain the data recognition model.

[0060] This step aims to train the model to obtain a data recognition model for data identification. This data recognition model is a network model based on a pre-defined neural network. Specifically, an initial neural network model can be built first, including each neural network layer (such as input layer, hidden layer, and output layer), the number of neurons in each neural network layer, the neuron activation function, the loss function, and other initial model parameters. Then, the initial neural network model is trained using sample data corresponding to each pre-defined data type to obtain a neural network model that meets the requirements (such as the model loss reaching a pre-defined range), i.e., the aforementioned data recognition model.

[0061] S103: Use a data recognition model to process the operational data of the target information system to obtain recognition results. The operational data includes operational data corresponding to each preset data type.

[0062] This step aims to achieve data processing based on a data recognition model to obtain corresponding recognition results. It should be noted that the model training used sample data of various preset data types. Similarly, when performing data recognition processing on the target information system, it is necessary to collect operational data of these preset data types, then input this operational data into the data recognition model for processing. The output of the data recognition model is the final recognition result. In essence, the recognition result of the data recognition model represents the degree of influence of the aforementioned preset data types on the target information system, and is used to determine the key influencing factors of the target information system among all preset data types based on this recognition result.

[0063] S104: Based on the identification results, determine the key influencing factors of the target information system among all preset data types.

[0064] This step aims to determine the key influencing factors, which are the data types among all preset data types that have the most critical impact on the actual operation of the target information system. As mentioned above, the identification result of the data identification model is the degree of influence of each preset data type on the target information system. Therefore, based on this identification result, the data type with the most critical impact (the greatest degree of influence) on the target information system can be determined, which is to say, the key influencing factors of the target information system can be determined.

[0065] As can be seen, the data identification method provided in this application firstly collects data from the target information system according to preset data types to obtain corresponding sample data. Each preset data type represents a potential influencing factor in the target information system. Then, a neural network model is trained using the sample data corresponding to each preset data type to obtain a data identification model. Therefore, for the target information system, the trained data identification model can be directly used to identify and process the operational data of each preset data type to obtain identification results. Based on these identification results, the key influencing factors of the target information system are determined among all preset data types. Clearly, compared to DOE technology, this technical solution does not require extensive experimentation, effectively simplifying its operation process. It can perform more efficient data identification on the target information system to quickly find its key influencing factors, demonstrating better practicality and application prospects.

[0066] Based on the above embodiments:

[0067] In one embodiment of this application, the above-mentioned training of the initial neural network model using sample data to obtain a data recognition model may include the following steps:

[0068] The sample data is divided according to a preset ratio to obtain training samples, validation samples, and test samples.

[0069] The initial neural network model is iteratively trained using training and validation samples to obtain the initial data recognition model.

[0070] The initial data recognition model was evaluated using test samples to obtain evaluation results;

[0071] When the evaluation result is that the evaluation fails, return to the step of iteratively training the initial neural network model using training samples and validation samples to obtain the initial data recognition model;

[0072] When the evaluation result is "passed," the initial data recognition model is determined as the data recognition model. Specifically, the sample data can be divided according to a preset ratio to obtain training samples, validation samples, and test samples. The training samples are used for model training, the validation samples are used for cross-validation of the model in each iteration during model training, and the test samples are used to evaluate the accuracy of the final trained model. Therefore, the initial neural network model can be iteratively trained using the training and validation samples to obtain the initial data recognition model. Then, the initial data recognition model can be evaluated using the test samples to obtain the evaluation result. This evaluation result indicates whether the initial data recognition model obtained in this training meets the preset requirements (such as whether the model accuracy reaches the preset range). If the preset requirements are met, the evaluation is considered passed, and the currently trained initial data recognition model can be directly used as the final data recognition model to complete the model training. Conversely, if the preset requirements are not met, the evaluation is considered failed, and the iterative training steps can be returned to retrain the model until a data recognition model that passes the evaluation is obtained.

[0073] This application provides a method for training a data recognition model.

[0074] In one embodiment of this application, before training the initial neural network model using the sample data to obtain the data recognition model, the process may further include: preprocessing the sample data; the preprocessing includes one or more of the following combinations: data cleaning, outlier handling, data standardization, and missing value imputation.

[0075] The data recognition method provided in this application embodiment can also implement data preprocessing functions to effectively ensure the integrity and accuracy of sample data, thereby ensuring the high precision of the data recognition model and improving the accuracy of the recognition results. The preprocessing of sample data can include one or more combinations of data cleaning, outlier handling, data standardization, and missing value imputation, which can be set by technical personnel according to actual needs.

[0076] In one embodiment of this application, the process of determining the key influencing factors of the target information system based on the identification results among all preset data types may include the following steps:

[0077] The influence weight of each preset data type is determined based on the identification results;

[0078] The preset data type corresponding to the largest influence weight is taken as the key influence factor of the target information system.

[0079] This application provides a method for determining key influencing factors among all preset data types. Specifically, the identification result of the data recognition model for the running data can be the influence weight of each preset data type on the target information system. The larger the influence weight, the greater the influence of the corresponding preset data type on the target information system; the smaller the influence weight, the smaller the influence of the corresponding preset data type on the target information system. Therefore, the influence weight of each preset data type can be determined based on the identification result of the data recognition model, and then the preset data type with the largest influence weight can be taken as the key influencing factor of the target information system.

[0080] In one embodiment of this application, the data identification method may further include the following steps:

[0081] Filter the operational data corresponding to key influencing factors to obtain abnormal operational data;

[0082] Calculate the failure rate of the target information system based on the amount of abnormal operation data;

[0083] Output key influencing factors, abnormal operation data, and defect rate.

[0084] The data identification method provided in this application embodiment can further realize the output function of the identification results. Specifically, after determining the key influencing factors of the target information system, the operating data corresponding to the key influencing factors can be filtered to obtain abnormal operating data. The abnormal operating data refers to the operating data that does not meet the corresponding standard threshold range among all operating data of this type. The number of abnormal operating data is counted, and then the failure rate of the target information system is calculated. Finally, when outputting the data, the output content can include three types of data information: key influencing factors, abnormal operating data, and failure rate.

[0085] In one embodiment of this application, the data identification method may further include the following steps:

[0086] Monitor the target information system based on key influencing factors and obtain monitoring data;

[0087] When the monitoring data does not exceed the first threshold range, return to the step of monitoring the target information system based on key influencing factors to obtain monitoring data;

[0088] When the monitored data exceeds the first threshold range but does not exceed the second threshold range, obtain the log information of the target information system.

[0089] If no abnormal information is found in the log information, return to the steps of monitoring the target information system based on key influencing factors and obtaining monitoring data;

[0090] When abnormal information is found in the log, an alarm message is output;

[0091] When the monitoring data exceeds the second threshold range, the monitoring data will be sent to the target terminal so that the target terminal user can remotely debug the target information system through the target terminal.

[0092] The data identification method provided in this application can further achieve data monitoring of key influencing factors. Specifically, since key influencing factors that will have a key impact on the operation of the target information system have been identified, the actual operating data of the key influencing factors can be monitored to obtain monitoring data. This monitoring process can be real-time monitoring or timed monitoring, which is not limited in this application. Then, the monitoring data is subjected to threshold evaluation to determine the actual operating status of the target information system based on its value, and then corresponding operation and maintenance strategies are adopted to ensure the normal operation of the target information system.

[0093] In the implementation process, a first threshold range and a second threshold range can be preset. Different threshold ranges correspond to different operating conditions of the target information system. Specifically, the first threshold range can be a warning range, and the second threshold range can be an alarm range. First, it is determined whether the monitoring data exceeds the first threshold range. If not, it indicates that the target information system is operating normally, and the monitoring process can be returned to continue collecting monitoring data. If yes, it indicates that the target information system may be experiencing an anomaly, so monitoring of the monitoring data can continue. Further, if it is determined that the monitoring data exceeds the first threshold range, it is determined whether the monitoring data exceeds the second threshold range. If not, it indicates that the target information system may be experiencing an anomaly. At this time, the log information of the target information system can be used for further determination. If there is no abnormal log information, it can be determined that the target information system is operating normally, and the monitoring process can be returned to continue. The monitoring data collection process involves collecting data and identifying any abnormal log entries. If these log entries indicate an anomaly in the target information system, an alarm can be triggered to alert technical personnel and facilitate timely maintenance. Finally, if the monitoring data exceeds the second threshold, indicating a relatively serious anomaly, the monitoring data can be sent to the target terminal. This allows technical personnel (users) to remotely control the system, ensuring its continued operation and preventing service interruptions caused by system malfunctions.

[0094] In one embodiment of this application, before the step of returning to monitoring the target information system based on key influencing factors and obtaining monitoring data when the monitoring data does not exceed the first threshold range, the following steps may also be included:

[0095] Perform normality verification on the monitoring data and obtain the verification results;

[0096] If the verification result does not meet the preset requirements, return to the step of monitoring the target information system based on key influencing factors and obtaining monitoring data.

[0097] The data identification method provided in this application embodiment can also perform a normality check on the monitoring data before threshold evaluation. This normality check aims to avoid inaccurate threshold evaluations due to monitoring data collection errors. Specifically, after obtaining the monitoring data, a normality check can be performed to determine whether the monitoring data conforms to a normal distribution. If it does, it indicates that the monitoring data collection process has not encountered any abnormalities, and the accuracy of the monitoring data can be confirmed. In this case, the threshold evaluation process can continue. Conversely, if the monitoring data does not conform to a normal distribution, it indicates that there may be an abnormality in the collection process, and the collected monitoring data is inaccurate. Therefore, the data monitoring process can be returned to re-collect the monitoring data to effectively ensure the accuracy of the collected monitoring data, thus providing accurate assurance for subsequent abnormal operation and maintenance work.

[0098] Based on the above embodiments, this application provides another data identification method using an MES system as an example.

[0099] Please refer to Figure 2 , Figure 2 The following is a flowchart illustrating another data recognition method provided in this application.

[0100] First, for the MES system, the following ten data types to be identified can be set: 6 preheating temperature zones, 2 wave soldering temperature zones, and 2 solder wave heights (all conforming to a normal distribution). Furthermore, sample data from a certain historical period (e.g., 10 months) can be exported from the MES system for neural network analysis to identify the key influencing factors of the MES system.

[0101] Assume the defect rate of wave soldering is f(x), and using a desired function, the goal is to minimize f(x). Simultaneously, set 10 potential factors for the wave soldering production line: preheating zone 1 = x1, preheating zone 2 = x2, preheating zone 3 = x3, preheating zone 4 = x4, preheating zone 5 = x5, preheating zone 6 = x6, wave soldering zone 1 = x7, wave soldering zone 2 = x8, solder wave height 1 = x9, and solder wave height 2 = x10. Correspondingly, the implementation steps may include:

[0102] (1) Write a Python (a high-level programming language) script to import the requests library (a third-party library in Python), and set the URL (Uniform Resource Locator, an address used to locate and access Internet resources) and the required API key for the MES system API (Application Programming Interface).

[0103] (2) Use the request.get function and API key to call the MES API, obtain the request response, and check the request response to confirm whether the sample data has been successfully obtained.

[0104] (3) When it is confirmed that the sample data has been successfully obtained, use the json library to process the API request and response, and convert the obtained sample data into a list of Python data structures.

[0105] (4) Save the list of data structures to a CSV file and then save it to the server.

[0106] (5) Preprocess the sample data in the data structure list, including but not limited to outlier identification, missing value filling, data cleaning, and data standardization.

[0107] (6) Divide the preprocessed sample data into training set, validation set and test set, and save them as corresponding NumPy arrays.

[0108] (7) Import the deep learning library TensorFlow (an open-source machine learning framework) and related modules.

[0109] (8) Design the initial neural network model structure in TensorFlow, and determine the number of neurons in the input layer, hidden layer, and output layer, as well as the neuron activation function TanH (hyperbolic tangent function).

[0110] (9) Set initial model parameters: gradient descent optimizer, initial learning rate, loss function, etc. Among them, the loss function can be the mean squared error.

[0111] (10) Combine the designs in (8) and (9) into a conformal model:

[0112] model.compile(loss=loss_fn,optimizer=optimizer,metrics=["mse"]).

[0113] (11) Define the number of training epochs and the batch size in each training epoch: epochs = 100; batch_size = 32.

[0114] (12) Implement training of the model in (10) on the training set:

[0115] history=model.fit(X_train,y_train,batch_size=batch_size,epochs=epochs,validation_data=(X_val,y_val),verbose=1);

[0116] Where X_train and y_train represent the features and labels of the training set, X_val and y_val represent the features and labels of the validation set, and setting the verbose parameter to 1 indicates that the progress bar and log information during the training process will be output.

[0117] (13) During each training round, output the training loss and validation loss:

[0118] model_history=history.historytrain_loss=model_history["loss"]

[0119] val loss=model_history["val_loss"]

[0120] for epoch in range(epochs):

[0121] print(f"Epoch{epoch+1} / {epochs}")

[0122] print(f"Train loss:{train_loss[epoch]:.4f}")

[0123] print(f"Validation loss:{val_loss[epoch]:.4f}").

[0124] (14) The training process can be visualized by using matplotlib to plot loss curves for visualization.

[0125] (15) When the loss no longer decreases, hyperparameters such as the learning rate can be adjusted; for example, the ReduceLROnPlateau callback function can be used to automatically reduce the learning rate:

[0126] lr_scheduler=keras.callbacks.ReduceLROnPlateau(factor=0.5,patience=10,verbose=1) history=model.fit(X_train,y_train,batch_size=batch_size,epochs=epochs,validation_data=(X_val,y_val),callbacks=[lr_scheduler],verbose=1);

[0127] Here, the factor parameter represents the factor for decreasing the learning rate, the patience parameter indicates how many epochs to observe before decreasing the learning rate if there is no progress, and the verbose parameter is set to 1 to output the log information of the callback function.

[0128] (16) Adjust the structure of the neural network model, such as the number of layers, activation function, and loss function, and evaluate the performance of the model using cross-validation on a validation set.

[0129] (17) Train the model until the validation loss flattens out and record the best model parameters; the EarlyStopping callback function can be used to determine when to stop training when the validation loss no longer improves.

[0130] (18) Use the test set to evaluate the model, calculate the mean square error and R2 value of the model, and determine the final neural network model when both meet the actual requirements.

[0131] (19) Using the neural network model in (18), analyze which factors are most important among the ten types of data mentioned above:

[0132] importance=np.abs(model.layers[0].get_weights()[0]).sum(axis=1);

[0133] importance=importance / importance.sum().

[0134] (20) Extract the influence weights and bias terms of the influencing factors from the model output, and analyze these values ​​to find the patterns:

[0135] weights, biases=model.layer2.weight.squeeze(),model.layer2.bias

[0136] print("Weight:", weights)

[0137] print("Bias term:", biases)

[0138] # Index for finding the most important factors

[0139] key_factor_index=torch.argmax(weights)

[0140] print("Key factor index:", key_factor_index).

[0141] (21) Output results (taking the tin wave height 2 as the key influencing factor as an example):

[0142] #The key factor is the tin wave height 2 (x10).

[0143] print("Key factor: Tin wave height 2 (x10)");

[0144] #Values ​​in the sample data where the tin wave height 2 (x10) is greater than 1500.

[0145] above_1500=data[:,9][data[:,9]>1500];

[0146] print("Values ​​where the height of the tin wave 2 (x10) is greater than 1500:", len(above_1500));

[0147] #Calculate the defect rate when the solder wave height 2 (x10) is greater than 1500.

[0148] above_1500_tensor=torch.Tensor(above_1500.reshape(-1,10))output=model(above_1500_tensor)average_f=torch.mean(output).item()

[0149] print("Defect rate when tin wave height 2 (x10) is greater than 1500:", average_f).

[0150] (22) Based on the output results, automatic judgment is made, and the key influencing factors identified are processed as follows:

[0151] A. Read a preset number (e.g., 50) of tin wave height 2 data points at preset time intervals (e.g., 50 seconds). This process can be implemented using the pyModbusTCP library for ModbusTCP communication.

[0152] client=ModbusTcpClient(host='192.168.1.1',port=502);

[0153] client.connect();

[0154] B. Perform a normality test on the above operating data. When the P-value is greater than 0.05, it is determined that the data conforms to a normal distribution.

[0155] C. When the running data of the tin wave height 2 does not conform to the normal distribution, it indicates that there is an abnormality in the data acquisition process. Return to A and re-run the data acquisition.

[0156] D. When the running data of the tin wave height 2 conforms to a normal distribution, determine whether these running data exceed the first threshold range: upper control line 1300 and lower control line 1100;

[0157] E. If the first threshold range is not exceeded, it indicates that there is no abnormality in the MES system, and the process returns to A to continue data collection.

[0158] F. If the first threshold range is exceeded, continue to determine whether these running data exceed the second threshold range: upper control line 1400 and lower control line 1000;

[0159] G. If the second threshold range is not exceeded, read the log information of the MES system and check whether there are any abnormal logs.

[0160] H. If no abnormal logs are found, it means that the MES system is experiencing normal fluctuations. Return to A to continue collecting running data.

[0161] 1. If abnormal logs are found, an alarm will be output to notify the equipment engineer; of course, if the operating data recovers to the first threshold range, this automatic handling process will stop.

[0162] J. If the error exceeds the second threshold range, the abnormal operation data will be sent to the quality engineer's terminal so that the quality engineer can remotely debug the MES system.

[0163] Based on the above A to J, the processing procedure for the operating data of tin wave height 2 is as follows:

[0164]

[0165]

[0166] This completes the identification and post-processing of key influencing factors in the MES system.

[0167] As can be seen, the data identification method provided in this application firstly collects data from the target information system according to preset data types to obtain corresponding sample data. Each preset data type represents a potential influencing factor in the target information system. Then, a neural network model is trained using the sample data corresponding to each preset data type to obtain a data identification model. Therefore, for the target information system, the trained data identification model can be directly used to identify and process the operational data of each preset data type to obtain identification results. Based on these identification results, the key influencing factors of the target information system are determined among all preset data types. Clearly, compared to DOE technology, this technical solution does not require extensive experimentation, effectively simplifying its operation process. It can perform more efficient data identification on the target information system to quickly find its key influencing factors, demonstrating better practicality and application prospects.

[0168] This application provides a data identification device.

[0169] Please refer to Figure 3 , Figure 3 This application provides a schematic diagram of the structure of a data identification device, which may include:

[0170] Data acquisition module 1 is used to collect data from the target information system according to preset data types and obtain sample data corresponding to each preset data type.

[0171] Training module 2 is used to train the initial neural network model using the sample data to obtain the data recognition model;

[0172] Processing module 3 is used to process the operational data of the target information system using a data recognition model to obtain recognition results. The operational data includes operational data corresponding to each preset data type.

[0173] Module 4 is used to determine the key influencing factors of the target information system among all preset data types based on the identification results.

[0174] As can be seen, the data identification device provided in this application first collects data from the target information system according to preset data types to obtain corresponding sample data. Each preset data type represents a potential influencing factor in the target information system. Then, a neural network model is trained using the sample data corresponding to each preset data type to obtain a data identification model. Therefore, for the target information system, the trained data identification model can be directly used to identify and process the operational data of each preset data type to obtain identification results. Based on these identification results, the key influencing factors of the target information system are determined among all preset data types. Clearly, compared to DOE technology, this technical solution does not require extensive experimentation, effectively simplifying its operation process. It can perform more efficient data identification on the target information system to quickly find its key influencing factors, demonstrating better practicality and application prospects.

[0175] In one embodiment of this application, the training module 2 can be specifically used to divide the sample data according to a preset ratio to obtain training samples, validation samples, and test samples; to iteratively train the initial neural network model using the training samples and validation samples to obtain an initial data recognition model; to evaluate the initial data recognition model using the test samples to obtain an evaluation result; when the evaluation result is that the evaluation fails, the process returns to the step of iteratively training the initial neural network model using the training samples and validation samples to obtain the initial data recognition model; when the evaluation result is that the evaluation passes, the initial data recognition model is determined to be a data recognition model.

[0176] In one embodiment of this application, the data recognition device may further include a preprocessing module for preprocessing the sample data before training the initial neural network model using the sample data to obtain the data recognition model; the preprocessing includes one or more of the following combinations: data cleaning, outlier handling, data standardization, and missing value imputation.

[0177] In one embodiment of this application, the determination module 4 can be specifically used to determine the influence weight of each preset data type based on the identification result; and to take the preset data type corresponding to the largest influence weight as the key influence factor of the target information system.

[0178] In one embodiment of this application, the data identification device may further include an output module for filtering the operational data corresponding to key influencing factors to obtain abnormal operational data; calculating the defect rate of the target information system based on the number of abnormal operational data; and outputting the key influencing factors, abnormal operational data, and defect rate.

[0179] In one embodiment of this application, the data identification device may further include a monitoring module for monitoring the target information system based on key influencing factors and obtaining monitoring data;

[0180] When the monitoring data does not exceed the first threshold range, return to the step of monitoring the target information system based on key influencing factors to obtain monitoring data; when the monitoring data exceeds the first threshold range but does not exceed the second threshold range, obtain the log information of the target information system; when the log information does not contain abnormal information, return to the step of monitoring the target information system based on key influencing factors to obtain monitoring data; when the log information contains abnormal information, output an alarm prompt; when the monitoring data exceeds the second threshold range, send the monitoring data to the target terminal so that the target terminal user can remotely debug the target information system through the target terminal.

[0181] In one embodiment of this application, the monitoring module can also be used to perform a normality check on the monitoring data and obtain a check result before returning to the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data when the monitoring data does not exceed the first threshold range; when the check result does not meet the preset requirements, the module returns to the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data.

[0182] For a description of the apparatus provided in the embodiments of this application, please refer to the above method embodiments; further details will not be repeated here.

[0183] This application provides an electronic device.

[0184] Please refer to Figure 4 , Figure 4 This application provides a schematic diagram of the structure of an electronic device, which may include:

[0185] Memory, used to store computer programs;

[0186] A processor, used to execute computer programs, can implement the steps of any of the data recognition methods described above.

[0187] like Figure 4 The diagram shows the structural composition of an electronic device, which may include a processor 10, a memory 11, a communication interface 12, and a communication bus 13. The processor 10, memory 11, and communication interface 12 all communicate with each other through the communication bus 13.

[0188] In this embodiment, the processor 10 may be a central processing unit (CPU), an application-specific integrated circuit, a digital signal processor, a field-programmable gate array, or other programmable logic devices.

[0189] The processor 10 can call programs stored in the memory 11. Specifically, the processor 10 can execute operations in the embodiments of the data recognition method.

[0190] The memory 11 is used to store one or more programs. The programs may include program code, which includes computer operation instructions. In this embodiment, the memory 11 stores at least a program for implementing the following functions:

[0191] Data is collected from the target information system according to the preset data types to obtain sample data corresponding to each preset data type.

[0192] The initial neural network model is trained using the sample data to obtain the data recognition model;

[0193] The data recognition model is used to process the operational data of the target information system to obtain the recognition results. The operational data includes the operational data corresponding to each preset data type.

[0194] Based on the identification results, the key influencing factors of the target information system are determined among all preset data types.

[0195] In one possible implementation, the memory 11 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and applications required for at least one function; and the data storage area may store data created during use.

[0196] In addition, memory 11 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device or other volatile solid-state storage device.

[0197] Communication interface 12 can be an interface for the communication module, used to connect with other devices or systems.

[0198] Of course, it should be noted that, Figure 4 The structure shown does not constitute a limitation on the electronic device in the embodiments of this application. In practical applications, the electronic device may include more than Figure 4 More or fewer components as shown, or combinations of certain components.

[0199] This application provides a computer-readable storage medium.

[0200] The computer-readable storage medium provided in this application embodiment stores a computer program, which, when executed by a processor, can implement the steps of any of the data recognition methods described above.

[0201] The computer-readable storage medium may include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0202] For a description of the computer-readable storage medium provided in the embodiments of this application, please refer to the above method embodiments; further details will not be repeated here.

[0203] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0204] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0205] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0206] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the methods and core ideas of this application. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A data recognition method characterized by, include: Data is collected from the target information system according to preset data types to obtain sample data corresponding to each preset data type; the target information system is specifically a MES system, and the preset data types include preheating temperature zone temperature, wave soldering temperature zone temperature, and solder wave height. The initial neural network model is trained using the sample data to obtain a data recognition model; The data recognition model is used to process the operational data of the target information system to obtain recognition results. The operational data includes operational data corresponding to each of the preset data types. Based on the identification results, the key influencing factors of the target information system are determined from all the preset data types; The operational data corresponding to the key influencing factors are filtered to obtain abnormal operational data; Calculate the failure rate of the target information system based on the amount of abnormal operation data; Output the key influencing factors, the abnormal operation data, and the defect rate; The target information system is monitored based on the key influencing factors to obtain monitoring data; When the monitoring data does not exceed the first threshold range, return to the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data; When the monitoring data exceeds the first threshold range but does not exceed the second threshold range, the log information of the target information system is obtained; when the log information does not contain any abnormal information, the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data is returned; when the log information contains any abnormal information, an alarm prompt is output. When the monitoring data exceeds the second threshold range, the monitoring data is sent to the target terminal so that the target terminal user can remotely debug the target information system through the target terminal; The step of determining the key influencing factors of the target information system based on the identification results among all the preset data types includes: The influence weight of each preset data type is determined based on the identification results; The preset data type corresponding to the largest influence weight is taken as the key influence factor of the target information system.

2. The data recognition method of claim 1, wherein, The step of training the initial neural network model using the sample data to obtain the data recognition model includes: The sample data is divided according to a preset ratio to obtain training samples, validation samples, and test samples. The initial neural network model is iteratively trained using the training samples and the validation samples to obtain an initial data recognition model; The initial data recognition model is evaluated using the test samples to obtain evaluation results; When the evaluation result is that the evaluation fails, return to the step of iteratively training the initial neural network model using the training samples and the validation samples to obtain the initial data recognition model; When the evaluation result is "evaluation passed", the initial data recognition model is determined to be the data recognition model.

3. The data recognition method of claim 1, wherein, Before training the initial neural network model using the sample data to obtain the data recognition model, the method further includes: The sample data is preprocessed; the preprocessing includes one or more of the following: data cleaning, outlier handling, data standardization, and missing value imputation.

4. The data recognition method of claim 1, wherein, Before the step of returning to the step of monitoring the target information system based on the key influencing factors and obtaining monitoring data when the monitoring data does not exceed the first threshold range, the method further includes: Perform a normality check on the monitoring data and obtain the check result; When the verification result does not meet the preset requirements, return to the step of monitoring the target information system based on the key influencing factors and obtaining monitoring data.

5. A data recognition apparatus characterized by comprising: include: The data acquisition module is used to acquire data from the target information system according to preset data types, and obtain sample data corresponding to each preset data type; the target information system is specifically a MES system, and the preset data types include preheating temperature zone temperature, wave soldering temperature zone temperature, and solder wave height. The training module is used to train the initial neural network model using the sample data to obtain a data recognition model; The processing module is used to process the operating data of the target information system using the data recognition model to obtain recognition results. The operating data includes operating data corresponding to each of the preset data types. A determination module is used to determine the key influencing factors of the target information system among all the preset data types based on the identification results; The output module is used to filter the operational data corresponding to the key influencing factors and obtain abnormal operational data. Calculate the failure rate of the target information system based on the amount of abnormal operation data; Output the key influencing factors, the abnormal operation data, and the defect rate; The monitoring module is used to monitor the target information system based on the key influencing factors and obtain monitoring data; When the monitoring data does not exceed the first threshold range, return to the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data; When the monitoring data exceeds the first threshold range but does not exceed the second threshold range, the log information of the target information system is obtained; when the log information does not contain any abnormal information, the step of monitoring the target information system according to the key influencing factors and obtaining monitoring data is returned; when the log information contains any abnormal information, an alarm prompt is output. When the monitoring data exceeds the second threshold range, the monitoring data is sent to the target terminal so that the target terminal user can remotely debug the target information system through the target terminal; Specifically, the determining module is used to determine the influence weight of each preset data type based on the identification result; The preset data type corresponding to the largest influence weight is taken as the key influence factor of the target information system.

6. An electronic device, comprising: include: Memory, used to store computer programs; A processor for executing the computer program to implement the steps of the data identification method as described in any one of claims 1 to 4.

7. A computer readable storage medium characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the data identification method as described in any one of claims 1 to 4.

Images