Risk prediction method and device for part change task

Abstract

The invention provides a part change task risk prediction method and device, and the method comprises the steps: receiving a risk prediction instruction of a target part change task, and enabling the risk prediction instruction to carry to-be-tested task data; under the condition that the to-be-tested task data lacks the target data item, supplementing the target data item of the to-be-tested task data according to a set supplementing rule to obtain updated to-be-tested task data; the updated to-be-tested task data are input into a risk prediction model, a corresponding risk prediction result is obtained, the risk prediction model is obtained based on risk prediction samples and corresponding sample label training, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules. The target data item is supplemented according to the set supplement rule, the prediction of the part change task of the missing data item is realized, the accuracy of the prediction result is improved, risks can be effectively avoided in each stage of production and manufacturing, and the overall stability of a production line is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a risk prediction method for a part change task. The present application also relates to a risk prediction device for a part change task, a computing device, and a computer-readable storage medium. Background Art

[0002] In the manufacturing process, people tend to pay more attention to the risks caused by product quality, and most risk judgments only stay at the judgment of risks related to product quality. In the actual manufacturing process, the risks of changing tasks often come from multiple stages and aspects of manufacturing. The current assessment method is relatively single and one-sided, and risk judgment is basically carried out in the second half of the production line, which is not conducive to timely discovery and avoidance of risks.

[0003] At present, most of the risk judgments for change tasks are made by experts based on the data information in historical change tasks and their experience. Experts need to spend a lot of time to formulate strategies and unify judgment standards, resulting in low efficiency in risk judgment. In addition, when there is partial missing data information in the change task, accurate judgment results cannot be obtained. Summary of the invention

[0004] In view of this, the embodiment of the present application provides a risk prediction method for a part change task. The present application also relates to a risk prediction device for a part change task, a computing device, and a computer-readable storage medium to solve the above problems existing in the prior art.

[0005] According to a first aspect of an embodiment of the present application, a risk prediction method for a part change task is provided, comprising:

[0006] receiving a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested;

[0007] In the case where the task data to be tested lacks a target data item, the target data item of the task data to be tested is supplemented according to a set supplement rule to obtain updated task data to be tested;

[0008] The updated task data to be tested is input into the risk prediction model to obtain the corresponding risk prediction results, wherein the risk prediction model is trained based on risk prediction samples and corresponding sample labels, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules.

[0009] According to a second aspect of an embodiment of the present application, a risk prediction device for a part change task is provided, comprising:

[0010] A receiving module is configured to receive a risk prediction instruction of a target part change task, wherein the risk prediction instruction carries task data to be tested;

[0011] A supplementing module is configured to supplement the target data item of the task data to be tested according to a set supplementing rule when the target data item is missing from the task data to be tested, so as to obtain updated task data to be tested;

[0012] The input module is configured to input the updated task data to be tested into the risk prediction model to obtain the corresponding risk prediction results, wherein the risk prediction model is obtained based on risk prediction samples and corresponding sample label training, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules.

[0013] According to a third aspect of an embodiment of the present application, a computing device is provided, comprising a memory, a processor, and computer instructions stored in the memory and executable on the processor, wherein the processor implements the steps of the risk prediction method for the part change task when executing the computer instructions.

[0014] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores computer instructions, and when the computer instructions are executed by a processor, the steps of the risk prediction method for the part change task are implemented.

[0015] The risk prediction method for a part change task provided in the present application receives a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested; in the case where the task data to be tested lacks a target data item, the target data item of the task data to be tested is supplemented according to set supplementary rules to obtain updated task data to be tested; the updated task data to be tested is input into a risk prediction model to obtain a corresponding risk prediction result, wherein the risk prediction model is obtained through training based on risk prediction samples and corresponding sample labels, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules.

[0016] In this way, by receiving the risk prediction instruction of the target part change task, it is possible to predict whether the target part change task has risks based on the task data to be tested, which is convenient for predicting whether the task has risks during the production and manufacturing process, and is conducive to taking timely measures to avoid risks; by supplementing the target data items of the task data to be tested according to the set supplementary rules when the target data items are missing in the task data to be tested, and obtaining updated task data to be tested, it is possible to make normal predictions of the target part change tasks with missing target data items, thereby improving the accuracy of the prediction results; by inputting the updated task data to be tested into the risk prediction model to obtain the corresponding risk prediction results, it is possible to make more accurate and efficient predictions of the risks of the target part change tasks, so that risk predictions can be made in a timely manner for the part change tasks at various stages of production and manufacturing, which can effectively avoid risks of different dimensions at different stages and improve the overall stability of the production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of a risk prediction method for a part change task provided by an embodiment of the present application;

[0018] Figure 2 It is a schematic diagram of the model structure of a risk prediction model in a risk prediction method for a part change task provided in one embodiment of the present application;

[0019] Figure 3 It is a processing flow chart of a risk prediction method for a parts change task applied to a vehicle production line provided by an embodiment of the present application;

[0020] Figure 4 It is a structural schematic diagram of a risk prediction device for a part change task provided by an embodiment of the present application;

[0021] Figure 5 It is a structural block diagram of a computing device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0022] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application, so the present application is not limited by the specific implementation disclosed below.

[0023] The terms used in one or more embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present application. The singular forms of "a", "said" and "the" used in one or more embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of the present application refers to and includes any or all possible combinations of one or more associated listed items.

[0024] It should be understood that, although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present application, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0025] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0026] First, the terms involved in one or more embodiments of the present application are explained.

[0027] One-Hot encoding: also known as unique hot encoding or single-bit effective encoding, uses an N-bit state register to encode N states. Each state has an independent register bit, and at any time, only one of the bits is valid and can be used to classify data.

[0028] At present, in industrial manufacturing, most risk assessments are still focused on product quality-related risks. Compared with the risks that exist at each stage of the entire manufacturing life cycle, product quality-related risks focus more on the risks in the second half of the production line. The amount of information contained is relatively single and one-sided, and it is impossible to detect and avoid risks in a timely manner.

[0029] In the factory-based vehicle manufacturing process, a large number of parts change tasks are involved. Part change tasks at different production stages often have risks from different sources, including but not limited to production cycle, logistics cycle, process allocation, etc. Since there are often complex logical connections between businesses on the production line, it is often difficult to determine the source of risk and the root cause of the risk based on expert experience alone, resulting in the risk prediction of parts change tasks is often not accurate enough and the risk cannot be avoided in time.

[0030] Based on this, the present application provides a risk prediction method for a part change task, which receives a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested; in the case where the task data to be tested lacks a target data item, the target data item of the task data to be tested is supplemented according to set supplementary rules to obtain updated task data to be tested; the updated task data to be tested is input into a risk prediction model to obtain a corresponding risk prediction result, wherein the risk prediction model is obtained through training based on risk prediction samples and corresponding sample labels, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules.

[0031] In this way, by receiving the risk prediction instruction of the target part change task, it is possible to predict whether the target part change task has risks based on the task data to be tested, which is convenient for predicting whether the task has risks during the production and manufacturing process, and is conducive to taking timely measures to avoid risks; by supplementing the target data items of the task data to be tested according to the set supplementary rules when the target data items are missing in the task data to be tested, and obtaining updated task data to be tested, it is possible to make normal predictions of the target part change tasks with missing target data items, thereby improving the accuracy of the prediction results; by inputting the updated task data to be tested into the risk prediction model to obtain the corresponding risk prediction results, it is possible to make more accurate and efficient predictions of the risks of the target part change tasks, so that risk predictions can be made in a timely manner for the part change tasks at various stages of production and manufacturing, which can effectively avoid risks of different dimensions at different stages and improve the overall stability of the production line.

[0032] In the present application, a risk prediction method for a part change task is provided. The present application also relates to a risk prediction device for a part change task, a computing device, and a computer-readable storage medium, which are described in detail one by one in the following embodiments.

[0033] Figure 1 A flow chart of a risk prediction method for a part change task provided according to an embodiment of the present application is shown, which specifically includes the following steps:

[0034] Step 102: Receive a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested.

[0035] In actual applications, when a part change task is detected in the production process, a risk prediction instruction for the target part change task can be generated, so that the risk prediction system receives the risk prediction instruction for the target part change task and predicts whether there is a risk in the target part change task.

[0036] Specifically, the target part change task can be understood as a change task for the target part. Part change tasks can specifically include replacing parts, changing part inventory, changing part models, changing part application scenarios, etc. Part change tasks will involve many different types of data, including but not limited to part models, associated factories, associated models, process owners, change start time, interval time, planned arrival time, and actual arrival time, etc. Part change tasks can be generated at different production and manufacturing stages, including but not limited to the vehicle production stage, maintenance stage, recycling stage, etc.

[0037] The risk prediction instruction can be used to instruct the risk prediction system to perform risk prediction on the target part change task. The task data to be tested can be understood as at least one item of data corresponding to the target part change task. Furthermore, whether there is a risk in the target part change task can be predicted based on the task data to be tested. The task data to be tested can be various data involved in the execution of the target part change task, including but not limited to part model, associated factory, associated vehicle model, process owner, change start time, interval time, planned arrival time and actual arrival time, etc.

[0038] By receiving the risk prediction instruction of the target part change task, it is possible to predict whether there is a risk in the target part change task based on the task data to be tested carried by the risk prediction instruction. The change task can be predicted in time during the production process, and risks can be avoided in time to ensure the stability of the overall production line.

[0039] Step 104: when the target data item is missing from the task data to be tested, the target data item of the task data to be tested is supplemented according to the set supplement rule to obtain updated task data to be tested.

[0040] In practical applications, the task data to be tested may include complete data items required for risk prediction, or may include partial data items required for risk prediction. In the case where the task data to be tested includes partial data items required for risk prediction, it means that the task data to be tested has missing data items, and the target data items of the task data to be tested may be supplemented according to the set supplementation rules, and after obtaining the updated task data to be tested, risk prediction may be performed on the updated task data to be tested.

[0041] Specifically, the target data item can be understood as the data item missing from the task data to be tested. The supplementary rules are set to supplement the missing data items of the task data to be tested, so that the task data to be tested has the complete data items required for risk prediction. The updated task data to be tested can be understood as the task data to be tested that has been supplemented with the target data item according to the set supplementary rules.

[0042] It should be noted that, when the task data to be tested includes complete data items required for risk prediction, risk prediction can be performed directly on the task data to obtain the risk prediction result of the target part change task. When there are missing data items in the task data to be tested, the missing target data items can be supplemented first, and then risk prediction can be performed on the updated task data to be tested after supplementation.

[0043] In an optional implementation of the present application, the supplementation rule is set to supplement the approximate value; accordingly, supplementing the target data item according to the set supplementation rule to obtain updated task data to be tested may include the following steps:

[0044] Determining at least one historical task data, wherein the historical task data includes a target data item;

[0045] Determining target historical task data according to the similarity between the task data to be tested and at least one historical task data;

[0046] The data content of the target data item in the target historical task data is supplemented to the task data to be tested to obtain updated task data to be tested.

[0047] Specifically, the supplement rule may be set to supplement the approximate value. The historical task data may be understood as the change task data involved in the historical part change task. The target data item may be understood as the data item included in the historical task data.

[0048] During the actual implementation process, historical records can be obtained from the database, and the similarity between the data items included in the task data to be tested and the other data items in the historical task data except the target data items can be calculated. When the similarity is close, the target data items in the historical task data can be directly used as the target data items of the task data to be tested.

[0049] It should be noted that in the process of calculating the similarity, the similarity weight can be set according to the type of data item. For example, the data items corresponding to the part model, planned arrival time, and actual arrival time have higher similarity weights, and the data items corresponding to the associated vehicle model and associated factory model have the second highest similarity weights. The historical task data may have missing data items, but it must contain the target data items.

[0050] In another optional implementation manner of the present application, setting the supplementary rule may also be to supplement a fixed value; accordingly, supplementing the target data item according to the set supplementary rule to obtain updated task data to be tested may also include the following steps:

[0051] The target data item is supplemented with a preset fixed value to obtain updated task data to be tested.

[0052] Specifically, a fixed value for supplementing missing data items may be set according to actual application conditions. For example, the number 0 or other numbers may be uniformly used to supplement missing data items, or fixed values ​​with different values ​​may be used to supplement missing data items according to the data type corresponding to the target data item.

[0053] By supplementing the target data items of the task data to be tested according to the set supplement rules when the target data items are missing in the task data to be tested, updated task data to be tested can be obtained. The task data to be tested with missing data items can be supplemented, so that the risk prediction system can make normal predictions on the input data and output accurate prediction results, thereby avoiding the inability to identify the task data to be tested or the inability to accurately predict risks, and improving the accuracy and flexibility of risk prediction.

[0054] Step 106: Input the updated task data to be tested into the risk prediction model to obtain the corresponding risk prediction result, wherein the risk prediction model is trained based on risk prediction samples and corresponding sample labels, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules.

[0055] In practical applications, after supplementing the target data items of the task data to be tested according to the set supplementary rules and obtaining updated task data to be tested, the updated task data to be tested can be input into the risk prediction model to obtain the risk prediction results output by the risk prediction model.

[0056] Specifically, the risk prediction result can be understood as the result output by the risk prediction model. The risk prediction model is a model obtained through pre-training for performing risk prediction on the task data to be tested. The risk prediction model can be obtained by training based on risk prediction samples and corresponding sample labels. The risk prediction sample can be historical task data that supplements historical task data with missing data items based on set supplementary rules.

[0057] It should be noted that the data items in the risk prediction sample are complete data items required for risk prediction. If there are no missing data items in the historical task data, the model can be directly trained as a risk prediction sample. If the historical task data includes missing data items, the missing data items can be filled in based on the set supplementary rules, and then used as a risk prediction sample to train the model.

[0058] In this way, the trained risk prediction model has the ability to accurately predict the task data to be tested with missing data items, which can improve the accuracy of the prediction results of the risk prediction model.

[0059] like Figure 2 As shown, Figure 2 A schematic diagram of the model structure of a risk prediction model in a risk prediction method for a part change task provided in an embodiment of the present application is shown. The risk prediction model includes a coding layer and a fully connected layer, wherein the coding layer can receive input task data to be tested, generate coding features of the task data to be tested through feature extraction and feature representation, and then input the coding features into the fully connected layer. The fully connected layer can specifically include at least one fully connected sublayer, and the coding features are fused through the fully connected sublayer to obtain fusion information. Finally, the result output by the fully connected layer can be normalized to obtain a risk prediction result and output it.

[0060] Exemplarily, the fully connected layer can be set to two fully connected sublayers, the output dimension (hidden_size) of the first layer can be set to 64, the output dimension (hidden_size) of the second layer can be set to 16, and the activation functions of both layers can be set to relu.

[0061] In the actual implementation process, after the task data to be tested is input into the risk prediction model, it can be first input into the coding layer of the risk prediction model. Since the task data to be tested is based on the data related to the part change task in the factory's production and manufacturing process, including data of different dimensions such as parts, staff, factories, warehouses, assembly lines, time, etc., in order to facilitate feature processing of the data, the coding layer can first discretize the input task data to be tested to obtain discrete data items, and then perform One-Hot encoding on the discrete data items to obtain the coding features corresponding to the task data to be tested. The coding features are then output and input into the fully connected layer. The fully connected layer processes the coding features to obtain the processing results, and the final processing results are subjected to softmax normalization to obtain the risk prediction results and output.

[0062] In practical applications, the risk prediction result can be a risk value or a result that characterizes whether the task data to be tested has a risk. In the case where the risk prediction result is a risk value, it is possible to determine whether the risk value exceeds the corresponding risk threshold according to the risk threshold corresponding to the task data to be tested, and determine whether the task data to be tested has a risk according to the determination result. In the case where the risk prediction result is a result that characterizes whether the task data to be tested has a risk, it is possible to determine whether the task data to be tested has a risk directly according to the result.

[0063] The risk prediction system can display the risk prediction results of the target part change task in the interactive interface based on the risk prediction results output by the risk prediction model. In practical applications, there can be multiple display forms. For example, if the risk prediction result is that there is a risk, a red light can be turned on in the prediction result display box corresponding to the target part change task; if the risk prediction result is that there is no risk, a green light can be turned on in the prediction result display box corresponding to the target part change task. The risk prediction results can also be displayed in the form of text or other symbols.

[0064] By inputting the updated data of the task to be tested into the risk prediction model, obtaining the corresponding risk prediction results, and displaying the risk prediction results, the generated part change tasks can be risk predicted in real time during the actual production and manufacturing process, and the risk prediction results can be displayed, so as to notify the relevant responsible personnel to handle the risky part change tasks in a timely manner, effectively avoid different risks in different stages, and improve the overall stability of the production line.

[0065] In addition, predicting the task data to be tested based on the pre-trained risk prediction model can effectively improve the risk prediction efficiency and accuracy. It can also solve the problem that humans cannot accurately predict the task data with missing data items.

[0066] In order to improve the accuracy of the prediction results output by the risk prediction model, in one or more embodiments of the present application, the risk prediction model can be trained by the following steps:

[0067] Obtaining risk prediction samples, wherein the risk prediction samples carry corresponding sample labels;

[0068] Inputting the risk prediction sample into the initial risk prediction model to obtain the sample prediction result output by the initial risk prediction model;

[0069] According to the sample labels and sample prediction results, the loss value of the initial risk prediction model is calculated, the model parameters of the initial risk prediction model are reversely adjusted according to the loss value, and the step of obtaining the risk prediction sample is returned to execute until the training stop condition is reached to obtain the trained risk prediction model.

[0070] Specifically, the risk prediction sample carries the corresponding sample label, and the sample label is obtained by annotating the risk prediction sample. The initial risk prediction model can be understood as a model that has not been trained based on the risk prediction sample. The training stop condition can be understood as the condition for the initial risk prediction model to complete the training. Specifically, it can be judged whether the model has reached the training stop condition based on the accuracy of the model output result, the average recall rate of positive and negative samples, etc. Exemplarily, the accuracy of the model output result can be set to 96.1%, and the average recall rate of positive and negative samples can be set to 96.8%. According to the adjustment of the training stop condition, a risk prediction model with more accurate prediction results can be obtained.

[0071] In practical applications, risk prediction samples with corresponding sample labels can be obtained from the training set, and the risk prediction samples can be input into the initial risk prediction model to obtain the sample prediction results output by the initial risk prediction model. The loss value of the initial risk prediction model can be calculated based on the sample labels and the sample prediction results. The model parameters of the initial risk prediction model can be adjusted in reverse based on the loss value, and the step of obtaining the risk prediction samples can be returned to execute. New risk prediction samples with corresponding sample labels can be obtained from the training set again and input into the initial risk prediction model for training until the model reaches the training stop condition to obtain the trained risk prediction model.

[0072] Furthermore, in order to improve the accuracy of the risk prediction model, a test set can be constructed, and the test samples can be input into the trained risk prediction model. The accuracy of the model prediction results can be tested based on the prediction results output by the risk prediction model and the sample labels corresponding to the test samples. If the model fails to meet the test conditions, the model can be used as the initial risk prediction model and returned to the training stage for model training until the model passes the test.

[0073] In an optional implementation of the present application, obtaining a risk prediction sample may include the following steps:

[0074] Get historical task data;

[0075] In the case where there are missing data items in the historical task data, the missing data items are supplemented by setting supplementary rules to obtain updated historical task data;

[0076] Filter risk prediction samples from updated historical task data.

[0077] Specifically, the historical task data can be understood as the task data involved in the historical part change task. The updated historical task data can be understood as the data containing the complete data items required for risk prediction after supplementing the missing data items.

[0078] In practical applications, the historical task data in the historical part change task may include complete data items required for risk prediction, or may have missing data items. In the case where there are missing data items in the historical task data, the missing data items can be supplemented by setting supplementary rules to obtain updated historical task data. In the case where there are no missing data items in the historical task data, the historical task data can be directly used as updated historical task data for screening risk prediction samples.

[0079] In the actual implementation process, the update history task data obtained can be screened out with a first data volume of update history task data to construct a training set of the model, and the remaining update history task data can be constructed as a test set of the model.

[0080] By obtaining historical task data, when there are missing data items in the historical task data, the missing data items are supplemented by setting supplementary rules to obtain updated historical task data, and risk prediction samples are screened from the updated historical task data, the model can be trained according to real data to improve the accuracy of the model's risk prediction.

[0081] In another optional implementation of the present application, when risk prediction samples have been stored historically, the risk prediction model can be trained directly using the risk prediction samples that have been used historically.

[0082] In another optional implementation of the present application, risk prediction samples can also be constructed by experts based on expert experience and real data from actual production and manufacturing processes.

[0083] In an optional embodiment of the present application, the sample label can be obtained by the following steps:

[0084] According to the preset labeling rules, the risk prediction samples are labeled to obtain the sample labels of the risk prediction samples.

[0085] Specifically, the preset labeling rules can be understood as rules for labeling risk prediction samples.

[0086] In practical applications, the preset labeling rules can be used to generate labeling instructions to instruct professional staff to manually label risk prediction samples based on expert experience, or they can be used to automatically label risk prediction samples based on preset labeling formulas.

[0087] Optionally, labeling risk prediction samples according to preset labeling rules may include the following steps:

[0088] Determine the change type of the historical parts change task corresponding to the risk prediction sample;

[0089] Generate formulas and risk thresholds based on labels matching risk prediction samples according to change types;

[0090] Label the risk prediction samples according to the label generation formula and risk threshold.

[0091] Specifically, the change type can be understood as the task type corresponding to the historical part change task, which can be the stage of generating the historical part change task, the stage of actual application of the part change task in the manufacturing process, or the importance and urgency of the task, etc. The risk threshold can be understood as the standard for judging whether the risk prediction sample has risks. If the risk value exceeds the risk threshold, the risk prediction sample is judged to have risks. If the risk value does not exceed the risk threshold, the risk prediction sample is judged to have no risks.

[0092] According to the preset labeling rules, the risk prediction samples are labeled by the label generation formula, and the labeling result obtained is the presence of risk or the absence of risk. The label generation formula can be understood as a formula for labeling whether the risk prediction sample has a risk, which is used to automatically calculate the risk result of the risk prediction sample.

[0093] By determining the change type of the historical part change task corresponding to the risk prediction sample, matching the label generation formula and risk threshold of the risk prediction sample according to the change type, and annotating the risk prediction sample according to the label generation formula and risk threshold, it is possible to automatically generate sample labels for the risk prediction sample, improve the efficiency of sample label generation, and thus improve the efficiency of model training.

[0094] Furthermore, labeling risk prediction samples according to the label generation formula and the risk threshold may include the following steps:

[0095] According to the label generation formula, the target sample data in the risk prediction sample is operated to generate the operation result;

[0096] Determine whether the calculation result exceeds the risk threshold;

[0097] If yes, the risk prediction sample is marked as risky;

[0098] If not, the risk prediction sample is marked as risk-free.

[0099] Specifically, the target sample data can be understood as the data included in the risk prediction sample and used for risk prediction. According to the label generation formula corresponding to the risk prediction sample, the target sample data can specifically include at least one of the change start time, interval time, planned change to warehouse time, and actual change to warehouse time.

[0100] Exemplarily, when the change type is an emergency change, the label generation formula for the corresponding matched risk prediction sample can be: interval time-actual change to library time, and the risk threshold can be 30 minutes. In this case, the target sample data can be the interval time and the actual change to library time. When the calculation result is calculated according to the label generation formula, the calculation result is used as the risk value and compared with the risk threshold. When the risk value does not exceed 30 minutes, the risk prediction sample is marked as risk-free, and the label can be set to 0; when the risk value exceeds 30 minutes, the risk prediction sample is marked as risk-free, and the label can be set to 1.

[0101] In another example, the change type is a non-emergency change, and the label generation formula for the corresponding matched risk prediction sample can be: change start time - actual change to the library time and interval time - actual change to the library time. At this time, the corresponding risk judgment condition is when change start time - actual change to the library time > interval time - actual change to the library time, it is judged as risky; when change start time - actual change to the library time ≤ interval time - actual change to the library time, it is judged as no risk.

[0102] Specifically, the label generation formula may be set based on expert experience, or may be set based on task data of historical risk change tasks.

[0103] By generating formulas based on labels, calculating the target sample data in the risk prediction samples, generating calculation results, judging whether the calculation results exceed the risk threshold, and labeling the risk prediction samples according to the judgment results, it can be combined with the change types in actual applications. For the vehicle production line in the factory, the change of each vehicle part will evaluate the risk points from multiple dimensions such as whether the parts can arrive in the warehouse, whether the profit in the warehouse is sufficient, whether the parts can arrive in the warehouse according to the planned time, and whether the parts changes will conflict, so as to generate more accurate sample labels, thereby improving the accuracy of the risk prediction results of the risk prediction model.

[0104] An embodiment of the present application realizes that, by receiving a risk prediction instruction for a target part change task, it is possible to predict whether a target part change task has a risk based on the task data to be tested, thereby facilitating the prediction of whether a task has a risk during the production process, and facilitating timely measures to avoid risks; by supplementing the target data items of the task data to be tested according to set supplementary rules when the target data items are missing in the task data to be tested, and obtaining updated task data to be tested, it is possible to perform normal prediction of the target part change task with the missing target data items, thereby improving the accuracy of the prediction results; by inputting the updated task data to be tested into a risk prediction model to obtain the corresponding risk prediction results, it is possible to perform more accurate and efficient prediction of the risks of the target part change task, thereby enabling timely risk prediction of the part change task at various stages of production and manufacturing, thereby effectively avoiding risks of different dimensions at different stages, and improving the overall stability of the production line.

[0105] The following combination Figure 3 Taking the application of the risk prediction method for part change tasks provided by the present application in a vehicle production line as an example, the risk prediction method for part change tasks is further described. Figure 3 A processing flow chart of a risk prediction method for a part change task applied to a vehicle production line provided by an embodiment of the present application is shown, which specifically includes the following steps:

[0106] Step 302: Receive a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested.

[0107] Step 304: In the case that the target data item is missing in the task data to be tested, determine at least one historical task data, determine the target historical task data based on the similarity between the task data to be tested and at least one historical task data, and supplement the data content of the target data item in the target historical task data to the task data to be tested to obtain updated task data to be tested.

[0108] Step 306: The updated task data to be tested is input into the coding layer in the risk prediction model. The coding layer discretizes the updated task data to be tested, obtains discrete data items, and then encodes the discrete data items to obtain coding features.

[0109] Step 308: Input the encoded features output by the encoding layer to the fully connected layer to obtain the fusion information output by the fully connected layer.

[0110] Step 310: normalize the fused information and output the risk prediction result corresponding to the target part change task.

[0111] An embodiment of the present application realizes that, by receiving a risk prediction instruction for a target part change task, it is possible to predict whether a target part change task has a risk based on the task data to be tested, thereby facilitating the prediction of whether a task has a risk during the production process, and facilitating timely measures to avoid risks; by supplementing the target data items of the task data to be tested according to set supplementary rules when the target data items are missing in the task data to be tested, and obtaining updated task data to be tested, it is possible to perform normal prediction of the target part change task with the missing target data items, thereby improving the accuracy of the prediction results; by inputting the updated task data to be tested into a risk prediction model to obtain the corresponding risk prediction results, it is possible to perform more accurate and efficient prediction of the risks of the target part change task, thereby enabling timely risk prediction of the part change task at various stages of production and manufacturing, thereby effectively avoiding risks of different dimensions at different stages, and improving the overall stability of the production line.

[0112] Corresponding to the above method embodiment, the present application also provides a risk prediction device embodiment for a part change task, Figure 4 FIG. 1 is a schematic diagram showing a risk prediction device for a part change task provided by an embodiment of the present application. Figure 4 As shown, the device comprises:

[0113] Receiving module 402: receiving a risk prediction instruction of a target part change task, wherein the risk prediction instruction carries task data to be tested;

[0114] Supplementation module 404: when the target data item is missing from the task data to be tested, the target data item of the task data to be tested is supplemented according to the set supplementation rule to obtain updated task data to be tested;

[0115] Input module 406: input the updated task data to be tested into the risk prediction model to obtain the corresponding risk prediction result, wherein the risk prediction model is obtained based on risk prediction samples and corresponding sample label training, and the risk prediction samples are obtained by processing historical task data based on set supplementary rules.

[0116] Optionally, set the supplementation rule to supplemental approximation;

[0117] The supplement module 404 is further configured to:

[0118] Determining at least one historical task data, wherein the historical task data includes a target data item;

[0119] Determining target historical task data according to the similarity between the task data to be tested and at least one historical task data;

[0120] The data content of the target data item in the target historical task data is supplemented to the task data to be tested to obtain updated task data to be tested.

[0121] Optionally, the risk prediction device for a part change task may further include a training module configured to:

[0122] Obtaining risk prediction samples, wherein the risk prediction samples carry corresponding sample labels;

[0123] Inputting the risk prediction sample into the initial risk prediction model to obtain the sample prediction result output by the initial risk prediction model;

[0124] According to the sample labels and sample prediction results, the loss value of the initial risk prediction model is calculated, the model parameters of the initial risk prediction model are reversely adjusted according to the loss value, and the step of obtaining the risk prediction sample is returned to execute until the training stop condition is reached to obtain the trained risk prediction model.

[0125] Optionally, the training module is further configured to:

[0126] Get historical task data;

[0127] In the case where there are missing data items in the historical task data, the missing data items are supplemented by setting supplementary rules to obtain updated historical task data;

[0128] Filter risk prediction samples from updated historical task data.

[0129] Optionally, the training module is further configured to:

[0130] According to the preset labeling rules, the risk prediction samples are labeled to obtain the sample labels of the risk prediction samples.

[0131] Optionally, the training module is further configured to:

[0132] Determine the change type of the historical parts change task corresponding to the risk prediction sample;

[0133] Generate formulas and risk thresholds based on labels matching risk prediction samples according to change types;

[0134] Label the risk prediction samples according to the label generation formula and risk threshold.

[0135] Optionally, the training module is further configured to:

[0136] According to the label generation formula, the target sample data in the risk prediction sample is operated to generate the operation result;

[0137] Determine whether the calculation result exceeds the risk threshold;

[0138] If yes, the risk prediction sample is marked as risky;

[0139] If not, the risk prediction sample is marked as risk-free.

[0140] An embodiment of the present application realizes that, by receiving a risk prediction instruction for a target part change task, it is possible to predict whether a target part change task has a risk based on the task data to be tested, thereby facilitating the prediction of whether a task has a risk during the production process, and facilitating timely measures to avoid risks; by supplementing the target data items of the task data to be tested according to set supplementary rules when the target data items are missing in the task data to be tested, and obtaining updated task data to be tested, it is possible to perform normal prediction of the target part change task with the missing target data items, thereby improving the accuracy of the prediction results; by inputting the updated task data to be tested into a risk prediction model to obtain the corresponding risk prediction results, it is possible to perform more accurate and efficient prediction of the risks of the target part change task, thereby enabling timely risk prediction of the part change task at various stages of production and manufacturing, thereby effectively avoiding risks of different dimensions at different stages, and improving the overall stability of the production line.

[0141] The above is a schematic scheme of a risk prediction device for a part change task of this embodiment. It should be noted that the technical scheme of the risk prediction device for the part change task and the technical scheme of the risk prediction method for the part change task mentioned above belong to the same concept. For details not described in detail in the technical scheme of the risk prediction device for the part change task, please refer to the description of the technical scheme of the risk prediction method for the part change task mentioned above.

[0142] Figure 5 The block diagram of a computing device according to an embodiment of the present application is shown. The components of the computing device 500 include but are not limited to a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and the database 550 is used to store data.

[0143] The computing device 500 also includes an access device 540 that enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 540 may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, a near field communication (NFC) interface, and the like.

[0144] In one embodiment of the present application, the above components of the computing device 500 and Figure 5 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 5 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of the present application. Those skilled in the art may add or replace other components as needed.

[0145] The computing device 500 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 500 may also be a mobile or stationary server.

[0146] Among them, when the processor 520 executes the computer instructions, the steps of the risk prediction method of the part change task are implemented.

[0147] The above is a schematic scheme of a computing device of this embodiment. It should be noted that the technical scheme of the computing device and the technical scheme of the risk prediction method of the above-mentioned part change task belong to the same concept, and the details not described in detail in the technical scheme of the computing device can be referred to the description of the technical scheme of the risk prediction method of the above-mentioned part change task.

[0148] An embodiment of the present application further provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the steps of the risk prediction method for the part change task as described above.

[0149] The above is a schematic scheme of a computer-readable storage medium of this embodiment. It should be noted that the technical scheme of the storage medium and the technical scheme of the risk prediction method of the above-mentioned part change task belong to the same concept, and the details not described in detail in the technical scheme of the storage medium can be referred to the description of the technical scheme of the risk prediction method of the above-mentioned part change task.

[0150] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0151] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0152] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described action sequence, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0153] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0154] The preferred embodiments of the present application disclosed above are only used to help explain the present application. The optional embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the present application. The present application selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can understand and use the present application well. The present application is only limited by the claims and their full scope and equivalents.

Claims

1. A risk prediction method for a part change task, characterized in that: include: Receiving a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested; In the case that the task data to be tested lacks a target data item, supplementing the target data item of the task data to be tested according to a set supplement rule to obtain updated task data to be tested; The updated task data to be tested is input into the risk prediction model to obtain a corresponding risk prediction result, wherein the risk prediction model is trained based on risk prediction samples and corresponding sample labels, and the risk prediction samples are obtained by processing historical task data based on the set supplementary rules.

2. The method according to claim 1, characterized in that The setting supplementary rule is to supplement the approximate value; The step of supplementing the target data item according to the set supplementing rule to obtain updated task data to be tested includes: Determining at least one historical task data, wherein the historical task data includes the target data item; Determining target historical task data according to the similarity between the task data to be tested and the at least one historical task data; The data content of the target data item in the target historical task data is supplemented to the task data to be tested to obtain updated task data to be tested.

3. The method according to claim 1, characterized in that The risk prediction model is obtained by training through the following steps: Acquire a risk prediction sample, wherein the risk prediction sample carries a corresponding sample label; Inputting the risk prediction sample into an initial risk prediction model to obtain a sample prediction result output by the initial risk prediction model; According to the sample label and the sample prediction result, the loss value of the initial risk prediction model is calculated, the model parameters of the initial risk prediction model are reversely adjusted according to the loss value, and the step of obtaining the risk prediction sample is returned to execute until the training stop condition is reached to obtain the trained risk prediction model.

4. The method according to claim 3, characterized in that The obtaining of risk prediction samples comprises: Get historical task data; In the case where there are missing data items in the historical task data, the missing data items are supplemented by the set supplementary rules to obtain updated historical task data; The risk prediction samples are screened from the updated historical task data.

5. The method according to claim 3, characterized in that: The sample labels are obtained by the following steps: The risk prediction sample is labeled according to a preset labeling rule to obtain a sample label of the risk prediction sample.

6. The method according to claim 5, characterized in that The step of labeling the risk prediction sample according to a preset labeling rule includes: Determining a change type of a historical part change task corresponding to the risk prediction sample; A label generation formula and a risk threshold for matching the risk prediction sample according to the change type; The risk prediction sample is labeled according to the label generation formula and the risk threshold.

7. The method according to claim 6, characterized in that The step of labeling the risk prediction sample according to the label generation formula and the risk threshold comprises: According to the label generation formula, the target sample data in the risk prediction sample is operated to generate an operation result; Determining whether the calculation result exceeds the risk threshold; If yes, marking the risk prediction sample as risky; If not, the risk prediction sample is marked as risk-free.

8. A risk prediction device for a part change task, characterized in that: include: A receiving module is configured to receive a risk prediction instruction for a target part change task, wherein the risk prediction instruction carries task data to be tested; A supplementing module, configured to supplement the target data item of the task data to be tested according to a set supplementing rule to obtain updated task data to be tested when the target data item is missing from the task data to be tested; The input module is configured to input the updated task data to be tested into the risk prediction model to obtain the corresponding risk prediction result, wherein the risk prediction model is obtained based on risk prediction samples and corresponding sample label training, and the risk prediction samples are obtained by processing historical task data based on the set supplementary rules.

9. A computing device, characterized in that The method comprises a memory, a processor and computer instructions stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 7 when executing the computer instructions.

10. A computer-readable storage medium, characterized in that: It stores computer instructions, which, when executed by a processor, implement the steps of the method described in any one of claims 1 to 7.

Images