Method and device for predicting service life of part, processor and storage medium

Abstract

The invention discloses a method and device for predicting the service life of a part, a processor and a storage medium, and belongs to the field of engineering machinery. The method for predicting the service life of the part comprises the following steps: acquiring a first characteristic parameter value of a first preset characteristic parameter of a to-be-predicted part, the first preset characteristic parameters comprise a part model, a working time length, a production time length, a part material, a machine type to which the part belongs, a contact attribute parameter representing whether the part is in direct contact with a foreign object or not, and an environment attribute parameter representing an environment attribute of a use place; and determining the predicted service life of the to-be-predicted part according to the first characteristic parameter value based on a pre-trained first part service life prediction model. The accuracy of predicting the service life of the part can be improved.

Description

Technical Field

[0001] The present application relates to the field of engineering machinery, and in particular to a method, device, processor and storage medium for predicting the life of a part. Background Art

[0002] With the continuous improvement of industrialization and manufacturing technology, the demand for life prediction of mechanical parts of complete machines is also increasing. Since each mechanical part will experience different degrees of wear during use, when a certain mechanical part is worn to a certain extent, it will affect the use of the whole machine, thereby affecting production efficiency. Therefore, in order to avoid affecting the use of the whole machine, it is necessary to replace or process the mechanical parts of the whole machine in a timely manner. In the prior art, usually, technicians determine the average usage time of multiple parts of the same type as the service life of parts of this type. However, the service life of parts determined by the above-mentioned method for predicting the service life of parts in the prior art has the problem of low accuracy. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a method, device, processor and storage medium for predicting the life of a part, so as to solve the problem of low accuracy of the service life of a part determined in the prior art.

[0004] In order to achieve the above-mentioned purpose, a first aspect of an embodiment of the present application provides a method for predicting the life of a component, the method comprising:

[0005] Obtaining a first characteristic parameter value of a first preset characteristic parameter of the part to be predicted, the first preset characteristic parameter including part model, working time, production time, part material, type of machine to which the part belongs, a contact attribute parameter representing whether the part directly contacts an external object, and an environmental attribute parameter representing an environmental attribute of a place of use;

[0006] Based on the pre-trained first part life prediction model, the predicted service life of the part to be predicted is determined according to the first characteristic parameter value.

[0007] In an embodiment of the present application, the training process of the first part life prediction model includes: obtaining a first historical data set of historical parts, the first historical data set including historical data and actual service life corresponding to a first preset characteristic parameter of the historical parts; determining a first training sample set and a first test sample set based on the first historical data set; training the initial part life prediction model according to the first training sample set to obtain a trained model; testing the trained model according to the first test sample set to obtain a model test result of the trained model, the model test result including an error between the actual service life and the predicted service life output by the trained model; when the model test result meets a preset condition, using the trained model as the first part life prediction model, wherein the preset condition is that the error is within a preset error range; when the model test result does not meet the preset condition, continuing to train the initial part life prediction model according to the first training sample set until the model test result of the trained model meets the preset condition, so as to obtain the first part life prediction model, wherein the model parameters of the first part life prediction model include a first coefficient corresponding to the first preset characteristic parameter.

[0008] In an embodiment of the present application, the method also includes: obtaining a second characteristic parameter value of a second preset characteristic parameter of the part to be predicted, wherein the second preset characteristic parameter includes an economic attribute parameter characterizing the economic attributes of the place of use; based on a pre-trained second part life prediction model, determining the service life of the part to be predicted according to the first characteristic parameter value and the second characteristic parameter value.

[0009] In an embodiment of the present application, the training process of the second part life prediction model includes: obtaining a second historical data set of historical parts, the second historical data set including historical data corresponding to a first preset characteristic parameter of the historical parts, historical data corresponding to a second preset characteristic parameter, and an actual service life; determining a second training sample set and a second test sample set based on the second historical data set; training the first part life prediction model based on the second training sample set to obtain a trained model; testing the trained model based on the second test sample set to obtain a model test result of the trained model, the model test result including an error between the actual service life and the predicted service life output by the trained model; when the model test result meets a preset condition, using the trained model as the second part life prediction model, wherein the preset condition is that the error is within a preset error range; when the model test result does not meet the preset condition, retraining the first part life prediction model, and retraining the second part life prediction model based on the first part life prediction model obtained after retraining, wherein the model parameters of the first part life prediction model include a first coefficient corresponding to the first preset characteristic parameter, and the model parameters of the second part life prediction model include the model parameters of the first part life prediction model and a second coefficient corresponding to the second preset characteristic parameter.

[0010] In an embodiment of the present application, the method also includes: determining the difference between the predicted service life of the part to be predicted and the working hours of the part to be predicted to obtain the remaining life of the part to be predicted; when the remaining life of the part to be predicted is lower than a preset life value, obtaining the target machine identification of the machine to which the part to be predicted belongs; determining the target user corresponding to the machine to which the part to be predicted belongs based on the target machine identification and a pre-built information database, wherein the information database stores the correspondence between the machine identification and the user information; and issuing a warning message to the target user, wherein the warning message includes the remaining life of the part to be predicted.

[0011] In an embodiment of the present application, the method also includes data preprocessing, which includes: numerical transformation processing for converting non-numeric variables into numerical variables and data elimination processing, wherein the data elimination processing includes similar feature parameter elimination processing and / or extreme value elimination processing.

[0012] In an embodiment of the present application, the first part life prediction model and / or the second part life prediction model is a linear regression model.

[0013] A second aspect of the present application provides a device for predicting the service life of a part, comprising:

[0014] A parameter acquisition module, used to acquire a first characteristic parameter value of a first preset characteristic parameter of a part to be predicted;

[0015] The life prediction module is used to determine the predicted service life of the part to be predicted based on the pre-trained first part life prediction model and the first characteristic parameter value.

[0016] A third aspect of the present application provides a processor configured to execute the above-mentioned method for predicting the service life of a part.

[0017] The fourth aspect of the present application provides a storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the method for predicting the service life of a part is implemented according to the above-mentioned method.

[0018] The above technical scheme determines the predicted service life of the part to be predicted according to the pre-trained first part life prediction model and the first characteristic parameter value of the first preset characteristic parameter by obtaining the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted, the first preset characteristic parameter may include part model, working time, production time, part material, type of machine to which the part belongs, contact parameters characterizing whether it is in direct contact with external objects, and environmental parameters characterizing the place of use. Compared with the prior art of determining the service life of a part according to the average value of the service life of multiple parts of the same type, the influence of the first preset characteristic parameter on the service life of the part to be predicted can be comprehensively considered, thereby improving the accuracy of the service life prediction of the part to be predicted.

[0019] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present application, but do not constitute a limitation on the embodiments of the present application. In the accompanying drawings:

[0021] Figure 1 A schematic diagram of a process for predicting the service life of a part according to an embodiment of the present application is schematically shown;

[0022] Figure 2 A schematic diagram of a flow chart of a method for predicting the service life of a part according to another embodiment of the present application is shown;

[0023] Figure 3 A schematic diagram of the structure of a device for predicting the service life of a part according to an embodiment of the present application is schematically shown. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the specific implementation methods described herein are only used to illustrate and explain the embodiments of the present application, and are not used to limit the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0025] It should be noted that if the embodiments of the present application involve directional indications (such as up, down, left, right, front, back...), the directional indications are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0026] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.

[0027] Figure 1 The flowchart of a method for predicting the service life of a part according to an embodiment of the present application is schematically shown. Figure 1 As shown, the embodiment of the present application provides a method for predicting the service life of a part. The method is described by taking the application of the method to a processor as an example. The method may include the following steps:

[0028] Step S101, obtaining the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted, the first preset characteristic parameter including part model, working time, production time, part material, type of machine to which the part belongs, contact attribute parameters representing whether it is in direct contact with external objects, and environmental attribute parameters representing the environmental attributes of the place of use.

[0029] It can be understood that the part to be predicted is a part whose service life is to be predicted, the first preset characteristic parameter is a characteristic parameter preset by the user that can affect the service life of the part to be predicted, the first characteristic parameter value is a value used to represent the corresponding state of the first preset characteristic parameter, and the first preset characteristic parameter may include part model, working time, production time, part material, type of machine to which the part belongs, contact attribute parameters characterizing whether it is in direct contact with external objects, and environmental attribute parameters characterizing the environmental attributes of the place of use.

[0030] Among them, the part model is used to indicate the position and usage of the part to be predicted in the machine to which it belongs. Different positions and different usage methods in the machine will affect the service life of the part to be predicted; the working time is used to indicate the length of time the part to be predicted is used. The working time can, to a certain extent, indicate the degree of wear and tear of the part to be predicted, and the degree of wear and tear can indirectly affect the service life of the part to be predicted; the production time is used to indicate the length of time from the time the part to be tested leaves the factory to the present. Due to the particularity of the material of some parts to be predicted, as the production time continues to increase, the material of the parts will change to a certain extent, thus affecting the service life of the parts to be predicted; the part material indicates the main raw materials used to produce the part to be predicted. The service life of the predicted part varies depending on the material of the part; the type of machine to which the part belongs indicates the type of machine to which the predicted part belongs. When the same part is used in different types of machines, its service life is different; the contact attribute parameter that characterizes whether it directly contacts foreign objects is used to indicate whether the predicted part needs to contact foreign objects in addition to the parts of the machine to which it belongs. During the contact with foreign objects, the predicted part will be damaged to a certain extent due to the contact with foreign objects, thereby affecting the service life of the predicted part; the environmental attribute parameter that characterizes the environmental attributes of the place of use is used to indicate that during the use of the predicted part, the climate, working environment and the average technical level of the operators of the place of use will affect the service life of the predicted part.

[0031] In an embodiment of the present application, the processor can obtain the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted, that is, obtain the first characteristic parameter value of the first preset characteristic parameter that affects the service life of the part to be predicted, and determine the influence of the first preset characteristic parameter of the part to be predicted on the service life of the part to be predicted. The service life of the part to be predicted can be further determined based on the first preset characteristic parameter of the part to be predicted, thereby improving the accuracy of the prediction of the service life of the part to be predicted.

[0032] Step 102: Based on the pre-trained first part life prediction model, determine the predicted service life of the part to be predicted according to the first characteristic parameter value.

[0033] It can be understood that the pre-trained first part life prediction model is a model prepared for training for predicting the service life of parts. The input of the pre-trained first part life prediction model is the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted, and the output is the predicted service life of the part to be predicted. The predicted service life is the prediction result of the first part prediction model on the service life of the part to be predicted, and the predicted service life is the predicted usable (workable) life duration of the part to be predicted.

[0034] In an embodiment of the present application, the processor may input the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted into a pre-trained first part life prediction model, that is, in the first part life prediction model, the influence of the first preset characteristic parameter of the part to be predicted on the service life of the part to be predicted may be comprehensively considered. Furthermore, the influence of the first preset characteristic parameter of the part to be predicted on the service life of the part to be predicted may be summarized, so that the predicted service life of the part to be predicted may be obtained, thereby improving the accuracy of the service life prediction of the part to be predicted.

[0035] By obtaining the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted, the first preset characteristic parameter may include part model, working time, production time, part material, type of machine to which the part belongs, contact parameters characterizing whether it is in direct contact with external objects, and environmental parameters characterizing the place of use, that is, comprehensively statistically analyzing the characteristic parameters that affect the service life of the part to be predicted, the first preset characteristic parameter can be obtained. Furthermore, according to the first part life model, the influence of the first preset characteristic parameter on the service life of the part to be predicted caused by the wear of the part to be predicted can be comprehensively considered, so as to accurately determine the service life of the part to be predicted, and further improve the accuracy of the service life prediction of the part to be predicted.

[0036] In an embodiment of the present application, the training process of the first part life prediction model may include: obtaining a first historical data set of historical parts, the first historical data set including historical data and actual service life corresponding to a first preset characteristic parameter of the historical parts; determining a first training sample set and a first test sample set based on the first historical data set; training the initial part life prediction model according to the first training sample set to obtain a trained model; testing the trained model according to the first test sample set to obtain a model test result of the trained model, the model test result including an error between the actual service life and the predicted service life output by the trained model; when the model test result meets a preset condition, using the trained model as the first part life prediction model, wherein the preset condition is that the error is within a preset error range; when the model test result does not meet the preset condition, continuing to train the initial part life prediction model according to the first training sample set until the model test result of the trained model meets the preset condition, so as to obtain the first part life prediction model, wherein the model parameters of the first part life prediction model include a first coefficient corresponding to the first preset characteristic parameter.

[0037] It can be understood that historical parts are multiple parts that have reached their service life, and the actual service life is the length of time the historical parts are used from the beginning to the end of their use. The first historical data set may include historical data and actual service life corresponding to the first preset characteristic parameter of the historical parts. The first historical data set may include a certain proportion of the first training sample set and the first test sample set. The first training sample set is used to train the initial part life prediction model to obtain a trained model, and the first test sample set is used to test the trained model to obtain a first part life prediction model that passes the test; the test is passed when the error between the predicted service life predicted by the first part life prediction model and the actual service life meets the preset conditions, that is, it is determined that the first part life prediction model is successfully trained, that is, it can meet the conditions for passing the test. The initial part life prediction model is a model pre-built by the user to predict the service life of the parts, and the model test result is the test result obtained by inputting the first test sample set into the trained model.

[0038] Specifically, first, the processor obtains the first historical data set of historical parts, which can be divided into a first training sample set and a first test sample set according to a certain ratio. For example, the first historical data set is divided according to 7:3 to obtain a first training sample set of 70% and a first test sample set of 30%. Then, the first training sample set is input into the initial part life prediction model for training, and a trained model can be obtained.

[0039] Secondly, by inputting the first test sample set into the trained model, the predicted service life of multiple historical parts in the first test sample set can be obtained, and the multiple differences between the multiple predicted service lives of the multiple historical parts and the multiple actual service lives corresponding to the multiple historical parts are calculated to obtain the average value of the multiple differences. The average value is the error of the predicted service life output by the model after actual service life training. Therefore, the average value is the model test result of the trained model.

[0040] Finally, it is necessary to determine whether the model test results meet the preset conditions, that is, whether the error between the actual service life and the predicted service life output by the trained model is within the preset error range, and whether the trained model is a first part life prediction model that has been successfully trained and tested. If the model test results meet the preset conditions, that is, the error between the actual service life and the predicted service life output by the trained model is within the preset error range, then the trained model is determined to be a model that has passed the test or successfully trained. It can be considered that the accuracy of the part service life predicted by this model is high, and the trained model can be used as the first part life prediction model; if the model test results do not meet the preset conditions, that is, the error between the actual service life and the predicted service life output by the trained model is not within the preset error range, then the trained model is determined to be a model that has been successfully trained and unqualified in the test. It can be considered that the accuracy of the part service life predicted by this model is low, and the initial part life prediction model needs to be trained until a model that has passed the test and successfully trained is obtained. The trained model is used as the first part life prediction model to obtain the first part life prediction model. It can be seen that training and testing the initial part life prediction model can improve the accuracy of the model, thereby obtaining a first part life prediction model with higher accuracy.

[0041] In a specific embodiment, the first part life prediction model can be expressed by the following formula:

[0042] y(i)=w1*x1+w2*x2+w3*x3+w4*x4+w5*x5+w6*x6+w7*x7+b

[0043] Among them, i is the i-th part to be predicted, y(i) represents the service life of the i-th part to be predicted, x1~x7 can respectively represent the part model, working time, production time, part material, type of machine to which the part belongs, contact attribute parameters representing whether it is in direct contact with external objects, and environmental attribute parameters representing the environmental attributes of the place of use in the first preset characteristic parameters, w1~w7 respectively represent the first coefficients corresponding to the first preset characteristic parameters, and b is the bias coefficient.

[0044] It can be seen from this that the historical data and actual service life corresponding to the first preset characteristic parameters of the historical parts are input into the initial part life prediction model to determine the model parameters of the initial part life prediction model. The model parameters may also include a bias coefficient, that is, the first coefficient and the bias coefficient corresponding to the first preset characteristic parameters in the above formula are determined to obtain the trained model. The trained model is tested using the first test sample set to obtain the first part life prediction model after the model parameters of the initial part life prediction model are determined.

[0045] In an embodiment of the present application, the method for predicting the service life of a part may further include: obtaining a second characteristic parameter value of a second preset characteristic parameter of the part to be predicted, wherein the second preset characteristic parameter includes an economic attribute parameter characterizing the economic attributes of the place of use; based on a pre-trained second part life prediction model, determining the service life of the part to be predicted according to the first characteristic parameter value and the second characteristic parameter value.

[0046] It can be understood that the second preset characteristic parameter is a characteristic parameter other than the first preset characteristic parameter. The second preset characteristic parameter may include characteristic parameters that affect the service life of the part and are different from the first preset characteristic parameter. It can be considered that the second preset characteristic parameter is a newly added characteristic parameter different from the first preset characteristic parameter. For example, the second preset characteristic parameter may include economic attribute parameters that characterize the economic attributes of the place where the part is used. The number of the second preset characteristic parameters may be one or more. The economic attribute parameters that characterize the economic attributes of the place where the part is used are economic attribute parameters used to represent the economic development level of the place where the part is used. The pre-trained second part life prediction model is the second part life prediction model to be trained. The second part life prediction model is a part life prediction model different from the first part life prediction model. The input of the second part life prediction model is the first characteristic parameter value of the first preset characteristic parameter and the second characteristic parameter value of the second preset characteristic parameter, and its output is the service life of the part to be predicted.

[0047] In many cases, other newly added feature parameters besides the first preset feature parameters, namely the second preset feature parameters, will be found. At this time, if the second preset feature parameters are added to the first preset feature parameters for full parameter training, it will consume a considerable computing cost. In order to add the second preset feature parameters to the model parameters for predicting the service life of parts and reduce costs, the first coefficient corresponding to the trained first preset feature parameters can be fixed unchanged, and only the second preset feature parameters are trained to obtain the second coefficient corresponding to the second preset feature parameters, thereby obtaining a second part life prediction model containing the first coefficient corresponding to the first preset feature parameters and the coefficient corresponding to the second preset feature parameters.

[0048] Specifically, the second characteristic parameter value of the second preset characteristic parameter of the part to be predicted is obtained, the second characteristic parameter value of the second preset characteristic parameter is first determined, and the first characteristic parameter value and the second characteristic parameter value of the second preset characteristic parameter are input into a pre-trained second part life prediction model. Secondly, the characteristic parameter values ​​of each characteristic parameter that affects the service life of the part to be predicted are input into the second part life prediction model. On the one hand, the influence of the first preset characteristic parameter on the service life of the part to be predicted is taken into account, and on the other hand, the influence of the newly added second preset characteristic parameter on the service life of the part to be predicted is also taken into account. The factors that may affect the service life of the part to be predicted are further considered from various dimensions, thereby further improving the accuracy of the model in predicting the service life of the part to be predicted.

[0049] In an embodiment of the present application, the training process of the second part life prediction model includes: obtaining a second historical data set of historical parts, the second historical data set including historical data corresponding to a first preset characteristic parameter of the historical parts, historical data corresponding to a second preset characteristic parameter, and an actual service life; determining a second training sample set and a second test sample set based on the second historical data set; training the first part life prediction model based on the second training sample set to obtain a trained model; testing the trained model based on the second test sample set to obtain a model test result of the trained model, the model test result including an error between the actual service life and the predicted service life output by the trained model; when the model test result meets a preset condition, using the trained model as the second part life prediction model, wherein the preset condition is that the error is within a preset error range; when the model test result does not meet the preset condition, retraining the first part life prediction model, and retraining the second part life prediction model based on the first part life prediction model obtained after retraining, wherein the model parameters of the first part life prediction model include a first coefficient corresponding to the first preset characteristic parameter, and the model parameters of the second part life prediction model include the model parameters of the first part life prediction model and a second coefficient corresponding to the second preset characteristic parameter.

[0050] It can be understood that the second historical data set may include historical data corresponding to the first preset characteristic parameters of the historical parts, historical data corresponding to the second preset characteristic parameters, and the actual service life. Compared with the first historical data set, the second historical data set adds historical data corresponding to the second preset characteristic parameters and the actual service life. The second historical data set includes a second training sample set and a second test sample set. The second training sample set is used to train the first part life prediction model to obtain the trained model, and the second test sample set is used to test the trained model to obtain the second part life prediction model.

[0051] Specifically, first, a second historical data set of historical parts is obtained, which can be divided into a second training sample set and a second test sample set according to a certain ratio. For example, the second historical data set is divided according to 7:3 to obtain a first training sample set of 70% and a second test sample set of 30%. Then, the second training sample set is input into the initial part life prediction model for training, and a trained model can be obtained.

[0052] Secondly, by inputting the second test sample set into the trained model, the predicted service life of multiple historical parts in the second test sample set can be obtained, and the multiple differences between the multiple predicted service lives of the multiple historical parts and the multiple actual service lives corresponding to the multiple historical parts are calculated to obtain the average value of the multiple differences. The average value is the error between the actual service life and the predicted service life output by the trained model. Therefore, the average value is the model test result of the trained model.

[0053] Finally, it is necessary to determine whether the model test results meet the preset conditions, that is, whether the error between the actual service life and the predicted service life output by the trained model is within the preset error range, so as to determine whether the trained model is a second part life prediction model that has passed the test and successfully trained. If the model test results meet the preset conditions, that is, the error between the actual service life and the predicted service life output by the trained model is within the preset error range, then the trained model is determined to be a model that has passed the test or successfully trained. It can be considered that the accuracy of the part service life predicted by this model is high, and the trained model can be used as the second part life prediction model; if the model test results do not meet the preset conditions, that is, the error between the actual service life and the predicted service life output by the trained model is not within the preset error range, then the trained model is determined to be a model that has failed the test or failed the training. It can be considered that the accuracy of the part service life predicted by this model is low, and the initial part life prediction model needs to be trained until a model that has passed the test and successfully trained is obtained. The trained model is used as the second part life prediction model to obtain the second part life prediction model. It can be seen that training and testing the initial part life prediction model can improve the accuracy of the model, and a second part life prediction model with higher accuracy can be obtained.

[0054] In a specific embodiment, the second part life prediction model can be expressed by the following formula: y(i)=w1*x1+w2*x2+w3*x3+w4*x4+w5*x5+w6*x6+w7*x7+b+w8*x8+b1

[0055] Among them, i is the i-th part to be predicted, y(i) represents the service life of the i-th part to be predicted, x1~x7 can respectively represent the part model, working time, production time, part material, type of machine to which the part belongs, contact attribute parameters representing whether it is in direct contact with external objects, and environmental attribute parameters representing the environmental attributes of the place of use in the first preset characteristic parameters, x8 is the economic attribute parameter representing the economic attributes of the place of use in the second preset characteristic parameters, w1~w7 respectively represent the first coefficient corresponding to the first preset characteristic parameters, w8 is the second coefficient corresponding to the second preset characteristic parameters, b is the first bias coefficient, and b2 is the second bias coefficient.

[0056] It can be seen from this that the model parameters (w1~w7 and b) of the first part life prediction model are fixed unchanged, so that on the basis of the first part life prediction model, the first part life prediction model is trained to determine the second coefficient and the second bias coefficient corresponding to the second preset characteristic parameter, and then the second part life prediction model can be determined.

[0057] In an embodiment of the present application, the method for predicting the service life of a part may further include: determining the difference between the predicted service life of the part to be predicted and the working hours of the part to be predicted to obtain the remaining service life of the part to be predicted; when the remaining service life of the part to be predicted is lower than a preset service life value, obtaining the target machine identification of the machine to which the part to be predicted belongs; determining the target user corresponding to the machine to which the part to be predicted belongs based on the target machine identification and a pre-built information database, wherein the information database stores the correspondence between the machine identification and the user information; and issuing a warning message to the target user, wherein the warning message includes the remaining service life of the part to be predicted.

[0058] It can be understood that the remaining life of the part to be predicted is the life that the part to be predicted can still be used, the target machine identifier is the identification information of the machine to which the part to be predicted belongs, and the pre-built information database stores the correspondence between the machine identifier and the user information. The correspondence between the machine identifier and the user information may include ownership relationships and lease relationships. The warning information may include the remaining life of the part to be predicted, and may also include the name of the part to be predicted, the target machine identifier of the part to be predicted, and the warning level, etc.

[0059] Specifically, the difference between the predicted service life of the part to be predicted output by the part life prediction model (the first part life prediction model or the second part life prediction model) and the working time of the part to be predicted is calculated, so that the remaining life of the part to be predicted can be determined. If the remaining life of the part to be predicted is lower than the preset life value, it means that the part to be predicted will be scrapped or will soon be unusable. Since the correspondence between the machine identification and the user information is stored in the information database, the target machine identification of the machine to which the part to be predicted belongs can be obtained, so as to know the target user corresponding to the machine to which the part to be predicted belongs, and determine the contact information of the target user, which may include an email address and a telephone number, etc., and information such as the part name and part model of the part to be predicted can also be obtained, and the remaining life is compared with the preset life threshold, and relevant information about the warning level can also be obtained. Finally, a warning template can also be obtained, and relevant information such as the remaining life, part name, part model and warning level of the part to be predicted can be integrated into the warning template to obtain warning information, so as to send the warning information to the target customer in a timely manner according to the contact information of the target user, so that the target customer can handle the part to be predicted in a timely manner, reduce the economic loss caused by the scrapping of parts, and thus ensure the production efficiency to a certain extent.

[0060] In an embodiment of the present application, the method also includes data preprocessing, which includes: numerical transformation processing for converting non-numeric variables into numerical variables and data elimination processing, wherein the data elimination processing includes similar feature parameter elimination processing and / or extreme value elimination processing.

[0061] It can be understood that data preprocessing may include value transformation processing and data elimination processing, and data elimination processing may also include similar feature parameter elimination processing and / or extreme value elimination processing.

[0062] Specifically, before training the parts life prediction model (the first part life prediction model and the second part life prediction model), data preprocessing is performed on the acquired first historical data set and the second historical data set. First, the machine learning sklearn library is used to perform one-hot encoding conversion on the first historical data set and the second historical data set, that is, non-numeric variables are converted into numerical variables to facilitate subsequent analysis and model training.

[0063] In a specific embodiment, the historical data of the historical parts in the first historical data set can be shown in the following Table 1, and the data after the historical data of the historical parts in the first historical data set are numerically transformed can be shown in the following Table 2:

[0064] Table 1

[0065]

[0066] Table 2

[0067]

[0068]

[0069] Secondly, the machine learning sklearn library is used again to perform correlation analysis on the first preset characteristic parameter and the second preset characteristic parameter that affect the service life of the part to be predicted. If there is a correlation degree value of two characteristic parameters that is higher than the correlation threshold preset by the user, it is determined that the similarity of the two characteristic parameters is high, that is, the impact on the service life of the part to be predicted is not much different. Therefore, only one of the two characteristic parameters can be used as the characteristic parameter in the part life prediction model. Thus, the redundant characteristic parameters can be reduced without affecting the prediction of the service life of the part to be predicted by the part life prediction model, thereby reducing the data set of the characteristic parameter, increasing the data storage space, and improving the prediction speed of the part life prediction model.

[0070] Finally, the machine learning sklearn library can also be used to perform extreme value elimination processing on the characteristic parameter values ​​in the first preset characteristic parameters and the second preset characteristic parameters that affect the service life of the part to be predicted, that is, data standardization operation, and eliminate the extreme values ​​in the characteristic parameter values ​​in the first preset characteristic parameters and the second preset characteristic parameters that affect the service life of the part to be predicted, and do not participate in model training, thereby avoiding the influence of individual extreme values ​​on the prediction of the service life of the part to be predicted by the part life prediction model, so as to improve the accuracy of model prediction.

[0071] In an embodiment of the present application, the first part life prediction model and / or the second part life prediction model is a linear regression model.

[0072] It can be understood that the linear regression model is a model built with linear regression as the core idea. Linear regression is one of the basic and important prediction methods in the field of statistics and machine learning. Its core idea is to realize the prediction of unknown data by fitting the linear relationship between the independent variable and the dependent variable. The present application adopts a linear regression model to realize the prediction of the service life of the part to be predicted, wherein the independent variable of the linear regression model is the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted and / or the second characteristic parameter value of the second preset characteristic parameter, and the dependent variable is the service life of the part to be predicted.

[0073] Specifically, the classic machine learning method is used to simulate and predict the service life of parts with a linear regression model. In addition, the least squares method is used as the loss assessment function, and gradient descent is used for parameter replacement training to optimize the linear regression model, thereby improving the accuracy of predicting the service life of the parts to be predicted.

[0074] like Figure 2 As shown, the embodiment of the present application obtains the first historical data set or the second historical data set, and performs data preprocessing on the obtained first historical data set or the second historical data set. The data preprocessing may include numerical transformation processing and data elimination processing. The data transformation processing is to convert the non-numerical variables in the first historical data set or the second historical data set into numerical variables. The data elimination processing may include similar feature parameter elimination processing and / or extreme value elimination processing, that is, to perform correlation analysis on each two first preset feature parameters in the first historical data set. If the correlation between each two first preset feature parameters is not high, it can be considered that the two first preset feature parameters have different effects on the service life of the parts, and the two first preset feature parameters can be retained. If the correlation between each two first preset feature parameters is not high, it can be considered that the two first preset feature parameters have the same effect on the service life of the parts, and one of the two first preset feature parameters can be retained. In addition, if it is determined that there are individual extreme value feature parameters in the first historical data set, the data set corresponding to the feature parameter is deleted. In this way, the first historical data set after data preprocessing can be obtained, and the accuracy of predicting the life of the parts can be avoided due to data factors. The second preset feature parameter in the second historical data set is processed in the same way as above.

[0075] The first part life prediction model can be expressed by the following formula:

[0076] y(i)=w1*x1+w2*x2+w3*x3+w4*x4+w5*x5+w6*x6+w7*x7+b

[0077] Among them, i is the i-th part to be predicted, y(i) represents the service life of the i-th part to be predicted, x1~x7 can respectively represent the part model, working time, production time, part material, type of machine to which the part belongs, contact attribute parameters representing whether it is in direct contact with external objects, and environmental attribute parameters representing the environmental attributes of the place of use in the first preset characteristic parameters, w1~w7 respectively represent the first coefficients corresponding to the first preset characteristic parameters, and b is the first bias coefficient.

[0078] If it is determined that there are no new feature parameters, the first historical data set after data preprocessing is used. The first historical data set can be divided into a first training sample set and a first test sample set in a ratio of 7:3. The first training sample set is used to train the initial part life prediction model, that is, the feature parameter values ​​x1~x7 corresponding to the first preset feature parameters and the actual service life y(i) are input, and the first coefficients (w1~w7) and the first bias coefficient (b) corresponding to the first preset feature parameters can be determined to obtain the trained model. The trained model is tested using the second test sample set to obtain the model test results. When the model test results meet the preset conditions, the model parameters of the first part life prediction model are determined. When the model test results do not meet the preset conditions, the initial part life prediction model needs to be retrained until the model test results meet the preset conditions.

[0079] Furthermore, the second part life prediction model can be expressed by the following formula:

[0080] y(i)=w1*x1+w2*x2+w3*x3+w4*x4+w5*x5+w6*x6+w7*x7+b+w8*x8+b1

[0081] Among them, i is the i-th part to be predicted, y(i) represents the service life of the i-th part to be predicted, x1~x7 can respectively represent the part model, working time, production time, part material, type of machine to which the part belongs, contact attribute parameters representing whether it is in direct contact with external objects, and environmental attribute parameters representing the environmental attributes of the place of use in the first preset characteristic parameters, x8 is the economic attribute parameter representing the economic attributes of the place of use in the second preset characteristic parameters, w1~w7 respectively represent the first coefficient corresponding to the first preset characteristic parameters, w8 is the second coefficient corresponding to the second preset characteristic parameters, b is the first bias coefficient, and b2 is the second bias coefficient.

[0082] If it is determined that there are newly added characteristic parameters, the second historical data set after data preprocessing is used, and the second historical data set can be divided into a second training sample set and a second test sample set in a ratio of 7:3. The second training sample set is used to train the first part life prediction model, that is, the characteristic parameter values ​​x1~x8 of the second preset characteristic parameters and the actual service life y(i) are input, and the second coefficient w8 and the first bias coefficient b1 corresponding to the second preset characteristic parameters can be determined while ensuring that the first coefficients w1~w7 and the first bias coefficient b corresponding to the first preset characteristic parameters in the first part life prediction model remain unchanged, so as to obtain the trained model, and the trained model is tested using the second test sample set to obtain the model test result. When the model test result meets the preset conditions, the model parameters of the second part life prediction model are determined, and the model parameters of the second part life prediction model may include the first coefficient corresponding to the first preset characteristic parameters, the second coefficient corresponding to the second preset characteristic parameters, and the second bias coefficient, so as to determine the second part life prediction model. When the model test result does not meet the preset conditions, the first part life prediction model needs to be retrained until the model test result of the second part life prediction model meets the preset conditions. In summary, the first historical data set or the second historical data set can be used to obtain the first part life prediction model or the second part life prediction model respectively, so as to predict the part to be predicted.

[0083] Furthermore, the remaining service life of the part to be predicted can be obtained based on the difference between the service life of the part to be predicted and its working time. When the remaining service life of the part to be predicted is lower than the preset life value, it means that the part to be predicted is about to be scrapped and needs to be processed in time. The target machine identification of the machine to which the part to be predicted belongs can be obtained, and the target machine identification can be used in a pre-built information database to find the corresponding information of the target user of the machine to which the part to be predicted belongs, and send an early warning message to the user to remind the user to process the part to be predicted.

[0084] In addition, according to the first life prediction model or the second life prediction model proposed in the present application, the service life of the part can be predicted, the service life of the part can be recorded, and the correspondence between the machine identification of the machine to which the predicted part belongs and the user information can be further recorded, so that the correspondence between the machine identification and the user information is stored in the information library. The intended customer can query the machine identification of the machine to which the used part belongs in the information library by grabbing the order, and then find the target user information corresponding to the machine identification, and check the user information with the remaining life of its parts. The intended customer can contact the target user to whom the part belongs through the obtained user information to reach a purchase intention. The intended customer can also sign a relevant reservation agreement with the target user to obtain the required used parts, thereby realizing the rapid transaction of parts.

[0085] like Figure 3 As shown, the embodiment of the present application provides a device 300 for predicting the service life of a part, comprising:

[0086] The parameter acquisition module 301 is used to acquire a first characteristic parameter value of a first preset characteristic parameter of a part to be predicted.

[0087] The life prediction module 302 is used to determine the predicted service life of the part to be predicted based on the pre-trained first part life prediction model and the first characteristic parameter value.

[0088] The above-mentioned device for predicting the service life of a part determines the predicted service life of the part to be predicted according to a pre-trained first part life prediction model and the first characteristic parameter value of the first preset characteristic parameter by obtaining the first characteristic parameter value of the first preset characteristic parameter of the part to be predicted, where the first preset characteristic parameter may include part model, working time, production time, part material, type of machine to which the part belongs, contact parameters characterizing whether it is in direct contact with external objects, and environmental parameters characterizing the place of use. Compared with the prior art of determining the service life of a part according to the average service life of multiple parts of the same type, the device can comprehensively consider the influence of the first preset characteristic parameter on the service life of the part to be predicted, thereby improving the accuracy of the service life prediction of the part to be predicted.

[0089] In one embodiment, the device 300 for predicting the service life of a part may also include a first model training module, which is used to obtain a first historical data set of historical parts, the first historical data set including historical data and actual service life corresponding to a first preset characteristic parameter of the historical parts; determining a first training sample set and a first test sample set based on the first historical data set; training an initial part life prediction model based on the first training sample set to obtain a trained model; testing the trained model based on the first test sample set to obtain a model test result of the trained model, the model test result including an error between the actual service life and the predicted service life output by the trained model; when the model test result meets a preset condition, using the trained model as a first part life prediction model, wherein the preset condition is that the error is within a preset error range; when the model test result does not meet the preset condition, continuing to train the initial part life prediction model based on the first training sample set until the model test result of the trained model meets the preset condition, so as to obtain a first part life prediction model, wherein the model parameters of the first part life prediction model include a first coefficient corresponding to the first preset characteristic parameter.

[0090] In one embodiment, the parameter acquisition module 301 is also used to obtain a second characteristic parameter value of a second preset characteristic parameter of the part to be predicted, wherein the second preset characteristic parameter includes an economic attribute parameter characterizing the economic attributes of the place of use; the life prediction module 302 is also used to determine the service life of the part to be predicted based on the first characteristic parameter value and the second characteristic parameter value based on a pre-trained second part life prediction model.

[0091] In one embodiment, the device 300 for predicting the service life of a part may further include a second model training module, which is used to obtain a second historical data set of historical parts, wherein the second historical data set includes historical data corresponding to the first preset characteristic parameter of the historical part, historical data corresponding to the second preset characteristic parameter, and actual service life; determine a second training sample set and a second test sample set according to the second historical data set; train the first part life prediction model according to the second training sample set to obtain a trained model; test the trained model according to the second test sample set to obtain a model test result of the trained model, wherein the model test result includes the actual service life and the trained model input. the error of the predicted service life; when the model test result meets the preset condition, the trained model is used as the second part life prediction model, wherein the preset condition is that the error is within the preset error range; when the model test result does not meet the preset condition, the first part life prediction model is retrained, and the second part life prediction model is retrained based on the first part life prediction model obtained after the retraining, wherein the model parameters of the first part life prediction model include the first coefficient corresponding to the first preset characteristic parameter, and the model parameters of the second part life prediction model include the model parameters of the first part life prediction model and the second coefficient corresponding to the second preset characteristic parameter.

[0092] In one embodiment, the device 300 for predicting the service life of a part may further include an early warning module, which is used to: determine the difference between the predicted service life of the part to be predicted and the working hours of the part to be predicted to obtain the remaining service life of the part to be predicted; when the remaining service life of the part to be predicted is lower than a preset service life value, obtain the target machine identification of the machine to which the part to be predicted belongs; determine the target user corresponding to the machine to which the part to be predicted belongs based on the target machine identification and a pre-built information library, wherein the information library stores the correspondence between the machine identification and the user information; and send an early warning message to the target user, wherein the early warning message includes the remaining service life of the part to be predicted.

[0093] In one embodiment, the device 300 for predicting the service life of parts may also include a data preprocessing module for numerical transformation processing and data elimination processing for converting non-numeric variables into numerical variables, wherein the data elimination processing includes similar characteristic parameter elimination processing and / or extreme value elimination processing.

[0094] In one embodiment, the first part life prediction model and / or the second part life prediction model is a linear regression model.

[0095] An embodiment of the present application provides a processor configured to execute the method for predicting the service life of a part according to the above-mentioned implementation.

[0096] An embodiment of the present application also provides a storage medium having instructions stored thereon, the instructions being used to enable a machine to execute the above-mentioned method for predicting the service life of a part.

[0097] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0098] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0099] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0101] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0102] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0103] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0104] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0105] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for predicting the service life of a component, characterized in that: The method comprises: Obtaining a first characteristic parameter value of a first preset characteristic parameter of the part to be predicted, wherein the first preset characteristic parameter includes a part model, working time, production time, part material, type of machine to which the part belongs, a contact attribute parameter representing whether the part directly contacts an external object, and an environmental attribute parameter representing an environmental attribute of a place of use; Based on a pre-trained first part life prediction model, the predicted service life of the part to be predicted is determined according to the first characteristic parameter value.

2. The method according to claim 1, characterized in that The training process of the first part life prediction model includes: Acquire a first historical data set of a historical part, wherein the first historical data set includes historical data corresponding to the first preset characteristic parameter of the historical part and an actual service life; Determine a first training sample set and a first test sample set according to the first historical data set; Training the initial part life prediction model according to the first training sample set to obtain a trained model; Testing the trained model according to the first test sample set to obtain a model test result of the trained model, wherein the model test result includes an error between the actual service life and a predicted service life output by the trained model; In the case where the model test result meets a preset condition, using the trained model as the first part life prediction model, wherein the preset condition is that the error is within a preset error range; When the model test result does not meet the preset condition, continue to train the initial part life prediction model according to the first training sample set until the model test result of the trained model meets the preset condition, so as to obtain the first part life prediction model, wherein the model parameters of the first part life prediction model include the first coefficient corresponding to the first preset characteristic parameter.

3. The method according to claim 1, characterized in that The method further comprises: Obtaining a second characteristic parameter value of a second preset characteristic parameter of the part to be predicted, wherein the second preset characteristic parameter includes an economic attribute parameter characterizing an economic attribute of a place of use; Based on the pre-trained second part life prediction model, the service life of the part to be predicted is determined according to the first characteristic parameter value and the second characteristic parameter value.

4. The method according to claim 3, characterized in that: The training process of the second part life prediction model includes: Acquire a second historical data set of a historical part, wherein the second historical data set includes historical data corresponding to the first preset characteristic parameter, historical data corresponding to the second preset characteristic parameter, and an actual service life of the historical part; Determine a second training sample set and a second test sample set according to the second historical data set; Training the first part life prediction model according to the second training sample set to obtain a trained model; Testing the trained model according to the second test sample set to obtain a model test result of the trained model, wherein the model test result includes an error between the actual service life and a predicted service life output by the trained model; In the case where the model test result meets a preset condition, using the trained model as the second part life prediction model, wherein the preset condition is that the error is within a preset error range; When the model test result does not meet the preset conditions, the first part life prediction model is retrained, and the second part life prediction model is retrained based on the retrained first part life prediction model, wherein the model parameters of the first part life prediction model include the first coefficient corresponding to the first preset characteristic parameter, and the model parameters of the second part life prediction model include the model parameters of the first part life prediction model and the second coefficient corresponding to the second preset characteristic parameter.

5. The method according to claim 1, characterized in that The method further comprises: Determine the difference between the predicted service life of the part to be predicted and the working time of the part to be predicted to obtain the remaining service life of the part to be predicted; When the remaining life of the part to be predicted is lower than a preset life value, obtaining a target machine identifier of the machine to which the part to be predicted belongs; Determine the target user corresponding to the machine to which the part to be predicted belongs according to the target machine identifier and a pre-built information database, wherein the information database stores the correspondence between the machine identifier and the user information; A warning message is issued to the target user, wherein the warning message includes the remaining life of the part to be predicted.

6. The method according to claim 2 or 4, characterized in that: The method further includes data preprocessing, which includes: Numerical transformation processing and data elimination processing for converting non-numeric variables into numerical variables, wherein the data elimination processing includes similar feature parameter elimination processing and / or extreme value elimination processing.

7. The method according to claim 3, characterized in that The first part life prediction model and / or the second part life prediction model is a linear regression model.

8. A device for predicting the service life of a part, characterized in that: The device comprises: A parameter acquisition module, used to acquire a first characteristic parameter value of a first preset characteristic parameter of a part to be predicted; The life prediction module is used to determine the predicted service life of the part to be predicted based on the pre-trained first part life prediction model and the first characteristic parameter value.

9. A processor, characterized in that: The method is configured to execute the method for predicting the service life of a component as claimed in any one of claims 1 to 7.

10. A storage medium having instructions stored thereon, characterized in that: The instructions are used to cause the machine to execute the method for predicting the service life of a component as described in any one of claims 1 to 7.

Images