Method, device, electronic equipment and medium for determining the cause of frequent engine failures
By inputting engine failure data into pre-trained models, determining the cause of high-incidence engine failure and its probability, the problem of inaccurate diagnosis in the prior art is solved, and the success rate and accuracy of engine failure repair are improved.
Patent Information
- Application Number
- CN202210657654.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-10
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2042-06-10
AI Technical Summary
The prior art is difficult to accurately diagnose the causes of high-incidence engine failures, resulting in a low success rate of engine failure repair, and limited maintenance data and deviations from the mechanism.
By obtaining the current dispatch order and engine sales files, enter the pre-trained engine high-incidence fault cause determination model, output the fault cause and its probability, and determine the target fault cause and component troubleshooting information based on this information.
It realizes more accurate fault diagnosis based on market maintenance experience and engine mechanism, improves the accuracy and operability of determining the cause of high-incidence engine failures, and improves the success rate of engine failure repair.
Smart Images

Figure CN115077906B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of engine fault diagnosis, and in particular to a method, device, electronic equipment and medium for determining a cause of a common engine fault. Background Art
[0002] There are relevant studies on engine fault diagnosis in various directions of industry, academia and research, and in-depth research has been conducted on engine fault diagnosis technology from both the mechanism and data aspects. From the perspective of after-sales service of corporate products, companies have accumulated a large amount of historical after-sales maintenance data. The maintenance work orders contain many years of actual maintenance experience from maintenance engineers, and the value of millions of historical maintenance data is yet to be mined.
[0003] Its data-based fault diagnosis mainly uses sensor data and is applied to the fault diagnosis of rotating machinery that is convenient for installing vibration sensors. However, in actual scenarios, many faults are difficult to diagnose through sensor signal collection and are costly. The research results of the academic community are difficult to implement and remain only at the theoretical level. At the same time, the maintenance experience of maintenance engineers at various service stations varies, misjudgment occurs, and the one-time success rate of fault repair needs to be improved. Summary of the invention
[0004] The present invention provides a method, device, electronic device and medium for determining the cause of common engine failures, so as to solve the problem that the actual engine failure maintenance data on the market is limited and deviates from the engine mechanism, making it difficult to accurately diagnose the cause of common engine failures, thereby leading to a low success rate in engine failure maintenance.
[0005] According to one aspect of the present invention, a method for determining a cause of a common engine failure is provided, the method comprising:
[0006] Obtaining a current work order and a current engine sales file, and determining current engine basic information and current engine fault information based on the current work order;
[0007] Input the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and output the current engine high-incidence fault cause and the corresponding fault probability;
[0008] The target engine's frequent failure causes are determined based on the current engine's frequent failure causes and the corresponding failure probabilities, and corresponding component fault troubleshooting information is determined based on the target engine's frequent failure causes.
[0009] Optionally, before inputting the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, the method further includes:
[0010] Preprocessing the current engine fault information, the current engine basic information and the current engine sales file, and performing feature extraction, type feature coding and dimensionless processing on the preprocessed current engine fault information, the current engine basic information and the current engine sales file;
[0011] The preprocessing includes standardizing data types, processing outliers, and completing missing values.
[0012] Optionally, the method for determining the cause of frequent engine failures further includes:
[0013] Preprocess at least one set of historical work orders, historical engine basic information, historical engine maintenance reports, and historical engine sales files to obtain a model building data set;
[0014] Inputting the model building data set into a pre-established engine high-incidence failure cause determination model to obtain historical engine high-incidence failure causes and corresponding failure probabilities;
[0015] Model parameters of the engine frequent failure cause determination model are retrained based on the historical engine frequent failure causes and corresponding failure probabilities and expected engine frequent failure causes and corresponding failure probabilities.
[0016] Optionally, the method for determining the cause of frequent engine failures further includes:
[0017] Obtaining a current engine maintenance report, and storing the current engine fault information, the current engine basic information, and the current engine sales file into the model building data set;
[0018] The model parameters of the engine frequent failure cause determination model are retrained according to the updated model building data set.
[0019] Optionally, before inputting the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, the method further includes:
[0020] If the current engine fault information does not contain fault description information, a missing necessary parameter error instruction is generated, wherein the missing necessary parameter instruction is used to indicate that the current engine fault information has an error; or,
[0021] If it is detected that the fault description information included in the current engine fault information does not belong to the preset fault range, a missing fault description error instruction is generated, and the missing fault description error instruction is used to prompt that the fault description information does not belong to the preset fault range.
[0022] Optionally, determining the target engine's high-incidence failure cause according to the current engine's high-incidence failure cause and the corresponding failure probability includes:
[0023] The current engine high-incidence failure cause parts are sorted according to the failure probability, and a set number of engine high-incidence failure cause parts are selected in sequence as target engine high-incidence failure cause parts.
[0024] Optionally, the method for determining the cause of frequent engine failures further includes:
[0025] Obtaining the failure probability corresponding to the high-incidence failure-causing component of the target engine;
[0026] The target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are fed back to a service terminal, so that the target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are displayed on the service terminal.
[0027] According to another aspect of the present invention, a device for determining a cause of a frequent engine failure is provided, the device comprising:
[0028] An information acquisition module, used to acquire the current work order and the current engine sales file, and determine the current engine basic information and the current engine fault information according to the current work order;
[0029] a cause component output module, used to input the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause component determination model, and output the current engine high-incidence fault cause components and corresponding fault probabilities;
[0030] The fault troubleshooting module is used to determine the target engine's frequent fault causes according to the current engine's frequent fault causes and the corresponding fault probabilities, and to determine corresponding component fault troubleshooting information according to the target engine's frequent fault causes.
[0031] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:
[0032] at least one processor; and
[0033] a memory communicatively connected to the at least one processor; wherein,
[0034] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for determining the cause of frequent engine failures described in any embodiment of the present invention.
[0035] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for determining the cause of frequent engine failures described in any embodiment of the present invention when executed.
[0036] The technical solution of the embodiment of the present invention is to obtain the current work order, the current engine maintenance report and the current engine sales file, and determine the current engine basic information and the current engine fault information according to the current work order; input the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and output the current engine high-incidence fault cause and the corresponding fault probability; determine the target engine high-incidence fault cause according to the current engine high-incidence fault cause and the corresponding fault probability, and determine the corresponding component fault troubleshooting information according to the target engine high-incidence fault cause. The invention solves the problem that the actual engine failure maintenance data in the market is limited and deviates from the engine mechanism, making it difficult to accurately diagnose the causes of common engine failures, which in turn leads to a low success rate in engine failure maintenance. The invention forms a more accurate model for determining the causes of common engine failures based on market maintenance experience and engine mechanism, so as to efficiently and accurately diagnose the causes of common engine failures. At the same time, the invention has business interpretability for the accuracy of the model for determining the causes of common engine failures, and the prediction results of the model for determining the causes of common engine failures have the beneficial effect of higher operability.
[0037] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0039] Figure 1 is a flow chart of a method for determining a cause of a high-incidence engine failure provided in accordance with a first embodiment of the present invention;
[0040] Figure 2 is a schematic diagram of processing a wide meter applicable to Embodiment 1 of the present invention;
[0041] Figure 3 is a flow chart of a method for determining a cause of a common engine failure provided in accordance with a second embodiment of the present invention;
[0042] Figure 4 This is a schematic diagram of a specific result implementation of a variant one-hot encoding provided according to an embodiment of the present invention;
[0043] Figure 5 is a schematic diagram of a specific implementation result of label encoding provided according to an embodiment of the present invention;
[0044] Figure 6 2 is a schematic diagram of the structure of a device for determining a cause of a common engine failure provided by Embodiment 3 of the present invention;
[0045] Figure 7 It is a schematic diagram of the structure of an electronic device for implementing the method for determining the cause of frequent engine failures according to an embodiment of the present invention. DETAILED DESCRIPTION
[0046] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.
[0047] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0048] Embodiment 1
[0049] Figure 1 A flowchart of a method for determining the cause of frequent engine failures is provided for the first embodiment of the present invention. This embodiment can be applied to the situation where a large amount of historical engine maintenance data is used to fully explore the characteristics of engine failure influencing factors in the maintenance data to perform overall modeling and efficient and accurate diagnosis of the cause of frequent engine failures. The method for determining the cause of frequent engine failures can be performed by a device for determining the cause of frequent engine failures. The device for determining the cause of frequent engine failures can be implemented in the form of hardware and / or software, and the device for determining the cause of frequent engine failures can be configured in an electronic device. Figure 1 As shown, the method for determining the cause of frequent engine failures includes:
[0050] S110, obtaining a current work order and a current engine sales file, and determining current engine basic information and current engine fault information according to the current work order.
[0051] The work order is a data table generated when a user reports a fault in an engine, which is used to record the fault description, fault location and other fault conditions, as well as basic engine information such as the engine number. In this embodiment, the current work order is the work order for the current engine, and the current engine basic information and current engine fault information can be obtained according to the current work order.
[0052] The engine fault description is a standard description selected by the user based on the engine fault symptoms he or she experiences when reporting a problem with the engine.
[0053] The engine sales file is a file of recording engine sales information generated when the engine is first compulsorily maintained. In the present embodiment, the current engine sales file is the engine sales file of the engine at that time.
[0054] In this embodiment, in order to associate the engine maintenance report, the current engine sales file, the current engine basic information and the current engine fault information, a wide table for further analysis and processing is generated. Figure 2 Schematic diagram of processing a wide table provided in an embodiment of the present invention. Figure 2 As shown, the engine maintenance report may include information such as the work order number, engine number and accessory name, the engine sales file may include information such as the engine number, production date and equipment purpose, the basic engine information may include information such as the engine number, engine series and manufacturer, and the work order may include information such as the work order number, fault description and fault date. Furthermore, it can be known that according to the business logic, the variables that can be used to construct the characteristics of the model for training the causes of high-incidence engine failures can be specifically selected, which may include information such as fault description, engine series, engine model, sales date, fault date, manufacturer, and OEM.
[0055] It should be noted that the cleaning logic of the maintenance report is: select the maintenance report with the process approval status of "accounted", "to be accounted" or "accounting" as the current engine maintenance report; select the sales file with the status of "expired" or "valid" as the current engine sales file; if the fault description does not contain maintenance and rectification related fields, the logic of sales files, work orders, and engine basic information processing is the maintenance report. It should also be noted that in the case of multiple engine repairs, if the same engine with the same fault phenomenon is repaired multiple times within 15 days, only the most recent date is taken as the fault date.
[0056] For details, see Figure 2 , with the maintenance report as the main table, the work order number as the associated field, associated with the work order, and the engine number as the associated engine basic information and sales files.
[0057] S120, inputting the current engine fault information, the current engine basic information, and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and outputting the current engine high-incidence fault cause and the corresponding fault probability.
[0058] It can be understood that before the current engine fault information, the current engine basic information and the current engine sales file are input into the pre-trained engine high-incidence fault cause determination model, it also includes: preprocessing the current engine fault information, the current engine basic information and the current engine sales file, and extracting features, encoding type features and dimensionless processing for the preprocessed current engine fault information, the current engine basic information, the current engine maintenance report and the current engine sales file; wherein the preprocessing includes standardizing data types, processing outliers and completing missing values.
[0059] It should be noted that when performing preprocessing, it is necessary to judge whether the current engine fault information, the current engine basic information and the current engine sales file have problems with whether the data type is standardized, whether there are abnormal values and missing values, and then perform corresponding preprocessing operations on the existing problems. When the current engine fault information, the current engine basic information and the current engine sales file do not have problems with irregular data types, abnormal values and missing values, there is no need to perform preprocessing operations on the current engine fault information, the current engine basic information and the current engine sales file.
[0060] In this embodiment, the model for determining the causes of common engine failures can adopt a classification model, for example, an SVM (Support Vector Machine) model, an XGBoost model (Extreme GradientBoosting), a deep learning model, etc. The deep learning model can be a deep neural network DNN model, a convolutional neural network CNN model, a BP (Back Propagation) neural network BPNN model, a probabilistic neural network PNN model, or a machine learning model. Optionally, the model for determining the causes of common engine failures adopts a random forest model.
[0061] On the basis of the above, before the current engine fault information, the current engine basic information and the current engine sales file are input into the pre-trained engine high-incidence fault cause determination model, the engine high-incidence fault cause determination model needs to be trained, specifically including: pre-processing at least one set of historical work orders, historical engine basic information, historical engine maintenance reports and historical engine sales files to obtain a model building data set; inputting the model building data set into the pre-established engine high-incidence fault cause determination model to obtain historical engine high-incidence fault causes and corresponding fault probabilities; adjusting the model parameters of the engine high-incidence fault cause determination model based on the historical engine high-incidence fault causes and corresponding fault probabilities and the expected engine high-incidence fault causes and corresponding fault probabilities.
[0062] The engine maintenance report is a data sheet that records engine failures after the engine is repaired, and includes information such as the engine maintenance process and results, and the cause of the failure. In this embodiment, the historical engine maintenance report is an engine maintenance report generated by the engine in the past.
[0063] It is understandable that R&D engineers and service-related personnel regularly monitor data quality, reduce misjudgments in after-sales maintenance, ensure the quality of maintenance reports and work order data, explore new engine series, models and market feedback, and promptly update the combination table of common faults and causal parts, that is, update the model for determining the causal parts of common engine failures.
[0064] In one embodiment, the model for determining the causes of common engine failures is updated at any time, specifically: the current engine failure information, the current engine basic information and the current engine sales file are stored in the model building data set; and the model parameters of the model for determining the causes of common engine failures are adjusted according to the updated model building data set.
[0065] In one embodiment, the model for determining the cause of common engine failures is updated regularly, that is, the accuracy of the model for determining the cause of common engine failures is continuously monitored, and error records are analyzed. The model for determining the cause of common engine failures can be regularly, optionally, every 2 months, selected from the data 2 years before the current time point as a training model building data set, and the model for determining the cause of common engine failures can be retrained, and a new model for determining the cause of common engine failures can be deployed to recommend a method for troubleshooting the cause of common engine failures.
[0066] The engine common fault troubleshooting solutions accumulated from market maintenance experience are mainly expressed in the form of text descriptions, some of which are accompanied by pictures, but the existing market maintenance experience cannot be passed on to more maintenance engineers. In addition, due to the current shortage of maintenance personnel in society, the after-sales maintenance market regulations need to be continuously improved. If the engine fault description terminology standards are not unified, it will increase the difficulty of maintenance and increase the manpower input of maintenance personnel. In this embodiment, in order to solve this problem, before the current engine fault information, the current engine basic information, the current engine maintenance report and the current engine sales file are trained on the model, it is also necessary to standardize the fault description, and before inputting into the pre-trained engine common fault cause determination model, determine whether there is fault description information in the current engine fault information, and determine whether the existing fault description is within the pre-stored fault range.
[0067] Specifically, if there is no fault description information in the current engine fault information, a missing necessary parameter error instruction is generated, and the missing necessary parameter error instruction is used to prompt that there is an error in the current engine fault information; or, if it is detected that the fault description information contained in the current engine fault information does not belong to the preset fault range, a missing fault description error instruction is generated, and the missing fault description error instruction is used to prompt that the fault description information does not belong to the preset fault range.
[0068] S130, determining a target engine high-incidence failure cause according to the current engine high-incidence failure cause and the corresponding failure probability, and determining corresponding component fault troubleshooting information according to the target engine high-incidence failure cause.
[0069] Specifically, the current engine high-incidence failure cause parts are sorted according to the failure probability, and a set number of engine high-incidence failure cause parts are selected in sequence as target engine high-incidence failure cause parts.
[0070] The number of target engine high-incidence failure cause parts can be selected and set according to actual needs, and this embodiment does not impose any restrictions on this. Optionally, three engine high-incidence failure cause parts are selected in sequence as target engine high-incidence failure cause parts, and further, three engine high-incidence failure cause parts with the highest failure probability corresponding to the current engine high-incidence failure cause parts are selected as target engine high-incidence failure cause parts.
[0071] In order to provide maintenance technicians with maintenance guidance and suggestions, and to be able to view them at any time through a service terminal, in this embodiment, the failure probability corresponding to the target engine's frequent failure cause is obtained, and then the target engine's frequent failure cause, the corresponding failure probability and the component failure troubleshooting information are fed back to the service terminal, so that the target engine's frequent failure cause, the corresponding failure probability and the component failure troubleshooting information are displayed on the service terminal.
[0072] The technical solution of the embodiment of the present invention is to obtain the current work order and the current engine sales file, and determine the current engine basic information and the current engine fault information according to the current work order; input the current engine fault information, the current engine basic information, the current engine maintenance report and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and output the current engine high-incidence fault cause and the corresponding fault probability; determine the target engine high-incidence fault cause according to the current engine high-incidence fault cause and the corresponding fault probability, and determine the corresponding component fault troubleshooting information according to the target engine high-incidence fault cause. The invention solves the problem that the actual engine failure maintenance data in the market is limited and deviates from the engine mechanism, making it difficult to accurately diagnose the causes of common engine failures, which in turn leads to a low success rate in engine failure maintenance. A more accurate model for determining the causes of common engine failures is formed based on market maintenance experience and engine mechanism, so as to efficiently and accurately diagnose the causes of common engine failures. At the same time, the business explainability of the accuracy of the model for determining the causes of common engine failures is achieved, and the prediction results of the model for determining the causes of common engine failures are more operational.
[0073] Embodiment 2
[0074] Figure 3 This is a flow chart of a method for determining a cause of a high-incidence engine failure provided in Embodiment 2 of the present invention, which provides an optional implementation method based on the above embodiment. Figure 2 As shown, the method for determining the cause of frequent engine failures includes:
[0075] S310, obtaining a current work order and a current engine sales file, and determining current engine basic information and current engine fault information according to the current work order.
[0076] S320, preprocessing the current engine fault information, the current engine basic information and the current engine sales file, and performing feature extraction, type feature coding and dimensionless processing on the preprocessed current engine fault information, the current engine basic information, the current engine maintenance report and the current engine sales file.
[0077] The preprocessing includes standardizing data types, processing outliers, and completing missing values.
[0078] Among them, in this embodiment, the engine fault descriptions recorded in the work order are standardized, and according to the frequency of occurrence of the engine fault descriptions, the high-incidence engine faults are counted, and the causal parts with lower occurrence frequencies are filtered. Optionally, the TOP30 high-incidence faults to be repaired and their corresponding causal part combinations (many-to-many) are exported. Each engine fault description may be a fault condition manifested by the same causal part fault caused by different causal parts under different circumstances.
[0079] R&D engineers screen engine fault diagnosis mechanisms, common engine faults and their corresponding causal parts, eliminate unreasonable engine fault maintenance plans based on engine mechanism inferences, and form a screened combination table of common engine faults and causal parts. Based on this table, they screen engine maintenance reports, that is, pre-process current engine fault information, current engine basic information, current engine maintenance report and current engine sales file.
[0080] Normalizing the data types is done by removing the symbols present in the mileage and time columns and converting all values in both columns to numbers.
[0081] Outlier processing is to correct the sales date to the failure date for records where the engine's sales date is later than the failure date, and delete the production date and treat it as a missing value for records where the production date is later than the failure date.
[0082] Missing value completion is to count the proportion of missing values in each column and each row in the above information, and use statistical values such as median and mode to complete the missing values according to the proportion of missing values and feature types.
[0083] It should be noted that preprocessing operations are performed on the current engine fault information, the current engine basic information, the current engine maintenance report and the current engine sales file, but it is not limited to completing each preprocessing operation, and preprocessing can be performed according to actual needs.
[0084] S330, inputting the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and outputting the current engine high-incidence fault cause and the corresponding fault probability.
[0085] It can be understood that before the current engine fault information, the current engine basic information and the current engine sales file are input into the pre-trained engine high-incidence fault cause determination model, the engine high-incidence fault cause determination model is trained, specifically including: pre-processing at least one set of historical work orders, historical engine basic information, historical engine maintenance reports and historical engine sales files to obtain a model building data set; inputting the model building data set into the pre-established engine high-incidence fault cause determination model to obtain historical engine high-incidence fault causes and corresponding fault probabilities; and adjusting the model parameters of the engine high-incidence fault cause determination model based on the historical engine high-incidence fault causes and corresponding fault probabilities and the expected engine high-incidence fault causes and corresponding fault probabilities.
[0086] In this embodiment, before training the model for determining the causes of common engine failures, it is also necessary to construct the current engine failure information, the current engine basic information and the current engine sales file as features according to the business logic, and perform type feature encoding and variable dedimensionalization to facilitate the subsequent generation of a data set for model building.
[0087] Specifically, the difference between the failure date and the sales date is extracted as the engine running time, the production date is treated as a type variable, and the impact of different production conditions is characterized to complete feature construction.
[0088] The encoding method of type features needs to consider the business scenario and the structure of the subsequent model. In the dispatch maintenance scenario, a single work order has one to four fault descriptions (fault description 1 to fault description 4). For the case where there are multiple fault descriptions in the same record, all fault descriptions come from a unified list. The same fault description, no matter which fault description it appears in, currently tries two encoding methods: one is a variant one-hot encoding, which uses the unique fault description that appears in all fault description related columns as the vocabulary. When the number of fault descriptions in the work order is greater than 1, the encoding will be set to 1 when the number of fault descriptions is the same as the number of fault descriptions. The specific implementation result diagram is shown in Figure 4 , other types of features are concatenated after conventional one-hot encoding; another way is to generate new features from all fault description columns, unordered fault description word combinations, and then label encode them together with other types of features. The specific implementation result of label encoding is shown in Figure 5 .
[0089] It should be noted that type feature encoding can also perform embedding learning (similar to word2vec, matrix factorization) based on the improved one-hot encoding.
[0090] The magnitude of the type feature variable and the quantity feature after dimensionless processing is quite different. The magnitude of the type feature after type feature encoding is smaller. After using the variant one-hot encoding, the corresponding virtual variables are all 0 or 1. After using the second encoding method, the magnitude of the corresponding virtual variables is hundreds, while the mileage feature value can reach hundreds of thousands. The difference in magnitude will have an adverse effect on the convergence of subsequent model training. Here, a standardized method is used to handle the difference in magnitude. The specific formula is as follows:
[0091]
[0092] Among them, x′ ij is the normalized value of the current feature of the current record; x ih The actual value of the current feature for the current record; is the average value of the current feature; n is the number of records; i is the current number of records; j represents the current feature.
[0093] Furthermore, the causal factors are labeled and used as target variables for model training.
[0094] Exemplarily, taking the implementation of the model for determining the cause of frequent engine failures using a random forest model as an example, the data in the model building data set is randomly extracted in stratified layers according to the cause of the work order number, of which 70% is used as a training set and 30% is used as a test set. The preprocessing operations are first calculated on the training set, and the results are used to test on the test set. In the case where there are multiple cause parts in the same engine maintenance report, each cause part is a record line, and the accuracy of the multiple cause parts business is determined as long as the recommended cause part is any cause part in the maintenance report. For the data of the TOP30 frequent engine failures, based on the performance of the model for determining the cause of frequent engine failures on the test set (the performance measurement indicators of the model for determining the cause of frequent engine failures include recall, precision, F-1 and the custom indicator L1), the random forest model has the highest accuracy, and the random forest algorithm can be selected for modeling.
[0095] The main hyperparameters in the random forest algorithm are the number of trees n_estimators, the minimum number of samples required for leaf nodes min_samples_leaf, the maximum depth max_depth, etc. The hyperparameters are adjusted to check their performance on the validation set. After selecting the best performing hyperparameter combination, training is performed on the entire data set (combined training set, validation set, and test set) to generate the final engine high-incidence failure cause determination model.
[0096] On the basis of the above, the current engine fault information, the current engine basic information and the current engine sales file can also be stored in the model building data set, and the model parameters of the engine common fault cause determination model can be adjusted according to the updated model building data set to make real-time adjustments to the engine common fault cause determination model.
[0097] In one embodiment, the model for determining the cause of common engine failures is regularly maintained and updated, that is, within the verification period of the model for determining the cause of common engine failures, optionally three months is selected as the model verification period, and the accuracy of the cause of common engine failures determination model is monitored. The accuracy of the model for determining the cause of common engine failures continues to reach more than 85% according to the custom evaluation index L1, and is integrated into the service terminal for receiving work dispatches used by maintenance technicians, and is continuously monitored.
[0098] S340, sorting the current engine high-incidence failure cause parts according to the failure probability, and sequentially selecting a set number of engine high-incidence failure cause parts as target engine high-incidence failure cause parts.
[0099] S350: Obtain the failure probability corresponding to the frequently occurring failure-causing component of the target engine.
[0100] S360, feeding back the target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information to a service terminal, so that the target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are displayed on the service terminal.
[0101] The technical solution of the embodiment of the present invention realizes the construction of a model for determining the cause of common engine failures based on fault information such as fault description and basic engine information, incorporates the engine mechanism knowledge of R&D engineers into data processing and model construction, realizes engine fault diagnosis combining engine mechanism with market feedback, and in view of the situation where there are multiple engine fault descriptions, modifies the category variable encoding algorithm for the characteristic information of engine fault descriptions in different combinations, constructs an accuracy index based on the work order for the phenomenon of multiple cause parts, and realizes a business explanation of the accuracy of the model for determining the cause of common engine failures.
[0102] Embodiment 3
[0103] Figure 6 This is a schematic diagram of the structure of a device for determining the cause of frequent engine failures provided by the third embodiment of the present invention. Figure 6 As shown, the device for determining the cause of frequent engine failures includes:
[0104] The information acquisition module 610 is used to acquire the current work order and the current engine sales file, and determine the current engine basic information and the current engine fault information according to the current work order;
[0105] The cause output module 620 is used to input the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and output the current engine high-incidence fault cause and the corresponding fault probability;
[0106] The fault troubleshooting module 630 is used to determine the target engine's frequent fault causes based on the current engine's frequent fault causes and the corresponding fault probabilities, and to determine corresponding component fault troubleshooting information based on the target engine's frequent fault causes.
[0107] Optionally, the device for determining the cause of frequent engine failures further includes:
[0108] Preprocessing the current engine fault information, the current engine basic information and the current engine sales file, and performing feature extraction, type feature coding and dimensionless processing on the preprocessed current engine fault information, the current engine basic information, the current engine maintenance report and the current engine sales file;
[0109] The preprocessing includes standardizing data types, processing outliers, and completing missing values.
[0110] Optionally, the device for determining the cause of frequent engine failures further includes:
[0111] Preprocess at least one set of historical work orders, historical engine basic information, historical engine maintenance reports, and historical engine sales files to obtain a model building data set;
[0112] Inputting the model building data set into a pre-established engine high-incidence failure cause determination model to obtain historical engine high-incidence failure causes and corresponding failure probabilities;
[0113] Model parameters of the engine frequent failure cause determination model are retrained based on the historical engine frequent failure causes and corresponding failure probabilities and expected engine frequent failure causes and corresponding failure probabilities.
[0114] Optionally, the device for determining the cause of frequent engine failures further includes:
[0115] Storing the current engine fault information, the current engine basic information and the current engine sales file into the model building data set;
[0116] The model parameters of the engine frequent failure cause determination model are retrained according to the updated model building data set.
[0117] Optionally, the device for determining the cause of frequent engine failures further includes:
[0118] If the current engine fault information does not contain fault description information, a missing necessary parameter error instruction is generated, wherein the missing necessary parameter instruction is used to indicate that the current engine fault information has an error; or,
[0119] If it is detected that the fault description information included in the current engine fault information does not belong to the preset fault range, a missing fault description error instruction is generated, and the missing fault description error instruction is used to prompt that the fault description information does not belong to the preset fault range.
[0120] Optionally, determining the target engine's high-incidence failure cause according to the current engine's high-incidence failure cause and the corresponding failure probability includes:
[0121] The current engine high-incidence failure cause parts are sorted according to the failure probability, and a set number of engine high-incidence failure cause parts are selected in sequence as target engine high-incidence failure cause parts.
[0122] Optionally, the device for determining the cause of frequent engine failures further includes:
[0123] Obtaining the failure probability corresponding to the high-incidence failure-causing component of the target engine;
[0124] The target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are fed back to a service terminal, so that the target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are displayed on the service terminal.
[0125] The device for determining the cause of frequent engine failures provided in an embodiment of the present invention can execute the method for determining the cause of frequent engine failures provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method for determining the cause of frequent engine failures.
[0126] Embodiment 4
[0127] Figure 7 A schematic diagram of an electronic device 710 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.
[0128] like Figure 7 As shown, the electronic device 710 includes at least one processor 711, and a memory connected to the at least one processor 711 in communication, such as a read-only memory (ROM) 712, a random access memory (RAM) 713, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 711 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 712 or the computer program loaded from the storage unit 718 to the random access memory (RAM) 713. In the RAM 713, various programs and data required for the operation of the electronic device 710 can also be stored. The processor 711, the ROM 712, and the RAM 713 are connected to each other via a bus 714. An input / output (I / O) interface 715 is also connected to the bus 714.
[0129] A number of components in the electronic device 710 are connected to the I / O interface 715, including: an input unit 716, such as a keyboard, a mouse, etc.; an output unit 717, such as various types of displays, speakers, etc.; a storage unit 718, such as a disk, an optical disk, etc.; and a communication unit 719, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 719 allows the electronic device 710 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0130] The processor 711 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 711 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 711 executes the various methods and processes described above, such as a method for determining the cause of a high-incidence engine failure.
[0131] In some embodiments, the method for determining the cause of a high-incidence engine failure may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 718. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 710 via the ROM 712 and / or the communication unit 719. When the computer program is loaded into the RAM 713 and executed by the processor 711, one or more steps of the method for determining the cause of a high-incidence engine failure described above may be performed. Alternatively, in other embodiments, the processor 711 may be configured to execute the method for determining the cause of a high-incidence engine failure in any other appropriate manner (e.g., by means of firmware).
[0132] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0133] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.
[0134] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0135] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).
[0136] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.
[0137] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.
[0138] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.
[0139] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for determining the cause of frequent engine failures, It is characterized in that include: Obtaining a current work order and a current engine sales file, and determining current engine basic information and current engine fault information based on the current work order; Input the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause determination model, and output the current engine high-incidence fault cause and the corresponding fault probability; Determine the target engine's high-incidence failure cause according to the current engine's high-incidence failure cause and the corresponding failure probability, and determine corresponding component fault troubleshooting information according to the target engine's high-incidence failure cause; Wherein, before the current engine fault information, the current engine basic information and the current engine sales file are input into the pre-trained engine high-incidence fault cause determination model, it also includes: if there is no fault description information in the current engine fault information, then a missing necessary parameter error instruction is generated, and the missing necessary parameter instruction is used to prompt that there is an error in the current engine fault information; or, if it is detected that the fault description information included in the current engine fault information does not belong to the preset fault range, then a missing fault description error instruction is generated, and the missing fault description error instruction is used to prompt that the fault description information does not belong to the preset fault range; Determining target engine high-incidence failure causes according to the current engine high-incidence failure causes and corresponding failure probabilities includes: sorting the current engine high-incidence failure causes according to failure probabilities, and selecting three engine high-incidence failure causes in sequence as target engine high-incidence failure causes.
2. The method for determining the cause of frequent engine failures according to claim 1, It is characterized in that Before inputting the current engine fault information, the current engine basic information and the current engine sales file into the pre-trained engine high-incidence fault cause determination model, the method further includes: Preprocessing the current engine fault information, the current engine basic information and the current engine sales file, and performing feature extraction, type feature coding and dimensionless processing on the preprocessed current engine fault information, the current engine basic information and the current engine sales file; The preprocessing includes standardizing data types, processing outliers, and completing missing values.
3. The method for determining the cause of frequent engine failures according to claim 1, It is characterized in that The method for determining the cause of frequent engine failures also includes: Preprocess at least one set of historical work orders, historical engine basic information, historical engine maintenance reports, and historical engine sales files to obtain a model building data set; Inputting the model building data set into a pre-established engine high-incidence failure cause determination model to obtain historical engine high-incidence failure causes and corresponding failure probabilities; Model parameters of the engine frequent failure cause determination model are retrained based on the historical engine frequent failure causes and corresponding failure probabilities and the expected engine frequent failure causes and corresponding failure probabilities.
4. The method for determining the cause of frequent engine failures according to claim 3, It is characterized in that The method for determining the cause of frequent engine failures also includes: Obtaining a current engine maintenance report, and storing the current engine fault information, the current engine basic information, the current engine maintenance report, and the current engine sales file into the model building data set; The model parameters of the engine frequent failure cause determination model are retrained according to the updated model building data set.
5. The method for determining the cause of frequent engine failures according to claim 1, It is characterized in that The method for determining the cause of frequent engine failures also includes: Obtaining the failure probability corresponding to the high-incidence failure-causing component of the target engine; The target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are fed back to a service terminal, so that the target engine's frequent failure causes, the corresponding failure probabilities, and the component failure troubleshooting information are displayed on the service terminal.
6. A device for determining the cause of frequent engine failures, It is characterized in that include: An information acquisition module, used to acquire the current work order and the current engine sales file, and determine the current engine basic information and the current engine fault information according to the current work order; a cause component output module, used to input the current engine fault information, the current engine basic information and the current engine sales file into a pre-trained engine high-incidence fault cause component determination model, and output the current engine high-incidence fault cause components and corresponding fault probabilities; A fault troubleshooting module, used to determine the target engine's high-incidence fault cause according to the current engine's high-incidence fault cause and the corresponding fault probability, and determine corresponding component fault troubleshooting information according to the target engine's high-incidence fault cause; Wherein, before the current engine fault information, the current engine basic information and the current engine sales file are input into the pre-trained engine high-incidence fault cause determination model, it also includes: if there is no fault description information in the current engine fault information, then a missing necessary parameter error instruction is generated, and the missing necessary parameter instruction is used to prompt that there is an error in the current engine fault information; or, if it is detected that the fault description information included in the current engine fault information does not belong to the preset fault range, then a missing fault description error instruction is generated, and the missing fault description error instruction is used to prompt that the fault description information does not belong to the preset fault range; Determine the target engine high-incidence failure causes according to the current engine high-incidence failure causes and the corresponding failure probabilities, specifically for: sorting the current engine high-incidence failure causes according to the failure probabilities, and selecting three engine high-incidence failure causes in turn as the target engine high-incidence failure causes.
7. An electronic device, It is characterized in that The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for determining the cause of frequent engine failures according to any one of claims 1-5.
8. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for determining the cause of frequent engine failures according to any one of claims 1 to 5 when executed.
