Health management method, device and equipment for engine and medium

Abstract

The invention discloses an engine health management method, device and equipment and a medium. The method comprises the steps that after it is detected that the current execution moment meets the current management cycle starting condition of an engine, operation and maintenance log information corresponding to the current execution moment of the engine is determined; determining a health prediction model corresponding to the engine according to the operation and maintenance log information; input data of the health prediction model are determined, a prediction result output by processing the input data through the health prediction model is obtained, and the prediction result is used for health assessment and fault early warning of the engine. According to the method, the health prediction model corresponding to the engine is determined according to the operation and maintenance log information, health assessment and fault early warning can be carried out on the engine through the corresponding health prediction model before and after maintenance, and effective management on engine health management under the condition of state transition before and after maintenance is achieved.

Description

Technical Field

[0001] The present invention relates to the field of engine technology, and in particular to an engine health management method, device, equipment and medium. Background Art

[0002] Diesel engines are widely used in railway locomotives, construction machinery, and heavy-duty vehicles. They are core equipment that ensures vehicle power output and operational stability. Over time, the performance of various diesel engine components gradually degrades. Therefore, diesel engine health management is necessary to ensure regular maintenance and troubleshooting to ensure reliable operation.

[0003] In practical applications, existing diesel engine health management implementations often experience internal state transitions after repairs or component replacements. Existing engine health management approaches don't adequately consider the impact of these transitions on diesel engine health management, impacting engine health assessments and fault warnings. Summary of the Invention

[0004] The present invention provides an engine health management method, device, equipment and medium to solve the problem in the prior art that the impact of state transitions before and after maintenance on engine health management is not well considered, resulting in the inability to effectively manage the engine health in the case of state transitions before and after maintenance.

[0005] In a first aspect, an embodiment of the present invention provides an engine health management method, comprising:

[0006] After detecting that the current execution time satisfies the current management cycle start condition of the engine, determining the operation and maintenance log information corresponding to the engine at the current execution time;

[0007] Determining a health prediction model corresponding to the engine according to the operation and maintenance log information;

[0008] Determine input data of the health prediction model, and obtain a prediction result output by the health prediction model after processing the input data, wherein the prediction result is used for health assessment and fault warning of the engine.

[0009] In a second aspect, an embodiment of the present invention provides an engine health management device, comprising:

[0010] A log determination module, configured to determine operation and maintenance log information corresponding to the engine at the current execution moment after detecting that the current execution moment satisfies the current management cycle start condition of the engine;

[0011] A model determination module, configured to determine a health prediction model corresponding to the engine according to the operation and maintenance log information;

[0012] The evaluation and warning module is used to determine the input data of the health prediction model and obtain the prediction results output by the health prediction model after processing the input data. The prediction results are used for health evaluation and fault warning of the engine.

[0013] In a third aspect, an embodiment of the present invention provides an electronic device, comprising:

[0014] at least one processor;

[0015] and a memory communicatively coupled to the at least one processor;

[0016] The memory stores a computer program that can be executed 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 engine health management method described in any embodiment of the present invention.

[0017] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the engine health management method described in any embodiment of the present invention when executed.

[0018] The technical solution of an embodiment of the present invention determines the operation and maintenance log information corresponding to the engine at the current execution time after detecting that the current execution time meets the start conditions of the current management cycle of the engine; determines the health prediction model corresponding to the engine based on the operation and maintenance log information; determines the input data of the health prediction model, and obtains the prediction result output by the health prediction model after processing the input data, thereby performing a health assessment and fault warning on the engine. This method determines the health prediction model corresponding to the engine based on the operation and maintenance log information, and can perform health assessment and fault warning on the engine using the corresponding health prediction model before and after maintenance, respectively, to achieve effective management of engine health management in the case of state transitions before and after maintenance.

[0019] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] 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.

[0021] Figure 1 This is a flow chart of an engine health management method provided in Example 1 of the present invention;

[0022] Figure 2 This is a flow chart of an engine health management method provided by Embodiment 2 of the present invention;

[0023] Figure 3 This is a schematic structural diagram of an engine health management device provided by a third embodiment of the present invention;

[0024] Figure 4 It is a structural diagram of an electronic device for implementing an engine health management device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions 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 embodiments described 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 making creative efforts should fall within the scope of protection of the present invention.

[0026] It should be noted that the terms "first", "second", etc. in the description 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 numbers 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 clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] It should be noted that as the use time increases, the performance of various components in engines such as diesel engines or gasoline engines will gradually degrade. Regular maintenance and fault repair become necessary means to ensure its reliable operation. In actual applications, after the engine undergoes maintenance or component replacement, its internal system state often changes significantly. For example, key data such as sensor parameters, operating pressure, temperature curve, load response, etc. remain in a relatively stable state before maintenance, but will show obvious changes in the short period after maintenance. This phenomenon can be called "pre-maintenance state transition". In the case of a state transition before and after maintenance, how to accurately assess the health of the engine and provide fault warnings is of great significance.

[0028] Figure 1 This is a flow chart of an engine health management method provided by the first embodiment of the present invention. This embodiment of the present invention can be applied to perform engine health management and fault warning in the case of state transition before and after engine maintenance. This method can be executed by an engine health management device, which can be implemented in the form of hardware and / or software. The engine health management device can be configured in a vehicle health status assessment and fault prediction system or platform. Figure 1 As shown, the method may include:

[0029] S101: After detecting that the current execution time satisfies the current management cycle start condition of the engine, determining the operation and maintenance log information corresponding to the engine at the current execution time.

[0030] Among them, the current execution moment can be any time point in the entire life cycle of the engine. For example, the current execution moment can be any time point before or after the engine is repaired. The current management cycle start condition can be used to determine whether the current execution moment is within the management cycle for health management of the engine. Exemplarily, the current management cycle start condition can be that the current execution moment is after the time point when the operation and maintenance personnel instructs to perform health management, or the current moment reaches the health management time period preset by the operation and maintenance personnel in the health status assessment and fault prediction system or platform. The operation and maintenance log information can be a record of the actual maintenance behavior and maintenance time of the operation and maintenance personnel.

[0031] For example, when it is detected that the current execution moment is within the management cycle for engine health management, it can be determined that the start conditions of the current management cycle of the engine are met. At this time, the maintenance behavior and corresponding maintenance time of the engine before the current execution moment can be obtained from the background database to determine the operation and maintenance log information corresponding to the engine at the current execution moment. It should be noted that during the entire life cycle of the engine, the operation and maintenance personnel can repair the engine multiple times at different times and record multiple maintenance behavior data. The maintenance behavior data existing before different time points may be different, so the operation and maintenance log information corresponding to different time points is also different.

[0032] Optionally, determining the operation and maintenance log information corresponding to the engine at the current execution time may include:

[0033] Obtaining maintenance behavior data recorded relative to the engine before the current execution time, the maintenance behavior including: engine component replacement and fault repair, and operation and maintenance debugging of the operation and maintenance platform;

[0034] The maintenance behavior data is structured to generate operation and maintenance log information in a set data format, wherein the operation and maintenance log information includes the maintenance time of the generated maintenance behavior.

[0035] Maintenance behavior data can represent actual maintenance operations performed by maintenance personnel, such as engine component replacement and troubleshooting, as well as maintenance and commissioning of the maintenance platform. It should be noted that maintenance behavior data can be sourced from a pre-set standardized information system. After performing actual maintenance operations, maintenance personnel can register their maintenance actions in the system and generate corresponding data records.

[0036] The set data format may be a pre-set unified data format stored in a background database or a log file, and the data format includes fields related to the maintenance date.

[0037] For example, after determining the corresponding operating time point of the current execution moment in the entire life cycle of the engine, the maintenance behavior data before the current execution moment can be determined in the preset standardized information system, and the maintenance behavior data can be structured in combination with the maintenance date corresponding to the maintenance behavior to generate operation and maintenance log information that can represent the maintenance time corresponding to the maintenance behavior through the corresponding field.

[0038] S102: Determine a health prediction model corresponding to the engine according to the operation and maintenance log information.

[0039] The health prediction model can be used to analyze engine status data and predict engine deviation trends to manage engine health and provide fault warnings. Engine status data can include engine speed, pressure, temperature, power, vibration, etc.

[0040] It should be noted that existing technologies usually use fixed models to perform health management and fault prediction for the entire life cycle of the engine, and regard all status data collected before and after maintenance (such as engine speed, pressure, temperature, power, vibration, etc.) as homogeneous data. The impact of significant jumps in status data before and after maintenance caused by maintenance behavior is not taken into account, which greatly reduces the effectiveness of health management and fault diagnosis after maintenance.

[0041] Based on this, the health prediction model set up by the present invention can include a long-term trend prediction model applicable to the case where the engine status data is stable, and a short-term prediction model applicable to the case where the engine status data is unstable and the sample data volume is insufficient, so that different health prediction models can be used according to the stability of the status data at different stages of the engine, realizing the staged management and intelligent switching of the model, improving the continuity, accuracy and engineering practicality of fault identification, and thus effectively realizing health management and fault warning throughout the life cycle of the engine.

[0042] For example, after determining the operation and maintenance log information corresponding to the current execution time, the maintenance time recorded in the operation and maintenance log information can be used to determine whether there are any maintenance records in the recent period before the current execution time, or whether the current execution time has just undergone maintenance operations. If so, the current execution time is determined to be in a post-maintenance data instability phase, and a short-term prediction model suitable for unstable engine status data and insufficient sample data can be determined as the corresponding health prediction model. If not, the current execution time is determined to be in a state data stability phase, and a long-term trend prediction model suitable for stable engine status data before maintenance can be determined as the corresponding health prediction model.

[0043] Optionally, determining the health prediction model corresponding to the engine according to the operation and maintenance log information may include:

[0044] Parsing the operation and maintenance log information to obtain the maintenance time of the engine included in the operation and maintenance log information;

[0045] Determine the shortest time interval between the current execution time and the maintenance time, and compare the shortest time interval with a set interval threshold;

[0046] If the comparison result shows that the shortest time interval is less than or equal to the interval threshold, the pre-built short-cycle prediction model is determined as the health prediction model; otherwise,

[0047] A pre-built long-term trend prediction model is determined as the health prediction model.

[0048] The engine maintenance time may correspond to the maintenance behavior recorded in the operation and maintenance log information. The shortest time interval between the current execution time and the maintenance time may refer to the time interval between the current execution time and the most recent maintenance time.

[0049] The set interval threshold can be a pre-set duration of instability in the post-repair engine status data, and can be used to determine whether the current execution moment is in a phase of instability in the post-repair engine status data. It should be noted that after engine maintenance, the engine's status data will experience significant transitions within a certain period of time. However, as the engine's operating time progresses, the engine will return to a stable operating phase, at which point the engine's operating status data will return to a stable state until the next maintenance. Therefore, this embodiment can determine the stability of the engine status data corresponding to the current execution moment by determining whether the current execution moment is within a pre-set time period starting from the most recent maintenance time.

[0050] For example, after obtaining the operation and maintenance log information corresponding to the current execution moment, the corresponding engine maintenance time can be obtained by parsing the maintenance time field representing the maintenance behavior in the operation and maintenance log information. Then, the maintenance time corresponding to the most recent maintenance behavior before the current execution moment is obtained from the engine maintenance time, and the time difference between the maintenance time and the current execution moment is used as the shortest time interval between the current execution moment and the maintenance time. The shortest time interval is compared with the set interval threshold to determine whether the current execution moment is within the time period when the engine maintenance status data is unstable. If the comparison result is that the shortest time interval is less than or equal to the interval threshold, it is determined that the engine status data corresponding to the current execution moment is unstable, and the pre-built short-term prediction model can be determined as a healthy prediction model. On the contrary, it is determined that the engine status data corresponding to the current execution moment is stable, and the pre-built long-term trend prediction model can be determined as a healthy prediction model.

[0051] S103: Determine input data of the health prediction model, and obtain a prediction result output by the health prediction model after processing the input data, wherein the prediction result is used for health assessment and fault warning of the engine.

[0052] The input data may be data required by the health prediction model to perform engine health management and fault warning. For example, the input data required by the short-term prediction model may be short-term historical data, and the input data required by the long-term trend prediction model may be long-term stable historical data.

[0053] It should be noted that different health prediction models utilize different modeling logic to meet different application scenarios, and therefore require different input data. For example, the long-term trend prediction model is suitable for modeling and trending the health of an engine during periods of continuous and stable operation without maintenance intervention. Therefore, the engine's long-term accumulated, high-quality, stable operating data can be used as the prior knowledge foundation for establishing the long-term trend prediction model, thereby constructing a trend model that characterizes the evolution of system performance. Regarding data processing, to enhance the stability and generalization of the long-term trend prediction model, the raw operating data accumulated over a long period of time can be periodically aggregated and structured to extract representative statistical features and trend signals. This extraction process includes, but is not limited to, metrics such as the median, rate of change, and fluctuation amplitude of data at daily, weekly, and monthly scales. The resulting time series has a high degree of trend expression. This modeling process, supported by long-term, stable historical data, constructs a health reference curve for equipment operation, which is used to continuously track operating status, identify deviations from trends, and assess degradation rates. This ensures that long-term, stable historical data serves as input for the long-term trend prediction model. The above examples are merely illustrative of the modeling logic of the long-term trend prediction model, and do not specifically limit the modeling method of the long-term trend prediction model.

[0054] For example, the corresponding input data can be determined based on the health prediction model corresponding to the determined engine. If the determined health prediction model is a long-term trend prediction model, it means that the current execution moment is in the stable stage of the engine status data. At this time, the long-term stable historical data of the engine before the current execution moment can be obtained and determined as the input data of the long-term trend prediction model. The input data is analyzed and processed by the long-term trend prediction model, and the result of the predicted engine status data trend is output. Based on the status data trend, it is determined whether the offset of the status data exceeds the preset threshold. The health status of the engine (such as sub-health or health, etc.) can be diagnosed to perform a health assessment and fault warning on the engine. If the engine is determined to be in a sub-healthy state, intervention suggestions can also be provided to the operation and maintenance personnel to repair or replace the components in the engine.

[0055] If the determined health prediction model is a short-cycle prediction model, it means that the current execution moment is in an unstable stage of status data after engine maintenance. At this time, the continuous amount of historical data obtained in a short period before the previous execution moment can be determined as the input data of the short-cycle prediction model, and the input data is analyzed and processed by the short-cycle prediction model to predict the result of the status data at the current execution moment. Then, based on the comparison between the predicted status data at the current execution moment and the actual status data, it is identified whether there is a deviation from the trend, so as to diagnose the health status of the engine (such as sub-health or health, etc.) to perform health assessment and fault warning on the engine.

[0056] This embodiment provides a method for managing the health of an engine. After detecting that the current execution moment satisfies the start conditions of the current management cycle of the engine, the method determines the operation and maintenance log information corresponding to the engine at the current execution moment, determines the health prediction model corresponding to the engine based on the operation and maintenance log information, determines the input data of the health prediction model, and obtains the prediction result output by the health prediction model after processing the input data. The prediction result is used for health assessment and fault warning of the engine. By determining the health prediction model corresponding to the engine through the operation and maintenance log information, the health assessment and fault warning of the engine can be performed using the corresponding health prediction model before and after maintenance, thereby achieving effective management of the engine health management in the case of state transitions before and after maintenance.

[0057] Based on the above embodiment, the present invention further provides an optional embodiment, which may further improve step S103 in the above embodiment when the health prediction model is the long-term trend prediction model, and may include:

[0058] Obtaining key operating parameter data of the engine summarized within a set time interval, obtaining health trend data of the engine within a set historical period, and recording the key operating parameter data and the health trend data as the input data, wherein the set time interval is the time interval from the end of the previous management cycle to the current execution time;

[0059] determining statistical data characteristics of the engine in the current management cycle according to the key operating parameter data;

[0060] The statistical data feature is compared with the health trend data to determine the trend deviation data of the engine, and the trend deviation data is determined as the prediction result.

[0061] It's important to note that the engine's key operating parameters are crucial indicators for evaluating a vehicle's powertrain. They directly impact the vehicle's driving performance and fuel economy, and play a crucial role in a user's purchasing decision. For example, these parameters might include engine power, torque, speed, fuel consumption, and emissions.

[0062] The key operating parameter data may be the total data of all key operating parameters aggregated within a preset time period. For example, the key operating parameter data may be the total speed data aggregated from all speed data within a preset time period. The set historical duration may be a preset historical duration prior to the current execution time.

[0063] The health trend data may be data reflecting the changing trend of key operating parameters of the engine over time, for example, a linear data graph of the engine's speed data.

[0064] The statistical data characteristics of the current management cycle can be statistical characteristics (such as mean, variance, standard deviation, etc.) determined based on the key operating parameters of all engines during the period from the start time of the current management cycle to the current execution time. The trend deviation data can be the offset between the healthy trend data and the determined statistical characteristics, which can be used to determine whether the key operating parameters of the engine are gradually deteriorating.

[0065] For example, taking the key operating parameter as the engine speed data as an example, when the health prediction model is a long-term trend prediction model, it can be determined that the current execution moment is in the state data stable stage in the current management cycle. At this time, the engine speed data can be summarized in the time interval from the end of the previous management cycle to the current execution moment to form a total speed data (i.e., key operating parameter data) with a sufficient data sample size. At the same time, the trend of the engine speed data changing over time can be determined within the historical length before the current execution moment to form health trend data. Then, the total speed data and health trend data are input as input data into the long-term trend prediction model.

[0066] Furthermore, the long-term trend prediction model can calculate the standard deviation of the engine during the current management cycle based on the total speed data. This standard deviation can then be compared with the health trend data to determine the offset between the health trend data and the determined statistical characteristics, thereby generating trend offset data, which is then used as the prediction result. It is understood that after the long-term trend prediction model outputs the prediction result, it can be used to determine whether there is a sustained, directional offset in the engine's key operating parameters, thereby determining whether the engine is experiencing gradual degradation. This allows for health management and fault warnings while maintaining stable engine status data.

[0067] The advantage of this setting is that it can accurately identify the phenomenon of slow deviation of engine status data and provide maintenance personnel with earlier intervention suggestions.

[0068] Based on the above embodiment, the present invention further provides another optional embodiment, which can further improve step S103 in the above embodiment when the health prediction model is the short-term prediction model, and can include:

[0069] Obtaining historical parameter data corresponding to consecutive historical execution moments set before the current execution moment, and historical statistical data features determined before the health prediction model switches from the long-term trend prediction model to the short-term prediction model;

[0070] Using the historical parameter data and the historical statistical data features as the input data, processing the input data through the short-term prediction model to obtain the prediction parameter data of the current execution moment relative to the engine prediction;

[0071] Acquire actual parameter data corresponding to the engine at the current execution moment, compare the predicted parameter data with the actual parameter data, determine a deviation trend of the engine according to the comparison result, and determine the deviation trend as the prediction result.

[0072] The set continuous historical execution time may refer to a historical time point at which the continuous execution state data is measured in the historical record, and the historical parameter data may be a key operating parameter corresponding to the corresponding historical time point.

[0073] The historical statistical data features may be statistical data features of key operating parameters of the engine determined and recorded by a long-term trend prediction model before the engine is repaired (eg, the standard deviation of the engine speed data before the repair).

[0074] The predicted parameter data may be the predicted value of the key operating parameter corresponding to the current execution time by the short-term prediction model. The actual parameter data may be the key operating parameter actually measured at the current execution time, for example, the actual speed data value measured at the current execution time.

[0075] For example, let's take the key operating parameter as the engine speed data. First, when the health prediction model is a short-term prediction model, it can be determined that the current execution time is at a stage where the engine state data is unstable after maintenance. At this time, the corresponding n historical speed data can be determined at n historical time points of continuous execution state data measurement before the current execution time as historical parameter data, which is recorded as x. t+1-iAt the same time, obtain the statistical data characteristics of the engine speed data determined and recorded by the long-term trend prediction model before maintenance, and record them as historical statistical data characteristics σ pre .

[0076] Secondly, the historical speed data and historical statistical data features are used as input data and input into the short-term prediction model so that the short-term prediction model can process the input data and obtain the key speed data relative to the engine prediction at the current execution moment. The shutdown speed data can be used as prediction parameter data. The key speed data of the engine prediction at the current execution time It can be determined by the following formula:

[0077]

[0078] Among them, x t+1-i It can be the nth historical speed data before the current execution time; It can be the local trend center of the historical time point corresponding to the nth historical speed data before the current execution moment; It can be the weight coefficient of each lag term, obtained by fitting the local data; Can be a model bias term; Can be a state change correction term, and Δ t It can be the difference between the two most recent time points, indicating the speed of state change.

[0079] It can be understood that the local trend center of the historical time point corresponding to the nth historical speed data before the current execution time The determination can be made by determining the k historical speed data before and after the nth historical speed data before the current execution time, and calculating the corresponding statistical data characteristics (such as median, mean, variance and standard deviation, etc.). On this basis, the present invention further proposes to use the historical statistical data characteristics σ of the engine speed data determined and recorded by the long-term trend prediction model before maintenance pre , as the tolerance limit, to calculate the local trend center of the historical time point corresponding to the nth historical speed data before the current execution time Tolerance adjustment is performed, and the corresponding implementation formula is as follows:

[0080]

[0081] Among them, Median(x t+1-i±k ) can be the statistical data feature of the k adjacent historical speed data before and after the nth historical speed data before the current execution time; α can be the adjustment coefficient set by the operation and maintenance personnel based on experience;

[0082] Therefore, the short-term prediction module can obtain the key speed data of the current execution time relative to the engine prediction based on the historical speed data corresponding to the continuous historical execution time set before the current execution time and the historical statistical data characteristics determined by the long-term trend prediction model before maintenance.

[0083] Finally, after obtaining the predicted parameter data, the actual engine speed data corresponding to the current execution time can be obtained through real-time measurement or observation, and this actual speed data can be used as the actual parameter data. The predicted parameter data can then be compared with the actual parameter data to identify whether there is a deviation trend in the key operating parameters of the engine, and the result of the determination can be determined as the predicted result.

[0084] The advantage of this setting is that it fully considers the offset characteristics between the short-term historical speed data and its trend baseline to achieve fault prediction in the unstable stage of the engine status data after maintenance. At the same time, the state change rate correction term is introduced to achieve dynamic adaptation to the small sample and high disturbance stage, so that the model has rapid self-adaptation capabilities when facing short-term jumps in the engine status data, thereby achieving more accurate health management and fault warning. In addition, after the model is switched, by introducing the historical statistical data characteristics in the long-term trend prediction model before maintenance, the tolerance of the historical trend center value in the short term is adjusted to ensure that even if the key operating data characteristics fluctuate after maintenance, the trend center of the key operating data still retains the statistical data characteristics before maintenance as a reference, which can avoid discontinuity in judgment or sudden increase in deviation during the model replacement process, and help to smooth the judgment during the transition period.

[0085] Based on the above embodiment, the present invention also provides another optional embodiment, which can further optimize the deployment of the long-term trend prediction model and the short-term prediction model, and may include:

[0086] Different health prediction models involve different reading paths for engine parameter data and different model configuration information.

[0087] For example, to ensure the independence and deployment flexibility of the two models, the system architecture can be designed to structurally decouple the two types of models. Each model can have an independent configuration file, data reading path, feature processing logic, and threshold setting standards, so as to avoid parameter crossover or state coupling problems. It is understandable that during deployment, users can only enable one of the models for monitoring based on the actual working conditions, or they can choose to enable the automatic switching function. The system will intelligently judge the current stage based on the maintenance time and operation time of the maintenance behavior and automatically switch to the optimal prediction model.

[0088] Example 2

[0089] Figure 2 This is a flowchart of a method for health management of an engine provided by the second embodiment of the present invention. This embodiment is further optimized based on the above-mentioned first embodiment. The method also includes: when the health prediction model is switched from a long-term trend prediction model to a short-term prediction model, recording the engine parameter data corresponding to the transmitter after starting the health management of the engine; when it is detected that the accumulated value of the transmitter parameter data meets the update condition of the long-term trend prediction model, rebuilding the long-term trend prediction model based on the transmitter parameter data. Figure 2 As shown, the method includes:

[0090] S201: After detecting that the current execution time satisfies the current management cycle start condition of the engine, determining the operation and maintenance log information corresponding to the engine at the current execution time;

[0091] S202, determining a health prediction model corresponding to the engine according to the operation and maintenance log information;

[0092] S203: Determine input data of the health prediction model, and obtain a prediction result output by the health prediction model after processing the input data, wherein the prediction result is used for health assessment and fault warning of the engine.

[0093] S204 , when the health prediction model is switched from the long-term trend prediction model to the short-term prediction model, recording engine parameter data corresponding to the transmitter after starting the health management of the engine.

[0094] The engine parameter data may be any one or more parameter data of the engine's shutdown operating parameters.

[0095] For example, consider engine speed data as an example. When the health prediction model switches from a long-term trend prediction model to a short-term prediction model, it can be determined that maintenance occurred at that time point. At this point, a data accumulation strategy can be used to continuously monitor and assess sample quality of the post-repair engine to obtain engine speed data corresponding to the repaired engine, which can be used to retrain the long-term trend prediction model.

[0096] S205 : When it is detected that the accumulated value of the transmitter parameter data meets the updating condition of the long-term trend prediction model, rebuild the long-term trend prediction model based on the transmitter parameter data.

[0097] The accumulated value of the transmitter parameter data may represent the amount of data of the key operating parameters, and the updating condition may be that the number of samples required by the model reaches a preset scale.

[0098] For example, taking the engine parameter data as engine speed data, when it is detected that the accumulated amount of engine speed data corresponding to a repaired engine reaches the sample size required for rebuilding the long-term trend prediction model, the long-term trend prediction model can be rebuilt using the newly acquired engine speed data corresponding to the repaired engine.

[0099] This embodiment provides a method for engine health management. When the health prediction model is switched from a long-term trend prediction model to a short-cycle prediction model, the engine parameter data corresponding to the transmitter after starting the engine health management is recorded, and when the cumulative value of the transmitter parameter data is detected to meet the update conditions of the long-term trend prediction model, the long-term trend prediction model is reconstructed based on the transmitter parameter data. A long-term trend prediction model that is more suitable for the post-maintenance status can be reconstructed, and the reconstructed long-term trend prediction model can seamlessly replace the short-cycle model, thereby achieving improved prediction accuracy and stability, and significantly improving the health management capabilities of the engine throughout its life cycle and the intelligence level of practical applications.

[0100] Based on the above embodiment, the present invention further provides another optional embodiment, which can further improve step S205 in the above embodiment and may include:

[0101] Dividing the accumulated recorded engine parameter data according to a set data division dimension to obtain at least one set of divided parameter data;

[0102] Reconstructing the model structure according to the set model configuration to obtain an initial health prediction model having the model structure, extracting feature trends from each of the divided parameter data to obtain corresponding feature extraction results;

[0103] The initial health prediction model is trained based on the feature extraction results, and a reconstructed health prediction model is obtained after the training is completed.

[0104] The data partitioning dimension can be time, engine duty cycle, or other dimensions, which can be used to divide engine parameter data into different stages. The pre-defined model configuration can be a general model configuration for a long-term trend prediction model, which can be used to initially construct a long-term trend prediction model for health management and fault warning. The feature extraction result can be a data trend result determined by calculating the statistical data features of the engine parameter data. For example, the center of the data trend can be determined by calculating the median of the statistical data of the speed data and used as the feature extraction result.

[0105] For example, consider engine speed data as engine parameter data. After acquiring enough speed data to meet the sample size required for reconstructing a long-term trend prediction model, the accumulated recorded engine speed data can be segmented by time to obtain at least one group of segmented speed data. For each group of segmented speed data, statistical data features can be calculated to determine feature extraction results that represent data trends. This can then be combined with an initial health prediction model constructed using the general model configuration of the long-term trend prediction model to train a health prediction model that meets the requirements for post-maintenance stability.

[0106] Example 3

[0107] Figure 3 This is a schematic diagram of the structure of an engine health management device provided by the third embodiment of the present invention. Figure 3 As shown, the device includes:

[0108] The log determination module 31 may be configured to determine the operation and maintenance log information corresponding to the engine at the current execution moment after detecting that the current execution moment satisfies the current management cycle start condition of the engine;

[0109] A model determination module 32 may be configured to determine a health prediction model corresponding to the engine based on the operation and maintenance log information;

[0110] The evaluation and warning module 33 can be used to determine the input data of the health prediction model and obtain the prediction results output by the health prediction model after processing the input data. The prediction results are used for health evaluation and fault warning of the engine.

[0111] This embodiment provides an engine health management device, which, after detecting that the current execution moment satisfies the start conditions of the current management cycle of the engine, determines the operation and maintenance log information corresponding to the engine at the current execution moment, determines the health prediction model corresponding to the engine based on the operation and maintenance log information, determines the input data of the health prediction model, and obtains the prediction result output by the health prediction model after processing the input data. The prediction result is used for health assessment and fault warning of the engine. By determining the health prediction model corresponding to the engine through the operation and maintenance log information, the health assessment and fault warning of the engine can be performed using the corresponding health prediction model before and after maintenance, thereby achieving effective management of engine health management in the case of state transitions before and after maintenance.

[0112] Optionally, the log determination module 31 may be specifically configured to obtain maintenance behavior data recorded relative to the engine before the current execution time, wherein the maintenance behavior includes: engine component replacement and fault repair, and operation and maintenance debugging of the operation and maintenance platform;

[0113] The maintenance behavior data is structured to generate operation and maintenance log information in a set data format, wherein the operation and maintenance log information includes the maintenance time of the generated maintenance behavior.

[0114] Optionally, the model determination module 32 may be specifically configured to parse the operation and maintenance log information to obtain the maintenance time of the engine included in the operation and maintenance log information;

[0115] Determine the shortest time interval between the current execution time and the maintenance time, and compare the shortest time interval with a set interval threshold;

[0116] If the comparison result shows that the shortest time interval is less than or equal to the interval threshold, the pre-built short-cycle prediction model is determined as the health prediction model; otherwise,

[0117] A pre-built long-term trend prediction model is determined as the health prediction model.

[0118] Optionally, the evaluation and early warning module 33 may be specifically configured to, when the health prediction model is the long-term trend prediction model, obtain key operating parameter data of the engine summarized within a set time interval, obtain health trend data of the engine within a set historical period, and record the key operating parameter data and the health trend data as the input data, wherein the set time interval is the time interval from the end of the previous management cycle to the current execution moment;

[0119] determining statistical data characteristics of the engine in the current management cycle according to the key operating parameter data;

[0120] Comparing the statistical data features with the monitored trend data to determine trend deviation data of the engine, and determining the trend deviation data as a prediction result;

[0121] Perform health assessment and fault warning on the engine based on the prediction results.

[0122] Optionally, the evaluation and early warning module 33 may further, when the health prediction model is the short-cycle prediction model, obtain historical parameter data corresponding to consecutive historical execution moments set before the current execution moment, and obtain historical statistical data features determined before the health prediction model is switched from the long-term trend prediction model to the short-cycle prediction model;

[0123] Using the historical parameter data and historical statistical data features as the input data, processing the input data through the short-term prediction model to obtain prediction parameter data relative to the engine prediction at the current execution moment;

[0124] Acquire actual parameter data corresponding to the engine at the current execution moment, compare the predicted parameter data with the actual parameter data, determine a deviation trend of the engine according to the comparison result, and determine the deviation trend as the prediction result.

[0125] Optionally, the device further comprises: a recording module and an update module;

[0126] a recording module, which may be used to record engine parameter data corresponding to the transmitter after the health management of the engine is started when the health prediction model is switched from the long-term trend prediction model to the short-term prediction model;

[0127] The updating module may be configured to rebuild the long-term trend prediction model based on the transmitter parameter data when detecting that the accumulated value of the transmitter speed data satisfies the updating condition of the long-term trend prediction model.

[0128] Optionally, the updating module may be specifically configured to divide the accumulated recorded engine parameter data according to a set data division dimension to obtain at least one set of divided parameter data;

[0129] Reconstructing the model structure according to the set model configuration to obtain an initial health prediction model having the model structure, extracting feature trends from each of the divided parameter data to obtain corresponding feature extraction results;

[0130] The initial health prediction model is trained based on the feature extraction results, and a reconstructed health prediction model is obtained after the training is completed.

[0131] The engine health management device provided in the embodiment of the present invention can execute the engine health management method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0132] Example 4

[0133] Figure 4 A schematic diagram of the structure of an electronic device 40 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 can 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 claimed herein.

[0134] like Figure 4 As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., which is communicatively connected to the at least one processor 41. The memory stores a computer program that can be executed by the at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 into the random access memory (RAM) 43. Various programs and data required for the operation of the electronic device 40 can also be stored in the RAM 43. The processor 41, ROM 42, and RAM 43 are connected to each other via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0135] Multiple components in the electronic device 40 are connected to the I / O interface 45, including an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a magnetic disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0136] Processor 41 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any other suitable processor, controller, microcontroller, etc. Processor 41 executes the various methods and processes described above, such as the engine health management method.

[0137] In some embodiments, the engine health management method can be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the engine health management method described above can be performed. Alternatively, in other embodiments, processor 41 can be configured to execute the engine health management method in any other suitable manner (e.g., via firmware).

[0138] Various embodiments of the systems and techniques described 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), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes 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.

[0139] 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, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0140] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0141] To provide interaction with a user, the systems and techniques described herein can 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 pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0142] The systems and techniques described herein can be implemented in a computing system that includes back-end 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 front-end components (e.g., a user computer with a graphical user interface or 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 back-end components, middleware components, or front-end components. The components of the system can 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.

[0143] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0144] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0145] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for engine health management, characterized in that: include: After detecting that the current execution time satisfies the current management cycle start condition of the engine, determining the operation and maintenance log information corresponding to the engine at the current execution time; Determining a health prediction model corresponding to the engine according to the operation and maintenance log information; Determine input data of the health prediction model, and obtain a prediction result output by the health prediction model after processing the input data, wherein the prediction result is used for health assessment and fault warning of the engine.

2. The method according to claim 1, characterized in that The determining of the operation and maintenance log information corresponding to the engine at the current execution time includes: Obtaining maintenance behavior data recorded relative to the engine before the current execution time, the maintenance behavior including: engine component replacement and fault repair and operation and maintenance debugging of the operation and maintenance platform; The maintenance behavior data is structured to generate operation and maintenance log information in a set data format, wherein the operation and maintenance log information includes the maintenance time of the generated maintenance behavior.

3. The method according to claim 1, characterized in that Determining the health prediction model corresponding to the engine according to the operation and maintenance log information includes: Parsing the operation and maintenance log information to obtain the maintenance time of the engine included in the operation and maintenance log information; Determine the shortest time interval between the current execution time and the maintenance time, and compare the shortest time interval with a set interval threshold; If the comparison result shows that the shortest time interval is less than or equal to the interval threshold, the pre-built short-cycle prediction model is determined as the health prediction model; otherwise, A pre-built long-term trend prediction model is determined as the health prediction model.

4. The method according to claim 3, characterized in that When the health prediction model is the long-term trend prediction model, determining the input data of the health prediction model and obtaining the prediction result output by the health prediction model after processing the input data includes: Obtaining key operating parameter data of the engine summarized within a set time interval, obtaining health trend data of the engine within a set historical period, and recording the key operating parameter data and the health trend data as the input data, wherein the set time interval is the time interval from the end of the previous management cycle to the current execution time; determining statistical data characteristics of the engine in the current management cycle according to the key operating parameter data; The statistical data feature is compared with the health trend data to determine the trend deviation data of the engine, and the trend deviation data is determined as the prediction result.

5. The method according to claim 3, characterized in that When the health prediction model is the short-cycle prediction model, determining the input data of the health prediction model and obtaining the prediction result output by the health prediction model after processing the input data includes: Obtaining historical parameter data corresponding to consecutive historical execution moments set before the current execution moment, and obtaining historical statistical data features determined before the health prediction model switches from the long-term trend prediction model to the short-term prediction model; Using the historical parameter data and historical statistical data features as the input data, processing the input data through the short-term prediction model to obtain prediction parameter data relative to the engine prediction at the current execution moment; Acquire actual parameter data corresponding to the engine at the current execution moment, compare the predicted parameter data with the actual parameter data, determine a deviation trend of the engine according to the comparison result, and determine the deviation trend as the prediction result.

6. The method according to claim 1, characterized in that Also includes: When the health prediction model is switched from a long-term trend prediction model to a short-term prediction model, recording engine parameter data corresponding to the transmitter after starting the health management of the engine; When it is detected that the accumulated value of the transmitter parameter data satisfies the updating condition of the long-term trend prediction model, the long-term trend prediction model is rebuilt based on the transmitter parameter data.

7. The method according to claim 6, characterized in that The reconstructing the health prediction model based on the transmitter speed data includes: Dividing the accumulated recorded engine parameter data according to a set data division dimension to obtain at least one set of divided parameter data; Reconstructing the model structure according to the set model configuration to obtain an initial health prediction model having the model structure, extracting feature trends from each of the divided parameter data to obtain corresponding feature extraction results; The initial health prediction model is trained based on the feature extraction results, and a reconstructed health prediction model is obtained after the training is completed.

8. An engine health management device, characterized in that: include: A log determination module, configured to determine operation and maintenance log information corresponding to the engine at the current execution moment after detecting that the current execution moment satisfies the current management cycle start condition of the engine; A model determination module, configured to determine a health prediction model corresponding to the engine according to the operation and maintenance log information; The evaluation and warning module is used to determine the input data of the health prediction model and obtain the prediction results output by the health prediction model after processing the input data. The prediction results are used for health evaluation and fault warning of the engine.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively coupled to the at least one processor; 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 engine health management method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the engine health management method according to any one of claims 1 to 7 when executed.

Images