Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the application.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this disclosure, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.
It should be understood that although the terms first, second, third, etc. may be used herein to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the application. The term "if" as used herein may be interpreted as "at..once" or "when..once" or "in response to a determination", depending on the context.
Specific examples are given below to describe the technical solution of the present application in detail.
Fig. 1 is a flowchart of a first embodiment of a method for constructing a hydraulic pump life prediction model according to the present application. Referring to fig. 1, the method provided in this embodiment may include:
S101, acquiring sectional life test data and continuous life test data of the hydraulic pump.
Specifically, the segment life test of the hydraulic pump refers to dividing the whole life cycle of the hydraulic pump into a plurality of stages, respectively enabling the hydraulic pump under each stage to enter a specified duration test under a rated operation condition, recording operation data of the hydraulic pump under each stage in real time, wherein a set of the operation data of the hydraulic pump under each stage is the segment life test data of the hydraulic pump, and the segment life test can acquire short-time degradation characteristics of the hydraulic pump under different stages along with time. The specified duration of the operation of the hydraulic pump in the segment life test is set according to actual needs, and in this embodiment, the operation is not limited. For example, in one embodiment, the specified duration is 50 hours.
Further, the continuous life test of the hydraulic pump refers to that according to the rated operation working condition of the hydraulic pump, the hydraulic pump is enabled to continuously operate under the rated operation working condition until the hydraulic pump fails, the hydraulic pump data in the period from the start of the test to the failure is the continuous life test data, and the continuous life test can describe the degradation state of the hydraulic pump along with time through the complete working process of the hydraulic pump.
Further, when the continuous life test and the sectional life test are performed on the hydraulic pump, various data of the hydraulic pump during the test are obtained in real time through a sensor which is arranged on the hydraulic pump in advance, and the sensor which is arranged on the hydraulic pump in advance can be a pressure sensor, a temperature sensor, a flow sensor, a vibration sensor, a rotating speed sensor and the like, so that signals of pressure pulsation, temperature, flow, vibration, rotating speed and the like of the hydraulic pump during operation are obtained through the sensor.
S102, respectively carrying out degradation characteristic analysis on the sectional life test data and the continuous life test data to obtain a first fusion characteristic and a second fusion characteristic, wherein the first fusion characteristic is a characteristic corresponding to the sectional life test data, and the second fusion characteristic is a characteristic corresponding to the continuous life test data.
Specifically, the degradation characteristic refers to a characteristic parameter that is displayed by the hydraulic pump during use and can reflect the decline, aging or approaching fault state of the hydraulic pump with time.
Further, the step of performing degradation characteristic analysis on the segment life test data to obtain a first fusion characteristic includes the following specific implementation steps:
(1) And extracting the characteristics of the segmented life test data to obtain first parameter characteristics.
Specifically, in this step, feature extraction may be performed on the segment life test data by methods such as time domain analysis, frequency domain analysis, time-frequency domain analysis, and statistical analysis, to extract useful degradation features, for example, effective values and peak-to-peak values of the time domain are extracted by time domain analysis for high-frequency information such as pressure pulsation and vibration, and failure features and frequency amplitude features of the frequency domain are extracted by frequency domain analysis, and features such as oil return temperature rise, outlet temperature rise, and amplitude trend are extracted for low-frequency information such as temperature and flow.
Furthermore, in the step, a large model can be constructed by using methods such as machine learning, deep learning and the like, and degradation characteristics related to the degradation of the hydraulic pump can be automatically identified and extracted through the learning of a large number of segment life test data, so as to obtain first parameter characteristics.
(2) And screening the first parameter characteristics to obtain first sensitive characteristics.
It can be understood that among the first parameter features, different first parameter features have different importance on the prediction of the health state of the hydraulic pump, and for better modeling, the first parameter features need to be screened, the features sensitive to the health state of the hydraulic pump in the first parameter features are selected, and for convenience of explanation, the features sensitive to the health state of the hydraulic pump in the first parameter features are the first sensitive features.
In the specific implementation, the change trend of each first parameter feature along with time is analyzed, if the feature is increased or decreased along with time, the feature is a sensitive feature, otherwise, the feature does not have the change trend along with the time, and the feature is an insensitive feature. Thus, the first parameter features are screened to obtain first sensitive features.
(3) And performing dimension reduction processing on the first sensitive feature by using a principal component analysis method to obtain a first dimension reduction feature.
Specifically, although the degradation feature extraction from multiple angles can fully express the health degradation state, the first sensitive features inevitably contain some redundant, useless and even ambiguous features, so that the calculation burden of the prediction algorithm is heavy, the prediction performance robustness is poor, and the complexity of subsequent analysis is increased. A dimension-reduction compression process is therefore required for the first sensitive feature.
Specifically, the principal component analysis method is mainly used in the fields of data dimension reduction, feature extraction, data visualization and the like, and converts an original variable into a group of new uncorrelated variables, namely principal components, by performing linear transformation on the original variable. These principal components are linear combinations of the original variables that can preserve as much information as possible of the original data while reducing the dimensionality of the data. For a specific implementation process of performing the dimension reduction processing on the first sensitive feature by using the principal component analysis method to obtain the first dimension reduction feature, please refer to the description in the related art, and details are not repeated here.
(4) And fusing the first dimension reduction features to obtain first fused features.
Specifically, the fusion characteristic refers to that a plurality of different sensitive characteristics are integrated by a specific method to form a comprehensive characteristic capable of reflecting the health state of the product more comprehensively and accurately. In the health assessment of hydraulic pumps, there may be several sensitive features such as pressure pulsation, flow changes, vibration amplitude, temperature changes, etc. These features alone may only provide partial information, but their advantages may be combined by fusion to more effectively describe the overall health of the hydraulic pump.
In specific implementation, the first dimension reduction features can be fused by adopting methods such as weighted average, feature stitching and the like to obtain first fusion features.
Further, the degradation characteristic analysis is performed on the continuous life test data, and the specific implementation steps for obtaining the second fusion characteristic are as follows:
(1) And extracting the characteristics of the continuous life test data to obtain second parameter characteristics.
Specifically, the continuous life test data generally includes a large amount of information about the change of the performance of the hydraulic pump with time, and the feature extraction of the continuous life test data is to find out the parameter features which can represent the essential characteristics of the performance of the hydraulic pump from the complex and huge data, so as to provide a basis for subsequent analysis.
Further, in this step, feature extraction may be performed on the continuous life test data by methods such as time domain analysis, frequency domain analysis, time-frequency domain analysis, and statistical analysis, to extract useful degradation features, for example, effective values and peak-to-peak values of the time domain are extracted by time domain analysis for high-frequency information such as pressure pulsation and vibration, and failure features and frequency amplitude features of the frequency domain are extracted by frequency domain analysis, and features such as oil return temperature rise, outlet temperature rise, and amplitude trend are extracted for low-frequency information such as temperature and flow.
Furthermore, in the step, a large model can be constructed by using methods such as machine learning, deep learning and the like, and degradation characteristics related to the degradation of the hydraulic pump can be automatically identified and extracted through learning of a large number of continuous life test data, so as to obtain second parameter characteristics.
(2) And screening the second parameter characteristics to obtain second sensitive characteristics.
(3) And performing dimension reduction treatment on the second sensitive characteristic by using a principal component analysis method to obtain a second dimension reduction characteristic.
(4) And fusing the second dimension reduction features to obtain second fused features, and mapping the health state according to the second fused features.
Specifically, for the implementation process of obtaining the second fusion feature in the steps (2) to (4), please refer to the above description, and the detailed description is omitted here.
Specifically, when mapping the health status according to the second fusion feature, the health status level first needs to be determined. The health status level of the hydraulic pump is set according to actual needs, and is not limited in this embodiment. For example, in one possible implementation, the health state H ranges from 0 to 1, if H >0.8, the hydraulic pump is in a good state, no performance degradation or insignificant performance degradation occurs, no maintenance is required at this time, if 0.4< H <0.8, it indicates that the hydraulic pump is in a general state, task evaluation is required, and the remaining service life and maintenance schedule of the hydraulic pump are determined, and if H <0.4, it indicates that the hydraulic pump is in a dangerous state, and immediate maintenance and replacement are required.
Further, in this step, the mapping relationship between the second fusion feature and the health state may be constructed by a method such as mathematical analysis or machine learning. For example, in one embodiment, the relationship between the second fusion feature and the health state is found by performing data analysis on the second fusion feature and the health state, and a mathematical model between the second fusion feature and the health state is established. In the use process, a second fusion characteristic is input into the constructed mathematical model, and the mathematical model outputs a health state corresponding to the second fusion characteristic.
S103, constructing a first life prediction model according to the first fusion characteristic.
The specific implementation steps comprise:
(1) And determining the number of the health state classification clusters according to the characteristic differences of the hydraulic pump under different health state levels.
Specifically, in combination with the above description, the features of the hydraulic pump under different health status levels are the corresponding features of the first fusion feature under each health status, the number of classification clusters is determined by using a clustering evaluation index, the clustering evaluation index may be a profile coefficient (Silhouette Coefficient), a Calinski-Harabasz index, a Davies-Bouldin index, or the like, and the clustering evaluation index is selected according to actual needs, which is not limited in this embodiment.
(2) And clustering the first fusion features according to the determined health state classification cluster number.
Specifically, after the number of health status classification clusters is determined, a clustering algorithm may be used to cluster the first fusion features. The first fusion characteristic is obtained by characteristic extraction and fusion of the monitoring data of the hydraulic pump, integrates information of multiple dimensions, and can reflect the health state of the hydraulic pump more comprehensively. The first fusion features can be clustered through K-Means clustering, hierarchical clustering and other algorithms, and the features are divided into different clusters according to the similarity among the features, so that the features in the same cluster have higher similarity, and the features among different clusters have larger difference.
(3) Taking the cluster with the largest feature quantity as the target health state corresponding to the first fusion feature.
Specifically, after the clustering is completed, the feature quantity in each cluster is counted. It will be appreciated that the cluster with the greatest number of features represents the most common state of health of the hydraulic pump, and therefore the cluster with the last number of features is determined to be the target state of health corresponding to the first fusion feature.
(4) And adjusting the cluster number and the cluster center characteristic based on the difference between the target health state and the marked health state corresponding to the first fusion characteristic to obtain a first life prediction model.
Specifically, the marked health state corresponding to the first fusion feature can be obtained through methods such as manual marking, comparison of historical data or auxiliary machine learning marking, and if the target health state and the marked health state have a large difference, the cluster number and the cluster center feature can be considered to be adjusted. For example, if a feature in a cluster is found to not match the health status of the marker, the number of clusters may be readjusted or the central feature of the cluster may be adjusted to improve the accuracy of the clustering.
Further, after the first fusion feature data is obtained, the corresponding health state is known, namely sample data carrying the health state label is formed, the clustering model is trained based on the sample data, and the number of the health state classification clusters and the central feature data of each cluster are determined, so that the first life prediction model is obtained.
Further, methods such as distance evaluation and regression analysis can be adopted, a first life prediction model is constructed according to the first fusion characteristic, the input of the first life prediction model is the first fusion characteristic, and the output of the first life prediction model is the health state of the hydraulic pump.
S104, constructing a second life prediction model according to the second fusion characteristic.
The specific implementation steps comprise:
(1) And acquiring the first life prediction model to identify the real-time health state of the sample hydraulic pump.
Specifically, the first life prediction model is utilized to identify the sample hydraulic pump, so that the real-time health state of the sample hydraulic pump output by the first life prediction model is obtained, and the real-time health state reflects the running condition of the sample hydraulic pump at the current moment.
(2) And inputting the real-time health state of the sample hydraulic pump and the second fusion characteristic into BiLSTM, and training BiLSTM based on the residual life label of the sample hydraulic pump to obtain a second life prediction model.
Specifically, biLSTM is Bidirectional Long Short-Term Memory, the two-way Long-Short Term Memory network, and the BiLSTM model is composed of a forward LSTM model (Long Short-Term Memory network) and a backward LSTM model, wherein the forward LSTM model realizes tasks such as diagnosis and prediction through learning past information, and the backward LSTM model retains future information. During the backward run, it can capture the characteristics of the subsequent part of the sequence. By combining the forward LSTM model with the backward LSTM model, the BiLSTM model is built so that the BiLSTM model can save past and future information at any point in time, thereby more fully understanding the contextual relationship of the data.
Further, the remaining life label is the actual remaining life value of the known sample hydraulic pump. The residual life label can be obtained through actual monitoring and experiments or determined according to experience and expert knowledge, the real-time health state, the second fusion characteristic and the residual life label of the sample hydraulic pump are input BiLSTM, and the parameters of the model are adjusted, so that BiLSTM can learn the mapping relation between the input characteristic and the residual life. In the training process, the difference between the predicted value of the model and the actual residual life label is measured by using the loss function, the difference is continuously reduced by an optimization algorithm, and after repeated iterative training, when the performance of the model reaches the preset requirement, the second life prediction model can be obtained. The input of the second life prediction model is a second fusion characteristic value, and the output of the second life prediction model is the residual life of the hydraulic pump.
S105, fusing the first life prediction model and the second life prediction model to obtain a life prediction model, wherein the real-time health state of the hydraulic pump is identified based on the first life prediction model, and the residual life of the hydraulic pump in the real-time health state is predicted based on the second life prediction model.
Specifically, the input of the first life prediction model is a current state parameter of the hydraulic pump, the output of the first life prediction model is a current health state prediction result of the hydraulic pump, the first life prediction model is used for identifying monitoring data of the hydraulic pump and judging the current health state of the hydraulic pump, the input of the second life prediction model is a state parameter of the hydraulic pump in the current health state predicted by the first life prediction model, the output of the second life prediction model is the residual life of the hydraulic pump in the current health state, and the second life prediction model is used for predicting the residual life of the hydraulic pump according to the health state of the hydraulic pump predicted by the first life prediction model.
Furthermore, the first life prediction model and the second life prediction model may be fused by a hierarchical fusion, a parallel fusion, a feedback fusion, or the like, which is not limited in this embodiment.
For example, in one possible implementation manner, the first life prediction model and the second life prediction model are fused by adopting a hierarchical fusion method, the current health state of the hydraulic pump is firstly evaluated by using the first life prediction model, namely, the first life prediction model outputs the current health state of the hydraulic pump, and further, continuous test data in the current health state of the hydraulic pump output by the first life prediction model in continuous life test data is input into the second life prediction model, so that the second life prediction model predicts the residual life of the hydraulic pump and outputs the residual life of the hydraulic pump.
According to the method for constructing the hydraulic pump life prediction model, in the first aspect, through obtaining the continuous life test data and the sectional life test data of the hydraulic pump, the performance change and the degradation process of the hydraulic pump can be known from different angles and time scales, the actual working condition of the hydraulic pump can be more comprehensively reflected by combining the two types of data, and the deviation and the limitation possibly brought by a single data source are reduced. In the second aspect, the first life prediction model and the second life prediction model are constructed according to the first fusion characteristic and the second fusion characteristic, so that a plurality of indexes and factors can be synthesized, the health state and the degradation trend of the hydraulic pump can be described more accurately, and the accuracy and the reliability of the model on the life prediction of the hydraulic pump are improved. In the third aspect, the health degree of the hydraulic pump can be comprehensively evaluated by comprehensively using the model established by the continuous life test data and the sectional life test data. By predicting the operating state and remaining life of the product in advance, maintenance time and resources can be reasonably arranged, supporting predictive maintenance.
Fig. 2 is a flowchart of a second embodiment of a method for constructing a hydraulic pump life prediction model according to the present application. Referring to fig. 2, on the basis of the above embodiment, the method for acquiring continuous life test data of a hydraulic pump includes:
and S201, respectively performing a durability life test and an accelerated life test on the hydraulic pump.
Specifically, the step of performing a durability life test for the hydraulic pump includes:
(1) And determining the rated operation condition of the hydraulic pump, and building a durability test bed according to the rated operation condition.
Specifically, the nominal operating conditions of a hydraulic pump refer to a specific set of operating conditions, typically including pressure, flow, rotational speed, etc., under which the hydraulic pump is expected to operate stably and reliably in design and normal use. In the durability test of the hydraulic pump, the parameters should be as close as possible to the working conditions of the hydraulic pump in actual operation.
Further, according to the determined rated operation condition, a durability test bed is built.
(2) The endurance test bed builds the working environment of the rated operation working condition, so that the hydraulic pump continuously operates under the rated operation working condition until the hydraulic pump fails.
Specifically, the hydraulic pump is continuously operated on the test bed according to the determined rated operation conditions (pressure, flow rate and rotating speed), and the process simulates the long-term use condition of the hydraulic pump in practical application so as to test the durability and reliability of the hydraulic pump. And (3) formulating specific fault judgment standards according to actual conditions, continuously monitoring and evaluating the state of the hydraulic pump in the test process, and immediately stopping the test once the fault judgment standards are met.
For example, in one possible implementation, the fault determination criteria are:
1. The performance is seriously reduced, such as the flow is obviously reduced, the pressure can not be maintained at the rated value, etc.;
2. Abnormal noise, increased vibration, may indicate damage to or excessive wear of the internal components;
3. Leakage occurs, affecting the proper operation of the hydraulic system.
Further, the step of performing an accelerated life test on the hydraulic pump includes:
(1) Determining the operation condition of the hydraulic pump, and building an acceleration test bed according to the operation condition;
Specifically, in the accelerated life test, the aging and failure processes of the hydraulic pump should be accelerated by increasing the stress condition so as to evaluate the life characteristics thereof in a short time. The operation condition of the hydraulic pump can be determined by increasing the rotation speed, the pressure, the displacement and the like of the hydraulic pump, and also by increasing the working environment temperature, the environment pollution degree and the like, and in the embodiment, the operation condition is not limited.
Further, as in the endurance life test, an acceleration test stand is set up according to the determined operating conditions.
(2) The acceleration test bed builds the working environment of the operation working condition, so that the hydraulic pump continuously operates under the operation working condition, and the operation working condition is adjusted until the hydraulic pump fails.
Specifically, the hydraulic pump is enabled to continuously work on the acceleration test bed according to the determined operation working condition, specific fault judgment standards are formulated according to actual conditions, the state of the hydraulic pump is continuously monitored and evaluated in the test process, and once the fault judgment standards are met, the test is immediately stopped. For example, in one embodiment, the fault criteria are:
1. The performance is seriously reduced, such as the flow is obviously reduced, the pressure can not be maintained at the rated value, etc.;
2. Abnormal noise, increased vibration, may indicate damage to or excessive wear of the internal components;
3. Leakage occurs, affecting the proper operation of the hydraulic system.
S202, recording state parameters of the hydraulic pump under the endurance life test and the accelerated life test in real time.
Specifically, each state parameter of the hydraulic pump under the endurance life test and the accelerated life test is monitored and recorded in real time by a sensor installed on the hydraulic pump.
And S203, combining the state parameters of the endurance life test and the state parameters of the accelerated life test to obtain the continuous life test data.
Specifically, the start time and end time of the endurance life test and the accelerated life test, and the corresponding state parameters at each time point during the test, are determined. And converting the test duration of the accelerated life test by using the acceleration factor by taking the test duration of the endurance life test as a reference time period, so that the test duration of the accelerated life test is the same as the test duration of the endurance life test. For example, in one possible implementation, the acceleration factor is 2 and the test duration of the accelerated life test is 10 hours, corresponding to 20 hours in the endurance life test.
Further, according to the characteristics required to be analyzed, the parameter sequence of the combined data is determined. For example, the order of time, pressure, flow, temperature, vibration, etc. may be arranged. And merging the endurance life test data and the adjusted accelerated life test data row by row. For data at the same time point, if records exist in both tests, one or the average value can be selected to be reserved, and if only one test exists, the records are reserved. Thus, continuous life test data were obtained.
According to the method provided by the embodiment, the hydraulic pump is subjected to a endurance life test and an accelerated life test respectively, and the state parameters of the endurance life test and the state parameters of the accelerated life test are combined and determined to be continuous life test data. Therefore, the performance of the hydraulic pump under different working conditions can be comprehensively covered, a richer data base is provided for life prediction, and the comprehensiveness of test data of the hydraulic pump is ensured. In addition, a single endurance life test may take a long time to obtain enough data, and in practical application, the device may encounter various accidents, resulting in uncertainty of test results, and an accelerated life test may obtain results in a short time, but due to the fact that the acceleration conditions are different from the actual use conditions, a certain deviation may exist in the results, and the data of the two tests can be mutually supplemented and verified, so that reliability and accuracy of life prediction are improved.
Fig. 3 is a flowchart of a third embodiment of a method for constructing a hydraulic pump life prediction model according to the present application. Referring to fig. 3, on the basis of the above embodiment, the method for acquiring the segment life test data of the hydraulic pump includes:
s301, dividing the health state of the hydraulic pump according to the abrasion fault of the hydraulic pump.
Specifically, the health state of the hydraulic pump is divided into:
(1) The good state indicates that the hydraulic pump is not worn and is a new product to be put into use;
(2) The slight abrasion state indicates that the output power of the hydraulic pump is reduced to a certain extent, the function is reduced to a certain extent, but the hydraulic pump can work normally and can be degraded for use;
(3) The moderate abrasion state indicates that the function of the hydraulic pump is reduced, the performance requirement cannot be met, and the hydraulic pump needs to be disassembled and maintained;
(4) The heavy wear state indicates that the hydraulic pump is scrapped, cannot be used and needs to be replaced.
S302, selecting hydraulic pumps in different health states to continuously run for a preset time under the same running condition.
Specifically, for each hydraulic pump in a healthy state, the hydraulic pump is enabled to continuously run for a preset time under the same running condition, and the preset time is set according to actual needs, and in this embodiment, the hydraulic pump is not limited.
Further, the specific steps of the test are described above, and will not be repeated here.
S303, recording test data of the hydraulic pump in the test process under different health states in real time, and obtaining the sectional life test data.
Specifically, through the sensor of installing on the hydraulic pump, the test data of hydraulic pump under the different health conditions in the test of real-time supervision and record to the test data that different health conditions correspond is in the same place integrated, obtains segmentation life test data.
The method provided by the embodiment divides the health state of the hydraulic pump according to the abrasion faults of the hydraulic pump, selects the hydraulic pumps in different health states to continuously run for preset time under the same running working condition, and records the test data of the hydraulic pumps in different health states in the test process in real time to obtain the sectional life test data. The performance of the hydraulic pump at different wear stages can be more accurately known. Different wear failures correspond to different health degrees and can provide more targeted information for subsequent analysis. And by monitoring the abrasion fault, the potential problem can be found in time before the hydraulic pump has serious fault. The early fault signs such as slight abrasion can be recognized earlier by dividing according to different health states, and sufficient time is provided for taking preventive maintenance measures.
Fig. 4 is a flowchart of a fourth embodiment of a method for constructing a hydraulic pump life prediction model according to the present application. Referring to fig. 4, after the life prediction model is obtained, the method further includes:
s401, detecting real-time state data of the hydraulic pump to be evaluated.
Specifically, the real-time data of the hydraulic pump to be evaluated in the running process is collected through various sensors and monitoring equipment, and the real-time state data of the hydraulic pump comprise the current parameters of pressure pulsation, vibration information, temperature, flow and the like of the hydraulic pump.
S402, identifying the state category to which the real-time health state of the hydraulic pump to be evaluated belongs based on the real-time state data and the first life prediction model.
Specifically, in connection with the above description, the first life prediction model is used to identify a hydraulic pump health status category, which classifies the current health status of the hydraulic pump into different status categories according to the input real-time status data of the hydraulic pump. It should be noted that the health status of the hydraulic pump is preset, for example, in one possible implementation, the health status of the hydraulic pump includes a normal status, a light wear status, a moderate wear status, a heavy wear status, and a fault status.
S403, the second life prediction model predicts the state holding time of the hydraulic pump to be evaluated based on the belonging state category.
Specifically, the input of the second life prediction model is the state type of the hydraulic pump output by the first life prediction model, according to the state type, the state holding time of the hydraulic pump to be evaluated is predicted in combination with the historical data and the equipment characteristics of the hydraulic pump in the state type, and it is understood that the state holding time of the hydraulic pump refers to the time length of the hydraulic pump which is expected to be capable of continuing to normally run in the current state type.
S404, determining the experience time length of all states by taking the belonging state type as a starting point and the state holding time as an interval, and obtaining the residual life of the hydraulic pump to be evaluated.
Specifically, the state class to which the first life prediction model output belongs is taken as a starting point. Then, the time duration of the hydraulic pump in different states is estimated gradually and forwards with the state maintaining time output by the second life prediction model as an interval. By accumulating the time of each state, the total time of the hydraulic pump from the initial state to the current state can be obtained, and the residual life of the hydraulic pump to be evaluated can be obtained by subtracting the time of the elapsed life according to the rated life of the hydraulic pump.
According to the method provided by the embodiment, the real-time state data of the hydraulic pump is obtained, the health state of the hydraulic pump is classified by using the first life prediction model, so that maintenance personnel can quickly know the current state of equipment, the urgency and the priority of maintenance are determined, early warning can be sent out before serious faults of the hydraulic pump occur through recognition of the state types, clear time references are provided by using the second life prediction model based on the state retention time predicted by the state types, maintenance or replacement is ensured before the faults of the equipment, and an enterprise can be helped to plan production tasks better through accurate state retention time prediction. The production interruption caused by the sudden failure of equipment is avoided, and the production efficiency is improved. In sum, through real-time state data detection, the state type identification is predicted by using the first life prediction model, the state retention time is predicted by using the second life prediction model based on the state type, comprehensive and accurate information is provided for the maintenance and management of the hydraulic pump, the reliability and the production efficiency of the equipment are improved, and the maintenance cost is reduced.
Corresponding to the embodiment of the method for constructing the hydraulic pump life prediction model, the application also provides an embodiment of the device for constructing the hydraulic pump life prediction model.
Fig. 5 is a schematic structural diagram of an embodiment of a device for constructing a hydraulic pump life prediction model according to the present application. Referring to fig. 5, the apparatus provided in this embodiment includes an obtaining module 510, an analyzing module 520, and a constructing module 530, where:
The acquiring module 510 is configured to acquire segment life test data and continuous life test data of the hydraulic pump;
the analysis module 520 is configured to perform degradation feature analysis on the segment life test data and the continuous life test data to obtain a first fusion feature and a second fusion feature, where the first fusion feature is a feature corresponding to the segment life test data, and the second fusion feature is a feature corresponding to the continuous life test data;
the constructing module 530 is configured to construct a first life prediction model according to the first fusion feature;
the constructing module 530 is further configured to construct a second life prediction model according to the second fusion feature;
the construction module 530 is further configured to fuse the first life prediction model and the second life prediction model to obtain a life prediction model, where the real-time health state of the hydraulic pump is identified based on the first life prediction model, and the remaining life of the hydraulic pump in the real-time health state is predicted based on the second life prediction model.
The apparatus of this embodiment may be used to execute the steps of the method embodiment shown in fig. 1, and the specific implementation principle and implementation process are similar, and are not described herein again.
The implementation process of the functions and roles of each unit in the above device is specifically shown in the implementation process of the corresponding steps in the above method, and will not be described herein again.
For the device embodiments, reference is made to the description of the method embodiments for the relevant points, since they essentially correspond to the method embodiments. The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purposes of the present application. Those of ordinary skill in the art will understand and implement the present application without undue burden.
The foregoing description of the preferred embodiments of the application is not intended to be limiting, but rather to enable any modification, equivalent replacement, improvement or the like to be made within the spirit and principles of the application.