A method for health monitoring and fault diagnosis of a planetary gear reducer

Abstract

The present invention discloses a health monitoring and fault diagnosis method for a planetary gear reduction device, which relates to the technical field related to transmission devices. By identifying key parts of a target planetary gear reduction device, a set of key parts of the reduction device is obtained, and a sensor group is arranged to obtain a working perception module of the reduction device. The working data stream set of the key parts is monitored, and distributed preprocessing and fault feature extraction are performed to obtain a set of working features of key part faults. The reduction device fault data set is obtained by graph sampling, and logical classification and training are performed to generate a fault diagnosis model library. Based on the reduction device fault diagnosis model library, the working feature set of key part faults is matched, associated and fault diagnosed to obtain target reduction device fault diagnosis information. The technical problem that it is difficult for planetary gear reduction devices in the prior art to monitor and diagnose key component faults in real time and accurately, resulting in untimely fault identification and affecting the safety and reliability of equipment operation is solved.

Description

Technical Field

[0001] The present application relates to the technical field related to transmission devices, and in particular to a health monitoring and fault diagnosis method for a planetary gear reduction device. Background Art

[0002] Planetary gear reducers are important equipment widely used in wind power, machinery manufacturing, mining and other fields. Their health monitoring and fault diagnosis are crucial to the reliable operation of the equipment. However, due to the complex working environment, the key components of the reducer are easily affected by external factors such as high temperature and high load during long-term operation, causing equipment wear or failure. Existing monitoring methods often rely on single signal monitoring, which makes it difficult to conduct a comprehensive real-time assessment of the equipment's operating status, and the accuracy and timeliness of fault identification are low.

[0003] Therefore, in the prior art, it is difficult for planetary gear reduction devices to monitor and diagnose key component failures in real time and accurately, resulting in technical problems such as untimely fault identification and affecting the safety and reliability of equipment operation. Summary of the invention

[0004] This application provides a health monitoring and fault diagnosis method for a planetary gear reduction device, which solves the technical problem in the prior art that it is difficult to monitor and diagnose key component faults in real time and accurately, resulting in untimely fault identification and affecting the safety and reliability of equipment operation. By constructing application scenario calibration parameters, deploying multiple sensor groups, and combining an intelligent fault diagnosis model library, the working status and fault characteristics of key components of the reduction device can be obtained in real time and accurately, effectively improving the accuracy of fault identification and the reliability of equipment operation.

[0005] The present application provides a health monitoring and fault diagnosis method for a planetary gear reduction device, the method comprising: identifying key parts of a target planetary gear reduction device to obtain a set of key parts of the reduction device, and deploying a sensor group on the set of key parts of the reduction device to obtain a reduction device working perception module; monitoring and obtaining a set of working data streams of key parts of the reduction device through the reduction device working perception module, and performing distributed preprocessing and fault feature extraction on the set of working data streams of key parts of the reduction device to obtain a set of working features of key parts faults; obtaining a reduction device fault data set by graph sampling, logically classifying and training the reduction device fault data set by using application scenario influencing factor information, and generating a reduction device fault diagnosis model library; matching, associating and fault-diagnosing the set of working features of key parts faults based on the reduction device fault diagnosis model library to obtain target reduction device fault diagnosis information, and evaluating and obtaining a reduction device health coefficient based on the target reduction device fault diagnosis information.

[0006] In the implementation method, the obtaining of the key parts set of the reduction device includes: obtaining application scenario information of the target planetary gear reduction device, performing working condition analysis on the application scenario information, and obtaining a set of simulated working condition parameters of the reduction device; performing associated extraction of performance evaluation indicators based on the simulated working condition parameter set of the reduction device, and obtaining a set of working performance indicators of the reduction device; obtaining a reduction device working simulation parameter table according to the set of simulated working condition parameters of the reduction device and the set of working performance indicators of the reduction device; performing simulation test records on the target planetary gear reduction device according to the set of simulated working parameters of the reduction device, and obtaining a set of working test data of the reduction device; mapping and associating the set of working test data of the reduction device with a set of structural parts of the target planetary gear reduction device, and obtaining a set of test data of structural parts of the reduction device; performing influence degree analysis and screening on the set of structural parts based on the test data set of structural parts of the reduction device, and obtaining the set of key parts of the reduction device.

[0007] In the implementation method, the acquisition of the key part fault working feature set includes: performing noise characteristic extraction on the key part working data stream set of the deceleration device to obtain data noise characteristic information, and determining a data wavelet selection threshold value based on the data noise characteristic information; performing multi-scale decomposition on the key part working data stream set of the deceleration device to obtain a data multi-scale wavelet coefficient set; performing available signal interception and filtering reconstruction on the data multi-scale wavelet coefficient set according to the data wavelet selection threshold to obtain an available key part working data stream set; performing principal component analysis and cluster extraction on the available key part working data stream set to obtain the key part fault working feature set.

[0008] In the implementation method, the method of obtaining the key part fault working feature set includes: performing decentralized processing based on the available key part working data flow sets to obtain a key part covariance matrix set; performing feature decomposition and principal component solution selection on the key part covariance matrix set in turn to obtain a principal component information set, performing projection mapping transformation on the available key part working data flow set through the principal component information set to determine the key part working data flow set associated feature set; performing tomographic clustering analysis on the key part working data flow set associated feature set to obtain a key part working data clustering result set; performing fault mode identification and fault feature association based on the key part working data clustering result set to obtain the key part fault working feature set.

[0009] In the implementation method, the graph sampling obtains the reduction device fault data set, including: constructing a reduction device fault database, matching the attributes of the specification and model information of the target planetary gear reduction device with the reduction device fault database, and obtaining a first-order similar sample data set; taking the first-order similar sample data set as a benchmark, collecting neighboring sample data in the reduction device fault database as a second-order similar sample data set; traversing the second-order similar sample data set, performing N-order growth collection in the reduction device fault database, and obtaining an N-order similar sample data set, wherein 2≤N≤10; and integrating the graph structure based on the first-order similar sample data set, the second-order similar sample data set, and the N-order similar sample data set until the N-order similar sample data set, to obtain the reduction device fault data set.

[0010] In the implementation method, the generation of the reduction device fault diagnosis model library includes: performing fault diagnosis correlation analysis on each influencing factor in the application scenario influencing factor information to obtain a set of influencing factor correlation coefficients; screening a set of associated influencing factors in the application scenario influencing factor information that reaches a preset correlation coefficient threshold according to the set of influencing factor correlation coefficients; parsing the application scenario information of the target planetary gear reduction device according to the set of associated influencing factors to obtain an application scenario calibration parameter set; performing logical clustering and training on the reduction device fault data set based on the application scenario calibration parameter set to generate the reduction device fault diagnosis model library.

[0011] In the implementation method, the generation of the deceleration device fault diagnosis model library includes: logically clustering the deceleration device fault data set based on the application scenario calibration parameter set to obtain a fault calibration cluster data set; performing fault diagnosis analysis training on the fault calibration cluster data set to obtain a calibrated deceleration device fault diagnosis model set; performing performance verification on the calibrated deceleration device fault diagnosis model set to obtain diagnostic model performance parameters, and determining a model parameter tuning strategy based on the diagnostic model performance parameters; and tuning, updating and integrating the calibrated deceleration device fault diagnosis model set based on the model parameter tuning strategy to generate the deceleration device fault diagnosis model library.

[0012] In the implementation method, the target reduction device fault diagnosis information is obtained, including: classifying and identifying the target application scenario of the target planetary gear reduction device through the set of associated influencing factors to obtain the target application scenario calibration parameters; performing model matching and association with the reduction device fault diagnosis model library based on the target application scenario calibration parameters to obtain the target reduction device fault diagnosis model; using the target reduction device fault diagnosis model to perform fault diagnosis analysis on the key part fault working feature set, and outputting the target reduction device fault diagnosis information.

[0013] The health monitoring and fault diagnosis method of a planetary gear reducer proposed in this application is proposed. By identifying the key parts of the target planetary gear reducer, a set of key parts of the reducer is obtained, and a sensor group is arranged on the key parts of the reducer to obtain a working perception module of the reducer; the working data stream set of the key parts of the reducer is monitored and obtained by the working perception module of the reducer, and the distributed preprocessing and fault feature extraction of the working data stream set of the key parts of the reducer are performed to obtain a set of working features of key parts faults; the fault data set of the reducer is obtained by graph sampling, and the fault data set of the reducer is logically classified and trained using the information of the influencing factors of the application scenario to generate a fault diagnosis model library of the reducer; the key parts fault working feature set is matched, associated and fault diagnosed based on the fault diagnosis model library of the reducer to obtain the fault diagnosis information of the target reducer, and the health coefficient of the reducer is evaluated based on the fault diagnosis information of the target reducer. The technical problem that it is difficult for planetary gear reducers in the prior art to monitor and diagnose key component faults in real time and accurately, resulting in untimely fault identification and affecting the safety and reliability of equipment operation is solved. By constructing application scenario calibration parameters, deploying multiple sensor groups, and combining an intelligent fault diagnosis model library, the working status and fault characteristics of key components of the reduction device can be obtained in real time and accurately, effectively improving the accuracy of fault identification and the reliability of equipment operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0015] Figure 1 A schematic flow chart of a health monitoring and fault diagnosis method for a planetary gear reduction device provided in an embodiment of the present application;

[0016] Figure 2 A schematic flow chart of a health monitoring and fault diagnosis method for a planetary gear reduction device provided in an embodiment of the present application for obtaining a set of key parts of the reduction device. DETAILED DESCRIPTION

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0018] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0019] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server 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 modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0020] The present application embodiment provides a health monitoring and fault diagnosis method for a planetary gear reduction device, such as Figure 1 As shown, the method includes:

[0021] Identify key parts of a target planetary gear reduction device to obtain a set of key parts of the reduction device, and arrange a sensor group on the set of key parts of the reduction device to obtain a reduction device working perception module; monitor and obtain a set of working data streams of key parts of the reduction device through the reduction device working perception module, and perform distributed preprocessing and fault feature extraction on the set of working data streams of key parts of the reduction device to obtain a set of working features of key parts faults;

[0022] The key parts of the target planetary gear reduction device are identified to obtain a set of key parts of the reduction device. A sensor group is arranged on the set of key parts of the reduction device. The sensor group is used to monitor the status of the key parts in real time, including temperature sensors, vibration sensors and pressure sensors, to obtain a reduction device working perception module. Subsequently, the working data stream set of the key parts of the reduction device is monitored and obtained through the working perception module of the reduction device, and the working data stream set of the key parts of the reduction device is distributedly preprocessed and fault feature extraction is performed on the working data stream set of the key parts of the reduction device to obtain a set of key part fault working features.

[0023] like Figure 2 As shown, the method provided in the embodiment of the present application also includes: obtaining application scenario information of the target planetary gear reduction device, performing working condition analysis on the application scenario information, and obtaining a set of parameters for simulated working conditions of the reduction device; performing associated extraction of performance evaluation indicators based on the set of parameters for simulated working conditions of the reduction device, and obtaining a set of working performance indicators of the reduction device; obtaining a reduction device working simulation parameter table according to the set of parameters for simulated working conditions of the reduction device and the set of working performance indicators of the reduction device; performing simulation test records on the target planetary gear reduction device according to the set of parameters for simulated working conditions of the reduction device, and obtaining a set of working test data of the reduction device; mapping and associating the set of working test data of the reduction device with a set of structural parts of the target planetary gear reduction device, and obtaining a set of test data of structural parts of the reduction device; performing influence degree analysis and screening on the set of structural parts based on the set of test data of structural parts of the reduction device, and obtaining a set of key parts of the reduction device.

[0024] The application scenario information of the target planetary gear reduction device is obtained. The application scenario information refers to the specific conditions of the environment in which the planetary gear reduction device is located, including operating conditions, load characteristics, temperature range, etc. Subsequently, the application scenario information is analyzed for working conditions, and the application scenario information is decomposed into specific working parameters, such as ambient temperature, pressure, speed, and load, etc., to obtain a set of parameters for simulated working conditions of the reduction device. Further, based on the set of parameters for simulated working conditions of the reduction device, performance evaluation index association extraction is performed. The performance evaluation index is a quantitative standard used to evaluate the performance of the reduction device under specific conditions, such as efficiency, temperature rise, noise, etc., to obtain a set of working performance indicators of the reduction device.

[0025] Subsequently, according to the reduction device simulation working condition parameter set and the reduction device working performance index set, the working condition parameters and the reduction device working performance index are integrated to obtain a reduction device working simulation parameter table. Further, according to the reduction device working simulation parameter table, a simulation test record is performed on the target planetary gear reduction device, that is, the performance data of the reduction device under different working conditions obtained through actual testing or computer simulation is obtained to obtain a reduction device working test data set, and the reduction device working test data set contains performance evaluation index data corresponding to multiple working condition parameters.

[0026] Finally, the speed reducer working test data set is mapped and associated with the structural parts set of the target planetary gear speed reducer. During the mapping, the working performance indicators in the speed reducer working test data set can be mapped with the structural parts set according to professional and technical personnel, such as the mapping relationship between noise and bearings, lubrication systems, and gears. After the mapping and association, each working performance indicator corresponds to each structural part, thereby obtaining the speed reducer structural part test data set. Based on the speed reducer structural part test data set, the structural part set is analyzed and screened for influence, and the frequency of occurrence of the structural parts corresponding to each working performance indicator in the speed reducer structural part test data set is obtained. The structural parts with a number greater than or equal to the threshold number are obtained as key structural parts, and the speed reducer key part set is obtained. The threshold number is the frequency threshold number of occurrences of the structural parts corresponding to the pre-set working performance indicators.

[0027] The method provided in the embodiment of the present application also includes: extracting noise characteristics of the working data stream set of the key parts of the deceleration device to obtain data noise characteristic information, and determining a data wavelet selection threshold based on the data noise characteristic information; performing multi-scale decomposition on the working data stream set of the key parts of the deceleration device to obtain a data multi-scale wavelet coefficient set; performing available signal interception and filtering reconstruction on the data multi-scale wavelet coefficient set according to the data wavelet selection threshold to obtain a set of available working data streams of key parts; performing principal component analysis and cluster extraction on the available working data stream set of key parts to obtain a set of fault working characteristics of the key parts.

[0028] The spectral characteristics of the data of the working data stream set of the key parts of the reduction device are analyzed to generate the spectrum diagram of the signal, observe the amplitude of different frequency bands, and complete the noise characteristic extraction based on different acquisition objects. For example, in the vibration signal of the planetary gear reduction device, the frequency components above 50Hz mainly come from the environmental noise, so these components can be considered as noise, and the data noise characteristic information is obtained. And the data wavelet selection threshold is determined according to the data noise characteristic information. Subsequently, the working data stream set of the key parts of the reduction device is multi-scaled to obtain a data multi-scale wavelet coefficient set. Further, according to the data wavelet selection threshold, the data multi-scale wavelet coefficient set is subjected to available signal interception and filtering reconstruction to obtain an available key part working data stream set. Finally, the principal component analysis and cluster extraction are performed on the available key part working data stream set to obtain the key part fault working feature set.

[0029] The method provided by the embodiment of the present application also includes: performing decentralized processing based on the available key part working data flow sets respectively to obtain a key part covariance matrix set; performing feature decomposition and principal component solution selection on the key part covariance matrix set in turn to obtain a principal component information set, and performing projection mapping transformation on the available key part working data flow set through the principal component information set to determine the key part working data flow set associated feature set; performing tomographic clustering analysis on the key part working data flow set associated feature set to obtain a key part working data clustering result set; performing fault mode identification and fault feature association based on the key part working data clustering result set to obtain the key part fault working feature set.

[0030] The method of obtaining the key part fault working feature set includes: performing decentralized processing based on the available key part working data stream set, that is, subtracting the mean of the data set from each data point in the data set, so that the mean of the processed data is zero, eliminating the absolute offset of the data, so that the subsequent analysis can better reflect the relative change of the data, and after the decentralized processing, performing covariance calculation on the working data stream set of all key parts to obtain the key part covariance matrix set. Exemplarily, a vibration sensor monitors the vibration of the gear, and another monitors the temperature. If the calculated covariance is positive and large, then when the gear vibration increases, the temperature also increases accordingly, which can indicate that the vibration causes mechanical friction, thereby causing the temperature to rise.

[0031] The covariance matrix set of the key parts is subjected to eigendecomposition and principal component solution selection in turn, and the covariance matrix is ​​subjected to eigendecomposition to obtain eigenvalues ​​and eigenvectors. The larger the eigenvalue, the stronger the explanatory power of the direction corresponding to the eigenvector is for the data. The largest eigenvalue is selected as the corresponding principal component. A principal component information set is obtained, and the available key part working data stream set is projected and mapped through the principal component information set. The principal component information is applied to the available key part working data stream set, and the data is projected into a low-dimensional space defined by the principal component, thereby simplifying the processing and analysis of the data, so that only the most important characteristic information is retained, and the associated characteristic set of the key part working data stream set is determined. Further, hierarchical clustering analysis is performed on the associated characteristic sets of the key part working data stream set, and similar data features are classified into the same category, and a clustering result set of key part working data is obtained. Finally, fault mode identification and fault feature association are performed based on the clustering result set of the key parts working data, that is, the different clusters obtained by tomographic clustering are matched with the known fault modes. The fault mode is the monitoring data feature caused by the fault, so as to identify the specific fault type of the equipment and obtain the key parts fault working feature set.

[0032] The reduction device fault data set is obtained by graph sampling, and the application scenario influencing factor information is used to perform logical classification and training on the reduction device fault data set to generate a reduction device fault diagnosis model library; based on the reduction device fault diagnosis model library, the key parts fault working feature set is matched, associated and fault diagnosed to obtain the target reduction device fault diagnosis information; and based on the target reduction device fault diagnosis information, the reduction device health factor is evaluated and obtained.

[0033] The speed reducer fault data set is obtained by image sampling, and the speed reducer fault data set is logically classified and trained using the application scenario influencing factor information to generate a speed reducer fault diagnosis model library, which contains multiple fault diagnosis models for different scenarios. Finally, the key parts fault working feature set is matched, associated and fault diagnosed based on the speed reducer fault diagnosis model library to obtain the target speed reducer fault diagnosis information, which contains the diagnosis results of whether each reducer structure is faulty, and based on the target speed reducer fault diagnosis information, when there is a fault, the diagnostic coefficient of the corresponding structure is 1, otherwise it is 0, and the diagnostic coefficient is weighted and summed to evaluate the speed reducer health coefficient, where the weight parameter is set according to the importance of the structure by professional and technical personnel. The technical problem that it is difficult for planetary gear speed reducers in the prior art to monitor and diagnose key component faults in real time and accurately, resulting in untimely fault identification and affecting the safety and reliability of equipment operation is solved. By constructing application scenario calibration parameters, deploying multiple sensor groups, and combining an intelligent fault diagnosis model library, the working status and fault characteristics of key components of the reduction device can be obtained in real time and accurately, effectively improving the accuracy of fault identification and the reliability of equipment operation.

[0034] The method provided in the embodiment of the present application also includes: constructing a reduction device fault database, matching attributes with the reduction device fault database through specification and model information of the target planetary gear reduction device, and obtaining a first-order similar sample data set; using the first-order similar sample data set as a benchmark, collecting neighboring sample data in the reduction device fault database as a second-order similar sample data set; traversing the second-order similar sample data set, performing N-order growth collection in the reduction device fault database, and obtaining an N-order similar sample data set, where 2≤N≤10; and integrating graph structures based on the first-order similar sample data set, the second-order similar sample data set, and the N-order similar sample data set to obtain the reduction device fault data set.

[0035] The graph sampling obtains the speed reducer fault data set, including: collecting a large amount of historical data about the planetary gear speed reducer, including the fault records of various speed reducer structures of speed reducers of different specifications and models under various working conditions. These data can be obtained from the manufacturer's maintenance records, sensor data, maintenance reports and other channels to build a speed reducer fault database. Through the specification and model information of the target planetary gear speed reducer, such as equipment type, workload, environmental conditions, etc., attribute matching is performed in the speed reducer fault database to find those equipment fault records with the same or very similar attributes to form a first-order similar sample data set. Based on the first-order similar sample data set as a benchmark, the adjacent sample data, i.e., the sample data of similar equipment, such as speed reducers with only power differences, are collected in the speed reducer fault database as a second-order similar sample data set. Starting from the second-order similar sample data set, the similarity range is gradually expanded to collect third-order, fourth-order, and even N-order similar sample data sets. Compared with the previous order, the sample data set of each order will have greater differences in some attributes, but still maintain a certain similarity, where 2≤N≤10. Finally, the first-order, second-order, and N-order similar sample data sets are constructed into a graph structure according to the similarity relationship between the samples. In the figure, the distance between nodes represents the similarity between samples. The higher the similarity, the shorter the distance between nodes. The speed reducer fault data set is obtained.

[0036] The method provided in the embodiment of the present application also includes: performing fault diagnosis correlation analysis on each influencing factor in the application scenario influencing factor information to obtain a set of influencing factor correlation coefficients; screening a set of associated influencing factors in the application scenario influencing factor information that reaches a preset correlation coefficient threshold according to the set of influencing factor correlation coefficients; parsing the application scenario information of the target planetary gear reduction device according to the set of associated influencing factors to obtain an application scenario calibration parameter set; performing logical clustering and training on the reduction device fault data set based on the application scenario calibration parameter set to generate the reduction device fault diagnosis model library.

[0037] The generation of the deceleration device fault diagnosis model library includes: performing fault diagnosis correlation analysis on each influencing factor in the application scenario influencing factor information through a deceleration device fault data set, and when performing the correlation analysis, statistics are performed on each fault type in the deceleration device fault data set and the corresponding application scenario influencing factor information to obtain the statistical result of the application scenario influencing factor information corresponding to the fault type, obtain the ratio of the frequency of each influencing factor information to the sum of the frequencies of all influencing factor information, obtain the correlation coefficient corresponding to the influencing factor, obtain the correlation coefficients of all influencing factors, and obtain a set of influencing factor correlation coefficients.

[0038] Subsequently, the set of associated influencing factors that reach a preset correlation coefficient threshold in the application scenario influencing factor information is screened according to the set of influencing factor correlation coefficients. The preset correlation coefficient threshold is a threshold of the pre-set correlation coefficient. When the correlation coefficient of the influencing factor is higher than the threshold, it is considered that the factor has a significant impact on the fault and is included in the set of associated influencing factors. The application scenario information of the target planetary gear reduction device is parsed according to the set of associated influencing factors, and the corresponding set of associated influencing factors under each application scenario information is obtained to obtain an application scenario calibration parameter set. Finally, based on the application scenario calibration parameter set, the reduction device fault data set is logically clustered and trained, the fault data set is clustered according to the application scenario, and the machine learning model is trained using the data set after logical clustering to generate multiple independent fault diagnosis models, each of which corresponds to the analysis of multiple structural faults under a fault scenario. Multiple fault diagnosis models are integrated to generate the reduction device fault diagnosis model library.

[0039] The method provided in the embodiment of the present application also includes: logically clustering and identifying the deceleration device fault data set based on the application scenario calibration parameter set to obtain a fault calibration cluster data set; performing fault diagnosis analysis training on the fault calibration cluster data set respectively to obtain a calibrated deceleration device fault diagnosis model set; performing performance verification on the calibrated deceleration device fault diagnosis model set to obtain diagnostic model performance parameters, and determining a model parameter tuning strategy based on the diagnostic model performance parameters; and tuning, updating and integrating the calibrated deceleration device fault diagnosis model set respectively based on the model parameter tuning strategy to generate the deceleration device fault diagnosis model library.

[0040] Based on the application scenario calibration parameter set, the deceleration device fault data set is logically clustered and identified, the application scenario corresponding to the deceleration device fault data set is identified, clustering is performed based on the application scenario identification, and a fault calibration clustering data set is obtained. Fault diagnosis analysis training is performed on the fault calibration clustering data sets respectively, and the trained models are obtained to obtain a calibrated deceleration device fault diagnosis model set. Furthermore, using the fault calibration clustering data set, the performance of the calibrated deceleration device fault diagnosis model set is verified to obtain diagnostic model performance parameters, such as prediction accuracy, and the model parameter tuning strategy is determined based on the diagnostic model performance parameters. Based on the model parameter tuning strategy, the calibrated deceleration device fault diagnosis model set is tuned, updated and integrated respectively to generate the deceleration device fault diagnosis model library.

[0041] The method provided in the embodiment of the present application also includes: classifying and identifying the target application scenarios of the target planetary gear reduction device through the set of associated influencing factors to obtain the target application scenario calibration parameters; performing model matching and association with the reduction device fault diagnosis model library based on the target application scenario calibration parameters to obtain the target reduction device fault diagnosis model; using the target reduction device fault diagnosis model to perform fault diagnosis analysis on the key part fault working feature set, and outputting the target reduction device fault diagnosis information.

[0042] The method of obtaining the target reduction device fault diagnosis information includes: acquiring the collection of the target reduction device through the associated influencing factor set, classifying and identifying the target application scenario of the target planetary gear reduction device, and obtaining the target application scenario calibration parameters. Based on the target application scenario calibration parameters, the model matching association is performed with the reduction device fault diagnosis model library to obtain the target reduction device fault diagnosis model. Finally, the target reduction device fault diagnosis model is used to perform fault diagnosis analysis on the key part fault working feature set, and the target reduction device fault diagnosis information is output.

[0043] The technical solution provided by the embodiment of the present invention is to identify the key parts of the target planetary gear reducer, obtain the key parts set of the reducer, and deploy the sensor group to obtain the working perception module of the reducer. Monitor the working data stream set of the key parts, and perform distributed preprocessing and fault feature extraction to obtain the working feature set of the key parts fault. Graph sampling is used to obtain the reducer fault data set, and logical classification and training are performed to generate a fault diagnosis model library. Based on the reducer fault diagnosis model library, the key parts fault working feature set is matched, associated and fault diagnosed to obtain the target reducer fault diagnosis information. The technical problem that the planetary gear reducer in the prior art is difficult to monitor and diagnose the key component faults in real time and accurately, resulting in untimely fault identification and affecting the safety and reliability of equipment operation is solved. By constructing application scenario calibration parameters, deploying multiple sensor groups, and combining with the intelligent fault diagnosis model library, the working status and fault characteristics of the key components of the reducer can be obtained in real time and accurately, effectively improving the accuracy of fault identification and the technical effect of the reliability of equipment operation.

[0044] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A health monitoring and fault diagnosis method for a planetary gear reduction device, characterized in that: The method comprises: Identify key parts of a target planetary gear reduction device to obtain a reduction device key part set, and arrange a sensor group on the reduction device key part set to obtain a reduction device working perception module; The working data stream set of the key parts of the deceleration device is monitored and acquired through the deceleration device working perception module, and the working data stream set of the key parts of the deceleration device is distributedly preprocessed and fault feature extracted to acquire the working feature set of the key parts fault; The image sampling method is used to obtain a speed reduction device fault data set, and the application scenario influencing factor information is used to perform logical classification and training on the speed reduction device fault data set to generate a speed reduction device fault diagnosis model library; Based on the deceleration device fault diagnosis model library, matching, associating and fault diagnosis are performed on the key parts fault working feature set to obtain target deceleration device fault diagnosis information, and based on the target deceleration device fault diagnosis information, evaluating and obtaining the deceleration device health factor; The image sampling to obtain a speed reduction device fault data set includes: Constructing a reduction gear fault database, performing attribute matching between the specification and model information of the target planetary gear reduction gear and the reduction gear fault database, and obtaining a first-order similarity sample data set; Taking the first-order similar sample data set as a benchmark, collecting adjacent sample data in the speed reducer fault database as a second-order similar sample data set; Traversing the second-order similar sample data set, performing N-order growth collection in the speed reducer fault database, and obtaining an N-order similar sample data set, where 2≤N≤10; Based on the first-order similar sample data set, the second-order similar sample data set, and up to the N-order similar sample data set, a graph structure is integrated to obtain the speed reduction device fault data set.

2. A health monitoring and fault diagnosis method for a planetary gear reduction device according to claim 1, characterized in that: The step of obtaining a set of key parts of a deceleration device comprises: Acquire application scenario information of the target planetary gear reduction device, perform working condition analysis on the application scenario information, and acquire a set of simulated working condition parameters of the reduction device; Based on the reduction device simulated working condition parameter set, a performance evaluation index correlation extraction is performed to obtain a reduction device working performance index set; According to the reduction device simulation working condition parameter set and the reduction device working performance index set, a reduction device working simulation parameter table is obtained; Performing simulation test records on the target planetary gear reduction device according to the reduction device working simulation parameter table to obtain a reduction device working test data set; Mapping and associating the reduction gear working test data set with the structural parts set of the target planetary gear reduction gear to obtain a reduction gear structural parts test data set; Based on the test data set of the reduction gear structural parts, the influence degree of the structural parts set is analyzed and screened to obtain the key parts set of the reduction gear.

3. A health monitoring and fault diagnosis method for a planetary gear reduction device according to claim 1, characterized in that: The step of obtaining a key part fault working feature set includes: Extracting noise characteristics from a set of working data streams of key parts of the deceleration device to obtain data noise characteristic information, and determining a data wavelet selection threshold according to the data noise characteristic information; Performing multi-scale decomposition on a set of working data streams of key parts of the deceleration device to obtain a set of multi-scale wavelet coefficients of the data; According to the data wavelet selection threshold, the data multi-scale wavelet coefficient set is subjected to available signal interception and filtering reconstruction to obtain an available key part working data stream set; The principal component analysis and cluster extraction are performed on the available key part working data stream set to obtain the key part fault working feature set.

4. A health monitoring and fault diagnosis method for a planetary gear reduction device as claimed in claim 3, characterized in that: The step of obtaining the key part fault operating characteristic set comprises: Decentralizing the available key part work data stream sets to obtain a key part covariance matrix set; Performing feature decomposition and principal component solution selection on the key part covariance matrix set in turn to obtain a principal component information set, performing projection mapping transformation on the available key part work data stream set through the principal component information set, and determining a key part work data stream set associated feature set; Performing hierarchical clustering analysis on the associated feature sets of the key parts work data flow set, and obtaining a key parts work data clustering result set; Fault mode identification and fault feature association are performed based on the key part working data clustering result set to obtain the key part fault working feature set.

5. The health monitoring and fault diagnosis method of a planetary gear reduction device according to claim 1, characterized in that: The generating of the speed reducer fault diagnosis model library comprises: Performing a fault diagnosis correlation analysis on each influencing factor in the application scenario influencing factor information to obtain an influencing factor correlation coefficient set; According to the influencing factor correlation coefficient set, a set of associated influencing factors in the application scenario influencing factor information that reaches a preset correlation coefficient threshold is filtered; Analyzing the application scenario information of the target planetary gear reduction device according to the associated influencing factor set to obtain an application scenario calibration parameter set; Based on the application scenario calibration parameter set, the deceleration device fault data set is logically clustered and trained to generate the deceleration device fault diagnosis model library.

6. A health monitoring and fault diagnosis method for a planetary gear reduction device as claimed in claim 5, characterized in that: The generating of the speed reduction device fault diagnosis model library comprises: Based on the application scenario calibration parameter set, logical clustering identification is performed on the deceleration device fault data set to obtain a fault calibration cluster data set; Performing fault diagnosis analysis training on the fault calibration clustering data sets respectively to obtain a calibration speed reduction device fault diagnosis model set; Performing performance verification on the calibrated speed reducer fault diagnosis model set to obtain diagnostic model performance parameters, and determining a model parameter tuning strategy based on the diagnostic model performance parameters; Based on the model parameter tuning strategy, the calibrated speed reduction device fault diagnosis model set is tuned, updated and integrated respectively to generate the speed reduction device fault diagnosis model library.

7. A health monitoring and fault diagnosis method for a planetary gear reduction device as claimed in claim 6, characterized in that: The obtaining of target deceleration device fault diagnosis information includes: Classify and identify the target application scenario of the target planetary gear reduction device through the set of associated influencing factors to obtain calibration parameters of the target application scenario; Based on the target application scenario calibration parameters, model matching and association are performed with the deceleration device fault diagnosis model library to obtain a target deceleration device fault diagnosis model; The target deceleration device fault diagnosis model is used to perform fault diagnosis analysis on the key part fault working feature set, and the target deceleration device fault diagnosis information is output.

Images