A vibration analysis state index threshold value determination method, system, device and medium
By acquiring valid sample data from the aircraft state machine and using single-machine learning and swarm learning methods to determine thresholds, the problems of false alarms and missed alarms in helicopter vibration monitoring are solved, alarm accuracy is improved and maintenance costs are reduced.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU HANGXIN AVIATION TECH CO LTD
- Filing Date
- 2023-08-04
- Publication Date
- 2026-05-12
AI Technical Summary
The domestic helicopter health monitoring and management technology is weak. Vibration monitoring thresholds are determined by theoretical calculations by the design department, which lacks consideration of individual differences, leading to false alarms or missed alarms and increasing maintenance costs.
By acquiring valid sample data from the aircraft state machine, single-machine learning and swarm learning methods are used to determine the yellow attention threshold and the red alarm threshold. Considering individual differences, self-learning models and cluster analysis are used to evaluate indicators, calculate the mean and variance, and set the alarm threshold.
It improves the accuracy of alarms, reduces maintenance costs, and enhances the accuracy and practicality of health monitoring.
Smart Images
Figure CN117113172B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of aircraft technology, and in particular to a method, system, device and storage medium for determining the threshold of vibration analysis state index. Background Technology
[0002] With technological advancements, Helicopter Health Monitoring Systems (HUMS) are widely used in various types of helicopters and play an increasingly important role. HUMS utilizes data analysis, condition monitoring, fault diagnosis, and trend analysis to achieve vibration-based health monitoring of the helicopter's three main moving parts and fuselage structure, thereby improving safety and integrity while reducing maintenance costs.
[0003] To achieve the aforementioned health monitoring functions, setting appropriate health monitoring thresholds is crucial for effective health monitoring. The methods for formulating, validating, and updating HUMS thresholds can improve the accuracy and practicality of health monitoring, enabling HUMS to truly fulfill its role, which is of great significance.
[0004] Foreign HUMS technology and products are highly mature, possessing complete status indicator systems and threshold setting management methods. Eurocopter's helicopter HUMS systems (e.g., the M'arms system equipped on EC225 and EC155 helicopters) and Sikorsky's HUMS systems (e.g., the IMD-HUMS system equipped on the S92A) both have a complete method for determining and updating thresholds. Through statistical analysis, they calculate the statistical characteristics of each status indicator, and combine this with the characteristics of the monitored components and the variation patterns of CI (Statistical Indicators), setting different threshold types for each status indicator, ensuring the accuracy and practicality of the thresholds.
[0005] Domestic health monitoring and management technology started relatively late and has a weak foundation, particularly in the accumulation of basic data. Health monitoring functions, such as vibration monitoring thresholds for determining equipment normality, mainly rely on the design department. This method of threshold determination is primarily based on theoretical calculations and ground test results. The threshold types are limited, only considering group thresholds and neglecting individual differences, easily leading to false alarms or missed alarms. Furthermore, vibration over-limit status indicators do not differentiate between levels of importance, providing uniform maintenance recommendations, which can easily result in over-maintenance and increased maintenance costs. Therefore, a new method for determining vibration analysis status indicator thresholds is urgently needed. Summary of the Invention
[0006] The purpose of this application is to at least partially solve one of the technical problems existing in the prior art.
[0007] Therefore, one objective of this application is to provide a method, system, device, and storage medium for determining vibration analysis state index thresholds, which can increase the accuracy of alarms and reduce maintenance costs.
[0008] To achieve the above technical objectives, the technical solution adopted in this application includes: a method for determining vibration analysis state index thresholds, comprising: acquiring a plurality of valid sample datasets of aircraft state machines; the valid data being used to characterize the data set measured by the state machine when the aircraft operates under preset conditions for a preset time; determining a single-aircraft learning threshold and a fleet learning threshold based on the valid sample datasets; and determining a yellow attention threshold and a red alarm threshold based on the single-aircraft learning threshold and the fleet learning threshold.
[0009] In addition, the vibration analysis state index threshold determination method according to the above embodiments of the present invention may also have the following additional technical features:
[0010] Furthermore, in this embodiment of the application, the step of determining the single-machine learning threshold based on the effective sample dataset includes: calculating the mean of all effective sample data and the variance of the effective sample dataset; inputting the mean and the variance into a preset self-learning model to obtain the learning single-machine threshold.
[0011] Furthermore, in this embodiment of the application, the step of determining the swarm learning threshold based on the effective sample dataset includes: extracting all effective sample data for cluster analysis and determining the cluster analysis evaluation index for different numbers of categories; and determining whether a number of aircraft have a swarm learning threshold based on the cluster analysis evaluation index.
[0012] Further, in this embodiment, the step of determining whether a number of aircraft possess a swarm learning threshold based on the cluster analysis evaluation index specifically includes: extracting the maximum index among all the cluster analysis evaluation indices with different numbers of categories; if the number of categories corresponding to the maximum index is not 1, then the number of aircraft does not possess a swarm learning threshold and is instead a single-aircraft learning threshold; if the number of categories corresponding to the maximum index is 1, the valid samples of each aircraft are resampled using the quartile method, and the cluster analysis evaluation indices corresponding to different numbers of categories are recalculated. If the number of categories corresponding to the maximum index among the cluster analysis evaluation indices with different numbers of categories is 1, then the number of aircraft possesses a swarm learning threshold; otherwise, it is a single-aircraft learning threshold.
[0013] Further, in this embodiment of the application, the step of determining the yellow attention threshold and the red alarm threshold based on the single-machine learning threshold and the cluster learning threshold specifically includes: calculating the mean of all valid sample data and the variance of the valid sample dataset; inputting the mean and the variance into the alarm threshold determination formula to obtain the yellow attention threshold and the red alarm threshold; wherein the alarm threshold determination formula includes:
[0014] T=μ±N*σ
[0015] Where T is the yellow attention threshold or the red alarm threshold, μ is the mean, σ is the variance, and N is a positive integer.
[0016] Furthermore, in this embodiment of the application, the method further includes determining the health status of the aircraft based on the yellow attention threshold and the red alarm threshold.
[0017] Furthermore, in this embodiment of the application, the step of determining the health status of the aircraft based on the yellow attention threshold and the red alarm threshold specifically includes: acquiring the status indicators of each component of the aircraft; determining that the component is normal when the status indicators are less than or equal to the yellow threshold and the red threshold; issuing a yellow alarm when the status indicators are greater than the yellow threshold and less than the red threshold, wherein the yellow alarm is used to indicate an early warning of possible potential problems; and issuing a red alarm when the status indicators are greater than or equal to the red threshold, wherein the red alarm is used to indicate that the monitored component has relatively clear fault symptoms.
[0018] On the other hand, embodiments of this application also provide a vibration analysis state index threshold determination system, including:
[0019] An acquisition unit is used to acquire a number of valid sample datasets of aircraft state machines; the valid data is used to characterize the data set measured by the state machine when the aircraft operates under preset conditions for a preset time; a first processing unit is used to determine the single-machine learning threshold and the fleet learning threshold based on the valid sample dataset; a second processing unit is used to determine the yellow attention threshold and the red alarm threshold based on the single-machine learning threshold and the fleet learning threshold.
[0020] On the other hand, this application also provides a device for determining the threshold of vibration analysis state index, comprising:
[0021] At least one processor;
[0022] At least one memory for storing at least one program;
[0023] When the at least one program is executed by the at least one processor, the at least one processor implements a method for determining the threshold of a vibration analysis state index as described in any one of the inventions.
[0024] In addition, this application also provides a storage medium storing processor-executable instructions, which, when executed by a processor, are used to perform a vibration analysis state index threshold determination method as described in any of the preceding claims.
[0025] The advantages and beneficial effects of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application:
[0026] This application can obtain valid data from a set of data measured by several aircraft state machines under preset conditions for a preset time; determine the single-machine learning threshold and the fleet learning threshold based on the valid sample dataset; and determine the yellow attention threshold and the red alarm threshold based on the single-machine learning threshold and the fleet learning threshold. The determination method of this application takes into account individual differences, which can increase the accuracy of alarms and reduce maintenance costs. Attached Figure Description
[0027] Figure 1 This is a schematic diagram illustrating the steps of a method for determining the threshold of vibration analysis state index in a specific embodiment of the present invention;
[0028] Figure 2 This is a flowchart illustrating the process of determining the consistency of a certain state index using clustering and other analysis methods in a specific embodiment of the present invention.
[0029] Figure 3 This is a schematic diagram of a vibration analysis state index threshold determination system in a specific embodiment of the present invention;
[0030] Figure 4 This is a schematic diagram of a vibration analysis state index threshold determination device in a specific embodiment of the present invention. Detailed Implementation
[0031] The following detailed description, in conjunction with the accompanying drawings, illustrates the principles and processes of the vibration analysis state index threshold determination method, system, device, and storage medium in the embodiments of the present invention.
[0032] Reference Figure 1 This invention provides a method for determining the threshold of vibration analysis state indicators. The method may include, but is not limited to, steps S1-S3:
[0033] S1. Obtain a valid sample dataset of several aircraft state machines; the valid data is used to characterize the data set measured by the state machine when the aircraft operates under preset conditions for a preset time;
[0034] S2. Based on the effective sample dataset, determine the single-machine learning threshold and the cluster learning threshold;
[0035] S3. Determine the yellow attention threshold and the red alarm threshold based on the single-machine learning threshold and the cluster learning threshold.
[0036] Furthermore, in some embodiments of this application, the step of determining the single-machine learning threshold based on the effective sample dataset may include, but is not limited to, steps S21-S22:
[0037] S21. Calculate the mean of all valid sample data and the variance of the valid sample dataset;
[0038] S22. Input the mean and variance into a preset self-learning model to obtain the learning-type single-machine threshold.
[0039] Furthermore, in some embodiments of this application, the step of determining the swarm learning threshold based on the effective sample dataset may include, but is not limited to, steps S31-S32:
[0040] S31. Extract all valid sample data, perform cluster analysis, and determine the evaluation index for the cluster analysis of all valid sample data;
[0041] S32. Determine the swarm learning threshold based on the cluster analysis evaluation index.
[0042] Furthermore, in some embodiments of this application, the step of determining the swarm learning threshold based on the cluster analysis evaluation index may include, but is not limited to, steps S41-S42:
[0043] S41. Extract the maximum index from all the cluster analysis evaluation indicators with different numbers of categories;
[0044] S42. If the number of categories corresponding to the maximum index is not 1, then the plurality of aircraft do not have a group learning threshold and are instead a single-aircraft learning threshold.
[0045] If the maximum index corresponds to 1 number of categories, the effective samples of each aircraft are resampled using the quartile method, and the cluster analysis evaluation indexes corresponding to different numbers of categories are recalculated. If the maximum index in the cluster analysis evaluation indexes corresponding to different numbers of categories corresponds to 1 number of categories, the aircraft have a group learning threshold; otherwise, a single-aircraft learning threshold is used. Further, in some embodiments of this application, the step of determining the yellow attention threshold and the red alarm threshold based on the single-aircraft learning threshold and the group learning threshold may include, but is not limited to, steps S51-S52:
[0046] S51. Calculate the mean of all valid sample data and the variance of the valid sample dataset;
[0047] S52. Input the mean and the variance into the alarm threshold determination formula to obtain the yellow attention threshold and the red alarm threshold; wherein the alarm threshold determination formula includes:
[0048] T=μ±N*σ
[0049] Where T is the yellow attention threshold or the red alarm threshold, μ is the mean, σ is the variance, and N is a positive integer.
[0050] Furthermore, in some embodiments of this application, the vibration analysis state index threshold determination method may further include step S4 to determine the health status of the aircraft based on the yellow attention threshold and the red alarm threshold.
[0051] Furthermore, in some embodiments of this application, the step of determining the aircraft health status based on the yellow attention threshold and the red alarm threshold is described in steps S401-S404:
[0052] S401. Obtain the status indicators of each component of the aircraft;
[0053] S402. When the status indicator is less than or equal to the yellow threshold and the red threshold, the display component is determined to be normal.
[0054] S403. When the status indicator is greater than the yellow threshold and less than the red threshold, a yellow alarm is given. The yellow alarm is used to indicate that an early warning is given for possible potential problems.
[0055] S404. When the status indicator is greater than or equal to the red threshold, a red alarm is given. The red alarm is used to indicate that the monitored component has relatively clear fault symptoms.
[0056] The specific calculation principle of this application is explained below with reference to specific embodiments:
[0057] The threshold determination and diagnostic alarm process of this invention is as follows:
[0058] 1) Determine the security threshold for the status indicators.
[0059] Condition indicators are parameters that effectively reflect the health status of a helicopter; they are often abbreviated as CI. For example, the root mean square (RMS) value represents the dynamic signal strength, reflecting the signal's energy level; in engineering, it's called the effective value. When a fault or other abnormal frequency components occur, the overall signal energy increases, and the RMS value will increase. There is no fixed normal range for this value; it must be determined based on analysis of the actual flight conditions.
[0060] 2) Obtain valid data. When the single-machine learning threshold condition is met, obtain the single-machine learning threshold.
[0061] Valid data samples include: data acquired when the acquisition system and sensors are fault-free, the acquired signals are normal, and the data has been in a specified working state for a certain duration.
[0062] The conditions for a single-machine learning threshold are: 100 hours of flight time and 100 CIs. CI is short for Condition Indicator, such as the root mean square value reflecting the magnitude of signal energy.
[0063] 3) Obtain valid data. When the cluster learning threshold condition is met, use cluster analysis to evaluate the status indicators of each aircraft in the cluster. Use the clustering results to assess the consistency of the status indicators of each aircraft. Based on the importance of the status indicators, determine the attention thresholds for individual aircraft and the entire cluster, as well as the alarm thresholds.
[0064] The cluster condition is that, under the condition of meeting the single-machine learning threshold, the number of individual machines in the cluster is ≥5.
[0065] 4) Verify whether the threshold is reasonable. If it is not reasonable, collect data again and perform threshold learning.
[0066] Threshold rationality: For labeled normal / faulty datasets, alarm accuracy > 99%, false alarm rate < 1‰, and false negative rate < 1%. For unlabeled faulty normal datasets, the false alarm rate < 1%.
[0067] 5) Determine the yellow attention threshold and red alarm threshold based on the security threshold, single-machine learning threshold, and cluster learning threshold, and perform diagnostic alarms.
[0068] Safety thresholds refer to the maximum values set for the yellow alert threshold and the red alarm threshold. Individual learning thresholds refer to separate thresholds set for each aircraft, with individual yellow and red thresholds set according to the alarm level. Cluster learning thresholds refer to a unified threshold set for the entire cluster, with cluster yellow and red thresholds set according to the alarm level.
[0069] The calculation formulas for the yellow and red thresholds can be found in the threshold setting method.
[0070] 6) Threshold Update: When replacing onboard components, upgrading software, etc., and analysis indicates that a threshold update is needed, a threshold change process is initiated. The ground station threshold setting process may include (1) to (5).
[0071] (1) Method for determining HUMS security threshold
[0072] To ensure helicopter flight safety, certain parameters have strict vibration limits. HUMS sets safety thresholds for these parameters at the factory, according to the thresholds provided by the design department.
[0073] (2) Learning Threshold Determination Method
[0074] a) Learning-type standalone threshold
[0075] Due to the lack of historical fault flight data for each component in the helicopter's transmission system, a learning-based threshold strategy was adopted to monitor and alert on each component of the helicopter in the early stages of flight testing.
[0076] Because the mean and standard deviation of different CIs vary significantly over time, if a fixed and identical minimum flight hours or minimum CI number is specified for threshold learning, the learned threshold result will differ greatly from the actual value, easily leading to missed or false alarms. Therefore, an adaptive method is needed to determine the required data length for a single-aircraft learned threshold based on CI fluctuations.
[0077] An adaptive method is used to calculate the single-machine learning threshold. When the threshold self-learning algorithm determines that the single-machine CI is stable, the learning single-machine threshold is obtained.
[0078] b) Learning cluster threshold
[0079] The system collects fleet status indicators and analyzes their distribution using statistical learning methods such as clustering. The goal of the analysis is to assess the consistency of the status indicators. Consistency analysis of the status indicators is the primary basis for distinguishing between fleet thresholds and individual aircraft thresholds.
[0080] Reference Figure 2 The process of determining the consistency of a certain state index using analysis methods such as clustering is as follows:
[0081] i. Collect data on the CI (Indicators of State) of N aircraft, such as root mean square (RMS) values.
[0082] ii. Perform cluster analysis with K = 1 to N respectively. Commonly used clustering methods include hierarchical clustering, density clustering, and K-means clustering.
[0083] iii. Calculate the clustering evaluation index Xi (i = 1 to N) for K = 1 to N respectively. The evaluation index Xi is the silhouette coefficient and CH score.
[0084] iv. When indicator X1 is the largest among all Xi indicators, it indicates that the consistency of the cluster indicators is optimal.
[0085] v. Consistency result correction: In order to avoid the existence of outliers that lead to high dispersion of the dataset and thus cause the optimal consistency index, the cluster analysis is performed again according to the 25% to 75% value of the four-category method. When it is determined that the cluster is 1, the evaluation index is optimal.
[0086] vi. When the clustering is 1, the evaluation index is optimal, indicating that the indexes of the state are consistent; otherwise, they are inconsistent.
[0087] Consistent status indicators indicate good consistency among the status indicators of each aircraft, and the fleet can share a common threshold; therefore, the threshold for this indicator is the fleet threshold. Poor consistency indicates significant differences in the indicator data among the aircraft in the fleet, requiring the setting of a threshold for each aircraft; therefore, the threshold for this indicator is the individual aircraft threshold.
[0088] (3) Threshold setting method
[0089] The threshold calculation method based on the probabilistic model originates from the Three Sigma criterion. The Three Sigma criterion, also known as the Raida criterion, first assumes that a set of detection data contains only random errors. It then calculates the standard deviation and determines an interval based on a certain probability. Errors exceeding this interval are considered abnormal and should be flagged as data anomalies. In the example of the probabilistic analysis model (normal distribution), σ represents the standard deviation, and μ represents the mean. x = μ is the axis of symmetry of the image.
[0090] Based on the importance and alarm level of the status indicators, set yellow (attention) and red (alarm) thresholds for the status indicators. The determination method is as follows.
[0091] T = μ ± N * σ, where T is the threshold, μ is the mean, and σ is the variance. N is an integer obtained from actual flight data. When determining the yellow and red thresholds, μ and σ are the same, but N takes different values to obtain the yellow and red thresholds. The individual aircraft threshold refers to the individual threshold set for each aircraft; based on the alarm level, separate yellow and red thresholds are set for each aircraft. The fleet threshold refers to a unified threshold set for the entire fleet; based on the alarm level, separate yellow and red thresholds are set for the entire fleet. The safety threshold refers to the maximum value set for both the yellow and red thresholds.
[0092] (4) Diagnostic alarm logic
[0093] When the yellow threshold is exceeded, an early warning is given for potential problems. If no problems are found after completing the corresponding inspection, the flight mission can still be performed. However, maintenance personnel need to closely monitor the changes in the exceeding indicators after each flight. If the health status remains abnormal, they can analyze the situation with the remote technical support center and provide maintenance recommendations. When the red threshold is exceeded, it indicates that the monitored component has relatively clear signs of failure. Maintenance personnel must perform the corresponding maintenance work or test flight verification work according to the maintenance procedure before the next flight to check the integrity of the aircraft before the flight mission can continue.
[0094] a) Diagnostic alerts:
[0095] i) When the status indicator is below the yellow and red thresholds, the component is considered normal;
[0096] ii) When the status indicator is greater than the yellow threshold but less than the red threshold, a yellow alert is issued;
[0097] iii) When the status indicator is greater than the red threshold, a red alert is issued.
[0098] (5) Threshold setting and update process
[0099] When analysis determines that a single-machine threshold needs to be updated, the threshold calculation and update for both single-channel and full-channel operations are manually triggered, as follows:
[0100] 1) Replace the machine parts;
[0101] 2) The software upgrade affected the content related to learning thresholds;
[0102] 3) When manual analysis deems it necessary to update the threshold, then update the threshold.
[0103] After the host computer and other design units determine the updated security thresholds, the thresholds are updated.
[0104] In addition, refer to Figure 3 ,and Figure 1 Corresponding to the method described above, embodiments of this application also provide a vibration analysis state index threshold determination system. The system may include an acquisition unit 101, a first processing unit 102, and a second processing unit 103. The acquisition unit 101 can be used to acquire a plurality of valid sample datasets of aircraft state machines; the valid data is used to characterize the data set measured by the state machine when the aircraft operates under preset conditions for a preset time; the first processing unit 102 can be used to determine a single-aircraft learning threshold and a fleet learning threshold based on the valid sample dataset; the second processing unit 103 can be used to determine a yellow attention threshold and a red alarm threshold based on the single-aircraft learning threshold and the fleet learning threshold.
[0105] It should be noted that the content of the above-described vibration analysis state index threshold determination method embodiments is applicable to this vibration analysis state index threshold determination system embodiment. The specific functions implemented by this vibration analysis state index threshold determination system embodiment are the same as those of the above-described vibration analysis state index threshold determination method embodiments, and the beneficial effects achieved are also the same as those achieved by the above-described vibration analysis state index threshold determination method embodiments.
[0106] and Figure 1 Corresponding to the method, this application also provides a vibration analysis state index threshold determination device, the specific structure of which can be referred to Figure 4 ,include:
[0107] At least one processor;
[0108] At least one memory for storing at least one program;
[0109] When the at least one program is executed by the at least one processor, the at least one processor implements the vibration analysis state index threshold determination method.
[0110] The content of the above method embodiments is applicable to the device embodiments. The specific functions implemented by the device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.
[0111] and Figure 1 Corresponding to the method described above, this application also provides a storage medium storing processor-executable instructions, which, when executed by the processor, are used to perform the vibration analysis state index threshold determination method.
[0112] The content of the above-described vibration analysis state index threshold determination method embodiments is applicable to this storage medium embodiment. The specific functions implemented by this storage medium embodiment are the same as those of the above-described vibration analysis state index threshold determination method embodiments, and the beneficial effects achieved are also the same as those achieved by the above-described vibration analysis state index threshold determination method embodiments.
[0113] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this application are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.
[0114] Furthermore, although this application is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding this application. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional technology for an engineer. Therefore, those skilled in the art can implement the application set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of this application, which is determined by the full scope of the appended claims and their equivalents.
[0115] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several programs to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0116] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequential list of executable programs for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, a program execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can retrieve and execute a program from or in conjunction with such a program execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit a program for use by or in conjunction with a program execution system, apparatus, or device.
[0117] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0118] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable program execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0119] In the foregoing description of this specification, the references to terms such as "one embodiment," "another embodiment," or "some embodiments," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.
[0120] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.
[0121] The above is a detailed description of the preferred embodiments of this application, but this application is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application, and these equivalent modifications or substitutions are all included within the scope defined by the claims of this application.
Claims
1. A method for determining the threshold of vibration analysis state index, characterized in that, include: Acquire several valid sample datasets of aircraft state machines; valid data are used to characterize the data set measured by the state machine when the aircraft operates under preset conditions for a preset time. The samples of the valid data include the acquisition system, sensor without faults, and acquisition signals without abnormalities; wherein, the valid data corresponding to a single aircraft must meet the requirements of flight duration greater than or equal to 100 hours and the number of state indicators CI greater than or equal to 100, and the valid data corresponding to a fleet must meet the condition that the number of single aircraft is greater than or equal to 5 under the condition of single aircraft learning threshold; Based on the valid sample dataset, the single-aircraft learning threshold and the fleet learning threshold are determined. The determination of the fleet learning threshold includes: extracting all valid sample data for cluster analysis and determining cluster analysis evaluation indicators for different numbers of categories; extracting the maximum indicator among all the cluster analysis evaluation indicators for different numbers of categories; if the number of categories corresponding to the maximum indicator is not 1, then the aircraft do not have a fleet learning threshold, and the threshold is a single-aircraft learning threshold; if the number of categories corresponding to the maximum indicator is 1, the valid samples of each aircraft are resampled using the quartile method, and the cluster analysis evaluation indicators for different numbers of categories are recalculated; if the number of categories corresponding to the maximum indicator among the cluster analysis evaluation indicators for different numbers of categories is 1, then the aircraft have a fleet learning threshold; otherwise, the threshold is a single-aircraft learning threshold. Calculate the mean and variance of all valid sample data; input the mean and variance into the alarm threshold determination formula to obtain the yellow attention threshold and the red alarm threshold; wherein the alarm threshold determination formula includes: T = μ±N s Where T is the yellow attention threshold or red alarm threshold, μ is the mean, σ is the variance, and N is a positive integer; Acquire the status indicators of each component of the aircraft; When the status indicators are less than or equal to the yellow threshold and the red threshold, the display component is determined to be normal; When the status indicator is greater than the yellow threshold and less than the red threshold, a yellow alert is issued. The yellow alert is used to indicate that an early warning of potential problems is being given. When the status indicator is greater than or equal to the red threshold, a red alarm is issued. The red alarm is used to indicate that the monitored component has relatively clear fault symptoms. For labeled normal or faulty datasets, the alarm accuracy rate must be greater than 99%, the false alarm rate must be less than 1‰, and the missed alarm rate must be less than 1%. For unlabeled faulty normal datasets, the false alarm rate must be less than 1%. If the above standards are not met, a new valid sample dataset must be collected and the threshold values of the aforementioned status indicators must be redefined.
2. The method for determining the threshold of vibration analysis state index according to claim 1, characterized in that, The step of determining the single-machine learning threshold based on the effective sample dataset includes: Calculate the mean of all valid sample data and the variance of the valid sample dataset; The mean and variance are input into a preset self-learning model to obtain a learning-type single-machine threshold.
3. A vibration analysis state index threshold determination system, applied to the method described in any one of claims 1-2, characterized in that, include: The acquisition unit is used to acquire a number of valid sample datasets of aircraft state machines. The valid data is used to characterize the data set measured by the state machine when the aircraft operates under preset conditions for a preset time. The samples of the valid data include the acquisition system, the sensors are fault-free, and the acquisition signals are normal. Among them, the valid data corresponding to a single aircraft must meet the requirements of a flight duration of 100 hours or more and the number of state indicators (CI) greater than or equal to 100. The valid data corresponding to a fleet must meet the condition that the number of single aircraft is greater than or equal to 5 under the condition of a single aircraft learning threshold. The first processing unit is used to determine the single-machine learning threshold and the cluster learning threshold based on the effective sample dataset. The second processing unit is used to determine the yellow attention threshold and the red alarm threshold using the single-machine learning threshold and the cluster learning threshold; Acquire the status indicators of each component of the aircraft; When the status indicators are less than or equal to the yellow threshold and the red threshold, the display component is determined to be normal; When the status indicator is greater than the yellow threshold and less than the red threshold, a yellow alert is issued. The yellow alert is used to indicate that an early warning of potential problems is being given. When the status indicator is greater than or equal to the red threshold, a red alarm is issued. The red alarm is used to indicate that the monitored component has relatively clear fault symptoms. For labeled normal or faulty datasets, the alarm accuracy rate must be greater than 99%, the false alarm rate must be less than 1‰, and the missed alarm rate must be less than 1%. For unlabeled faulty normal datasets, the false alarm rate must be less than 1%. If the above standards are not met, a new valid sample dataset must be collected and the threshold values of the aforementioned status indicators must be redefined.
4. A device for determining the threshold of vibration analysis state index, characterized in that... include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the vibration analysis state index threshold determination method as described in any one of claims 1-2.
5. A computer-readable storage medium storing processor-executable instructions, characterized in that, The processor-executable instructions, when executed by the processor, are used to perform the vibration analysis state index threshold determination method as described in any one of claims 1-2.