A method for early warning of faults according to an aero-engine overall vibration trend limiting value

Abstract

The application discloses a kind of methods for early warning failure according to aero-engine whole machine vibration trend limit value, reach the limit value set when triggering aero-engine to carry out fault diagnosis, it is a kind of fault early warning method, comprising: obtaining vibration original signal, vibration original signal is preprocessed, analysis data are generated;The analysis data are made of multiple data times data;According to analysis data, vibration trend is judged, vibration trend includes setting vibration trend limit value Z ij ;According to the vibration trend limit value Z ij Realize fault early warning.According to the technical scheme, the characteristics of vibration data of aero-engine in use process, the characteristics of failure, the vibration trend limit value is used to propose the early warning failure scheme, effectively overcome the defect of using fixed vibration limit value to judge failure, both can discover potential failure earlier, and will not waste a lot of manpower, material resources, financial resources for data analysis maintenance.

Description

Technical Field

[0001] This invention relates to the field of aircraft engine health management, and more specifically, to a method for early warning of faults based on the overall vibration trend limit value of an aircraft engine. Background Technology

[0002] The vibration limit value for an aircraft engine refers to the engine's fault alarm value. Exceeding this value will trigger an alarm from the engine controller, indicating that the vibration has exceeded the limit. When an engine is first manufactured, its vibration is well below the vibration limit value. During engine use, the vibration will change over time, typically showing a period of stability followed by an increase in vibration over time. The vibration limit value, vibration trend limit value, and vibration change trend are as follows: Figure 1 As shown.

[0003] Because engine vibration values ​​change over time, the method for determining vibration limits and triggering fault alarms requires continuous adjustment to adapt to the engine's condition. Currently, when setting vibration limits, methods such as the "probability equality method" and the "3σ method" can be used. The "probability equality method" calculates the probability distribution of characteristic parameters under normal and fault conditions to determine the probability of the engine operating normally and the probability of a fault when vibration parameters are at a specific amplitude level. If the probability of normal operation is greater than the probability of fault, the engine is considered to be in normal operation; otherwise, a fault is considered to have occurred. This method is highly dependent on the prior distribution and is often insensitive to new engine models and new faults, requiring a certain amount of engine fault data. The "3σ method" depends on the fluctuation of the sample data. For newly manufactured engines with small data fluctuations, the vibration trend limit range set by this method is very small, easily generating false alarms and affecting the normal use of the engine.

[0004] Based on vibration limit values, vibration trend limit values, and vibration variation patterns, this invention proposes a method for judging faults based on vibration trend limit values, in order to overcome the shortcomings of traditional methods that judge faults based on vibration limit values. Summary of the Invention

[0005] To achieve the above objectives, this application provides a method for setting and judging fault warnings based on the overall vibration trend limit value of an aero-engine, comprising the following steps:

[0006] The raw vibration signal is acquired, preprocessed, and analyzed to generate analysis data; the analysis data consists of data from multiple data points.

[0007] Based on the analysis data, the vibration trend is determined. Determining the vibration trend includes setting a vibration trend limit value Z. ij ;

[0008] Based on the vibration trend limit value Z ij Implement fault early warning.

[0009] Preprocessing includes: total signal amount and component extraction;

[0010] Signal total and component extraction refers to extracting the total vibration and components from 60Hz to 350Hz in the original data, including the fundamental frequency components of the high and low voltage rotors.

[0011] Preprocessing also includes working condition screening;

[0012] The operating conditions include the first idle time (MC) before takeoff, the maximum speed (ZD) before takeoff, and the maximum vibration (ZDV) before takeoff. Operating condition screening refers to extracting the total amount and components corresponding to each operating condition. During extraction, n seconds of data before the operating condition changes are extracted as analysis data, and each n-second data point is considered a data point.

[0013] Determining vibration trends includes:

[0014] Determine if there is a vibration trend limit value Z. ij Alternatively, determine if the number of data points is less than or equal to n1: If there is a vibration component or total vibration limit value Z for a given operating condition. ij If the data does not exist, or the number of data occurrences is less than or equal to n1, set the vibration trend limit value Z. ij .

[0015] Furthermore, based on the vibration trend limit value Z ij Fault diagnosis is achieved by setting the vibration trend limit value Z. ij After completion, the average value of the total amount and components of the data for n seconds under a working condition is calculated and compared with the vibration trend limit value Z. ij For comparison, if the average value of one operating condition is greater than the vibration tendency limit value Z, then... ij If this occurs, a vibration alarm will be triggered, and fault diagnosis will be required.

[0016] Fault alarm implementation means: when the vibration of a certain operating condition exceeds the vibration trend of that condition, the fault handling process is initiated to determine the fault; if no fault is determined, a new vibration trend limit value Z is generated from the data of the most recent n1 data points. ij .

[0017] Furthermore, set the vibration trend limit value Z. ij This refers to: determining the value of the data point; if the value of the data point is less than or equal to n1, then taking the last analyzed data point. ijm Calculate the mean μ ijm Define the vibration trend limit value Z for the total quantity and components of each working condition. ij The calculation method is as follows: Where Z is the vibration limit value, i is the operating condition, and j is the total amount and component;

[0018] If the value of the data is greater than n1: merge the data from m analyses, represented as: data ijo Where o = m + 1;

[0019] Calculate data ijo mean μ ijo and standard deviation σ ijo ;

[0020] Obtain the vibration trend limit value Z at this time. ij The calculation method is as follows: Where h is a built-in parameter, which is a natural number;

[0021] According to Z ij The vibration trend limit value Z is updated with the magnitude of Z. ij .

[0022] Among them, according to Z ij The vibration trend limit value Z is updated with the magnitude of Z. ij Refers to: the limit value Z for judging the vibration trend. ij If the vibration limit value Z is greater than the vibration limit value Z, then update the vibration limit value Z to the vibration trend limit value Z. ij .

[0023] According to the present invention, a fault judgment scheme can be proposed based on the characteristics of vibration data and fault characteristics of aero-engines during use, and vibration trend limit values ​​can be used to effectively overcome the defects of using vibration limit values ​​to judge faults. This can detect potential faults earlier and avoid wasting a lot of manpower, material resources and financial resources on data analysis and maintenance. Attached Figure Description

[0024] Figure 1 This is a diagram showing the relationship between vibration trend, vibration limit value, and vibration trend limit value according to an embodiment of the present invention.

[0025] Figure 2 This is a flowchart of a method for locating faults based on the vibration trend of an aero-engine, according to an embodiment of the present invention.

[0026] Figure 3 This is a flowchart illustrating the vibration trend judgment process in the method for locating faults based on the overall vibration trend of an aero-engine, provided by an embodiment of the present invention. Detailed Implementation

[0027] This invention proposes a method for early warning of faults based on the overall vibration trend limit value of an aero-engine. When the set limit value is reached, fault diagnosis of the aero-engine is triggered, providing a method for early warning of faults. Fault warning can be initiated when engine vibration exceeds the vibration trend limit value. If the diagnosis determines that no fault is found, a new vibration trend limit value is updated. This process is repeated until the engine vibration exceeds the limit value.

[0028] The specific implementation of the present invention will now be described in detail with reference to the accompanying drawings.

[0029] Figure 2 A flowchart of a method for determining faults based on vibration trend limits is provided, as shown in the figure, including the following steps:

[0030] Step S200: Acquire the original vibration signal, preprocess the original vibration signal, and generate analysis data;

[0031] The preprocessing performed in this step includes two parts:

[0032] 1) Signal total and component extraction: This part of the preprocessing includes extracting the total vibration from 60Hz to 350Hz and the fundamental frequency components of the high and low voltage rotors from the original data, which are: total Z, high voltage component gF, and low voltage component dF, respectively.

[0033] 2) Operating condition screening;

[0034] The operating conditions involved in this step include the first idle speed (MC) before takeoff, the maximum speed (ZD) before takeoff, and the maximum vibration (ZDV) before takeoff.

[0035] During the operating condition screening process, after extracting the total signal amount and components, the corresponding total amount and components are then extracted from the three operating conditions respectively. When extracting the quantity, the data for the n seconds before the change of the operating condition is used as the analysis data, where n is a preset natural number, and one n-second data point is considered one data point. For example, the data extraction method is to use the data for the 10 seconds before the change of the operating condition as the analysis data, with one 10-second data point constituting one data point. The analysis data obtained during the acquisition process consists of data points from multiple data points.

[0036] Step S210: Determine the vibration trend based on the analysis data, and set a vibration trend limit value Z during the determination process. ij Where i represents the operating condition, j represents the total quantity and the high-pressure component gF, and the low-pressure component dF;

[0037] In this step, the analyzed data is evaluated to determine whether a vibration trend limit value Z exists. ij Or, the number of data points has reached n1, where n1 is a preset natural number: if there is a vibration trend limit value Z corresponding to a vibration component or total vibration amount under a certain working condition. ijIf the data does not exist, or the number of data iterations has reached n1, proceed to step S211 to set the vibration trend limit value Z. ij Otherwise, proceed to step S220 for further fault diagnosis.

[0038] In practical applications, n1 is usually set to 5, that is, the vibration trend limit value Z is set when the data is 5 times. ij The settings.

[0039] Step S211 is without vibration tendency limit value Z ij The data is n1 times, and the current vibration trend exceeds the original vibration limit value Z. ij Alternatively, if no fault is found after fault diagnosis, the vibration trend limit value Z can be set. ij The operation.

[0040] Specific implementation as follows Figure 3 As shown in step S330, firstly, the value of the data is determined to establish the vibration trend limit value Z. ij :

[0041] 1) If the value is less than or equal to n1 (e.g., n1 = 5): take the last analysis data. ijm Calculate the mean μ ijm Each data ijm Given a 10-second dataset, where m is the number of data points in the current data iteration;

[0042] Define the vibration trend limit value Z for the total quantity and components of each working condition. ij The calculation method is as follows: Where Z represents the vibration limit value, i represents the working condition, j represents the total Z / high pressure component gF / low pressure component dF, and m represents the number of data points.

[0043] 2) If the number of data points is greater than n1 (e.g., n1 = 5), merge the analysis data from n1 data points to form the total and component data for each operating condition (MC, ZD, and ZDV), represented as: data ijo Where o = n1 + 1; each data point is a dataset of n × n1 seconds (e.g., 50 seconds); calculate data ijo mean μ ijo and standard deviation σ ijo ;

[0044] Obtain the total and component vibration trend limit value Z for each working condition at this time. ij The calculation method is as follows:

[0045] Where h is a built-in parameter, which is a natural number.

[0046] Secondly, according to Z ijThe vibration trend limit value Z is updated with the magnitude of Z. ij At this point, determine the vibration trend limit value Z. ij If the vibration limit value Z is greater than the vibration limit value Z, then update the vibration limit value Z to the vibration trend limit value Z. ij .

[0047] Next, further fault diagnosis and judgment can be carried out.

[0048] Step S220: Based on the vibration trend limit value Z ij Implement fault early warning.

[0049] During this process, a vibration trend limit value Z is set. ij After completion, calculate the average of the total amount and components of the data for one working condition (MC, ZD, ZDV) over n seconds, and compare it with the set vibration trend limit value Z. ij For comparison, if the vibration values ​​for each working condition are all less than or equal to the vibration trend limit value Z for the corresponding working condition... ij If the average value of a given operating condition is greater than the vibration trend limit value Z, then continue monitoring the vibration data; ij If this occurs, a vibration alarm will be triggered, and fault diagnosis will be required.

[0050] During the fault early warning process: if the vibration of a certain operating condition exceeds the vibration trend of that condition, the fault handling procedure is initiated to determine the fault; if no fault is determined, a new vibration trend limit value Z is generated from the data of the most recent m data points. ij .

[0051] The present invention provides the following examples:

[0052] Taking the high-pressure rotor component gF vibration trend limit value of a certain engine under the maximum speed condition ZD before takeoff as an example:

[0053] In this step, vibration data preprocessing S1 is performed. After the first flight, the high-voltage rotor vibration component is extracted from the vibration data, and then all data in the 10 seconds before the end of the maximum speed condition before takeoff is extracted from the high-voltage rotor vibration component.

[0054] Next, as follows Figure 3 As shown, when judging the vibration trend, the first step is to determine whether there is a vibration trend limit value. On the first flight, there is no vibration trend limit value, so the process proceeds to step S330, setting the vibration trend limit value. During this process, it is first determined whether the number of data points is greater than or equal to 5. Since there is only one data point, the requirement of a number of data points greater than or equal to 5 is not met. Therefore, the data is calculated... ZD_gF_1 The mean μ of the data ZD_gF_1 ,by The vibration trend limit value Z of the high-pressure rotor vibration component under the maximum speed condition before takeoff is calculated by the formula.ZD_gF After calculating and setting the trend limit value, extract all data from the high-voltage rotor vibration component, specifically the data from the 10 seconds before the end of the maximum speed condition before takeoff. ZD_gF_1 Calculate the mean and the set vibration trend limit value Z. ZD_gF The comparison is performed. If the value is less than the limit, monitoring continues. If the value is greater than the limit, a fault warning is triggered. The first time the limit is exceeded, the limit will not be exceeded.

[0055] After the second flight, the high-voltage rotor vibration component was extracted from the vibration data first, and then all data from the 10 seconds before the end of the maximum speed condition before takeoff were extracted from the high-voltage rotor vibration component and recorded as data. ZD_gF_2 At this point, the vibration trend judgment S320 already contains a vibration trend limit value, therefore the data is calculated. ZD_gF_2 mean μ ZD_gF_2 With vibration trend limit value Z ZD_gF When comparing, when μ ZD_gF_2 Less than Z ZD_gF_1 If the value exceeds this limit, monitoring continues; if it exceeds this limit, a fault warning is triggered. If no fault is detected, the two most recent data points are recorded. ZD_gF_1 and data ZD_gF_2 The input vibration trend limit value is set to S330, and the number of data points is limited to two or no more than five. Therefore, the most recent data point is used. ZD_gF_2 According to the data, The vibration trend limit value Z of the high-pressure rotor vibration component under the new pre-takeoff maximum speed condition was calculated. ZD_gF_2 If the vibration trend limit is exceeded on the 3rd and 4th times, the same process applies. If it is not exceeded, the detection continues until the 5th time. On the 5th time, the vibration trend limit setting S330 is entered. If the number of data points is greater than or equal to 5, then the 5 vibration data points of the high-pressure rotor vibration component under the maximum speed condition before takeoff, i.e., data, will be collected. ZD_gF_1 to data ZD_gF_5 The data is spliced ​​together to form a 50-second data set. ZD_gF_6 Find the mean μ of this data. ZD_gF_6 and standard deviation σ ZD_gF_6 According to the formula: Generate the vibration trend limit value Z of the high-pressure rotor vibration component under the new pre-takeoff maximum speed condition. ZD_gF And compare it with the vibration limit value Z. If the vibration trend limit value Z ZD_gF If the vibration trend is greater than the vibration limit value Z, then the vibration limit value Z is used as the vibration trend limit value; otherwise, Z is used as the limit value. ZD_gF As a vibration trend limit, the average vibration value μ of the high-pressure rotor vibration component under maximum speed conditions up to a certain takeoff point. ZD_gF_m Exceeded ZZD_gF If a fault is detected, a fault warning will be issued. If a fault is detected, troubleshooting will be terminated. If no fault is detected, the data from the five most recent vibrations will be used, according to the formula:

[0056]

[0057] Calculate a new vibration trend limit value to replace the previous Z. ZD_gF Similarly, it is compared with the vibration limit value Z, and the troubleshooting process is terminated when a fault is detected.

[0058] This invention addresses the characteristics of vibration data and fault occurrence in aero-engines during operation, and proposes a fault warning scheme based on vibration trend limits. This effectively overcomes the shortcomings of judging faults based on fixed vibration limits, enabling the earlier detection of potential faults without wasting significant human, material, and financial resources on data analysis and maintenance.

[0059] The above-disclosed embodiments are merely a few specific examples of the present invention. However, the present invention is not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A method for warning a fault according to an aeroengine whole machine vibration trend limit value, characterized in that, Includes the following steps: The raw vibration signal is acquired, and the raw vibration signal is preprocessed to generate analysis data; the analysis data consists of data from multiple data points. determining a vibration trend based on the analysis data, the determining a vibration trend including setting a vibration trend limit value ; According to the vibration trend limit value Realize fault early warning; Among them, the set vibration trend limit value Refers to: judging the value of the data point; if the value of the data point is less than or equal to n1, the last analyzed data is taken. Calculate the mean ; Define the vibration trend limits for the total quantity and components of each working condition. The calculation method is as follows: Where Z is the vibration limit value, i is the working condition, and j is the total amount or component; If the value of the data is greater than n1: merging the analysis data from m times is represented as: Where o = m + 1; calculate mean and standard deviation ; Obtain the vibration trend limit value at this time. The calculation method is as follows: Where h is a built-in parameter, which is a natural number greater than or equal to 3; according to Update vibration trend limit value with the magnitude of Z .

2. The method according to claim 1, characterized in that, The preprocessing includes: total signal quantity and component extraction; The extraction of total signal and components refers to the extraction of the total vibration and components from 60Hz to 350Hz in the original data, including the fundamental frequency components of the high- and low-voltage rotors.

3. The method according to claim 2, characterized in that, The preprocessing also includes operating condition screening; The operating conditions include the first idle time (MC) before takeoff, the maximum speed (ZD) before takeoff, and the maximum vibration (ZDV) before takeoff; the operating condition screening refers to: extracting the total amount and component corresponding to each operating condition; during extraction, n seconds of data before the operating condition changes are extracted as analysis data, and one n-second data point is considered as one data point.

4. The method according to claim 1, characterized in that, The determination of vibration trend includes: Determine if there is a vibration trend limit. Alternatively, determine if the number of data points is less than or equal to n1: if there is a vibration component or total vibration limit value for a given operating condition. If the value does not exist, or the number of data occurrences is less than or equal to n1, set a vibration trend limit value. .

5. The method according to claim 1, characterized in that, According to the vibration trend limit value Fault diagnosis refers to setting the vibration trend limit value. After completion, calculate the average of the total amount and components of the data for a given operating condition over n seconds, and compare it with the vibration trend limit value. If, during the comparison, the average value for one operating condition is greater than the vibration trend limit value... If this occurs, a vibration alarm will be triggered, and fault diagnosis will be required.

6. The method according to claim 1, characterized in that, According to Update vibration trend limit value with the magnitude of Z Refers to: the limit value for judging the vibration trend. If the vibration trend limit is greater than the vibration limit value Z, update the vibration trend limit value with the value of Z. .

7. The method according to claim 5, characterized in that, The fault diagnosis process refers to the following: when the vibration of a certain operating condition exceeds the vibration trend of that condition, the fault handling procedure is initiated to determine the fault; if no fault is determined, a new vibration trend limit value is generated from the data of the most recent n1 data points. .

Images