Engine Vibration Fault Isolation Method

Abstract

The present invention relates to the field of engine vibration faults. To facilitate engine vibration fault isolation, an engine vibration fault isolation method is provided, comprising: Step 1: Establishing a basic database: Acquiring multiple repair parameters for multiple engines; Step 2: Determining key repair parameters; Step 3: Calculating the engine's vibration coefficient: The vibration coefficient is the number of key repair parameters for the engine falling within a desired repair range; Step 4: Vibration isolation: Classifying the vibration coefficients based on limit values; Step 5: Vibration probability prediction: Calculating the vibration coefficient of the engine to be tested and predicting the engine's vibration probability based on the classification results of Step 4. The above method facilitates engine vibration fault isolation.

Description

Technical Field

[0001] The invention relates to the field of engine vibration faults, and in particular to an engine vibration fault isolation method. Background Art

[0002] The prediction and research of aircraft engine vibration has long been a hot topic in industry research. Currently, the primary vibration analysis method involves studying parameters such as the engine's structural state and employing forward mechanistic research to suppress or predict engine vibration. Because the relationship between engine repair, rotor assembly, and overall engine vibration is not a simple linear one, such research often only analyzes and controls specific types of vibration faults and fails to effectively predict overall engine vibration. Furthermore, the application of neural network calculations to predict engine vibration has also been a hot topic in recent years. However, due to the complex structural parameters and extensive data associated with aircraft engines, effective prediction or isolation of vibration faults has not yet been possible. Summary of the Invention

[0003] In order to facilitate engine vibration fault isolation, the present application provides an engine vibration fault isolation method.

[0004] The technical solution adopted by the present invention to solve the above problems is:

[0005] Engine vibration fault isolation methods, including:

[0006] Step 1: Establish a basic database: obtain multiple repair parameters of multiple engines;

[0007] Step 2: Determine the key repair parameter: For each repair parameter, if the number of engines with excessive vibration within a certain repair range is less than a first preset value and the number of engines with acceptable vibration within the range is greater than a second preset value, then the repair parameter is determined to be the key repair parameter and the repair range is determined to be the expected repair range.

[0008] Step 3: Calculate the vibration coefficient of the engine: the vibration coefficient is the number of key repair parameters of the engine that fall within the corresponding expected repair range;

[0009] Step 4: Vibration isolation: Classify the vibration coefficient based on the limit value;

[0010] Step 5: Vibration probability prediction: Calculate the vibration coefficient of the engine to be tested and make a vibration probability prediction for the engine based on the classification result of step 4.

[0011] Furthermore, the step 2 obtains the key repair parameters by calculating the binary distribution of the repair parameters and the engine vibration amplitude.

[0012] Furthermore, the method further includes step 6, vibration condition prediction: predicting the vibration condition of the engine to be measured according to the category to which the vibration coefficient of the engine to be measured belongs.

[0013] Furthermore, in step 2, the first preset value and the second preset value are both determined based on the total number of engines and repair parameters.

[0014] Furthermore, in step 4, the limit value is determined by sorting the vibration coefficients by size. When the vibration coefficient reaches a certain value, the probability of the engine vibrating beyond the limit is less than a third preset value. At this time, the corresponding vibration coefficient size is the limit value.

[0015] Compared to existing technologies, this invention offers the following advantages: by statistically analyzing repair and test data, the engine vibration coefficient is calculated and an isolation model is established, facilitating the isolation of engine vibration faults. The engine vibration coefficient can also be used to predict engine vibration conditions. The vibration coefficient is calculated based on all repair items for all engines in the database, providing broader data coverage and more accurate isolation models. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart of the engine vibration fault isolation method;

[0017] Figure 2 is the binary distribution of a key repair item and engine vibration;

[0018] Figure 3 is the binary distribution of insensitive repair items and engine vibration;

[0019] Figure 4 This is the V2 vibration condition after the engine is isolated according to K≤8. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0021] like Figure 1 As shown, the engine vibration fault isolation method includes:

[0022] Engine vibration fault isolation methods, including:

[0023] Step 1: Establish a basic database: obtain multiple repair parameters of multiple engines;

[0024] Step 2: Determine the key repair parameter: For each repair parameter, if the number of engines with excessive vibration within a certain repair range is less than a first preset value and the number of engines with acceptable vibration within the range is greater than a second preset value, then the repair parameter is determined to be the key repair parameter and the repair range is determined to be the expected repair range.

[0025] Step 3: Calculate the vibration coefficient of the engine: the vibration coefficient is the number of key repair parameters of the engine that fall within the expected repair range;

[0026] Step 4: Vibration isolation: Classify the vibration coefficient based on the limit value;

[0027] Step 5: Vibration probability prediction: Calculate the vibration coefficient of the engine to be tested and make a vibration probability prediction for the engine based on the classification result of step 4.

[0028] Example

[0029] Engine vibration fault isolation methods, including:

[0030] Prepare basic engine repair and test data to form an m×n matrix [D], where m is the number of engines and n is the number of engine repair and test data items being counted. For a certain aircraft engine, its vibration parameters include B, A1, A4, V2, and V5. These vibration data are represented by D(:,1) to D(:,5), respectively. The remaining columns of the matrix represent the engine repair and assembly data. To improve model accuracy, as much engine repair data and number of engines as possible should be counted. In this example, 72 repair data items for 136 engines were counted.

[0031] Engine vibration failure is affected by many factors. When there are enough statistical samples, if only the relationship between engine vibration and a certain parameter is analyzed, it can be simplified to a binary distribution of engine vibration and current parameters. Taking the engine vibration parameter V2 as an example, the technical document requires that V2 vibration should not exceed 40. The typical binary distribution of V2 and repair parameters is calculated, and the binary distribution diagram is interpreted: if the number of engines with excessive vibration within a certain repair range is less than the first preset value and the number of engines with qualified vibration within this range is greater than the second preset value, then the parameter item is considered to be a key parameter item. Figure 2As shown, if the number of engines with excessive vibration within the dotted box is 0 and there are multiple engines with qualified vibration, then the parameter item is considered to be a key parameter item, and the corresponding repair range is the expected repair range (to make the engine tend to pass the test). The specific values of the first preset value and the second preset value can be set according to the total number of engines and actual conditions. The first preset value and the second preset value can be a specific number or a certain proportion of the total number of engines, such as the first preset value is 5, the second preset value is 6, or the first preset value is 5% of the total number of engines, the second preset value is 6% of the total number of engines, etc. The first preset value and the second preset value corresponding to different repair parameters can also be set to different, and the specific values are not limited here. The key repair parameter is named C i , the corresponding expected repair range is F i If the probability of engine vibration exceeding the standard is similar within the full repair range, such as Figure 3 If the parameter is shown in the figure, the repair parameter is considered to be an insensitive repair parameter and will not be calculated again.

[0032] Calculate the engine vibration coefficient: For the n-5 repair data of the j-th engine, compare each key repair parameter D(j,i) to see if it falls within the corresponding expected repair range F i Within the expected repair range F i The key repair parameter within the quantity is determined as the vibration coefficient K.

[0033] By classifying engines by K value, it is clear that as the K value increases, the number of engine vibration failures decreases significantly. Therefore, a limit value can be set to isolate engines with K values below the limit value.

[0034] When K reaches a certain value, the probability of an engine exceeding the vibration standard is less than the third preset value. At this time, the corresponding K is the selectable limit value. The specific value of the third preset value can be set according to actual needs. As shown in the following table, with K≤8 as the limit value, the original engine group is divided into 38:98, and the engines exceeding the V2 standard are isolated within the range of 27.9% of the statistical sample. The probability of the remaining engines exceeding the V2 standard is only 5.1%. Figure 4 shown.

[0035] K value table

[0036] K(V2) Exceeding the standard number of units Total visits Probability of exceeding the standard 2 1 1 100.00％ 3 1 2 50.00％ 4 1 1 100.00％ 5 3 3 100.00％ 6 8 13 61.54％ 7 3 7 42.86％ 8 3 11 27.27％ 9 1 16 6.25％ 10 3 22 13.64％ 11 1 20 5.00％ 12 0 17 0.00％ 13 0 6 0.00％ 14 0 10 0.00％ 15 0 4 0.00％ 16 0 2 0.00％ 18 0 1 0.00％

[0037] Based on the classification results from the previous step, the engine vibration distribution corresponding to different K values can be calculated. When predicting engine vibration, the engine vibration coefficient k is first calculated. The corresponding vibration distribution is the vibration condition that the engine may exhibit.

Claims

1. An engine vibration fault isolation method, characterized in that: include: Step 1: Establish a basic database: obtain multiple repair parameters of multiple engines; Step 2: Determine the key repair parameter: For each repair parameter, if the number of engines with excessive vibration within a certain repair range is less than a first preset value and the number of engines with acceptable vibration within the range is greater than a second preset value, then the repair parameter is determined to be the key repair parameter and the repair range is determined to be the expected repair range. Step 3: Calculate the vibration coefficient of the engine: the vibration coefficient is the number of key repair parameters of the engine that fall within the corresponding expected repair range; Step 4: Vibration isolation: Classify the vibration coefficient based on the limit value; Step 5: Vibration probability prediction: Calculate the vibration coefficient of the engine to be tested and make a vibration probability prediction for the engine based on the classification result of step 4.

2. The engine vibration fault isolation method according to claim 1, characterized in that: The step 2 obtains the key repair parameters by calculating the binary distribution of the repair parameters and the engine vibration amplitude.

3. The engine vibration fault isolation method according to claim 1, characterized in that: The method further includes step 6, vibration condition prediction: predicting the vibration condition of the engine to be measured according to the category to which the vibration coefficient of the engine to be measured belongs.

4. The engine vibration fault isolation method according to claim 1, characterized in that: In step 2, the first preset value and the second preset value are both determined according to the total number of engines and repair parameters.

5. The engine vibration fault isolation method according to any one of claims 1 to 4, characterized in that: In step 4, the limit value is determined by sorting the vibration coefficients by size. When the vibration coefficient reaches a certain value, the probability of the engine vibrating beyond the limit is less than a third preset value. At this time, the corresponding vibration coefficient size is the limit value.

Images