A multi-slice linkage aircraft structure state early warning method

Abstract

The application belongs to the technical field of electric data processing, and particularly relates to a multi-measuring piece linkage aircraft structure state early warning method, wherein each strain gauge is classified according to data characteristics; based on the classification result, at least one classification correlation method is applied to determine a three-level alarm position; based on the result of the three-level alarm position, a space or structure correlation rule is applied to determine a two-level alarm position; based on the result of the two-level alarm position, clustering determination is performed in a key area of the structure to determine a one-level alarm position; the application overcomes the problem of difficult positioning of a weak position after abnormal strain data recognition, realizes monitoring and early warning of a structure dangerous position, and effectively supports command and decision making on a test site.

Description

Technical Field

[0001] This application belongs to the field of electrical data processing technology, and specifically relates to a method for early warning of aircraft structural status using multi-test plate linkage. Background Technology

[0002] Static strength testing of aircraft structures is a method used in ground environments to finalize the design of aircraft structures and verify them as closely as possible to real-world operating conditions. The test results are crucial for determining whether an aircraft can be approved for flight testing and entry into service. During the test, displacement sensors and strain gauges are typically used to measure the displacement and strain response of the structure. Data such as stress and displacement changes are then analyzed and compared with design specifications to determine whether the test specimen meets the static strength requirements.

[0003] Traditional anomaly identification and condition assessment rely heavily on human experience. Analysts must meticulously observe the load-strain curves of a massive number of strain gauges to screen for anomalies and evaluate the structural condition. This method is not only inefficient and costly, but also highly susceptible to the subjective influence of analysts, making it difficult to guarantee accuracy. As the structures of new aircraft become increasingly complex and the number of strain gauges used for monitoring surges, traditional methods can no longer meet the urgent need for real-time and precise monitoring at test sites.

[0004] To improve efficiency, feature-based rule-based automatic classification or machine learning algorithms have emerged in recent years, enabling rapid initial screening of strain gauge data. However, these methods still have significant limitations: firstly, the number of anomalies identified is large and their spatial correlation is poor, making it difficult to quantitatively evaluate the intrinsic correlation of the mechanical states of various parts of the aircraft structure as a whole; secondly, they fail to effectively integrate multi-source information to form hierarchical early warning decisions, resulting in weak support for command and decision-making at the test site. On the other hand, strengthening the structure based solely on a single or a few scattered strain gauge anomalies would lead to an overly conservative design and an increase in structural weight. Therefore, there is an urgent need for an intelligent method that can integrate information from multiple strain gauges, quantitatively evaluate the structural state, and provide hierarchical early warnings to resolve the contradiction between the weak decision support of existing technologies and conservative design. Summary of the Invention

[0005] To address the aforementioned problems, this application provides a multi-test-segment linkage method for early warning of aircraft structural conditions, including:

[0006] Step S1: Read in the strain data of each strain gauge in the static strength test of the aircraft structure in real time, and preprocess the strain data;

[0007] Step S2: Strain gauges with strain values ​​greater than the preset defective gauge threshold are identified as defective and discarded;

[0008] Step S3: Classify each strain gauge according to the characteristics of the strain data;

[0009] Step S4: Based on the classification results of step S3, apply at least one classification association method to determine the location of the Level 3 alarm;

[0010] Step S5: Based on the results of the three-level alarm locations, apply spatial or structural association rules to determine the second-level alarm locations;

[0011] Step S6: Based on the results of the secondary alarm locations, perform clustering in the key areas of the structure to determine the primary alarm locations;

[0012] Step S7: Repeat steps S2 to S6 for the next batch of strain data until all strain data has been processed, and output the strain gauge number corresponding to each level of alarm.

[0013] Preferably, the strain gauges are classified as follows:

[0014] Invalid data clips, irregular jitter clips, strain exceeding limits clips, all constant value fault clips, segment constant value fault clips, large oscillation clips, creep-like clips, strain surge clips, high linearity clips, general linearity clips, kink clips, same-direction jump failure clips, and opposite-direction jump failure clips.

[0015] Preferably, the classification methods for strain gauges include:

[0016] Invalid data: The absolute value of the strain gauge output value is always less than the set low threshold; Irregularly fluctuating strain gauge: The strain fluctuates randomly over time, or is unrelated to the applied load; Strain gauge exceeding limits: The absolute value of the strain exceeds a preset upper limit threshold; Constant-value fault gauge: The strain output remains constant at 0 or a fixed constant value throughout the entire test. Constant-value fault gauge: The strain output is a constant value over multiple consecutive time segments; Large-amplitude oscillation plate: The amplitude of the strain is greater than the preset amplitude, and the frequency is greater than the preset upper limit frequency; Creep-like strain gauges: After loading, the strain exhibits a linear relationship with the load; once a certain load point is reached, the strain value remains within a preset range. Strain maxima: After loading, the slope of the load-strain curve is greater than the set slope at a certain moment; Linear strain: The strain and load are linearly related, and the linearity is higher than the preset high linearity threshold; Typical linear strain: strain and load have a linear relationship, and the linearity fit is between the high linearity threshold and the low linearity threshold; Bend plate: The load-strain curve shows a bend; Same-direction jump failure plate: The strain undergoes a step-like abrupt change, and the direction of the jump is the same as the direction of the data trend before the jump; Reverse jump failure plate: The strain undergoes a step change, and the direction of the jump is opposite to the overall trend of the data before the jump.

[0017] Preferably, in step S4, the classification association method includes the flower petal association method and the threshold method;

[0018] The strain gauge association method is as follows: for three strain gauges constituting a strain gauge group, if at least one of them contains a folded strain gauge and one strain-increased strain gauge, or two folded strain gauges, then the location of the strain gauge group is defined as a level three alarm location.

[0019] The threshold method is as follows: for folding plates, unidirectional jump-damping plates, and reverse jump-damping plates, if their strain values ​​reach a preset strain threshold, the location of the strain plate is defined as a level three alarm location.

[0020] Preferably, in step S5, the spatial or structural association rules include the nearest neighbor search method and the inner and outer lateral piece association method;

[0021] Nearest neighbor search method: Based on the preset spatial coordinates of all strain gauges, search for the N nearest neighbor strain gauges for each strain gauge; if M of the N neighbor strain gauges of a strain gauge have been defined as level 3 alarm locations, then the location of that strain gauge is newly defined as a level 2 alarm location; where N and M are preset positive integers;

[0022] The inner and outer strain gauge association method includes: for strain gauge pairs pasted on the inner and outer sides of the same structural location, if both inner and outer strain gauges are bent gauges and the bending directions of the load and strain curves are opposite, then the location is defined as a secondary alarm location.

[0023] Preferably, N=5 and M=2.

[0024] Preferably, in step S6, the clustering determination in the key regions of the structure includes a location-based keyword search method:

[0025] Based on the strain gauge configuration information, a set of strain gauges containing preset keywords in the description text is selected; if a secondary alarm location of group P or above appears in the set defined by the same keyword, the structural region related to the keyword is immediately defined as a primary alarm location; where P is a preset positive integer.

[0026] Preferably, the preset keywords include at least one of the fuselage frame and wing ribs, and P≥2.

[0027] Preferably, in step S2, the preset defective piece threshold is 8000 microstrain.

[0028] This application also overcomes the difficulty of locating weak points after identifying anomalies in strain data, enabling the monitoring and early warning of dangerous structural parts, and effectively supporting command and decision-making at the test site. Attached Figure Description

[0029] Figure 1 This is a flowchart of an aircraft structural condition early warning method that uses multiple test pieces in tandem. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be described in more detail below with reference to the accompanying drawings. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are only some, not all, of the embodiments of this application. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application. The embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0031] like Figure 1 This is a flowchart of a multi-test-segment linkage aircraft structural condition early warning method. According to this application, the multi-test-segment linkage aircraft structural condition early warning method achieves automatic pre-tensioning. The aforementioned multi-test-segment linkage aircraft structural condition early warning method includes:

[0032] Step S1: Read in strain data from the static strength test of the aircraft structure in real time. The strain data includes channel data collected by strain gauges, and preprocess the data.

[0033] Step S2: Strain gauges with strain values ​​greater than the preset defective gauge threshold are identified as defective and discarded;

[0034] Step S3: Classify each strain gauge according to the characteristics of the strain data;

[0035] Step S4: Based on the classification results of step S3, apply at least one classification association method to determine the location of the Level 3 alarm;

[0036] Step S5: Based on the results of the three-level alarm locations, apply spatial or structural association rules to determine the second-level alarm locations;

[0037] Step S6: Based on the results of the secondary alarm locations, perform clustering in the key areas of the structure to determine the primary alarm locations;

[0038] Step S7: Repeat steps S2 to S6 for the next batch of strain data until all strain data has been processed, and output the strain gauge number corresponding to each level of alarm.

[0039] In some alternative implementations, the strain gauges are classified as follows:

[0040] Invalid data clips, irregular jitter clips, strain exceeding limits clips, all constant value fault clips, segment constant value fault clips, large oscillation clips, creep-like clips, strain surge clips, high linearity clips, general linearity clips, kink clips, same-direction jump failure clips, and opposite-direction jump failure clips.

[0041] Among them, invalid data sheets and irregular jitter sheets are classified as invalid; strain over-limit sheets, all constant value fault sheets, segment constant value fault sheets and large oscillation sheets are classified as failure sheets; creep-like sheets, strain surge sheets, high linearity sheets and general linear sheets are classified as normal sheets; bend sheets are classified as buckling sheets; unidirectional jump failure sheets and reverse jump failure sheets are classified as failure sheets; failure sheets and buckling sheets belong to monitoring and early warning sheets.

[0042] In some alternative implementations, the classification methods for strain gauges include:

[0043] Invalid data: The absolute value of the strain gauge output value is always less than the set low threshold; according to the experience of business personnel, the test process does not focus on small values.

[0044] Irregularly fluctuating strain: The strain fluctuates randomly over time or is not related to the applied load; this is because the test data is affected by environmental interference and other test factors, and cannot truly reflect the test specimen's response to the load.

[0045] Strain gauge exceeding limits: The absolute value of the strain exceeds the preset upper limit threshold; the definition is based on the experience of business personnel, and an excessively large absolute value indicates a strain sensor malfunction.

[0046] Constant-value fault strain gauge: The strain output remains constant at 0 or a fixed constant value throughout the entire test; this type is defined as strain sensor channel not being turned on or malfunctioning.

[0047] Constant-value fault test piece: The strain output is a constant value over multiple consecutive time segments; the test data is affected by environmental interference and other test factors, and cannot truly reflect the test piece's response to the load.

[0048] Large-amplitude oscillation tester: The strain amplitude is greater than the preset amplitude and the frequency is greater than the preset upper limit frequency; the test data is affected by environmental interference and other test factors, and cannot truly reflect the test piece's response to the load.

[0049] Creep-like strain gauges: After being loaded for a period of time, the strain exhibits a linear relationship with the load; once a certain load point is reached, the strain value remains within a preset range, meaning the strain value no longer changes. Strain maxima: After loading, the slope of the load-strain curve is greater than the set slope at a certain moment;

[0050] Linear sheet: The strain and load are linearly related, and the linearity is higher than the preset high linearity threshold; In the static test of the aircraft under limited load, the deformation is in the linear elastic stage of the material, that is, the strain data and load are linearly related.

[0051] Typical linear strain gauges show a linear relationship between strain and load, with a linear fit between the high and low linearity thresholds. However, due to errors such as strain sensor bonding, some normal data may exhibit only moderate linearity.

[0052] Inflection point: The load-strain curve shows an inflection point; the reason is that the structural instability of the aircraft is a key area of ​​concern for the testers. Data mining of a large amount of historical data revealed that when the structure is unstable, the strain and load data will show obvious "peaks" or "troughs", showing an inflection point characteristic in the load-strain data.

[0053] Co-directional jump failure: The strain undergoes a step-like abrupt change, and the direction of the jump is the same as the trend of the data before the jump; the parts of the aircraft where the deformation exceeds the material's yield limit exhibit structural tearing, fracture, and other failure morphologies. Data mining of a large amount of historical data reveals that when failure occurs, the strain data and load curves change abruptly when loaded to high loads, mainly in two categories: co-directional jump and reverse jump.

[0054] Reverse jump failure plate: The strain undergoes a step change, and the direction of the jump is opposite to the overall trend of the data before the jump.

[0055] In some alternative implementations, in step S4, the classification association method includes the flower petal association method and the threshold method;

[0056] The strain gauge association method is as follows: for three strain gauges constituting a strain gauge group, if at least one of them contains a folded strain gauge and one strain-increased strain gauge, or two folded strain gauges, then the location of the strain gauge group is defined as a level three alarm location.

[0057] The threshold method is as follows: for folding plates, unidirectional jump-damping plates, and reverse jump-damping plates, if their strain values ​​reach a preset strain threshold, the location of the strain plate is defined as a level three alarm location.

[0058] In some alternative implementations, in step S5, the spatial or structural association rules include the nearest neighbor search method and the inner and outer lateral piece association method;

[0059] Nearest Neighbor Search Method: Based on the preset spatial coordinates of all strain gauges, search for the N nearest neighbor strain gauges for each strain gauge; if M of the N neighbor strain gauges of a strain gauge have been defined as Level 3 alarm locations, then the location of that strain gauge is newly defined as a Level 2 alarm location; where N and M are preset positive integers; the distance formulas for searching for the N nearest neighbor strain gauges include, but are not limited to: Euclidean distance, Harmanton distance, and cosine distance.

[0060] The inner and outer strain gauge association method includes: for strain gauge pairs pasted on the inner and outer sides of the same structural location, if both inner and outer strain gauges are bent gauges and the bending directions of the load and strain curves are opposite, then the location is defined as a secondary alarm location.

[0061] In some alternative implementations, N=5 and M=2.

[0062] In some alternative implementations, step S6, which involves clustering key regions of the structure, includes a location-based keyword search method.

[0063] Based on the strain gauge configuration information, a set of strain gauges containing preset keywords in the description text is selected; if a secondary alarm location of group P or above appears in the set defined by the same keyword, the structural region related to the keyword is immediately defined as a primary alarm location; where P is a preset positive integer.

[0064] In some optional implementations, the preset keywords include at least one of fuselage frame and wing rib, where P≥2. Taking a certain type of aircraft test as an example, it includes 7124 strain gauges and has built-in conventional component classifications such as fuselage, wing, tail, and upper wing panel. It can search for keywords containing fuselage frame, wing rib, etc. If two or more sets of secondary alarm gauges appear in the same keyword, a primary alarm will be triggered immediately.

[0065] In some alternative implementations, in step S2, the preset defective piece threshold is 8000 microstrains.

[0066] This application also overcomes the difficulty of locating weak points after identifying anomalies in strain data, enabling the monitoring and early warning of dangerous structural parts, and effectively supporting command and decision-making at the test site.

[0067] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for early warning of aircraft structural status using multi-test plate linkage, characterized in that, include: Step S1: Read in the strain data of each strain gauge in the static strength test of the aircraft structure in real time, and preprocess the strain data; Step S2: Strain gauges with strain values ​​greater than the preset defective gauge threshold are identified as defective and discarded; Step S3: Classify each strain gauge according to the characteristics of the strain data; Step S4: Based on the classification results of step S3, apply at least one classification association method to determine the location of the Level 3 alarm; Step S5: Based on the results of the three-level alarm locations, apply spatial or structural association rules to determine the second-level alarm locations; Step S6: Based on the results of the secondary alarm locations, perform clustering in the key areas of the structure to determine the primary alarm locations; Step S7: Repeat steps S2 to S6 for the next batch of strain data until all strain data has been processed, and output the strain gauge number corresponding to each level of alarm. Strain gauges can be classified as follows: Invalid data clips, irregular jitter clips, strain exceeding limits clips, all constant value fault clips, segment constant value fault clips, large oscillation clips, creep-like clips, strain surge clips, high linearity clips, general linearity clips, bend clips, same-direction jump failure clips, and opposite-direction jump failure clips. The classification methods for strain gauges include: Invalid data: The absolute value of the strain gauge output value is always less than the set low threshold; Irregularly fluctuating strain gauge: The strain fluctuates randomly over time, or is unrelated to the applied load; Strain gauge exceeding limits: The absolute value of the strain exceeds a preset upper limit threshold; Constant-value fault gauge: The strain output remains constant at 0 or a fixed constant value throughout the entire test. Constant-value fault gauge: The strain output is a constant value over multiple consecutive time segments; Large-amplitude oscillation plate: The amplitude of the strain is greater than the preset amplitude, and the frequency is greater than the preset upper limit frequency; Creep-like strain gauges: After loading, the strain exhibits a linear relationship with the load; once a certain load point is reached, the strain value remains within a preset range. Strain maxima: After loading, the slope of the load-strain curve is greater than the set slope at a certain moment; Linear strain: The strain and load are linearly related, and the linearity is higher than the preset high linearity threshold; Typical linear strain: strain and load have a linear relationship, and the linearity fit is between the high linearity threshold and the low linearity threshold; Bend plate: The load-strain curve shows a bend; Same-direction jump failure plate: The strain undergoes a step-like abrupt change, and the direction of the jump is the same as the direction of the data trend before the jump; Reverse jump failure plate: The strain undergoes a step change, and the direction of the jump is opposite to the overall trend of the data before the jump.

2. The aircraft structural condition early warning method with multi-measurement linkage as described in claim 1, characterized in that, In step S4, the classification association method includes the flower petal association method and the threshold method; The strain gauge association method is as follows: for three strain gauges constituting a strain gauge group, if at least one of them contains a folded strain gauge and one strain-increased strain gauge, or two folded strain gauges, then the location of the strain gauge group is defined as a level three alarm location. The threshold method is as follows: for folding plates, unidirectional jump-damping plates, and reverse jump-damping plates, if their strain values ​​reach a preset strain threshold, the location of the strain plate is defined as a level three alarm location.

3. The aircraft structural condition early warning method with multi-measurement linkage as described in claim 2, characterized in that, In step S5, the spatial or structural association rules include the nearest neighbor search method and the inner and outer lateral piece association method; Nearest neighbor search method: Based on the preset spatial coordinates of all strain gauges, search for the N nearest neighbor strain gauges for each strain gauge; if M of the N neighbor strain gauges of a strain gauge have been defined as level 3 alarm locations, then the location of that strain gauge is newly defined as a level 2 alarm location; where N and M are preset positive integers; The inner and outer strain gauge association method includes: for strain gauge pairs pasted on the inner and outer sides of the same structural location, if both inner and outer strain gauges are bent gauges and the bending directions of the load and strain curves are opposite, then the location is defined as a secondary alarm location.

4. The aircraft structural condition early warning method with multi-measurement linkage as described in claim 2, characterized in that, N=5, M=2.

5. The aircraft structural condition early warning method with multi-measurement linkage as described in claim 1, characterized in that, In step S6, clustering determination in key areas of the structure includes a location-based keyword search method: Based on the strain gauge configuration information, a set of strain gauges containing preset keywords in the description text is selected; if a set of P or more secondary alarm locations appear in the same keyword definition set, the structural region related to the keyword is defined as a primary alarm location; where P is a preset positive integer.

6. The aircraft structural condition early warning method with multi-test plate linkage as described in claim 5, characterized in that, The preset keywords include at least one of the fuselage frame and wing ribs, and P≥2.

7. The aircraft structural condition early warning method with multi-measurement linkage as described in claim 1, characterized in that, In step S2, the preset defective piece threshold is 8000 microstrain.

Images