Fault prediction method and system based on aircraft data envelope diagram

Abstract

The invention discloses a fault prediction method and system based on an aircraft data envelope diagram, and particularly relates to the field of aircraft fault prediction.The method comprises the steps that a large amount of aircraft flight data is imported as the basis, model building data and fault feature parameters are selected, specific event occurrence conditions are input, parameter data are intercepted according to the specific event occurrence conditions, and fault prediction is conducted; and drawing a data envelope diagram and fitting a baseline function after eliminating interference data, and storing the model. During prediction, the model is selected, to-be-detected data is intercepted and compared with the baseline function to calculate a fault index, the equipment state is judged according to the fault index, potential faults are predicted, and early warning information is generated. The system comprises a data acquisition module, a data interception module, a fusion module, a model creation module and a fault prediction module. The aircraft fault prediction accuracy can be improved, safe operation of the aircraft is guaranteed, the maintenance response time is shortened, the maintenance cost is reduced, and the skill level of maintenance personnel is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft fault prediction, and more specifically, to a fault prediction method and system based on aircraft data envelopment diagram. Background Art

[0002] The safe operation of aircraft is vital to the aviation industry, and regular maintenance and timely fault diagnosis are the key to ensuring aircraft safety.

[0003] Traditional aircraft maintenance training mainly relies on experienced maintenance personnel, but this method has disadvantages such as long training cycle, high cost and high risk. With the development of data-driven technology, especially envelope analysis based on data statistics, a new method is provided for aircraft fault prediction, which can significantly improve the efficiency and quality of maintenance training.

[0004] Therefore, a fault prediction method and system based on aircraft data envelopment diagram is proposed. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides a fault prediction method and system based on an aircraft data envelopment diagram, which extracts data features in a specific range from aircraft flight data, establishes a prediction model in combination with a data fusion algorithm, and realizes early warning of aircraft failures to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a fault prediction method based on aircraft data envelopment diagram, the method steps are as follows: AS1: Import a large amount of aircraft flight data as the data basis for the fault prediction model. The aircraft flight data includes but is not limited to the parameter information of the aircraft's flight altitude, speed, engine speed, and oil temperature during the flight; AS2: Select data for building the model; AS3: Select the fault characteristic parameters involved; AS4: Enter the conditions for a specific event to occur; AS5: intercepting the fault characteristic parameter data selected in AS3 according to the specific event occurrence conditions entered in AS4; AS6: Draw a parameter data envelope diagram based on the data intercepted in AS5; AS7: Users can selectively remove some special interference data; AS8: Perform data fitting on the data envelope diagram obtained in AS7 to form a data baseline function f(x); AS9: Baseline function and the characteristic parameter information of AS3 and the data condition information of AS4 are saved as a fault prediction model; AS10: Select the corresponding fault prediction model; AS11: Select a package of test data, and perform data interception according to the model conditions to obtain the data curve y(x); AS12: Bring the intercepted test data into the prediction model baseline function , draw the expected data curve ; AS13: and The deviation is used as the fault index. The higher the index, the higher the probability of fault occurrence. The fault index formula is expressed as: , judge the equipment operation status and predict potential failures based on the fault index, and generate early warning information.

[0007] Preferably, the specific events in AS4 include but are not limited to take-off and landing conditions.

[0008] Preferably, the fault prediction model generating step is: S1: First, select a large amount of data as the model data basis and save it to the data package list middle; S2: Traverse the data packets one by one ; S3: Read the Packets Data to memory, where " " is the index of the data list, and the initial value is ; S4: parse the conditions for the occurrence of specific events (such as aircraft takeoff and landing) that are recorded, where the event condition format example is " ",in" "and" " are all parameter names, " "and" " represents commonly used relational operators, "and" " indicates the values ​​of two parameters respectively," " indicates commonly used logical operators. This conditional format is compatible with commonly used relational operators and logical operators. S5: Traverse the data packets one by one Each frame of data in; S6: Judgment Whether the condition in S4 is met, if it is met, then execute S7, if not, then execute S8, where The data frame number, the initial value is ; S7: Add the satisfying data information to the queue In, and , continue to traverse the next frame of data; S8: Determine the queue Is the total time of the data in greater than the duration of the condition being met? , if it is established, execute S9; if it is not established, execute S10; S9: Clear the queue , , continue to traverse the next frame of data; S10: Queue The first data frame in is the start frame, and the last frame is the event end frame; S11: intercepting fault characteristic parameter data according to the start frame and end frame in S10; S12: After step S11, the intercepted fault characteristic parameter data is obtained; S13: Draw the fault characteristic parameter data curve in S12 on the curve display page; ; Repeat steps S3-S13 until all data packets are traversed; S14: Manually select and remove the interfering data curve; S15: performing data fitting on multiple groups of data curves to obtain a fitting function as a fault prediction baseline function.

[0009] S16: Save the baseline function and the data capture condition information as a fault prediction model.

[0010] A fault prediction system based on an aircraft data envelopment diagram is implemented using the above-mentioned fault prediction method based on an aircraft data envelopment diagram, comprising: Data acquisition module: used to import a large amount of aircraft flight data to provide a data basis for the fault prediction model; Data interception module: used to select fault characteristic parameters, enter the conditions for the occurrence of specific events, and intercept fault characteristic parameter data based on the conditions; Data fusion and model creation module: used to draw data envelope diagrams based on intercepted data, remove interfering data, and perform data fitting to obtain baseline functions , and save the fault prediction model; Fault prediction module: used to select the appropriate fault prediction model, process the test data, calculate the fault index, judge the equipment operation status and predict potential faults, and generate early warning information.

[0011] Preferably, the specific events recorded by the data interception module include but are not limited to the conditions for aircraft take-off and landing.

[0012] Preferably, when the data interception module determines whether the data meets the conditions for a specific event to occur, the conditional format is compatible with commonly used relational operators and logical operators.

[0013] Preferably, the fault prediction module is connected to an aircraft cockpit display screen and a ground maintenance personnel terminal via a wireless network, and the aircraft cockpit display screen and the ground maintenance personnel terminal are used to receive the warning information generated by the fault prediction module.

[0014] Technical effects and advantages of the present invention: By importing a large amount of aircraft flight data, combining the conditions for specific events to intercept relevant fault characteristic parameter data, drawing a data envelope diagram and performing data fitting to obtain a baseline function, and then using the model to analyze the test data to calculate the fault index, it is possible to more accurately judge the equipment operating status and predict potential faults, reducing misjudgments and missed judgments.

[0015] By timely monitoring the status of aircraft equipment, predicting failures in advance, and discovering safety hazards as early as possible, the failure rate can be reduced, reducing casualties and equipment losses caused by failures, thereby ensuring aircraft flight safety.

[0016] It can provide early warning before a fault occurs, allowing maintenance personnel to prepare for maintenance work in advance, promptly inspect and maintain equipment that may fail, avoid the fault from worsening, and effectively shorten the time interval from fault discovery to maintenance processing.

[0017] Accurate fault prediction can avoid unnecessary excessive maintenance. At the same time, timely maintenance can prevent major failures, reduce the high maintenance costs caused by equipment damage and replacement, and improve the economic efficiency of aircraft maintenance.

[0018] Maintenance personnel can gain an in-depth understanding of the operating conditions and failure patterns of aircraft equipment based on fault prediction results and related data, which helps accumulate maintenance experience, improve maintenance technical capabilities, and thus enhance the skill level of the entire maintenance team. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flow chart of the fault prediction method of the present invention; Figure 2 A process flow chart for generating a fault prediction model of the present invention; Figure 3 This is a schematic diagram of the fault index principle of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] As attached Figure 1 , Figure 2 and Figure 3 A fault prediction method based on aircraft data envelopment diagram is shown in FIG. Import flight data: Import a large amount of historical flight data packages from the aircraft flight data storage system ,These data contain various parameter information of the aircraft during ,flight, such as flight altitude, speed, engine speed, and oil temperature, which serve as the data basis of the ,fault prediction model.

[0022] Building a fault prediction model: Traverse each data packet in turn .

[0023] Read Packets Data to memory ( The data packet index, the initial value is ).

[0024] Parse specific event conditions, such as aircraft takeoff conditions" "(in" "and" " is the name of the relevant parameter, " "and" " is a relational operator, "and" " is the parameter value, " is a logical operator).

[0025] Traverse the data packets one by one Each frame of data ( The data frame number, the initial value is ).

[0026] judge Whether the take-off conditions are met, if so, the data information that meets the requirements will be added to the queue In, and , continue to traverse the next frame of data; if it is not satisfied, execute the next step.

[0027] Judgment Queue Is the total time of the data in greater than the duration of the condition being met? , if established, clear the queue , , continue to traverse the next frame of data; if not, the queue The first data frame in is taken as the starting frame of the aircraft takeoff event, and the last frame is the end frame of the event.

[0028] According to the start frame and the end frame, characteristic parameter data related to the fault are intercepted, such as temperature and pressure parameter data of key engine components.

[0029] After the intercepted fault characteristic parameter data is obtained, the data curve is drawn on the curve display page.

[0030] Repeat the above steps until all data packets are traversed.

[0031] Manually select and remove possible interfering data curves, such as abnormal data curves caused by temporary sensor abnormalities.

[0032] Perform data fitting on multiple sets of data curves to obtain the fitting function as the fault prediction baseline function .

[0033] The baseline function And the data capture conditions (such as parameters related to take-off conditions) information is saved as a fault prediction model.

[0034] Fault prediction process: Select the appropriate fault prediction model (e.g., a model for the takeoff phase).

[0035] Select a package of test data, and intercept the data according to the model conditions (takeoff conditions) to obtain the data curve .

[0036] Substitute the intercepted test data into the prediction model baseline function , draw the expected data curve .

[0037] Calculating the failure index , whose formula is ,in is the number of data points, calculated by and The higher the index, the higher the probability of a fault. The equipment operating status during the takeoff phase of the aircraft is judged based on the fault index. If the fault index exceeds the set threshold, potential faults are predicted and early warning information is generated to prompt maintenance personnel to conduct inspections and maintenance.

[0038] like Figure 3 As shown, The curve represents the actual curve of the data to be tested. It represents the data curve function that is finally fitted after data interception and fusion. The higher the index, the higher the probability of failure.

[0039] The present invention provides a fault prediction method based on an aircraft data envelopment diagram, which can predict and troubleshoot the probability of a fault occurring when no obvious fault occurs, thereby discovering potential safety hazards in advance and ensuring safe operation of the aircraft.

[0040] This embodiment provides a fault prediction system based on an aircraft data envelopment diagram, including: Data acquisition module: Obtain a large amount of aircraft flight data from the aircraft's flight data recording system or data repository. This module has a data interface that can connect to different data sources, such as the aircraft's onboard data recorder and ground data management system. The data acquisition module will initially organize and convert the acquired flight data to ensure that the data can be effectively used by subsequent modules, providing a rich data foundation for the fault prediction model.

[0041] Data interception module: The operator selects relevant fault characteristic parameters, such as engine parameters and flight attitude parameters, according to the requirements of aircraft fault analysis. At the same time, the conditions for the occurrence of specific events are entered, such as aircraft landing conditions (such as landing gear lowered, flight altitude close to runway altitude). Then, based on the entered landing conditions, the selected fault characteristic parameter data is accurately intercepted, and only the data related to the landing process is retained, so as to facilitate the subsequent analysis of the possibility of faults in the landing stage.

[0042] Data fusion and model creation module: A data envelopment diagram is drawn based on the data intercepted by the data interception module, and the data curves of the same characteristic parameters in different flight and landing processes are integrated into one chart to show the distribution range and trend of the data.

[0043] Operators can selectively remove some special interference data in the envelope diagram through the human-computer interaction interface, such as abnormal data points caused by severe weather factors.

[0044] The data envelope diagram after removing the interference data is fitted, and the data is fitted into a data baseline function using the least squares method. Finally, the baseline function As well as the relevant characteristic parameter information and data condition (landing condition) information, they are saved as a fault prediction model so that they can be called up for subsequent fault prediction.

[0045] Fault prediction module: Select an appropriate fault prediction model according to the current operation phase of the aircraft (such as takeoff, cruise, landing, etc.). For example, when the aircraft is about to land, select the fault prediction model of the landing phase.

[0046] Process the data to be tested (real-time data collected during the landing phase of the aircraft) and intercept the data according to the model requirements to obtain the data curve , and substitute it into the prediction model baseline function , draw the expected data curve .

[0047] Calculating the failure index , by comparison and If the fault index is high, it indicates that the aircraft landing system may have hidden faults, predict the potential fault types (such as landing gear system failure, brake system failure, etc.), and generate early warning information in time, and send alarms to relevant personnel through the aircraft cockpit display screen, ground maintenance personnel terminal, etc., so that appropriate measures can be taken to ensure the safe landing of the aircraft.

[0048] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A fault prediction method based on aircraft data envelopment diagram, characterized in that: The steps are as follows: AS1: Import a large amount of aircraft flight data as the data basis for the fault prediction model; AS2: Select data for building the model; AS3: Select the fault characteristic parameters involved; AS4: Enter the conditions for a specific event to occur; AS5: intercepting the fault characteristic parameter data selected in AS3 according to the specific event occurrence conditions entered in AS4; AS6: Draw a parameter data envelope diagram based on the data intercepted in AS5; AS7: Users can selectively remove some special interference data; AS8: Perform data fitting on the data envelope diagram obtained in AS7 to form a data baseline function ; AS9: Baseline function and the characteristic parameter information of AS3 and the data condition information of AS4 are saved as a fault prediction model; AS10: Select the corresponding fault prediction model; AS11: Select a package of test data, and perform data interception according to the model conditions to obtain the data curve y(x); AS12: Bring the intercepted test data into the prediction model baseline function , draw the expected data curve ; AS13: and The deviation is used as the fault index. The higher the index, the higher the probability of fault occurrence. The fault index formula is expressed as: , judge the equipment operation status and predict potential failures based on the fault index, and generate early warning information.

2. A fault prediction method based on aircraft data envelopment diagram according to claim 1, characterized in that: The specific events in AS4 include but are not limited to the conditions for aircraft take-off and landing.

3. The fault prediction method based on aircraft data envelopment diagram according to claim 1 is characterized in that: The aircraft flight data includes, but is not limited to, parameter information of the aircraft's flight altitude, speed, engine speed, and oil temperature during flight.

4. The fault prediction method based on aircraft data envelopment diagram according to claim 1 is characterized in that: The fault prediction model generation steps are: S1: First, select a large amount of data as the model data basis and save it to the data package list middle; S2: Traverse the data packets one by one ; S3: Read the Packets Data to memory, where " " is the index of the data list, and the initial value is ; S4: Analyze the conditions for the occurrence of the entered specific events; S5: Traverse the data packets one by one Each frame of data in; S6: Judgment Whether the condition in S4 is met, if it is met, then execute S7, if not, then execute S8, where The data frame number, the initial value is ; S7: Add the satisfying data information to the queue In, and , continue to traverse the next frame of data; S8: Determine the queue Is the total time of the data in greater than the duration of the condition being met? , if it is established, execute S9; if it is not established, execute S10; S9: Clear the queue , , continue to traverse the next frame of data; S10: Queue The first data frame in is the start frame, and the last frame is the event end frame; S11: intercepting fault characteristic parameter data according to the start frame and end frame in S10; S12: After step S11, the intercepted fault characteristic parameter data is obtained; S13: Draw the fault characteristic parameter data curve in S12 on the curve display page; ; Repeat steps S3-S13 until all data packets are traversed; S14: Manually select and remove the interfering data curve; S15: performing data fitting on the multiple sets of data curves to obtain a fitting function as a fault prediction baseline function; S16: Save the baseline function and the data capture condition information as a fault prediction model.

5. The fault prediction method based on aircraft data envelopment diagram according to claim 4 is characterized in that: The event conditional format example in S4 is " ",in" "and" "are all parameter names, " "and" " represents commonly used relational operators, " "and" "Represents the values ​​of two parameters respectively, " " represents commonly used logical operators.

6. A fault prediction method based on aircraft data envelopment diagram according to claim 5, characterized in that: The event condition format in S4 is compatible with commonly used relational operators and logical operators.

7. A fault prediction system based on aircraft data envelopment diagram, implemented by the fault prediction method based on aircraft data envelopment diagram according to any one of claims 1 to 6, characterized in that: include: Data acquisition module: used to import a large amount of aircraft flight data to provide a data basis for the fault prediction model; Data interception module: used to select fault characteristic parameters, enter the conditions for the occurrence of specific events, and intercept fault characteristic parameter data based on the conditions; Data fusion and model creation module: used to draw data envelope diagrams based on intercepted data, remove interfering data, and perform data fitting to obtain baseline functions , and save the fault prediction model; Fault prediction module: used to select the appropriate fault prediction model, process the test data, calculate the fault index, judge the equipment operation status and predict potential faults, and generate early warning information.

8. The fault prediction system based on aircraft data envelopment diagram according to claim 7, characterized in that: The specific events recorded by the data interception module include but are not limited to the conditions for aircraft take-off and landing.

9. The fault prediction system based on aircraft data envelopment diagram according to claim 8, characterized in that: When the data interception module determines whether the data meets the conditions for the occurrence of a specific event, the conditional format is compatible with commonly used relational operators and logical operators.

10. The fault prediction system based on aircraft data envelopment diagram according to claim 9, characterized in that: The fault prediction module is connected to an aircraft cockpit display screen and a ground maintenance personnel terminal via a wireless network, and the aircraft cockpit display screen and the ground maintenance personnel terminal are used to receive the warning information generated by the fault prediction module.

Images