Aviation big data analysis methods and tools
By using aviation big data analytics to filter and process historical system data, a performance baseline is established, and the system status is monitored in real time. This solves the problem of the inability to detect abnormal status in a timely manner in existing technologies, reduces risks and costs, and improves detection efficiency and system utilization.
Patent Information
- Application Number
- CN202110757795.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-07-05
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-07-05
AI Technical Summary
Existing system status detection methods cannot analyze system status in real time, resulting in the inability to detect status anomalies in a timely manner, which increases operational security risks and maintenance costs.
By employing aviation big data analysis methods, historical data of the system under test is filtered, preprocessed, and data feature values are extracted to establish a performance baseline. Linear or artificial intelligence models are then used to establish the performance baseline and monitor the system status in real time.
It enables real-time monitoring of system status, timely detection of anomalies, reduces system operation security risks and maintenance costs, and improves detection efficiency and system availability.
Smart Images

Figure CN113343364B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer technology, in particular to the field of civil aviation technology, and specifically to an aviation big data analysis method and tool. Background Art
[0002] During the system status detection process, it is necessary to monitor the performance indicators of important systems in real time so that possible abnormal status can be predicted in time. Taking aircraft maintenance as an example, the existing status detection adopts a passive response maintenance mode of routine work and regular maintenance. This backward maintenance mode cannot analyze and predict the real-time status of the system, and timely detect possible abnormal status. Reducing operational safety risks and reducing operation and maintenance costs are currently urgent issues to be solved. Summary of the invention
[0003] The present application embodiment provides an aviation big data analysis method and tool to solve at least one of the problems existing in the related technology. The technical solution is as follows:
[0004] In a first aspect, an embodiment of the present application provides an aviation big data analysis method, including:
[0005] Filter the historical data of the system to be detected to obtain the system parameters under the target scenario;
[0006] Preprocess the system parameters in the target scenario;
[0007] Extracting data feature values from the preprocessed system parameters in the target scenario;
[0008] Establish a performance baseline of the system to be tested based on the data feature values;
[0009] Perform status detection on the system to be detected based on the performance baseline.
[0010] In one embodiment, the system to be detected includes an aircraft; and the method further includes:
[0011] Construct a data storage framework corresponding to the aircraft configuration, which uses time series as the data storage form;
[0012] According to the aircraft configuration, the historical data of the system to be tested is stored in the corresponding data storage framework.
[0013] In one embodiment, the aircraft configuration includes at least one of an aircraft type enabling configuration, a stand-alone delivery configuration, and a stand-alone operational configuration.
[0014] In one embodiment, the preprocessing includes at least one of curve interpolation processing, filtering processing, curve smoothing processing, multi-parameter preprocessing and outlier processing.
[0015] In one embodiment, the system to be detected includes an aircraft; and the target scenario includes at least one of a specific time period, a specific flight segment, a specific flight phase, a specific parameter range, and a specific system type.
[0016] In one embodiment, establishing a performance baseline of the system to be detected based on the data characteristic value includes:
[0017] Analyzing the data characteristic values to obtain correlation between at least two system parameters;
[0018] A performance baseline of the system to be tested is established based on correlation between at least two system parameters.
[0019] In one embodiment, a performance baseline of the system to be detected is established based on the correlation between at least two system parameters, including: when the correlation between the system parameters is linear, using a linear model or an artificial intelligence model to establish the performance baseline of the system to be detected.
[0020] In one embodiment, a performance baseline of the system to be detected is established based on the correlation between at least two system parameters, including: when the correlation between the system parameters is nonlinear, using an artificial intelligence model to establish the performance baseline of the system to be detected.
[0021] In one embodiment, the artificial intelligence model includes at least one of a neural network model, a random forest model, and a support vector regression model.
[0022] In one embodiment, a performance baseline of the system to be detected is established based on the correlation between at least two system parameters, including: when the correlation between the system parameters is nonlinear, using a piecewise linear model to establish the performance baseline of the system to be detected.
[0023] In one implementation, performing status detection on a system to be detected based on a performance baseline includes:
[0024] Collect current data corresponding to system parameters of the system to be tested;
[0025] Compare current data with performance baselines;
[0026] Whether the state of the system to be detected is abnormal is determined according to the comparison result.
[0027] In a second aspect, the present application embodiment provides an aviation big data analysis tool, including:
[0028] A screening unit, used for screening the historical data of the system to be detected to obtain the system parameters under the target scenario;
[0029] A processing unit, used for preprocessing system parameters in a target scenario;
[0030] An extraction unit, used to extract data feature values from the preprocessed system parameters in the target scenario;
[0031] An analysis unit, used to establish a performance baseline of the system to be tested according to the data characteristic values;
[0032] The detection unit is used to perform status detection on the system to be detected based on the performance baseline.
[0033] In one embodiment, the system to be detected includes an aircraft; the above-mentioned tool also includes a storage unit for:
[0034] Construct a data storage framework corresponding to the aircraft configuration, which uses time series as the data storage form;
[0035] According to the aircraft configuration, the historical data of the system to be tested is stored in the corresponding data storage framework.
[0036] In one embodiment, the aircraft configuration includes at least one of an aircraft type enabling configuration, a stand-alone delivery configuration, and a stand-alone operational configuration.
[0037] In one embodiment, the preprocessing includes at least one of curve interpolation processing, filtering processing, curve smoothing processing, multi-parameter preprocessing and outlier processing.
[0038] In one embodiment, the system to be detected includes an aircraft; and the target scenario includes at least one of a specific time period, a specific flight segment, a specific flight phase, a specific parameter range, and a specific system type.
[0039] In one embodiment, the analysis unit comprises:
[0040] A first analysis subunit is used to analyze the data characteristic values to obtain the correlation between at least two system parameters;
[0041] The second analysis subunit is used to establish a performance baseline of the system to be detected according to the correlation between at least two system parameters.
[0042] In one embodiment, the second analysis subunit is used to: when the correlation between system parameters is linear, use a linear model or an artificial intelligence model to establish a performance baseline of the system to be tested.
[0043] In one embodiment, the second analysis subunit is used to: when the correlation between system parameters is nonlinear, use an artificial intelligence model to establish a performance baseline of the system to be tested.
[0044] In one embodiment, the artificial intelligence model includes at least one of a neural network model, a random forest model, and a support vector regression model.
[0045] In one embodiment, the second analysis subunit is used to establish a performance baseline of the system to be detected using a piecewise linear model when the correlation between system parameters is nonlinear.
[0046] In one embodiment, the detection unit is used to:
[0047] Collect current data corresponding to system parameters of the system to be tested;
[0048] Compare current data with performance baselines;
[0049] Whether the state of the system to be detected is abnormal is determined according to the comparison result.
[0050] In a third aspect, an embodiment of the present application provides an electronic device, the device comprising: a memory and a processor, wherein the memory stores instructions, and the instructions are loaded and executed by the processor to implement the method in any one of the above-mentioned aspects.
[0051] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, a method in any one of the above-mentioned embodiments is implemented.
[0052] The advantages or beneficial effects of the above technical solution include at least: being able to timely detect possible state anomalies and system failures in the target scenario, so as to reduce non-routine sudden failures of the system through preventive maintenance, reduce system operation safety risks, improve detection efficiency, reduce operation and maintenance costs, and improve system availability, thereby increasing profits.
[0053] The above summary is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments and features described above, further aspects, embodiments and features of the present application will be readily apparent by reference to the accompanying drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the multiple drawings represent the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some embodiments disclosed in the present application and should not be regarded as limiting the scope of the present application.
[0055] Figure 1 is a flow chart of an aviation big data analysis method according to an embodiment of the present application;
[0056] Figure 2 A schematic diagram of data acquisition of the aviation big data analysis method according to an embodiment of the present application;
[0057] Figure 3 A schematic diagram of data acquisition of the aviation big data analysis method according to an embodiment of the present application;
[0058] Figure 4 A schematic diagram of the data storage system architecture of the aviation big data analysis method according to an embodiment of the present application;
[0059] Figure 5 A schematic diagram of configuration management of an aviation big data analysis method according to an embodiment of the present application;
[0060] Figure 6 A schematic diagram showing the function of correlation analysis of the aviation big data analysis method according to an embodiment of the present application;
[0061] Figure 7 A schematic diagram showing the functions of the aviation big data analysis method according to an embodiment of the present application;
[0062] Figure 8 is a flow chart of an aviation big data analysis method according to an embodiment of the present application;
[0063] Fig. 9 is a flow chart of an aviation big data analysis method according to an embodiment of the present application;
[0064] Fig.10 A system architecture diagram of an aviation big data analysis method according to an embodiment of the present application;
[0065] Fig.11 is a schematic diagram of the structure of an aviation big data analysis tool according to an embodiment of the present application;
[0066] Fig.12 is a schematic diagram of the structure of an aviation big data analysis tool according to an embodiment of the present application;
[0067] Fig.13 A block diagram of an electronic device used to implement an embodiment of the present application. DETAILED DESCRIPTION
[0068] In the following, only some exemplary embodiments are briefly described. As those skilled in the art will appreciate, the described embodiments may be modified in various ways without departing from the spirit or scope of the present application. Therefore, the drawings and descriptions are considered to be exemplary and non-restrictive in nature.
[0069] Figure 1 FIG. 1 is a flow chart of an aviation big data analysis method according to an embodiment of the present application. Figure 1As shown, the aviation big data analysis method may include:
[0070] Step S110, screening the historical data of the system to be detected to obtain system parameters under the target scenario;
[0071] Step S120, preprocessing the system parameters in the target scene;
[0072] Step S122, extracting data feature values from the preprocessed system parameters in the target scenario;
[0073] Step S130, establishing a performance baseline of the system to be tested according to the data characteristic values;
[0074] Step S140: Perform status detection on the system to be detected based on the performance baseline.
[0075] In the process of system status detection, system parameters that can reflect or embody system performance can be obtained according to the historical data of the system to be detected, and then a performance baseline can be established based on the system parameters. By comparing the current data corresponding to the system parameters of the system to be detected with the performance baseline, possible state anomalies can be predicted in time. Among them, the system can be composed of multiple components. For example, a computer system can be composed of component-level devices such as a CPU (Central Processing Unit), a motherboard, a memory, a hard disk, an output device, and an input device. In one example, the system to be detected may include an aerospace system or a rail transit system. In another example, the system to be detected may also include a van, a subway, a light rail, and a tram. The performance baseline may include a standard value or a standard value range of a performance parameter of the system under normal conditions. In the embodiment of the present application, an aviation big data analysis tool can be used to collect, preprocess, and analyze the correlation of data and establish a performance baseline.
[0076] In the embodiment of the present application, historical data of the system to be detected can be acquired in advance before the system detection. Figure 2 and Figure 3 The historical data collected from the aircraft is shown in Figure 2 In the figure, the horizontal axis represents time and the vertical axis represents the inlet temperature of the air-conditioning component. Figure 2 It is also shown in FIG. 1 that the maximum value of the angle represented by the ordinate is 194, the minimum value is 62, and the median value is 159.46. Figure 3 In the figure, the horizontal axis represents time, and the vertical axis represents the temperature value collected by the temperature sensor. Figure 3 It is also shown that the maximum value of the temperature value represented by the ordinate is 73°C, the minimum value is 0°C, and the median value is 28.6°C.
[0077] In step S110, the system parameters under the target scenario are screened out from the historical data. For example, the target scenario may include a specific time period, a specific parameter range, and a specific system type. Taking aircraft maintenance as an example, the target scenario may also include a specific flight segment, a specific flight phase, etc. The historical data is preliminarily screened by a specific time period, a specific flight segment, a specific flight phase, and a specific parameter range to obtain the system parameters under the target scenario. Processing the screened system parameters under the target scenario can make the processing process more targeted, so that abnormal system status can be discovered through preventive maintenance, and sudden system failures under the target scenario can be reduced.
[0078] Still taking aircraft maintenance as an example, data segmentation and collection technology can be used to obtain system parameters in the target scenario. For example, data segmentation and collection may include selecting aircraft model and number, selecting time stage range, selecting fault data, custom parameter screening, etc. In one example, data slicing and segmentation can be achieved by using the aviation big data analysis tool provided in the embodiment of the present application. Among them, the segmentation method can be segmented by flight stage, and the segmentation method can be segmented by parameter range.
[0079] In one example, data segmentation can also be performed by custom parameter screening. For example, if an aircraft is used in a temperate zone, the data in tropical and frigid zones can be removed during data screening. In the above example, an aircraft is used in a temperate zone, and for SAT (Static Air Temperature), data below 10°C and data above 40°C can be removed.
[0080] In step S120 and step S122, the system parameters in the target scenario are preprocessed, and data feature values are extracted from the preprocessed system parameters in the target scenario. In step S130, a performance baseline of the system to be detected can be established based on the data feature values extracted in step S122. In one example, historical fault data can also be used for verification, and a model for fitting the performance baseline of the system to be detected is selected based on the verification results.
[0081] In the embodiment of the present application, the aviation big data analysis tool can be used to establish a performance baseline in the whole life cycle of the system to be detected or at any stage in the whole life cycle. After each acquisition of the actual data of the system to be detected, by comparing the actual data with the performance baseline, observing the deviation between the actual data and the data in the performance baseline, and determining whether the system to be detected has an abnormal state according to the deviation. Taking the maintenance of aircraft air conditioning components as an example, the performance baseline of the component may include the normal use state of the component when the outside temperature is 30°C, the inlet temperature is 100°C, and the heat exchanger outlet temperature is 60°C. For example, when the outside temperature is 30°C, the inlet temperature is 100°C, and the heat exchanger outlet temperature is 80°C, then its state is no longer consistent with the performance baseline, that is, the state of the component has deviated from the performance baseline. For example, the performance baseline of the component may include the corresponding relationship between the component outlet temperature and the outside temperature and the component inlet temperature. In one example, the performance baseline of the component is established based on the correlation between the component outlet temperature and the system parameters such as the outside temperature and the component inlet temperature. The performance baseline of a component may include: when the outside temperature is 20°C, the temperature difference between the inlet and outlet of the heat exchanger can reach 40°C, and the current thermal efficiency is calculated to be 0.85. If the temperature difference between the inlet and outlet of the heat exchanger reaches 20°C when the outside temperature is 20°C, and the current thermal efficiency is calculated to be 0.6, then the state of the component has deviated from the performance baseline. Although the component may still be in normal use and no abnormalities or failures have occurred for the time being, the current state of the component is no longer within the normal range. In step S140, current data corresponding to the system parameters of the system to be detected may be collected, such as collecting system parameters in the current state of the aircraft. The system parameters in the current state of the aircraft are then compared with the performance baseline, and the aircraft is tested based on the performance baseline.
[0082] In IT (Information Technology), "big data" requires new processing modes to have stronger decision-making power, insight discovery and process optimization capabilities to adapt to massive, high-growth and diverse information assets. In the embodiment of the present application, the historical data of the system to be tested can be used to build a database of system big data. In an example of aircraft maintenance, the aviation maintenance big data storage and management system can be used to manage historical data.
[0083] Taking the use of aviation big data analysis tools to perform aircraft engine early warning analysis as an example, the early warning analysis at the startup stage may include the following processes:
[0084] 1. First, extract the system parameters and related aircraft basic parameters during the engine startup phase, and then use QAR (Quick Access Recorder) data to perform trend analysis on the corresponding parameters. Among them, QAR is a data type. Trend analysis may include: continuous comparison of left and right engine data of the same aircraft; continuous comparison of data of a certain engine of the same aircraft for a period of time; and comparison of left engine data and right engine data of two aircraft of the same configuration on the same route. In an example, when checking the data of the aircraft engine, if abnormal values of speed, vibration, and temperature are found, the abnormal values can be removed through the outlier processing function in the aviation big data analysis tool.
[0085] 2. Based on the characteristics of data changes, obtain the law of data changes, and establish a performance baseline based on the law of data changes.
[0086] 3. Verify the performance baseline based on historical engine fault data. For example, the performance baseline can be verified through stable data, and the performance baseline alarm threshold can be adjusted through fault data.
[0087] 4. In the later stage, the performance baseline can be optimized through continuous big data analysis.
[0088] 5. Achieve accurate early warning of engine-related failures and component performance degradation.
[0089] In one example, when analyzing aircraft engine data, a specified number of aircraft in a fleet can be selected to analyze the data of the engine startup phase of a set time period (for example, 20 days). Aviation big data analysis tools can be used to extract and analyze data. In one example, the data extraction logic is: from the beginning of data recording to the first time the parking brake state changes from ON to OFF. Some abnormal data are deleted during data analysis, such as: the total duration of the startup phase is less than 30 seconds or greater than 5 minutes, and erroneous and abnormal segment data generated when the segment is split. The data analysis of aircraft engines may include: comparison of the difference data of the left and right engines on the same tail, comparison of different parameters of the same engine serial number, and comparison of parameters of different tails on the same route.
[0090] In aviation big data analysis tools, through the trend analysis of the above parameters related to the engine startup phase, it is possible to determine the time when the relevant engine parameters reach a relatively stable state under normal conditions and their changing patterns. Based on the analysis of a large amount of QAR data and the analysis and verification through the historical fault data of the relevant engines, an accurate engine early warning model can be established, so that early warning analysis of the engine startup phase can be performed based on the aircraft QAR data.
[0091] In the embodiment of the present application, aviation big data analysis tools can be used to monitor system performance, establish a normal value range, and issue an early warning for exceeding the normal range. The following uses the aircraft air conditioning system and the aircraft bleed air system as examples to illustrate how to perform performance monitoring and early warning.
[0092] Taking the performance monitoring of aircraft air conditioning system as an example, the following system parameters can be detected and monitored through aviation big data analysis tools: component outlet temperature TP, component flow PF, compressor outlet temperature COT, water separator temperature TW, bypass valve opening PBV, ram air inlet valve opening RI, precooler outlet temperature TPO, precooler inlet pressure PD. The detection and monitoring of aircraft air conditioning performance through aviation big data analysis tools can include the following processes:
[0093] 1) Screen the TPO, PF, and PD parameters with SAT greater than 20°C during the taxiing phase of the aircraft, process the outliers of the parameter set, filter out the mutation values, outliers, and null values other than the normal values, and obtain the parameter set within the normal range in summer.
[0094] 2) Select the frequency domain, time domain, and time-frequency feature extraction algorithms that meet the parameter characteristics, and apply the preprocessed data feature values.
[0095] 3) Analyze parameter correlation through correlation analysis function in combination with system principles. If it conforms to linear relationship, select regression model to establish performance baseline. If it does not conform to linear relationship, select AI tool to establish performance baseline.
[0096] 4) Import normal or fault historical data in summer through the condition monitoring function to verify the effectiveness of the performance baseline.
[0097] Taking the performance monitoring of the aircraft bleed air system as an example, the following system parameters can be detected and monitored through aviation big data analysis tools: precooler outlet temperature TPO, precooler inlet pressure PD, engine speed N2 (high-pressure rotor speed). The detection and monitoring of aircraft bleed air system performance through aviation big data analysis tools can include the following processes:
[0098] 1. Slide-out phase monitoring
[0099] 1.1) Screen the TPO, PF, and PD parameters when N2 is stable during the sliding phase, process the parameter set for outliers, filter out mutation values, outliers, and null values other than normal values, obtain the parameter set within the normal range for a specified period (such as summer), and draw a chart for monitoring. The N2 stable range is: median ± A, A = max (1, 1.5 * IQR), IQR = upper quartile - lower quartile.
[0100] Among them, IQR (Inter-Quartile Range) is a measure of the variability of the results. The interquartile range refers to the difference between the value on the third quartile and the value on the first quartile after arranging the values of each variable in order of size and then dividing the series into four equal parts.
[0101] 1.2) Select frequency domain, time domain, and time-frequency feature extraction algorithms that meet the parameter characteristics, and apply the preprocessed data feature values for subsequent processing.
[0102] 1.3) Analyze the correlation of parameters through the correlation analysis function in combination with the system principle. If it conforms to the linear relationship, select the regression model to establish the performance baseline; if it does not conform to the linear relationship, select the AI tool to establish the performance baseline.
[0103] 1.4) Import normal or fault historical data of a specified period through the condition monitoring function to verify the validity of the performance baseline.
[0104] 1.5) When performing status detection on the system to be detected, after collecting the actual data of the system to be detected, the actual data is compared with the performance baseline, and the deviation between the actual data and the data in the performance baseline is used to determine whether the system to be detected has an abnormal status. For example, an early warning is issued when the PD mean is lower than 20 and the standard deviation is higher than 5.
[0105] 2.1) Screen the PD parameters in the slide-out phase and execute a process similar to 1.1) to 1.4) above to process outliers, extract features, establish a performance baseline, and verify the effectiveness of the performance baseline.
[0106] In one example, peaks and troughs are identified when the first-order derivative is 0 and the second-order derivative is negative, or when the first-order derivative is 0 and the second-order derivative is positive, and a fluctuation cycle that includes a peak, a trough, and a peak in sequence or a trough, a peak, and a trough in sequence is taken as a fluctuation cycle, and the difference between the peak value and the trough value is taken as the amplitude.
[0107] When the status of the system to be detected is tested, the number of fluctuations, average fluctuation period, and average amplitude when the PD amplitude is greater than 8 are calculated, and a chart is drawn for monitoring. The actual data is compared with the performance baseline, and an early warning is issued when the number of fluctuations is greater than 20, the fluctuation period is less than 15S, and the average amplitude is greater than 10.
[0108] 2. Cruise phase monitoring
[0109] During the cruise phase, a process similar to the taxi-out phase is executed, using aviation big data analysis methods to perform parameter screening, outlier processing, feature extraction, establish a performance baseline, and verify the effectiveness of the performance baseline. When performing status detection on the system to be detected, the mean and standard deviation of TPO and PD in the cruise phase can be calculated, and a chart can be drawn for monitoring. The actual data is compared with the performance baseline, and an early warning is triggered for one of the following situations: 1) TPO mean is lower than 175; 2) TPO standard deviation is greater than 5; 3) PD mean is lower than 37; 4) PD standard deviation is greater than 5.
[0110] In one embodiment, the system to be detected includes an aircraft; and the target scenario includes at least one of a specific time period, a specific flight segment, a specific flight phase, a specific parameter range, and a specific system type.
[0111] In one example, aviation big data analysis tools can be used to filter data. The filtering conditions may include flight phases or flight routes. For aircraft, there may be corresponding different safety hazards in different target scenarios. Therefore, collecting system parameters for different target scenarios for analysis can make system detection more targeted and improve detection efficiency.
[0112] In summary, in the embodiments of the present application, aviation big data analysis tools can be used to collect, preprocess, and perform correlation analysis on data and establish a performance baseline.
[0113] In one embodiment, the system to be detected includes an aircraft; the method further includes:
[0114] Constructing a data storage framework corresponding to the aircraft configuration, wherein the data storage framework uses a time series as a data storage form;
[0115] According to the aircraft configuration, the historical data of the system to be tested is stored in the corresponding data storage framework.
[0116] Aircraft configuration is a general term for the main features of the aircraft's geometric shape and various loading arrangements of the aircraft. In the example of aircraft detection, the historical data of the system to be detected that has been collected in advance can be stored in the database. A data storage framework corresponding to the aircraft configuration can be built in the database in advance so that the historical data of the aircraft can be stored in the corresponding data storage framework. On this basis, a configured data access interface service can be provided for business applications.
[0117] In one example, the aviation maintenance big data storage management system mainly implements functions including QAR (Quick Access Recorder) data management and decoding processing, data storage and management, and configured data access interface services. The following means can be used to manage aviation maintenance big data:
[0118] 1) On the basis of decoding and structuring the QAR message data, the data is stored in a time series, and the processed and organized QAR message data is stored in a time series database. With the time series database as the core, the aircraft type configuration data and fleet basic information are used to effectively manage the massive aviation flight collection data. For example, QAR data can be used to establish a dynamic model of the aircraft engine, air source and air conditioning system for performance monitoring and subsequent early warning analysis. The QAR data exploration platform can be used to provide modeling support for the model operation platform.
[0119] Among them, fleet refers to the combination unit of aircraft. Airlines choose aircraft models and form fleets according to their own operational needs. QAR is an airborne device used to monitor and record a large number of flight parameters and data. For example, QAR can be used to collect hundreds of data at the same time, covering most of the parameters of aircraft operation quality. QAR monitoring is a scientific and effective technical means to ensure flight safety and improve operational efficiency. Its monitoring results are an important basis for flight technical inspections, safety assessments, safety incident investigations and aircraft maintenance. With the help of QAR, airlines can promptly discover problems in flight crew control, engine operating conditions, and aircraft performance, analyze and find causes, grasp safety trends, and take targeted measures to eliminate potential accidents and ensure flight safety.
[0120] In related data storage technologies, the acquired data can be packaged into folders for storage at preset time intervals. For example, the data acquired within 60 seconds is packaged in a folder. The drawback of the above data storage method is that the data of the entire folder needs to be parsed when acquiring data. For example, if it is necessary to acquire data within the first second, the data of the entire folder for 60 seconds must also be parsed. With this method, the efficiency of parsing and acquiring data is low and inconvenient to operate. In an embodiment of the present application, data is stored in a time series, for example, data can be stored once every 1 second, and a time series database is established based on this. In the time series database, data can be sorted according to time. This data storage method can facilitate data storage, search and parsing, and is more efficient.
[0121] 2) Flight parameters are managed in a unified manner in the form of a tag dictionary, and can also be managed in combination with the basic configuration information model of the aircraft model to be studied and constructed. Through the tag organization form and configuration information management, flight parameters with different attributes are connected in series, the flexibility of data access is improved, and the data access requirements of different business applications are met. For example, a data access interface for the maintenance system can be provided for airlines, and a data query interface for production and research and development can be provided for equipment manufacturers.
[0122] The data storage method in the embodiment of the present application is different from the method of packaging the entire flight segment into a file for storage in the prior art. In the embodiment of the present application, by storing the QAR data in a time sequence, it is helpful to extract the parameter group at a certain point in time, or extract the parameter group for a certain time period, so as to analyze the data more specifically. In the embodiment of the present application, the configuration information model of the aircraft model can be used as the basis, and by associating the flight parameters with the configuration entities (complete aircraft, system / subsystem, equipment, etc.), the organization and management of the flight parameter label dictionary library can be implemented in a configured "tree" structure. In an example, a label in the label dictionary can be selected, and then the custom parameters of the previous and / or next segment within the label range can be extracted for data exploration and processing.
[0123] Taking aircraft component maintenance as an example, the performance baseline of the component established through aviation big data analysis tools may include that the normal use state of the IDG component is a stable output of 115V 400Hz AC power. If the output of the IDG exceeds 420Hz for a period of time and generates a fault code, a "IDG overclocking" label can be added at the time of the fault. When repairing components, the label can be extracted from the time series database, and all recorded data at the time point corresponding to the label can be extracted. Relevant parameters can be found from these data to further find the cause of component failure. Through the time series database and the label organization of data, the flexibility of data access and data processing efficiency can be improved, thereby reducing the processing cost of resources.
[0124] Figure 4 FIG. 1 is a schematic diagram of a data storage system architecture of an aviation big data analysis method according to an embodiment of the present application. Figure 4 As shown, the aviation maintenance big data storage management system can obtain data from a third-party airborne data acquisition and recording system, and then process the acquired data through QAR data management and decoding, and store the processed data in the QAR original message repository. In addition, it is also necessary to perform structured flight data storage management on the data after QAR data management and decoding, store the data in time series, and store the organized data in the time series flight data repository. The embodiment of the present application supports data integration technology from data acquisition to data cleaning, collects and stores QAR data, builds a QAR database, and develops a data interface to support the flexible selection of different QAR parameters and feature values to generate new data sets.
[0125] The aviation maintenance big data storage management system can provide services for business applications through the configured data access interface service. In addition, in the aviation maintenance big data storage management system, it can also provide configured flight parameter configuration management for QAR data management and decoding processing, structured flight data storage management, and configured data access interface services.
[0126] Among them, the configurational data access interface service may include: 1) a parameter label dictionary library management interface based on the configuration model, which realizes the functions of configurational flight parameter query browsing and management and maintenance. 2) a time series data access management interface, which realizes the functions of querying time series flight status data, full parameter slice data, etc., and interpreting data maintenance.
[0127] Taking aircraft component maintenance as an example, the established performance baseline can be configured and managed. Configuration management can include establishing performance baselines for various types of components and then establishing corresponding relationships between each type of component and its respective performance baseline. In one example, if a component is replaced with a part during maintenance, such as a transformer, its performance baseline may change. In this case, it is necessary to re-establish the performance baseline for the component with the replaced parts.
[0128] In one embodiment, the aircraft configuration includes at least one of an aircraft type enabling configuration, a stand-alone delivery configuration, and a stand-alone operational configuration.
[0129] Figure 5 The figure is a schematic diagram of configuration management according to the aviation big data analysis method according to the embodiment of the present application. From the full life cycle of the aircraft, the changes in configuration management can be reflected in the three basic stages of aircraft introduction, delivery, and operation. The aircraft configurations corresponding to the three basic stages of aircraft introduction, delivery, and operation are the model installation configuration, the single-machine delivery configuration, and the single-machine operation configuration. A data storage framework corresponding to the aircraft configuration can be constructed in the aviation maintenance big data storage management system, and then the historical data of the system to be tested can be stored in the corresponding data storage framework.
[0130] like Figure 5As shown, in the data storage framework corresponding to the aircraft model installation configuration, the stored technical data may include: IPC (Illustrated Parts Catalogs), AMM (Aircraft Maintenance Manual), SSM (System Schematic Manual), and related data parameter lists. The configuration attributes in the aircraft model installation configuration may include: at least one of: part number, model, applicability, location, SB (service bulletins) modification status, upper and lower component structure, drawing number and replaceability. Among them, applicability may include whether the component is applicable to each aircraft, location may include the installation location of the component, and replaceability includes that the installed component can be replaced with components of different models or different manufacturers.
[0131] In the embodiment of the present application, the hierarchical structure of the parts (Part) can be organized based on the illustrated parts list (IPC), and the component configuration can be constructed by considering the replaceability of the sub-components in the hierarchical structure. The allowed configuration can check and verify the configuration, missing installation, wrong installation and other operations that affect the safety of the aircraft of the single machine configuration, and the allowed configuration of the aircraft model can be constructed by defining the allowed installation rules of each configuration element node.
[0132] In one example, when the performance baseline of a component includes that the components of the batch can be used normally for 1000 hours, it is assumed that some components in the batch fail around the 100th hour of use, and the compressor is replaced during maintenance. In this case, an SB operation can be performed to replace the compressors in all components of the batch. And the above SB modification status is mounted to the permissible configuration of the components of the batch, recording that the components of the batch have completed the above SB operation. In the example of aircraft maintenance, the change process of component configuration and permissible configuration can be sorted out, and the steps of changing component configuration, changing permissible configuration, and changing single-machine configuration can be clarified to achieve aircraft configuration change control, and form a change process for single-machine configuration change records.
[0133] like Figure 5As shown, in the data storage framework corresponding to the stand-alone delivery configuration, the stored delivery files may include: ARL (aircraft readiness log), onboard software list, etc. The supplementary attributes in the stand-alone delivery configuration may include: applicability screening, component serial number, whether it involves regular inspection, whether it involves prohibited installation, system maintenance strategy, etc. In one example, each component can be inspected regularly during the aircraft delivery phase. If a component is found to have a problem during the regular inspection, the problematic component will be recalled, and a prohibited installation label will be affixed to the component, and other modified components will be replaced. The system maintenance strategy may include replacing components, and the replaced components can be returned to the factory for disassembly and repair.
[0134] like Figure 5 As shown, in the data storage framework corresponding to the stand-alone operation configuration, the operation and maintenance records stored may include: records involving AD (Airworthiness Directive), records involving SB, and records of component replacement, etc. The management dimensions in the stand-alone operation configuration may include: at least one of: part number, model, region, SB execution status, upper and lower component structure, replaceability, and maintenance strategy.
[0135] In one example, data corresponding to various aircraft configurations can also be stored from a regional perspective and a functional perspective. Among them, the storage method based on the regional perspective can divide the data into regions according to actual needs. The storage method based on the functional perspective can store the data that implements related functions together, and can organize the system-level state parameters, QAR data and dynamic performance models in series based on the functional configuration of the aircraft, on this basis, data query and data analysis can be more convenient. Taking the aircraft as an example, facing the integration requirements of aircraft operation, maintenance and sensor data, the aircraft configuration data can be sorted out according to the configuration bus technical specifications to complete the aircraft configuration bus design. Configuration bus management software tools can also be deployed to integrate aircraft operation, maintenance and sensor data using the configuration bus to achieve multi-dimensional aircraft configuration management. In addition, the embodiment of the present application supports aircraft configuration change control, which can meet the needs of fine-grained control of aircraft configuration changes and the management requirements of one aircraft and one file technical data.
[0136] Taking air conditioning maintenance as an example, the hardware of the system can be associated with the corresponding baseline to effectively manage the performance baseline. For example, the hardware involved in the air conditioning flow control system may include air conditioning components, pipe installation, etc. The air conditioning components involve heat exchangers, compressors, condensers, regenerators, turbines, sensors at corresponding locations, etc. The established performance baseline can be mounted on the corresponding hardware, and a performance baseline can be further established for each hardware and its components.
[0137] Taking aircraft maintenance as an example, the aircraft can be functionally decomposed, and the decomposed elements can be hierarchically classified using the judgment method of "system, subsystem, functional group, functional part". An exemplary aircraft system functional classification may include: taking the aircraft as the root node, decomposing the aircraft into various systems at the lower level; then decomposing the system into subsystems and functional groups; and further decomposing the subsystem into functional parts (units). The functional configuration of the model can be organized according to the functional composition of the actual model; and the relevant parameters and dynamic performance models are associated with the functional configuration elements to form a unified management of data and models with the functional configuration as the core.
[0138] In an embodiment of the present application, before extracting data feature values from system parameters in the target scene in step S122, step S120 may be executed to perform data preprocessing. Data preprocessing may include: at least one of data interpolation processing, filtering processing, curve smoothing processing, multi-parameter preprocessing and outlier processing. Among them, multi-parameter preprocessing may include alignment or difference compensation for different parameters. For example, the outside temperature may be measured every 4 seconds, and the outlet temperature of the air-conditioning component may be measured every 1 second. In this case, a measurement value can be extracted from every 4 measured values of the outlet temperature of the air-conditioning component to make it correspond one to one with the measured value of the outside temperature. Outlier processing may include: removing outliers in system parameters that obviously deviate from the normal range. For example, when a system is used near 40 degrees north latitude, the ambient temperature measured is -50°C, then the above measurement data is an outlier that obviously deviates from the normal range and can be removed.
[0139] In one example, data for a target scenario is generated by preliminarily screening a specific time period, a specific flight segment, a specific flight phase, and a specific parameter range of historical data. Then, data preprocessing is performed on the system parameters for the target scenario. After data preprocessing, a preliminary understanding of the data characteristics can be obtained. For example, for a specific flight segment from Beijing to Sanya, data from a temperature sensor can be collected. After curve interpolation and curve smoothing processing of the collected temperature sensor data, a preliminary understanding of the data characteristics can be obtained: in a specific flight segment from Beijing to Sanya, the temperature value changes from low to high.
[0140] In the embodiment of the present application, the extraction of characteristic values can be for a single system parameter, or for a combination of multiple system parameters to generate a result. For example, in the above example, for a specific flight segment from Beijing to Sanya, the data collected by the temperature sensor and the flight altitude can be combined to generate a result.
[0141] In the embodiment of the present application, in step S122, the data characteristic value extracted from the system parameter may include at least one of a mean, a variance, an absolute value, a sum, a maximum value, a minimum value, and a median value. Figure 2 and Figure 3, which shows the maximum, minimum and median values of the system parameters. The method of extracting data feature values may include at least one of frequency domain features, time domain features, time-frequency features and custom features. Among them, frequency domain features may include: extracting frequency domain information using Fourier transform, such as extracting frequency domain features for data within a certain period. Time domain features may include extracting data within a preset time period, such as extracting the mean, root mean square or peak value. Time-frequency features include extracting data according to the time dimension. This method includes using short-time Fourier transform or wavelet analysis theory to characterize the frequency domain characteristics of a certain moment.
[0142] In one embodiment, Figure 1 Step S130 in the embodiment of the present invention is to establish a performance baseline of the system to be detected according to the data characteristic value, including:
[0143] Analyzing the data characteristic values to obtain correlation between at least two system parameters;
[0144] A performance baseline of the system to be tested is established based on correlation between at least two system parameters.
[0145] In the embodiment of the present application, the aviation big data analysis tool provided in the embodiment of the present application can be used to analyze the data feature values, select at least two system parameters from a variety of system parameters, and obtain the correlation between the at least two system parameters. Among them, the correlation may include linear correlation and nonlinear correlation. The correlation between the at least two system parameters may be linear correlation or nonlinear correlation. In the case of linear correlation, a linear model can be used to fit the performance baseline of the system to be detected. For a nonlinear model, an artificial intelligence model or a piecewise linear model can be used to fit the performance baseline of the system to be detected. In step S130, the aviation big data analysis tool provided in the embodiment of the present application can be used to select a model for establishing a performance baseline through correlation. In one example, a data analysis model corresponding to the correlation can be pre-set. For example, the data analysis model corresponding to the linear correlation can be a linear model or an artificial intelligence model, and the data analysis model corresponding to the nonlinear correlation can be at least one of an artificial intelligence model and a piecewise linear model. Then, the data analysis model corresponding to the correlation is used to establish the performance baseline of the system to be detected.
[0146] Figure 6 A schematic diagram showing the function of correlation analysis of the aviation big data analysis method according to an embodiment of the present application. Figure 6 The correlation analysis results of two system parameters, EGT MAX and MIN ALT, are shown. EGT MAX represents the maximum value of EGT (exhaust gas temperature) and MIN ALT represents the minimum value of ALT (altitude). Figure 6The correlation matrix (corMatrix) and heat map (corGramPath) of EGT MAX and MIN ALT are shown in FIG. The elements in the correlation matrix represent the values of the correlation coefficient between two system parameters. The values in the heat map represent the values of the correlation coefficient in the correlation matrix.
[0147] Figure 7 A schematic diagram showing the functions of the aviation big data analysis method according to an embodiment of the present application. Figure 7 The figure shows the correlation matrix of several system parameters including EGT, N1 (Low Rotor Speed), TAT (Total Air Temperature), ALT, MN (Mach Number), and SAT (Static Air Temperature). Figure 7 It can be seen that there is an obvious linear relationship between EGT and N1, and between TAT and SAT. Among them, TAT and SAT seem to be completely linearly correlated, so when doing regression, you can consider using one of TAT and SAT as the independent variable. The linear relationship between ALT and EGT is not obvious. In this case, in the aviation big data analysis tool, you can consider using a piecewise linear model or a nonlinear relationship model as a performance baseline.
[0148] In the embodiment of the present application, the correlation between system parameters can be analyzed by a predetermined mathematical algorithm. Data is extracted in different scenarios to obtain the correlation of data at different stages. Figure 7 , aviation big data analysis tools can be used to provide digital and visual visualization services.
[0149] By using the above method and using aviation big data analysis tools to analyze the correlation between multiple system parameters, the correlation between at least two system parameters can be obtained. For example, the correlation may include linear correlation and nonlinear correlation. In the case where the correlation between at least two system parameters is linearly correlated, a linear model or an artificial intelligence model can be used to fit the performance baseline of the system to be detected. In the case where there is a correlation between at least two system parameters, but the correlation is nonlinear, an AI (Artificial Intelligence) model or a piecewise linear model can be used to fit the performance baseline of the system to be detected.
[0150] In one embodiment, Figure 1 Step S130 in the method includes establishing a performance baseline of the system to be detected based on the correlation between at least two system parameters, including: when the correlation between the system parameters is a linear correlation, establishing a performance baseline of the system to be detected using a linear model or an artificial intelligence model.
[0151] Among them, linear model is a general term for a type of statistical model, and its production method is to connect various links with a certain process. Linear model can include linear regression model, variance analysis model, etc.
[0152] In one embodiment, Figure 1 Step S130 in the method includes establishing a performance baseline of the system to be detected based on the correlation between at least two system parameters, including: when the correlation between the system parameters is nonlinear, using an artificial intelligence model to establish the performance baseline of the system to be detected.
[0153] In one embodiment, the artificial intelligence model includes at least one of a neural network model, a random forest model, and a support vector regression model.
[0154] Among them, the random forest model is a classifier containing multiple decision trees, which can use multiple trees to train and predict samples. The support vector regression model can transform the actual problem into a high-dimensional feature space through nonlinear transformation, and after increasing the dimension, construct a linear decision function in the high-dimensional space to achieve linear regression.
[0155] In one embodiment, Figure 1 Step S130 in the method includes establishing a performance baseline of the system to be detected according to the correlation between at least two system parameters, including: when the correlation between the system parameters is nonlinear, using a piecewise linear model to establish the performance baseline of the system to be detected.
[0156] Among them, the piecewise linear model refers to a regression estimation method that is applicable when the data in a certain range obey a certain linear relationship and in other ranges obey a linear relationship with a different slope. This method uses indicator variables to simultaneously fit a unified regression model to the data in different ranges of each segment.
[0157] In one example, the performance baseline of the temperature of a component on an aircraft before modification may include: the component temperature increases as the outside temperature increases, and the two are in a linear relationship. A fan may be installed for the component during modification. After the fan is installed, the higher the outside temperature and the faster the fan speed, the more the component temperature drops. In other words, the performance baseline of the component may have changed after the modification. The component temperature and the outside temperature may no longer be in a linear relationship. The relationship between the three quantities of component temperature, outside temperature, and fan speed may become more complicated. In this case, it may be necessary to use an artificial intelligence model to establish a performance baseline for the component.
[0158] In an example of aircraft inspection, the aircraft system composition, the functions and working modes of each component of the aircraft, and the subsequent impact of the failure mode are first analyzed. In this example, the above-mentioned aviation big data analysis tool can be called an aircraft data exploration tool. Define the stable state and the unstable state, extract the data under these two states to establish the relationship between the system parameters, and use the data under the stable state to establish the performance baseline in the aircraft data exploration tool. For example, the air conditioning system uses the aircraft taxiing stage as the stable stage; the engine generator uses the cruising stage as the stable stage.
[0159] In one example, data in abnormal states can be removed during data screening so that data in normal and stable states can be used to establish a performance baseline in subsequent processing steps. Taking aircraft components as an example, data collected when the components are working normally can be selected. For example, when the components are working in a normal temperature environment, the components can be considered to be in a normal and stable state. Under the normal and stable state, the data set that most intuitively reflects the performance can be intercepted for the established performance baseline.
[0160] The aviation big data analysis tool applied to aircraft status detection is also called an aircraft data exploration tool. In the embodiment of the present application, the aircraft data exploration tool can be used to implement the function of aircraft status detection. The aircraft data exploration tool is a platform for processing, analyzing and exploring aviation big data, and can also be called a data exploration platform. The platform uses functions such as data preprocessing, data feature value extraction, correlation analysis, and data modeling to gradually strengthen the exploration and cognition of data, thereby strengthening the health management of the aircraft and ultimately achieving the goal of preventive maintenance.
[0161] In one example, the aviation big data analysis tool provided in the embodiment of the present application can be used to divide the historical data of the system to be detected into training data and verification data. After the performance baseline is established, the baseline model can also be verified by verification data to determine its accuracy. For example, the historical data can be compared with the established performance baseline. If the corresponding historical data deviates from the performance baseline when a system failure occurs, it means that the established performance baseline can verify the system failure; conversely, if the corresponding historical data does not deviate from the performance baseline when a system failure occurs, it means that the established performance baseline does not verify the system failure.
[0162] In one example, aviation big data analysis tools can be used to establish a performance baseline during the entire life cycle of the system to be tested or at any stage of the entire life cycle. After each acquisition of actual data of the system to be tested, the actual data is compared with the performance baseline, the deviation between the actual data and the data in the performance baseline is observed, and whether the system to be tested has an abnormal state is determined based on the deviation. Taking the maintenance of aircraft air conditioning components as an example, the performance baseline of the component may include the normal use state of the component when the outside temperature is 30°C, the inlet temperature is 100°C, and the heat exchanger outlet temperature is 60°C. For example, when the outside temperature is 30°C, the inlet temperature is 100°C, and the heat exchanger outlet temperature is 80°C, then its state is no longer consistent with the performance baseline, that is, the state of the component has deviated from the performance baseline. For another example, the performance baseline of the component may include the corresponding relationship between the component outlet temperature and the outside temperature and the component inlet temperature. In one example, the performance baseline of the component is established based on the correlation between the component outlet temperature and the system parameters such as the outside temperature and the component inlet temperature. The performance baseline of a component may include: when the outside temperature is 20°C, the temperature difference between the heat exchanger inlet and outlet can reach 40°C, and the current thermal efficiency is calculated to be 0.85. If the temperature difference between the heat exchanger inlet and outlet reaches 20°C when the outside temperature is 20°C, and the current thermal efficiency is calculated to be 0.6, the status of the component has deviated from the performance baseline. Although the component may still be in normal use and no abnormalities or failures have occurred for the time being, the current status of the component is no longer within the normal range.
[0163] In another example, different models may be used to establish performance baselines for the system to be tested, such as using a linear model, a piecewise linear model, and a support vector regression model to establish performance baselines respectively, and then comparing the various performance baselines to verify whether the model used is appropriate.
[0164] Figure 8 FIG. 1 is a flow chart of an aviation big data analysis method according to an embodiment of the present application. Figure 8 As shown, in one embodiment, Figure 1 Step S140 in the embodiment: performing status detection on the system to be detected based on the performance baseline, including:
[0165] Step S210, collecting current data corresponding to system parameters of the system to be detected;
[0166] Step S220, comparing the current data with the performance baseline to see if they are consistent;
[0167] Step S230: determine whether the system to be detected is in an abnormal state according to the comparison result.
[0168] In one example, in the historical data, the median of the temperature values collected from the temperature sensor of a certain component of the aircraft is 28.6°C. If the difference between the median of the current data of the temperature value collected from the temperature sensor of the component and the historical data exceeds the preset difference threshold, it can be determined that the state of the component is abnormal. For example, if the preset difference threshold of a component is set to 2°C, and the median of the current data of the temperature value of the temperature sensor is 38.6°C, it can be determined that the state of the component is abnormal. In this case, the component may be faulty.
[0169] Still taking the aircraft status detection as an example, in the embodiment of the present application, the obtainable aircraft system transmitter parameters can be sorted out based on the aircraft engineering technical manual and the analysis and understanding of the operating principles and composition of the system components. For the aircraft system transmitter parameters, aviation big data analysis tools can be used to pre-process the accumulated historical operation data, collect relatively stable normal operation data, extract characteristic data that directly reflects the system performance, analyze the correlation relationship between system parameters, and select an applicable algorithm to establish an expected performance baseline that describes the normal operation of the system. In the process of aircraft status detection, early warning of faults can be achieved by analyzing the deviation between the actual operation data and the expected performance baseline.
[0170] Fig. 9 FIG. 1 is a flow chart of an aviation big data analysis method according to an embodiment of the present application. The above method can be implemented using aviation big data analysis tools. Fig. 9 As shown, in step S310, current data corresponding to the system parameters of the system to be detected can be collected. In step S320, the current data collected in step S310 can be preprocessed. In step S330, data feature values can be extracted from the data preprocessed in step S320, and the data feature values can be compared with the performance baseline, and the state detection of the system to be detected is performed based on the performance baseline. For example, if the data feature values extracted from the current data corresponding to the system parameters do not match the baseline, it is considered that there may be a state abnormality. In step S340, a health assessment is performed on the system to be detected based on the state detection results in step S330. For example, a health assessment warning is given in the case where there may be a state abnormality.
[0171] In the embodiment of the present application, the following can be performed in an open state monitoring and diagnosis information structure: Fig. 9The process shown. The condition monitoring and diagnosis information structure may include a data acquisition module, a data operation module, a condition detection module, a health assessment module, a prediction assessment module, and a suggestion generation module. When each module in the system is correctly configured, the basic data is converted into digital form in the data sampling module, and then processed in different ways, and finally generates usable information in the suggestion generation module. In the process from the data sampling module to the suggestion generation module, the data of the previous module is transmitted to the subsequent module, and the input and output exchange with the data of the external system is carried out. Similarly, when the data is converted into information, a standard technical display and graphical expression format is required. In many cases, maintaining the output history of each module can be accomplished through data archiving. The data acquisition module, the data operation module, and the condition detection module are responsible for evaluating the data quality, and the evaluation conclusion can be good, bad, or uncertain.
[0172] Fig.10 : is a system architecture diagram of the aviation big data analysis method according to an embodiment of the present application. Fig.10 As shown, taking aircraft maintenance as an example, the above method is implemented using aviation big data analysis tools. The system for implementing the above aviation big data analysis method may include the following parts:
[0173] 1) Aircraft configuration bus management system. The above bus management system includes the following main functions:
[0174] a) Record the physical structure and / or functional characteristics of the aircraft, such as systems, subsystems, functional groups, functional parts, components, parts, etc. Identify the aircraft configuration through accurate identifiers, use the aircraft configuration as a configuration element, and manage the aircraft configuration. And record aircraft configuration changes.
[0175] b) Considering the entire life cycle of the aircraft, corresponding models or configurations are established for the aircraft according to the stages of design and development, production and manufacturing, and online service. The aircraft configuration in a certain life cycle transfers complete data to the next life cycle in the configuration data management and business data bus, and the next life cycle continues to supplement and improve the configuration data, and finally outputs the serialized configuration of the aircraft, such as the delivery configuration, operation and maintenance configuration, etc.
[0176] c) Supports the definition of technical parameters for configuration elements in each stage of aircraft configuration, which is used to manage the measurement points associated with the aircraft body.
[0177] d) Use the aircraft configuration as a carrier to record technical information, fault records, parameters, maintenance records and other data throughout the aircraft life cycle.
[0178] 2) Aviation maintenance big data storage and management system. The above storage management system can provide time series flight data storage and management functions, and provide a data interface for the data exploration platform prototype system. The specific implementation of the above storage management system can be found in Figure 4 The relevant description of the illustrated embodiment will not be repeated here.
[0179] 3) QAR Data Exploration Platform. The QAR Data Exploration Platform mainly provides data processing and analysis tools for aircraft health management business personnel to support business personnel in discovering the internal associations, patterns and potential laws of aircraft data, so as to provide a basis for subsequent modeling. The main functions of the data exploration platform include data access, data processing, feature extraction, visualization and other functions. The visualization display method can meet the user's various analysis and display requirements such as browsing, comparative analysis, correlation analysis, trend analysis and feature extraction of the data imported into the data exploration platform, and supports data export.
[0180] 4) Model operation platform. The model operation platform provides an operating environment for the aircraft dynamic performance model. It supports model developers to manage, test, and operate the aircraft dynamic models they develop through the platform, as well as configure the resources required for model operation. The model operation results can be freely exported through interfaces or intermediate files, and also supports visual display methods such as charts. The model operation platform provides functions including model management, model testing, model operation, and resource management.
[0181] The advantages or beneficial effects of the above technical solutions include at least: being able to timely detect possible abnormal status and system failures in the target scenario, so as to reduce sudden system failures through preventive maintenance, reduce system operation safety risks, reduce operation and maintenance costs, and improve detection efficiency. In the application scenario of the civil aviation system, using the aviation big data analysis method of the embodiment of the present application to perform status detection on the aircraft can improve the aircraft availability and thus increase revenue.
[0182] Fig.11 FIG. 1 is a schematic diagram of the structure of an aviation big data analysis tool according to an embodiment of the present application. Fig.11 As shown, the tool may include:
[0183] A screening unit 100 is used to screen the historical data of the system to be detected to obtain system parameters in a target scenario;
[0184] The processing unit 160 is used to pre-process the system parameters in the target scene;
[0185] An extraction unit 180, configured to extract data feature values from the preprocessed system parameters in the target scenario;
[0186] An analysis unit 200 is used to establish a performance baseline of the system to be detected according to the data characteristic values;
[0187] The detection unit 400 is used to perform status detection on the system to be detected based on the performance baseline.
[0188] Fig.12 FIG. 1 is a schematic diagram of the structure of an aviation big data analysis tool according to an embodiment of the present application. Fig.12 As shown, in one embodiment, the system to be detected includes an aircraft; the above tool also includes a storage unit 500 for:
[0189] Construct a data storage framework corresponding to the aircraft configuration, which uses time series as the data storage form;
[0190] According to the aircraft configuration, the historical data of the system to be tested is stored in the corresponding data storage framework.
[0191] In one embodiment, the aircraft configuration includes at least one of an aircraft type enabling configuration, a stand-alone delivery configuration, and a stand-alone operational configuration.
[0192] In one embodiment, the preprocessing includes at least one of curve interpolation processing, filtering processing, curve smoothing processing, multi-parameter preprocessing and outlier processing.
[0193] In one embodiment, the system to be detected includes an aircraft; and the target scenario includes at least one of a specific time period, a specific flight segment, a specific flight phase, a specific parameter range, and a specific system type.
[0194] In one embodiment, the analysis unit 200 includes:
[0195] A first analysis subunit is used to analyze the data characteristic values to obtain the correlation between at least two system parameters;
[0196] The second analysis subunit is used to establish a performance baseline of the system to be detected according to the correlation between at least two system parameters.
[0197] In one embodiment, the second analysis subunit is used to: when the correlation between system parameters is linear, use a linear model or an artificial intelligence model to establish a performance baseline of the system to be tested.
[0198] In one embodiment, the second analysis subunit is used to: when the correlation between system parameters is nonlinear, use an artificial intelligence model to establish a performance baseline of the system to be tested.
[0199] In one embodiment, the artificial intelligence model includes at least one of a neural network model, a random forest model, and a support vector regression model.
[0200] In one embodiment, the second analysis subunit is used to establish a performance baseline of the system to be detected using a piecewise linear model when the correlation between system parameters is nonlinear.
[0201] In one embodiment, the detection unit 400 is used to:
[0202] Collect current data corresponding to system parameters of the system to be tested;
[0203] Compare current data with performance baselines;
[0204] Whether the state of the system to be detected is abnormal is determined according to the comparison result.
[0205] Taking aviation big data analysis and exploration as an example, the aviation big data analysis tool provided in the embodiment of the present application can provide the ability to recognize, preprocess, extract features, establish performance models, and verify data results for aircraft data. And it can support the establishment of a health management system performance model, output analysis results, and ensure the reliability and quality of health management. According to time periods, flight segments, flight phases, and specific parameter ranges, historical data is preliminarily screened to generate parameters for specific scenarios. Preprocess the parameters in specific scenarios to gain a preliminary understanding of the data characteristics. For a single parameter, extract at least one of the frequency domain characteristics, time domain characteristics, and time-frequency characteristics of the parameter. Results can be calculated for multiple parameter combinations, while supporting existing algorithms such as variance, time-frequency analysis, and frequency analysis. The proposed parameter features are visualized by correlation analysis and potential laws. Finally, parameter modeling is performed through algorithms such as neural networks, linear models, and random forests, and the quality of the final performance model is effectively verified.
[0206] The functions of the various tools and units in the embodiments of the present application can be found in the corresponding descriptions in the above methods and will not be repeated here.
[0207] Fig.13 is a block diagram of an electronic device used to implement an embodiment of the present application. Fig.13 As shown, the electronic device includes: a memory 910 and a processor 920, wherein the memory 910 stores a computer program that can be run on the processor 920. When the processor 920 executes the computer program, the aviation big data analysis method in the above embodiment is implemented. The number of the memory 910 and the processor 920 can be one or more.
[0208] The electronic device also includes:
[0209] The communication interface 930 is used to communicate with external devices and perform data exchange transmission.
[0210] If the memory 910, the processor 920 and the communication interface 930 are implemented independently, the memory 910, the processor 920 and the communication interface 930 can be connected to each other through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Fig.13 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0211] Optionally, in a specific implementation, if the memory 910, the processor 920 and the communication interface 930 are integrated on a chip, the memory 910, the processor 920 and the communication interface 930 can communicate with each other through an internal interface.
[0212] An embodiment of the present application provides a computer-readable storage medium storing a computer program, which implements the method provided in the embodiment of the present application when the program is executed by a processor.
[0213] An embodiment of the present application also provides a chip, which includes a processor for calling and executing instructions stored in the memory from the memory, so that a communication device equipped with the chip executes the method provided by the embodiment of the present application.
[0214] An embodiment of the present application also provides a chip, including: an input interface, an output interface, a processor and a memory, wherein the input interface, the output interface, the processor and the memory are connected via an internal connection path, and the processor is used to execute the code in the memory. When the code is executed, the processor is used to execute the method provided in the embodiment of the application.
[0215] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the advanced RISC machines (ARM) architecture.
[0216] Further, optionally, the above-mentioned memory may include a read-only memory and a random access memory, and may also include a non-volatile random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct rambus RAM (DR RAM).
[0217] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.
[0218] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.
[0219] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0220] Any process or method description in the flow chart or otherwise described herein can be understood to represent a module, fragment or portion of a code including one or more executable instructions for implementing the steps of a specific logical function or process. And the scope of the preferred embodiment of the present application includes other implementations, in which the functions may not be performed in the order shown or discussed, including in a substantially simultaneous manner or in a reverse order according to the functions involved.
[0221] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus, tool or device (such as a computer-based system, a system including a processor or other system that can fetch instructions from an instruction execution system, apparatus, tool or device and execute instructions), or used in combination with these instruction execution systems, apparatuses, tools or devices.
[0222] It should be understood that the various parts of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented with software or firmware stored in a memory and executed by a suitable instruction execution system. All or part of the steps of the above embodiment method can be completed by instructing the relevant hardware through a program, which can be stored in a computer-readable storage medium, and when the program is executed, it includes one of the steps of the method embodiment or a combination thereof.
[0223] In addition, each functional unit in each embodiment of the present application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the above-mentioned integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a disk or an optical disk, etc.
[0224] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A method for analyzing aviation big data, characterized in that: include: Filter the historical data of the system to be detected to obtain the system parameters under the target scenario; Preprocessing system parameters under the target scenario; Extracting data feature values from the preprocessed system parameters in the target scenario; Establishing a performance baseline of the system to be detected according to the data characteristic values, including analyzing the data characteristic values to obtain correlations between at least two of the system parameters; According to the correlation between at least two of the system parameters, a performance baseline of the system to be detected is established; wherein, when the correlation between the system parameters is linear, a linear model or an artificial intelligence model is used to establish the performance baseline of the system to be detected; and when the correlation between the system parameters is nonlinear, a piecewise linear model is used to establish the performance baseline of the system to be detected; Performing status detection on the system to be detected based on the performance baseline includes collecting current data corresponding to the system parameters of the system to be detected; comparing whether the current data is consistent with the performance baseline; and determining whether the system to be detected is abnormal in status according to the comparison result.
2. The method according to claim 1, characterized in that The system to be detected includes an aircraft; the method further includes: Constructing a data storage framework corresponding to the aircraft configuration, wherein the data storage framework uses a time series as a data storage form; According to the aircraft configuration, the historical data of the system to be detected is stored in a corresponding data storage framework.
3. The method according to claim 2, characterized in that The aircraft configuration includes at least one of an aircraft type installation configuration, a single-machine delivery configuration, and a single-machine operation configuration.
4. The method according to any one of claims 1 to 3, characterized in that The preprocessing includes at least one of curve interpolation processing, filtering processing, curve smoothing processing, multi-parameter preprocessing and outlier processing.
5. The method according to any one of claims 1 to 4, characterized in that The system to be detected includes an aircraft; the target scene includes at least one of a specific time period, a specific flight segment, a specific flight phase, a specific parameter range and a specific system type.
6. The method according to claim 1, characterized in that Establishing a performance baseline of the system to be detected according to the correlation between the at least two system parameters includes: In the case where the correlation between the system parameters is nonlinear, an artificial intelligence model is used to establish a performance baseline of the system to be tested.
7. The method according to claim 6, characterized in that The artificial intelligence model includes at least one of a neural network model, a random forest model and a support vector regression model.
8. An aviation big data analysis tool, characterized in that: include: A screening unit, used for screening the historical data of the system to be detected to obtain the system parameters under the target scenario; A processing unit, used for preprocessing system parameters in the target scenario; An extraction unit, used to extract data feature values from the preprocessed system parameters in the target scenario; An analysis unit, used for establishing a performance baseline of the system to be detected according to the data characteristic value; The analysis unit includes a first analysis subunit, which is used to analyze the data characteristic values to obtain the correlation between at least two of the system parameters; A second analysis subunit is used to establish a performance baseline of the system to be detected based on the correlation between at least two of the system parameters; wherein the second analysis subunit is used to: when the correlation between the system parameters is linear, use a linear model or an artificial intelligence model to establish the performance baseline of the system to be detected; when the correlation between the system parameters is nonlinear, use a piecewise linear model to establish the performance baseline of the system to be detected; A detection unit is used to perform status detection on the system to be detected based on the performance baseline, including collecting current data corresponding to the system parameters of the system to be detected; comparing whether the current data is consistent with the performance baseline; and determining whether the system to be detected is in an abnormal state based on the comparison result.
9. The tool according to claim 8, characterized in that The system to be tested includes an aircraft; the tool also includes a storage unit for: Constructing a data storage framework corresponding to the aircraft configuration, wherein the data storage framework uses a time series as a data storage form; According to the aircraft configuration, the historical data of the system to be detected is stored in a corresponding data storage framework.
10. The tool according to claim 9, characterized in that The aircraft configuration includes at least one of an aircraft type installation configuration, a single-machine delivery configuration, and a single-machine operation configuration.
11. The tool according to claim 10, characterized in that The preprocessing includes at least one of curve interpolation processing, filtering processing, curve smoothing processing, multi-parameter preprocessing and outlier processing.
12. The tool according to claim 11, characterized in that The system to be detected includes an aircraft; the target scene includes at least one of a specific time period, a specific flight segment, a specific flight phase, a specific parameter range and a specific system type.
13. The tool according to claim 8, characterized in that The second analysis subunit is used to establish a performance baseline of the system to be detected by using an artificial intelligence model when the correlation between the system parameters is nonlinear.
14. The tool according to claim 13, characterized in that The artificial intelligence model includes at least one of a neural network model, a random forest model and a support vector regression model.
15. An electronic device, characterized in that: include: The system comprises a processor and a memory, wherein the memory stores instructions, and the instructions are loaded and executed by the processor to implement the method according to any one of claims 1 to 7. 16 . A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method according to claim 1 is implemented.
Citation Information
Patent Citations
Unmanned aerial vehicle flight parameter data fault detection and recovery method based on multivariable regression
CN110705658A
Method and device for analyzing running state of spacecraft
CN112257755A
Deep learning-based aero-engine nondestructive testing method, device, equipment and storage medium
CN112581430A
Aero-engine fault diagnosis method and system and electronic equipment
CN112733880A
Exception handling method and device for sensor data, electronic equipment and medium
CN112862127A
