Data management system and method for aviation operation safety

Abstract

The invention discloses a data management system and method for aviation operation safety, and relates to the technical field of data management, and the system comprises a data management module, a data preprocessing module and an early warning model. The data management module stores the aviation operation safety data in a database in a classified manner, and performs data management on the aviation operation safety data stored in the database; the data management module further comprises an index established according to a data logic relationship in the aviation operation safety data; the data management comprises the step of retrieving data in a database by using an index; the data preprocessing module is used for performing abnormal value elimination, data supplementation and data smoothing on the aviation operation safety data, and storing the processed aviation operation safety data in a database; the processed aviation operation safety data comprises real-time online data; and the early warning model is used for identifying abnormal data in the real-time online data and performing early warning. The effectiveness of multi-source data management can be improved.

Description

Technical Field

[0001] The present application relates to the field of data management technology, and in particular to a data management system and method for aviation operation safety. Background Art

[0002] With the rapid development of the civil aviation industry, the risks of aviation operations are growing exponentially. With the expansion of the application of sensors on aircraft and the popularization of mobile terminals, aircraft have accumulated a large amount of data during operation, such as Quick Access Recorder (QAR) data, Aircraft Communications Addressing and Reporting System (ACARS) data, maintenance reports, reliability data, etc. The Civil Aviation Administration and other relevant functional departments have also generated a large amount of data related to aviation operation safety during production and operation. With the rapid development of data monitoring and data analysis technology, data has become an important asset to ensure the safety of aviation operations. At present, most data management systems or platforms can only realize data collection and storage, with relatively single functions, and cannot effectively manage diversified data. Summary of the invention

[0003] The purpose of this application is to provide a data management system and method for aviation operation safety, which can improve the effectiveness of multi-source data management.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a data management system for aviation operation safety, including: a data management module, a data preprocessing module and an early warning model;

[0006] The data management module is used to classify and store the aviation operation safety data in a database, and to perform data management on the aviation operation safety data stored in the database; the data management module also includes an index established according to the logical relationship of the data in the aviation operation safety data; the data management includes retrieving, inserting, modifying and deleting the data in the database, and the data management also includes retrieving the data in the database using the index;

[0007] The data preprocessing module is used to remove abnormal values, supplement data and smooth data of the aviation operation safety data, and store the processed aviation operation safety data in the database; the processed aviation operation safety data includes real-time online data;

[0008] The early warning model is used to identify abnormal data in the real-time online data and issue an early warning.

[0009] Optionally, the aviation operation safety data includes safety policy data, risk management data, safety assurance data, safety promotion data and system management data;

[0010] The safety policy data includes a variety of policy document data; the risk management data includes aircraft basic data, aircraft operation data, aircraft failure data, aircraft maintenance data, early warning model input data, QAR data and analysis result data; the safety assurance data includes mandatory reports, employee voluntary reports, safety inspection contents and incident investigation files; the safety promotion data includes learning material data, learning plan data and learning personnel data; the system management data includes user authority table, user role table, role authority correspondence table, system user table and user role correspondence table.

[0011] Optionally, the database includes multiple data tables, and the same data fields in different data tables are associated with each other through primary keys.

[0012] Optionally, the network hardware in the data management system for aviation operational safety includes two 10 Gigabit switches, a management switch and a router connected to each 10 Gigabit switch, and the management switch is used for cascading between the two 10 Gigabit switches; the network hardware is used to provide wired network access and wireless network access.

[0013] Optionally, the data management system for aviation operation safety includes a cloud platform, which is used to provide infrastructure, resource pool, application support, autonomous security system, standard system and disaster recovery backup system.

[0014] Optionally, the data preprocessing module includes a QAR data missing value inference unit, an outlier detection unit and a smoothing processing unit;

[0015] The QAR data missing value inference unit is used to determine the missing values ​​in the QAR data using regression analysis or Bayesian calculation formula;

[0016] The outlier detection unit is used to determine outliers in the aviation operation safety data by using a statistical correlation algorithm and a median filtering algorithm, and remove the corresponding outliers;

[0017] The smoothing processing unit is used to smooth the aviation operation safety data by adopting a method combining a sliding window algorithm and a bilateral median.

[0018] Optionally, the data preprocessing module further includes a sample selection unit;

[0019] The sample selection unit is used to:

[0020] Randomly selecting a set number of data samples from the target data of the real-time online data, and labeling them according to the set rules corresponding to the target data to obtain a first data set;

[0021] An oversampling technology algorithm for synthesizing a minority class is used to add new samples to the original data set to obtain a second data set; the second data set is used to train a machine learning model to obtain an early warning model corresponding to the target data.

[0022] Optionally, the warning model includes a landing warning model;

[0023] The target data used in the training of the landing warning model include flight altitude, longitude and latitude, engine high-pressure rotor, engine low-pressure rotor, oil pressure, exhaust temperature, fuel flow, indicated airspeed and normal acceleration of the aircraft main wheel when it touches the ground, and the normal acceleration of the aircraft main wheel when it touches the ground is a label; when the normal acceleration of the aircraft main wheel when it touches the ground is greater than 1.7, the label is 1, and when the normal acceleration of the aircraft main wheel when it touches the ground is less than or equal to 1.7, the label is 0, a label of 1 indicates an abnormal landing, and a label of 0 indicates a normal landing.

[0024] Optionally, the data management system for aviation operation safety further includes: a visualization module, and the visualization module is used to visualize the real-time online data in the aviation operation safety data.

[0025] In a second aspect, the present application provides a data management method for aviation operation safety, the data management method for aviation operation safety applies the above-mentioned data management system for aviation operation safety, and the data management method for aviation operation safety includes:

[0026] The data management module stores the aviation operation safety data in a database by category, and performs data management on the aviation operation safety data stored in the database; the data management module also includes an index established according to the logical relationship of the data in the aviation operation safety data; the data management includes retrieving, inserting, modifying and deleting the data in the database, and the data management also includes retrieving the data in the database by using the index;

[0027] The data preprocessing module removes abnormal values, supplements data and smoothes data on the aviation operation safety data, and stores the processed aviation operation safety data in the database; the processed aviation operation safety data includes real-time online data;

[0028] The early warning model identifies abnormal data in the real-time online data and issues an early warning.

[0029] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0030] The present application provides a data management system and method for aviation operation safety. The aviation operation safety data is classified and stored in a database. An index is established according to the logical relationship of the data in the aviation operation safety data, and the data in the database is retrieved using the index, thereby improving the retrieval efficiency. The data in the database can be retrieved, inserted, modified and deleted, thereby improving the effectiveness of multi-source data management. In addition, the present application uses an early warning model to warn of abnormal data in real-time online data, thereby achieving rapid risk warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0032] Figure 1 A schematic diagram of the structure of a data management system for aviation operation safety provided in one embodiment of the present application;

[0033] Figure 2 A schematic diagram of a multi-source heterogeneous data structure provided in an embodiment of the present application;

[0034] Figure 3 A schematic diagram of basic functions of a data management system provided in one embodiment of the present application;

[0035] Figure 4 A schematic diagram of a data management system network architecture provided in one embodiment of the present application;

[0036] Figure 5 A schematic diagram of the overall architecture of a data management system provided in one embodiment of the present application;

[0037] Figure 6 A schematic diagram of a bilateral median and sliding window combination algorithm provided in one embodiment of the present application. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0039] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0040] This application provides a data management system for aviation operation safety, such as Figure 1 As shown, the data management system for aviation operation safety includes: a data management module, a data preprocessing module and an early warning model.

[0041] The data management module is used to classify and store the aviation operation safety data in a database, and to perform data management on the aviation operation safety data stored in the database; the data management module also includes an index established according to the logical relationship of the data in the aviation operation safety data; the data management includes retrieving, inserting, modifying and deleting the data in the database, and the data management also includes retrieving the data in the database using the index.

[0042] The data preprocessing module is used to remove abnormal values, supplement data and smooth data of the aviation operation safety data, and store the processed aviation operation safety data in the database; the processed aviation operation safety data includes real-time online data.

[0043] The early warning model is used to identify abnormal data in the real-time online data and issue an early warning.

[0044] The present application improves the retrieval efficiency by storing aviation operation safety data in a database according to classification, and uses the index established according to the logical relationship of the data in the aviation operation safety data to retrieve the data in the database. The application can also retrieve, insert, modify and delete the data in the database, thereby improving the effectiveness of multi-source data management. In addition, the present application uses an early warning model to warn of abnormal data in real-time online data, thereby achieving rapid risk warning.

[0045] In terms of classifying and storing aviation operation safety data in the database and managing the aviation operation safety data stored in the database, the data management module is specifically used to: design database table structures for different data sources according to the different characteristics of the data sources and establish corresponding indexes according to the logical relationship of the data, and design the basic functional module of the data management system.

[0046] Data sources include aircraft manufacturers, aircraft maintenance units, airlines, and portals.

[0047] Since there are many aviation operation data sources, it is necessary to design the database table structure for each data type. The design of the database table structure is to achieve fast data storage, which includes primary key settings, data type settings, etc., so as to form a corresponding database table structure for different types of data. As an example data table, see Table 1.

[0048] Table 1 A sample data table

[0049]

[0050]

[0051] The basic functions of data management are database creation, data organization and storage management, data manipulation functions, database transaction management and operation management, database maintenance, communication, such as Figure 3 As shown. The basic functions of the data management system are as follows 1-5.

[0052] 1: Data organization, storage and management. Data sources such as aircraft operation data, flight reliability data and aircraft maintenance data are classified, stored and managed according to offline data and real-time data, and a data dictionary is established and the corresponding storage path is set.

[0053] 2: Data manipulation function. Provide users with data operation functions to realize the retrieval, insertion, modification and deletion of database data, and ensure the realization of database update function.

[0054] 3: Database transaction management and operation management. Mainly includes transaction management functions and security integrity control functions in a multi-user environment; database recovery, concurrency control and deadlock detection, security check and access control, integrity check and execution, and organization and management of operation logs.

[0055] 4: Database maintenance. This includes data conversion, database dump and recovery, database reorganization and reconstruction, and performance monitoring and analysis.

[0056] 5: Communication function. It has the relevant interfaces with the online processing of the operating system, time-sharing system and remote operation input, and is responsible for processing data transmission.

[0057] The aviation operation safety data includes safety policy data, risk management data, safety assurance data, safety promotion data and system management data, such as Figure 2 shown.

[0058] The safety policy data includes a variety of policy document data; the risk management data includes aircraft basic data, aircraft operation data, aircraft failure data, aircraft maintenance data, early warning model input data, QAR data and analysis result data; the safety assurance data includes mandatory reports, employee voluntary reports, safety inspection content and incident investigation files; the safety promotion data includes learning material data, learning plan data and learning personnel data; the system management data includes user authority table, user role table, role authority correspondence table, system user table and user role correspondence table. The analysis result data includes early warning information, where the early warning information includes information such as the location of the fault, the severity of the fault, and the severity of the risk.

[0059] Mandatory reporting refers to all incident information that occurs within the airline and that is required to be reported by individuals or organizations within the prescribed time limit in accordance with international, national or company regulations. Such incident information is usually closely related to key areas such as flight safety and operational safety.

[0060] The contents of mandatory reports usually include but are not limited to the following categories:

[0061] 1) Safety incident information: unsafe incidents involving flight safety, such as aircraft failures, mid-air conflicts, etc.

[0062] 2) Dangerous goods incident information: unsafe incidents related to the transportation of dangerous goods, such as leakage of dangerous goods, incorrect loading, etc.

[0063] 3) Security incident information: unsafe incidents involving aviation security, such as illegal interference, terrorist threats, etc.

[0064] 4) Medical event information: medical events involving crew or passenger health issues, such as infectious disease outbreaks, emergency medical conditions, etc.

[0065] 5) Maintenance event information: unsafe events related to aircraft maintenance, such as maintenance errors, component failures, etc.

[0066] Voluntary employee reporting refers to any unsafe incidents, potential hazards and threats, and safety suggestions related to production and operation safety that occur or exist within the company and are reported proactively by flight crews, maintenance personnel, dispatch control personnel, cabin and ground support personnel, and other personnel. This reporting method encourages employees to actively participate in safety management and proactively report possible safety issues.

[0067] The report content usually includes but is not limited to the following categories:

[0068] 1) Deficiencies in rules, regulations, standards and work procedures: Possible problems or deficiencies in rules, regulations, standards and work procedures discovered by employees during work.

[0069] 2) Work environment issues: such as safety hazards in the workplace, defects in equipment and facilities, etc.

[0070] 3) Personnel training issues: employees’ understanding and mastery of training content and feedback on training effectiveness.

[0071] 4) Safety suggestions: Employees’ suggestions and opinions on improving safety management and raising safety levels.

[0072] Safety policy data refers specifically to policy document data stored by category. Safety policy data mainly includes: safety policy data on laws, regulations and rules issued by civil aviation management departments, safety policy data formulated by civil aviation enterprises, safety policy data formulated by airport units, etc.

[0073] In an exemplary embodiment, the aircraft manufacturer has basic data of the aircraft, such as aircraft series, aircraft manufacturing date, engine model, engine serial number and Accelerated Processing Unit (APU) model, etc.; the aircraft maintenance unit has fault data during the operation of the aircraft, such as fault part number, fault name, fault cause, fault occurrence stage, fault result, etc.; the airline has QAR data of aircraft operation, which contains data information such as the attitude, position, weather, etc. of the aircraft during operation. Data is stored in the form of characters, numerical values, logical values, etc., and different data sources establish associations by setting primary keys.

[0074] Data organization, storage and management in the basic functions of the data management system refer to the classification, storage and management of data sources such as aircraft operation data, flight reliability data and aircraft maintenance data according to offline data and real-time data. Among them, aircraft operation data is real-time data, and aircraft reliability data and maintenance data are offline data. Establish a data dictionary and set the corresponding storage path. The storage path in the data dictionary refers to the storage path of each type of data, such as the storage path of aircraft operation data. The data manipulation function refers to the retrieval, insertion, modification and deletion of database data to ensure the implementation of the database update function. Database transaction management and operation management include transaction management functions and security integrity control functions in a multi-user environment; database and maintenance include data conversion, database dump and recovery, database reorganization and reconstruction, and performance monitoring and analysis. It has relevant interfaces with the operating system's online processing, time-sharing system and remote job input, and is responsible for processing data transmission.

[0075] The database includes multiple data tables, and the same data fields in different data tables are associated with each other through primary keys.

[0076] The basic environment of the data management system for aviation operation safety includes network hardware, operating system, database and application server.

[0077] The network hardware includes two 10G switches, a management switch, and routers connected to the 10G switches. The data exchange channel between the system server (application server) and the client computer is completed through the configuration of the network hardware. Figure 4 shown.

[0078] The client computer is used to realize human-computer interaction for data management and provide a data management interface.

[0079] The management switch is used for cascading between two 10G switches; the 10G switch interconnects any nodes of the access switch to form a data path. The network hardware is used to provide wired network access and wireless network access.

[0080] The operating system uses Windows 10 environment, which enables the client to access the data management system. The client includes personal computer (PC) client and mobile client (Android system).

[0081] This application uses mature database software such as Oracle and MySQL to manage various data resources in the system. Tomcat application server based on Apache has advanced technology, stable performance and supports Servlet and JSP specifications. Choosing Tomcat as the application server can provide documents to Web clients such as browsers and manage data access services between clients and the platform.

[0082] This application deploys a cloud computing architecture, which builds the basic environment of the data management system and provides hardware environment support for the comprehensive management of multi-source heterogeneous data, such as Figure 5 shown.

[0083] The data management system for aviation operation safety includes a cloud platform, which is used to provide infrastructure, resource pool, application support, independent security system, standard system, disaster recovery system and comprehensive service capabilities. Infrastructure is a complex and dynamic environment, mainly composed of hardware resources, software resource service types, core components and mechanisms.

[0084] Hardware resources include servers, storage devices, network devices, and security devices.

[0085] The server is a core device used to provide computing power, including key components such as CPU, memory, and hard disk, which are used to execute user computing tasks.

[0086] Storage devices are used to store data and files, including hard disks, solid-state drives (SSDs), etc., providing safe, high-speed and convenient storage services.

[0087] Network devices include routers, switches, load balancers, etc., which are used to build the network architecture of the cloud platform and support connections and data transmission between different servers and applications.

[0088] Security equipment: such as firewalls, intrusion detection systems, etc., used to protect the cloud platform from network threats and unauthorized access.

[0089] Software resources include virtualization software, cloud management platform, and storage software.

[0090] Virtualization software: abstracts physical resources (such as servers and storage devices) into virtual resources, such as virtual machines and virtual storage, to achieve flexible allocation and management of resources.

[0091] Cloud management platform: It manages and dispatches computing resources, storage resources, network resources and security resources in the resource pool in a unified manner and provides services to users. Cloud management platforms usually have functions such as automated management, resource monitoring and fault recovery.

[0092] Storage software: software that provides data management and storage services, such as distributed file systems and object storage systems, to support efficient storage and access of data. The basic environment is composed of infrastructure, resource pools, application support, and various standard systems.

[0093] The main functions of the cloud platform are as follows.

[0094] 1) Data migration / virtual machine online migration: Based on the technology of continuous data protection (CDP), by real-time monitoring of changes in each block on the production environment disk, capture, read, write, copy and replay are completed, so as to achieve online migration of the original system and data while the business is running. Use the online migration tools vCenter and Hyper-VManager provided by the virtualization platform to achieve online migration of virtual machines from one host to another without interruption, or migrate virtual machine data between different storages.

[0095] 2) Automatic load balancing: The virtual machine automatically selects a host to run according to the host load, and can automatically migrate according to the policy as the host load changes, thereby achieving automatic load balancing. Specifically, the virtualization platform monitors the resource usage of all virtual machines in real time, and uses polling, least connection, weighted polling, source IP hashing and other algorithms to dynamically adjust the resource allocation of virtual machines according to the monitored load. The cluster management function of the vSphere Distributed Resource Scheduler (DRS) virtualization platform is used to automatically detect the health status and resource usage of each node in the cluster, and automatically trigger load balancing operations according to preset rules.

[0096] 3) Virtual machine snapshot: Supports virtual machine rolling snapshots, and uses snapshot rollback to defend against ransomware viruses such as WannaCry / Globelmposter.

[0097] 4) Graphics Processing Unit (GPU) virtualization: Supports multiple graphics card virtualization technologies such as GPU penetration, NvidiaGrid vGPU, AMD Multiuser GPU, and Intel Iris Pro Virtual GPU sharing (GVT-g). The parameters of the application server and GPU server are shown in Table 2.

[0098] Table 2 Parameters of application server and GPU service

[0099]

[0100] The servers in the first row in Table 2 are application servers.

[0101] 5) Platform migration: Supports migration from physical machines to this platform, supports migration from VMware vSphere to this platform; supports migration from this platform to physical machines, VMware, and public cloud platforms such as Microsoft Azure, Amazon AWS, and Alibaba Cloud.

[0102] 6) Multi-user management: Implement multi-user management with multiple roles such as admin and non-admin, and support mobile client management; conduct real-time monitoring and visual management of system CPU, memory, and storage space resource usage, and automatically generate operation reports, saving key operation data for a retrospective period of 72 hours.

[0103] 7) Supports node storage resource pool management, integrating all disks in the node into a unified storage pool, and evenly distributes data on all disks.

[0104] 8) Supports dual data protection mechanisms of intra-node redundant arrays of independent disks (RAID) protection and inter-node data block replication.

[0105] 9) It has an online data integrity verification mechanism, that is, each data block is equipped with a 256-bit data verification code; when the data is read out, it must match the verification code before it is delivered to the front-end application; otherwise, the data self-repair mechanism is activated to ensure the security of the front-end application data.

[0106] 10) It has online data compression function and supports LZ4, LZJB, GZIP1-9 compression encoding.

[0107] 11) Supports both horizontal and vertical expansion modes, which means increasing the system IOPS performance and storage capacity by increasing the number of nodes (horizontal expansion), and expanding capacity by adding JBOD expansion cabinets and disks within the node (vertical expansion).

[0108] 12) Supports read cache tiering mechanism. The hottest data, including metadata, is automatically allocated in DRAM, and the second hottest data is allocated in the solid-state drive. Common cache replacement strategies are used to distinguish the hottest data from the second hottest data, such as Least Recently Used (LRU) and Least Frequently Used (LFU).

[0109] 13) Automatically generate daily operation reports, including CPU, memory, disk, IOPS, network, usage and latency, operation and maintenance curves, etc.; and automatically send them to relevant personnel via email.

[0110] It should be noted that in the network resource pool, in order to solve the unified management and scheduling of data resources, the network resource pool data center module makes full use of the computing resource pool resources to build an Nginx-based application load balancer cluster, which saves bandwidth and improves access speed through HTTP compression, and accelerates WEB applications through optimization technologies such as static file and dynamic file cache, file compression, and browser-side file cache control, thereby improving user access speed.

[0111] The data preprocessing module realizes data supplementation, outlier removal, data smoothing and data sampling of preset types of data in aviation operation data. The data preprocessing module realizes the rapid processing of the entire process of initial aviation operation safety data.

[0112] The data preprocessing module includes a QAR data missing value inference unit, an outlier detection unit and a smoothing processing unit.

[0113] The QAR data missing value inference unit is used to determine the missing values ​​in the QAR data using regression analysis or Bayesian calculation formula. That is, the present application uses regression analysis, Bayesian calculation formula and decision tree to infer the maximum possible value of the missing value in view of the fact that the QAR data set has the characteristic of rapid changes.

[0114] Among them, for missing value interpolation using regression analysis, first select a prediction model based on the variable type (continuous or categorical) where the missing value is located and its relationship with other variables, such as linear regression, polynomial regression or logistic regression. Build a regression model based on complete data (observations without missing values), select other variables that are highly correlated with the variable where the missing value is located as predictors, use the established regression model to substitute other known variable values ​​corresponding to the missing value observations into the model, and calculate the predicted value of the missing variable. For missing value interpolation using the Bayesian calculation formula, first set the prior distribution of the missing variable based on domain knowledge or historical data, apply Bayes' theorem, combine the prior probability and the likelihood obtained through evidence, and calculate the posterior distribution of the missing value. Select the expected value or mode from the posterior distribution as the estimate of the missing value.

[0115] The outlier detection unit is used to determine the outliers in the aviation operation safety data by using statistical correlation algorithms and median filtering algorithms, and remove the corresponding outliers. That is, the present application uses statistical correlation algorithms and median filtering algorithms to process noise. In the statistical method, the corresponding data model is established by fitting according to the actual value of the data, and the 3σ criterion is used to detect singular points (outliers). In data mining, association rules and clustering algorithms are used to detect noise.

[0116] The smoothing processing unit is used to smooth the aviation operation safety data by using a method combining a sliding window algorithm with a bilateral median. Since both the data observed manually and the data acquired by the data acquisition system are inevitably superimposed with "noise" interference (reflected in the curve graph as some "burrs and spikes"), in order to improve the quality of the data, this application uses a method combining a sliding window algorithm with a bilateral median to smooth the data (remove noise interference). The specific implementation principle is as follows: Figure 6 As shown in the figure, in the time series, the data with close timestamps have a large correlation, while the data with far distances have a small correlation. A window is used to select a range with large data correlation, and the window is scanned from front to back along the entire data stream to calculate the degree to which each point deviates from the sliding window average. Based on the distance, it is determined whether the point is an outlier. τ is the threshold, m(a) is the mean of the data segment, and y(t) is the value of point a. If |y t -m(a)|<τ, then point a is judged as a normal value, otherwise it is an abnormal value.

[0117] The data preprocessing module also includes a sample selection unit. The present application considers that QAR data is time series data, and uses sampling methods to reduce its data. A small random sample (subset) can be used to represent a large data set. Assuming that the massive data D contains N objects, it can be sampled using methods such as overall sampling, stratified sampling, multi-stage sampling, and systematic sampling. Usually, the data changes greatly during the take-off and landing phases of an aircraft, and the data changes more steadily during the cruising phase. Sampling at different intervals in different phases can ensure the quality of its data reduction.

[0118] The data preprocessing module is also used to formulate preprocessing technical specifications based on QAR data, including data labeling (labeling abnormal segments and normal segments), sample imbalance processing during training, and processing of different sampling frequency parameters.

[0119] The sample selection unit is used to: randomly select a set number of data samples from the target data of the real-time online data, and label them according to the set rules corresponding to the target data to obtain a first data set; use an oversampling technology algorithm for synthesizing a minority class to add new samples to the original data set to obtain a second data set; the second data set is used to train a machine learning model to obtain an early warning model corresponding to the target data.

[0120] The warning model includes a landing warning model.

[0121] In an exemplary embodiment, the A320 aircraft heavy landing data preprocessing technical specifications are used as an example to illustrate that the target data used in the landing warning model training includes flight altitude, longitude and latitude, engine high-pressure rotor (N1), engine low-pressure rotor (N2), oil pressure, exhaust temperature, fuel flow, indicated airspeed and normal acceleration of the aircraft main wheel when it touches the ground (VRTG), and the normal acceleration of the aircraft main wheel when it touches the ground is a label; when the normal acceleration of the aircraft main wheel when it touches the ground is greater than 1.7, the label is 1, and when the normal acceleration of the aircraft main wheel when it touches the ground is less than or equal to 1.7, the label is 0, a label of 1 indicates an abnormal landing, and a label of 0 indicates a normal landing.

[0122] The oversampling technology algorithm for synthesizing minority classes is used to analyze minority class samples and artificially synthesize new samples based on the minority class samples and add them to the first data set to obtain the second data set, which is used as the input for subsequent early warning model training, specifically including:

[0123] 1) For each sample x in the minority class, calculate its distance to all samples in the minority class sample set using the Euclidean distance as the standard to obtain its k nearest neighbors.

[0124] 2) According to the sample imbalance ratio, a sampling ratio is set to determine the sampling rate N. For each minority class sample x, several samples are randomly selected from its k nearest neighbors. Assume that the selected nearest neighbors are x n .

[0125] This application uses the method of compressing frequency to solve the problem of different variables in the data. The flight parameters that affect landing safety are compressed by averaging. VRTG parameters (8HZ) are compressed by maximizing.

[0126] The real-time transmission of the simulator in this application mainly includes real-time collection of flight altitude, longitude and latitude, engine N1 and engine N2, oil pressure, exhaust temperature, fuel flow, indicated airspeed and other parameters, and uses the set early warning model to achieve real-time online data analysis (predicting the normal acceleration of the aircraft's main wheels when they touch the ground), and ultimately realize the flight safety early warning function.

[0127] The data management system for aviation operation safety also includes: a visualization module, which is used to visualize the real-time online data in the aviation operation safety data. That is, the data management system for aviation operation safety of the present application also includes data analysis and visualization function software, which displays aviation operation monitoring and early warning data in real time in a graphical manner, realizes rapid data collection, screening, analysis and induction, and can realize adaptive update according to the newly added data.

[0128] The data analysis and visualization software mainly includes data analysis management, visualization component calling, and plan display page design functions. It effectively integrates and sorts out all event information elements such as real-time event data, various data indicators, situation analysis, classification statistics, etc. in the form of visualization icons, platters, dynamic heat maps, etc., and performs comprehensive visualization and application on an ultra-high-resolution large screen. The visualization display includes fault location, email alarms, operation and maintenance reports, high-speed buffer (Cache) hit rate monitoring, capacity monitoring, performance monitoring, width monitoring, self-modification status monitoring, and CPU and disk temperature, which flexibly realizes information sharing and work collaboration.

[0129] The data management system for aviation operation safety in this application can realize the intelligent management of risk management data, safety policy data, safety assurance data, safety promotion data and system management data, and has the functions of storage, maintenance, preprocessing, data feature analysis and comprehensive data management of multi-source heterogeneous data. By building cloud computing servers and database servers, the database management functions (such as system configuration and management, data access and update management), database query and manipulation functions (such as data resource retrieval and modification, etc.), database maintenance functions (such as data import and export management, etc.), and database parallel operation functions (supporting multi-user simultaneous access, etc.) are realized; data preprocessing technology is used to realize the missing value supplement, anomaly monitoring, normalization, dynamic change feature analysis, data fusion and other functions of multi-source heterogeneous data. The technical specifications for aviation operation safety data preprocessing based on QAR data have been formulated to achieve the unification of data throughout the entire life cycle, and to provide the possibility for rapid risk warning of aviation operations through standardized data interfaces.

[0130] Based on the same inventive concept, the embodiment of the present application also provides a data management method for aviation operation safety for realizing the data management system for aviation operation safety involved above. The implementation scheme for solving the problem provided by the method is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the data management method for aviation operation safety provided below can refer to the limitations of the data management system for aviation operation safety above, and will not be repeated here.

[0131] In an exemplary embodiment, a data management method for aviation operation safety is provided. The data management method for aviation operation safety applies the data management system for aviation operation safety. The data management method for aviation operation safety includes:

[0132] Step 101: The data management module classifies and stores the aviation operation safety data in a database, and performs data management on the aviation operation safety data stored in the database; the data management module also includes an index established according to the logical relationship of the data in the aviation operation safety data; the data management includes retrieving, inserting, modifying and deleting the data in the database, and the data management also includes retrieving the data in the database using the index.

[0133] Step 102: The data preprocessing module removes abnormal values, supplements data, and smoothes data on the aviation operation safety data, and stores the processed aviation operation safety data in the database; the processed aviation operation safety data includes real-time online data.

[0134] Step 103: The early warning model identifies abnormal data in the real-time online data and issues an early warning.

[0135] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0136] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A data management system for aviation operation safety, characterized in that: The data management system for aviation operation safety includes: a data management module, a data preprocessing module and an early warning model; The data management module is used to classify and store the aviation operation safety data in a database, and to perform data management on the aviation operation safety data stored in the database; the data management module also includes an index established according to the logical relationship of the data in the aviation operation safety data; the data management includes retrieving, inserting, modifying and deleting the data in the database, and the data management also includes retrieving the data in the database using the index; The data preprocessing module is used to remove abnormal values, supplement data and smooth data of the aviation operation safety data, and store the processed aviation operation safety data in the database; the processed aviation operation safety data includes real-time online data; The early warning model is used to identify abnormal data in the real-time online data and issue an early warning.

2. The data management system for aviation operation safety according to claim 1, characterized in that: The aviation operation safety data includes safety policy data, risk management data, safety assurance data, safety promotion data and system management data; The safety policy data includes a variety of policy document data; the risk management data includes aircraft basic data, aircraft operation data, aircraft failure data, aircraft maintenance data, early warning model input data, QAR data and analysis result data; the safety assurance data includes mandatory reports, employee voluntary reports, safety inspection contents and incident investigation files; the safety promotion data includes learning material data, learning plan data and learning personnel data; the system management data includes user authority table, user role table, role authority correspondence table, system user table and user role correspondence table.

3. The data management system for aviation operation safety according to claim 1, characterized in that: The database includes multiple data tables, and the same data fields in different data tables are associated with each other through primary keys.

4. The data management system for aviation operation safety according to claim 1, characterized in that: The network hardware in the data management system for aviation operation safety includes two 10 Gigabit switches, a management switch and a router connected to each 10 Gigabit switch. The management switch is used for cascading between the two 10 Gigabit switches; the network hardware is used to provide wired network access and wireless network access.

5. The data management system for aviation operation safety according to claim 1, characterized in that: The data management system for aviation operation safety includes a cloud platform, which is used to provide infrastructure, resource pool, application support, independent security system, standard system and disaster recovery backup system.

6. The data management system for aviation operation safety according to claim 1, characterized in that: The data preprocessing module includes a QAR data missing value inference unit, an outlier detection unit and a smoothing processing unit; The QAR data missing value inference unit is used to determine the missing values ​​in the QAR data using regression analysis or Bayesian calculation formula; The outlier detection unit is used to determine outliers in the aviation operation safety data by using a statistical correlation algorithm and a median filtering algorithm, and remove the corresponding outliers; The smoothing processing unit is used to smooth the aviation operation safety data by adopting a method combining a sliding window algorithm and a bilateral median.

7. The data management system for aviation operation safety according to claim 1, characterized in that: The data preprocessing module also includes a sample selection unit; The sample selection unit is used to: Randomly selecting a set number of data samples from the target data of the real-time online data, and labeling them according to the set rules corresponding to the target data to obtain a first data set; An oversampling technology algorithm for synthesizing a minority class is used to add new samples to the original data set to obtain a second data set; the second data set is used to train a machine learning model to obtain an early warning model corresponding to the target data.

8. The data management system for aviation operation safety according to claim 7, characterized in that: The early warning model includes a landing early warning model; The target data used in the training of the landing warning model include flight altitude, longitude and latitude, engine high-pressure rotor, engine low-pressure rotor, oil pressure, exhaust temperature, fuel flow, indicated airspeed and normal acceleration of the aircraft main wheel when it touches the ground, and the normal acceleration of the aircraft main wheel when it touches the ground is a label; When the normal acceleration of the aircraft's main wheel when touching the ground is greater than 1.7, the label is 1; when the normal acceleration of the aircraft's main wheel when touching the ground is less than or equal to 1.7, the label is 0; a label of 1 indicates an abnormal landing, and a label of 0 indicates a normal landing.

9. The data management system for aviation operation safety according to claim 1, characterized in that: The data management system for aviation operation safety also includes: a visualization module, which is used to visualize the real-time online data in the aviation operation safety data.

10. A data management method for aviation operation safety, characterized in that: The data management method for aviation operation safety applies the data management system for aviation operation safety according to any one of claims 1 to 9, and the data management method for aviation operation safety includes: The data management module stores the aviation operation safety data in a database by category, and performs data management on the aviation operation safety data stored in the database; the data management module also includes an index established according to the logical relationship of the data in the aviation operation safety data; the data management includes retrieving, inserting, modifying and deleting the data in the database, and the data management also includes retrieving the data in the database by using the index; The data preprocessing module removes abnormal values, supplements data and smoothes data on the aviation operation safety data, and stores the processed aviation operation safety data in the database; the processed aviation operation safety data includes real-time online data; The early warning model identifies abnormal data in the real-time online data and issues an early warning.

Images