Safety management method and system based on aviation big data

Through the classification of aviation big data and multi-model scoring processing, combined with weight coefficient calculation, the problem of low accuracy of aircraft safety management results in the existing technology is solved, and more efficient safety management is achieved.

CN118134259BActive Publication Date: 2024-12-03河北临空集团有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410417919.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-09
Publication Date
2024-12-03
Estimated Expiration
2044-04-09

AI Technical Summary

Technical Problem

In the prior art, the data types of aircraft safety management are diverse and unbalanced, resulting in lack of precision in mining model processing, low accuracy of safety management results, and lack of targeted weight coefficient settings, which affects the accuracy of safety management.

Method used

Aviation big data is divided into numerical and text-based data, and different data mining models are input to score and process them. The weight coefficient is calculated through the third data mining model, and the aircraft's safety score is obtained comprehensively. Artificial neural network and cyclic nerves are used. Network model for data mining.

Benefits of technology

It improves the accuracy and pertinence of aircraft safety management and ensures the scientificity and accuracy of safety scores.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118134259B_ABST
    Figure CN118134259B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of safety management technology, and in particular to a safety management method and system based on aviation big data. The present invention first classifies the data types of the aviation big data into data type data and text type data, adopts an unpredictable data mining model for different data, adopts an artificial neural network model for data type data, and adopts a recurrent neural network model that is good at processing time series text information for text type data. Finally, the safety score of an aircraft to be managed is obtained based on the mining results of the two models, and safety management is achieved based on the safety score, which can effectively improve the accuracy of safety management. At the same time, when the safety score of the aircraft to be managed is obtained based on the mining results of the two models, the present invention inputs the data into a third data mining model, calculates and obtains the weight coefficient of the corresponding data, so that the setting of the weight coefficient is more scientific and more targeted, and the accuracy of safety management is further improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of safety management technology, and in particular to a safety management method and system based on aviation big data. Background Art

[0002] With the rapid development of emerging technologies such as big data, electronic information technology and cloud computing technology, aviation safety work faces many challenges, and negligence in any link may lead to safety accidents.

[0003] In the digital age, big data has begun to be and gradually applied in-depth in various businesses of airlines, and has played a positive role and influence. How to better develop and utilize the airlines' own big data resources, reasonably use production big data, and form a closed loop of risk control management while continuously optimizing and upgrading operational risk management, and enable the operational risk management team to grasp operational risks more quickly and accurately, and timely control and eliminate potential risks, is the direction pursued by all airlines on the research path of intelligent risk control and management.

[0004] In the prior art, there is a method of achieving aircraft safety management through data mining, but generally, all data of the aircraft to be managed are input into a mining model to obtain safety management results. However, there are a lot of aircraft safety-related data, and the data types are also different. When all the data are input into one model, due to the need to take into account the data type, the mining model needs to balance the impact of different data types on the model, and thus lacks in model processing accuracy, resulting in low accuracy of the mined safety management results; at the same time, when different mining models are used in the prior art to comprehensively obtain safety management conclusions, the safety management conclusions are obtained by setting weight coefficients for calculation, which is not very targeted and also leads to low accuracy of the mined safety management results. Summary of the invention

[0005] In view of this, the present invention proposes a safety management method and system based on aviation big data, which are used to improve the accuracy of aircraft safety management.

[0006] In order to achieve the above object, a safety management method based on aviation big data is provided, the method comprising the following steps:

[0007] S1: Obtain aviation big data of the aircraft to be managed;

[0008] S2: performing preprocessing operations on the aviation big data;

[0009] S3: dividing the aviation big data into numerical data and text data;

[0010] S4: inputting the numerical data into a first data mining model, thereby outputting a first score of the aviation big data;

[0011] S5: inputting the text data into a second data mining model, thereby outputting a second score of the aviation big data;

[0012] S6: Obtaining a safety score of the aircraft to be managed according to the first score and the second score;

[0013] The calculation formula of the safety score F of the aircraft to be managed is:

[0014]

[0015] In the formula, , is the weight coefficient, f1 is the first score value, and f2 is the second score value;

[0016] The method for determining the weight coefficient is as follows: the aviation big data of the aircraft to be managed, the first score value, and the second score value are input into the third data mining model, and the output of the third data mining model is and The value of

[0017] S7: Perform safety management on the aircraft to be managed according to the safety score of the aircraft to be managed.

[0018] Preferably, the aviation big data includes weather data of take-off and landing airports of the aircraft to be managed, aircraft failure and maintenance data, and pilot data.

[0019] Preferably, S2 is specifically:

[0020] S2.1: Process missing values ​​of the aviation big data;

[0021] S2.2: performing abnormal data processing on the aviation big data;

[0022] S2.3: Delete duplicate data from the aviation big data.

[0023] Preferably, the data cleaning includes missing value processing, abnormal data processing, and duplicate data deletion, and the missing value processing is to use the difference method to supplement the missing values;

[0024] The abnormal data processing is specifically as follows: abnormal data is identified and deleted by adopting the piecewise least square method, and then the deleted abnormal data is supplemented by the difference method.

[0025] Preferably, the first data mining model is an artificial neural network model;

[0026] The training of the artificial neural network model adopts the back propagation method, the activation function selects the nonlinear function Relu, the loss function selects Keras's Categorical crossentropy, the hyperparameters of the artificial neural network model are: the optimizer is Adam, the excitation function of each hidden layer is the Relu function, the learning rate is set to 0.003, the Batch size is set to 50, the Average is set to weighted, and the training process traverses all training samples several times according to the Batch size, and each round of traversal is called an Epochs.

[0027] Preferably, the training set for training the artificial neural network model includes numerical data collected from the historical aviation big data of the aircraft to be managed, and is manually labeled by experts to obtain a safety score; the input of the artificial intelligence model is the numerical data in the aviation big data of the aircraft to be managed; and the output of the artificial intelligence model is the safety score predicted by the model.

[0028] Preferably, S5 is specifically:

[0029] S5.1: Establish a second data mining model;

[0030] S5.2: Obtain a training set for training the second data mining model;

[0031] S5.3: Using the training set to train the second data mining model;

[0032] S5.4: Input the text data of the aircraft to be managed into a second data mining model, thereby outputting a second score of the aviation big data.

[0033] Preferably, the third data mining model is an artificial neural network model, and the training set of the third data mining model is the historical aviation big data of the aircraft to be managed, the first score obtained by S4, the second score obtained by S5, and the manually annotated safety score.

[0034] Preferably, the S7 specifically includes determining the safety management level of the aircraft to be managed according to a formula, and the formula is:

[0035] .

[0036] According to another aspect of the present invention, a safety management system based on aviation big data is provided, wherein the system adopts the above-mentioned safety management method based on aviation big data, wherein the system further comprises:

[0037] A big data acquisition module is used to acquire aviation big data of the aircraft to be managed;

[0038] A preprocessing module, connected to the big data acquisition module, for performing preprocessing operations on the aviation big data;

[0039] A data classification module, connected to the preprocessing module, for classifying the aviation big data into numerical data and text data;

[0040] A first score calculation module, connected to the data classification module, for inputting the numerical data into a first data mining model, thereby outputting a first score of the aviation big data;

[0041] A second score calculation module, connected to the first score calculation module, is used to input the text data into a second data mining model, thereby outputting a second score of the aviation big data;

[0042] a safety score calculation module, connected to the second score calculation module, and configured to obtain a safety score of the aircraft to be managed according to the first score and the second score;

[0043] A safety management module is connected to the safety score calculation module and is used to perform safety management on the aircraft to be managed according to the safety score of the aircraft to be managed.

[0044] The advantages and beneficial effects of the present invention are:

[0045] The present invention first classifies the data types of the aviation big data into data type data and text type data, adopts different data mining models for different data, adopts artificial neural network model for data type data, and adopts recurrent neural network model that is good at processing time series text information for text type data, and finally obtains the safety score of the aircraft to be managed according to the mining results of the two models, implements safety management according to the safety score, and can effectively improve the accuracy of safety management;

[0046] At the same time, when the safety score of the aircraft to be managed is obtained based on the mining results of the two models, the present invention inputs the data into the third data mining model to calculate the weight coefficient of the corresponding data, so that the setting of the weight coefficient is more scientific and more targeted, and the accuracy of safety management is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the description of the present invention or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0048] Figure 1A flow chart of a safety management method based on aviation big data provided by an embodiment of the present invention;

[0049] Figure 2 A flowchart of data preprocessing provided by an embodiment of the present invention;

[0050] Figure 3 A flowchart of inputting the text data into a second data mining model to output a second score of the aviation big data is provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Safety management is an important measure for airlines to ensure safety and no accidents. Safety management consists of four parts: risk identification, risk analysis, risk assessment and risk control. The safety management process of airlines is based on the actual operation and focus of the company. For different flight types and operation stages, the three aspects of people, aircraft and environment are analyzed for the aircraft, pilots, weather and other factors involved, and the identified danger sources are evaluated and assigned values. According to the size of the risk value, control measures are formulated for the aircraft to mitigate operational risks and achieve early warning and active control.

[0053] As attached Figure 1 As shown, a safety management method based on aviation big data comprises the following steps:

[0054] S1: Obtain aviation big data of the aircraft to be managed;

[0055] Among them, the aviation big data includes airport weather data, aircraft failure and maintenance data, pilot data and other information during the flight of the managed aircraft; the above data basically covers almost all aviation accident risk points. By comprehensively processing and analyzing the above data, the flight risk value can be evaluated to provide data support for safety management.

[0056] S2: performing preprocessing operations on the aviation big data;

[0057] When acquiring aviation big data, some data may be incomplete or invalid due to data collection errors or transmission errors. If this part of the data is not preprocessed, the final result of data mining using the data may be biased. Therefore, preprocessing the data can significantly improve the efficiency of the data mining algorithm and the accuracy of knowledge discovery.

[0058] Specifically, as attached Figure 2 As shown, S2 is specifically:

[0059] S2.1: Process missing values ​​of the aviation big data;

[0060] In this embodiment, the data preprocessing mainly includes missing value processing, abnormal data processing, duplicate data deletion, etc.;

[0061] Specifically, the missing value processing is to use the difference method to fill in the missing value;

[0062] S2.2: performing abnormal data processing on the aviation big data;

[0063] Generally, in aviation big data, for numerical data, there may be abnormal data, that is, "wild values" commonly seen in the field of data mining. In this embodiment, the abnormal data processing is specifically as follows: using the piecewise least squares method to identify and delete the abnormal data, and then using the difference method to supplement the deleted abnormal data;

[0064] S2.3: deduplicating the aviation big data;

[0065] Since the same parameter is collected repeatedly through different collection channels and the duplicate data is not deleted during aggregation, duplicate data exists in aviation big data. Therefore, it is necessary to delete the duplicate data to reduce the computing power occupied by invalid data during mining.

[0066] S3: dividing the aviation big data into numerical data and text data;

[0067] In fact, aviation big data includes a variety of data, including not only numerical data, such as airport temperature, humidity, pilot flight time, number of aircraft failures, etc., but also text data, such as airport wind direction, aircraft failure type, pilot mental state, etc.; when mining different types of data, due to the need to take into account the data type, the model parameters of the deep learning model need to balance the impact of different data types on the model, and thus the model processing accuracy is lacking. Therefore, in this embodiment, the data type of the aviation big data is first classified, and different models are used for data mining for different types of data, thereby realizing airline safety management;

[0068] S4: inputting the numerical data into a first data mining model, thereby outputting a first score of the aviation big data;

[0069] Wherein, the first data mining model is an artificial neural network model;

[0070] The training of the artificial neural network model adopts the back propagation method. The activation function uses the nonlinear function Relu, and the loss function uses Keras's Categorical crossentropy. In this embodiment, the hyperparameters of the artificial neural network model are: the optimizer is Adam, the excitation function of each hidden layer is the Relu function, the learning rate is set to 0.003, the batch size is set to 50, and the average is set to weighted. The training process traverses all training samples several times according to the batch size, and each round of traversal is called an epoch.

[0071] At the same time, the training set for training the artificial neural network model includes the numerical data in the historical aviation big data of the aircraft to be managed, which is manually annotated by experts to obtain a safety score; the input of the artificial intelligence model is the historical operation data of the aircraft to be managed, airport weather data, aircraft failure and maintenance data, and numerical data in pilot data; the output of the artificial intelligence model is the safety score predicted by the model;

[0072] S5: inputting the text data into a second data mining model, thereby outputting a second score of the aviation big data;

[0073] Specifically, if Figure 3 As shown, the S5 is specifically:

[0074] S5.1: Establish a second data mining model;

[0075] Among them, the text data contains more time series information, so the second data mining model is a recurrent neural network model, wherein the recurrent neural network model includes an input layer, a hidden layer and an output layer;

[0076] S5.2: Obtain a training set for training the second data mining model;

[0077] The training set includes text data from the historical aviation big data of the aircraft to be managed, which is manually annotated by experts to obtain a safety score;

[0078] S5.3: Using the training set to train the second data mining model;

[0079] Inputting the training sets into the second data mining model respectively, and judging whether the training of the second data mining model is completed by using a loss function;

[0080] S5.4: Inputting the text data of the aircraft to be managed into a second data mining model, thereby outputting a second score of the aviation big data;

[0081] S6: Obtaining a safety score of the aircraft to be managed according to the first score and the second score;

[0082] The calculation formula of the safety score F of the aircraft to be managed is:

[0083]

[0084] In the formula, , is the weight coefficient, f1 is the first score value, and f2 is the second score value;

[0085] Furthermore, the weight coefficient is determined by: inputting the aviation big data, the first score value, and the second score value of the aircraft to be managed into the third data mining model, and the output of the third data mining model is and The numerical value of .

[0086] Wherein, the third data mining model is an artificial neural network model; the parameter setting of the model is the same as that of the first data mining model;

[0087] It is worth emphasizing that the input of the third data mining model is different from that of the first data mining model, and therefore, the training set is also different; the training set of the third data mining model is the historical aviation big data of the aircraft to be managed and the first score, the second score and the manually annotated safety score obtained through S4 and S5;

[0088] Generally, the weight coefficients in the prior art are set manually based on experience or roughly based on the amount of data and the degree of importance. This setting method is relatively general and has a certain impact on the accuracy of various security scores. Therefore, compared with the prior art, this embodiment calculates the weight coefficients of the first score and the second score through a data mining model, so that different data obtain different weight coefficients. In other words, the weight coefficient contains the inherent knowledge of the big data, so that the security score finally obtained is more accurate.

[0089] S7: performing safety management on the aircraft to be managed according to the safety score of the aircraft to be managed;

[0090] Specifically, S7 is to determine the safety management level of the aircraft to be managed according to a formula, and the formula is:

[0091] .

[0092] Embodiment 2: This embodiment also includes a safety management system based on aviation big data, the system adopts the safety management method based on aviation big data of embodiment 1, wherein the system also includes:

[0093] A big data acquisition module is used to acquire aviation big data of the aircraft to be managed;

[0094] A preprocessing module, used for performing preprocessing operations on the aviation big data;

[0095] A data classification module, used for classifying the aviation big data into numerical data and text data;

[0096] A first score calculation module, used for inputting the numerical data into a first data mining model, thereby outputting a first score of the aviation big data;

[0097] A second score calculation module, used for inputting the text data into a second data mining model, thereby outputting a second score of the aviation big data;

[0098] A safety score calculation module, used for obtaining the safety score of the aircraft to be managed according to the first score and the second score;

[0099] The safety management module is used to perform safety management on the aircraft to be managed according to the safety score of the aircraft to be managed.

[0100] Embodiment 3: This embodiment includes a computer-readable storage medium on which a data processing program is stored. The data processing program is executed by a processor to implement the safety management method based on aviation big data of embodiment 1.

[0101] It should be understood by those skilled in the art that the embodiments of this document may provide methods, devices (equipment) or computer program products. Therefore, this document may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Including but not limited to RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassette, tape, disk storage or other magnetic storage device, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that communication media generally contain computer-readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier or other transmission mechanism, and may include any information transfer medium.

[0102] This document is described with reference to flowcharts and / or block diagrams of methods, apparatuses (devices), and computer program products according to the embodiments of this document. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0103] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0104] The embodiments and / or implementation methods described above are only used to illustrate the preferred embodiments and / or implementation methods of the technology of the present invention, and are not intended to impose any formal restrictions on the implementation methods of the technology of the present invention. Any person skilled in the art may make some changes or modifications to other equivalent embodiments without departing from the scope of the technical means disclosed in the content of the present invention, but they should still be regarded as the technology or embodiments that are essentially the same as the present invention. Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described here, and that various obvious changes, readjustments and substitutions can be made to those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention is described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may also include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the attached claims.

Claims

1. A safety management method based on aviation big data, It is characterized in that The method comprises the following steps: S1: Obtaining aviation big data of the aircraft to be managed; the aviation big data includes weather data of the take-off airport, weather data of the landing airport, aircraft failure and maintenance data, and pilot data of the aircraft to be managed; S2: Preprocessing the aviation big data; S2 specifically includes: S2.1: Process missing values ​​of the aviation big data; S2.2: performing abnormal data processing on the aviation big data; S2.3: deduplicating the aviation big data; S3: dividing the aviation big data into numerical data and text data; S4: Input the numerical data into the first data mining model, so as to output the first score of the aviation big data; the first data mining model is an artificial neural network model, the training of the artificial neural network model adopts the back propagation method, the activation function uses the nonlinear function Relu, the loss function uses Keras's Categoricalcrossentropy, and the hyperparameters of the artificial neural network model are: the optimizer is Adam, the excitation function of each hidden layer is the Relu function, the learning rate is set to 0.003, the Batch size is set to 50, the Average is set to weighted, and the training process traverses all training samples several times according to the Batch size; the training set of the artificial neural network model training includes the numerical data in the historical aviation big data of the aircraft to be managed, and is manually annotated by experts to obtain a safety score; the input of the artificial intelligence model is the numerical data in the aviation big data of the aircraft to be managed; the output of the artificial intelligence model is the safety score predicted by the model; S5: Input the text data into a second data mining model, thereby outputting a second score of the aviation big data; the second data mining model is a recurrent neural network model, wherein the recurrent neural network model includes an input layer, a hidden layer, and an output layer; The S5 is specifically: S5.1: Establish a second data mining model; S5.2: Obtain a training set for training the second data mining model; S5.3: Using the training set to train the second data mining model; S5.4: Inputting the text data of the aircraft to be managed into a second data mining model, thereby outputting a second score of the aviation big data; S6: Obtaining a safety score of the aircraft to be managed according to the first score and the second score; The calculation formula of the safety score F of the aircraft to be managed is: In the formula, , is the weight coefficient, f 1 is the first score value, f 2 is the second rating value; The method for determining the weight coefficient is as follows: the aviation big data of the aircraft to be managed, the first score value, and the second score value are input into the third data mining model, and the output of the third data mining model is and The value of S7: Perform safety management on the aircraft to be managed according to the safety score of the aircraft to be managed.

2. According to the safety management method based on aviation big data according to claim 1, It is characterized in that The missing value processing is to use the difference method to fill in the missing values; The abnormal data processing is specifically as follows: abnormal data is identified and deleted by adopting the piecewise least square method, and then the deleted abnormal data is supplemented by the difference method.

3. According to the safety management method based on aviation big data according to claim 1, It is characterized in that The third data mining model is an artificial neural network model, and the training set of the third data mining model is the historical aviation big data of the aircraft to be managed, the first score obtained by S4, the second score obtained by S5, and the manually labeled safety score.

4. The safety management method based on aviation big data according to claim 1, It is characterized in that Specifically, S7 is to determine the safety management level of the aircraft to be managed according to a formula, wherein the formula is: 。 5. A safety management system based on aviation big data, the system adopts the safety management method based on aviation big data according to any one of claims 1 to 4, the system include: A big data acquisition module is used to acquire aviation big data of the aircraft to be managed; A preprocessing module, connected to the big data acquisition module, for performing preprocessing operations on the aviation big data; A data classification module, connected to the preprocessing module, for classifying the aviation big data into numerical data and text data; A first score calculation module, connected to the data classification module, for inputting the numerical data into a first data mining model, thereby outputting a first score of the aviation big data; A second score calculation module, connected to the first score calculation module, is used to input the text data into a second data mining model, thereby outputting a second score of the aviation big data; a safety score calculation module, connected to the second score calculation module, and configured to obtain a safety score of the aircraft to be managed according to the first score and the second score; A safety management module is connected to the safety score calculation module and is used to perform safety management on the aircraft to be managed according to the safety score of the aircraft to be managed.