Method for guaranteeing whole-process task early warning model based on airplane

Abstract

The invention relates to the technical field of aviation guarantee, and discloses a method based on an aircraft guarantee whole-process task early warning model, and the method specifically comprises the steps: comprehensively collecting and preprocessing multi-source data, such as tasks, resources and historical data; deeply analyzing a task logic relationship, and modeling and quantifying a dependency relationship by using a directed acyclic graph; determining early warning indexes such as task progress deviation, resource utilization rate and task risk level; a machine learning algorithm is adopted to construct an early warning model, and an ensemble learning method is used for fusion to improve performance; setting an early warning threshold value, dynamically adjusting according to task types and priorities, triggering early warning when the threshold value is exceeded, and deploying emergency resources when high-grade early warning is triggered. According to the method based on the aircraft guarantee whole-process task early warning model, the guarantee task can be early warned in advance, the execution efficiency and quality are improved, safe operation of an aircraft is ensured, and the method further has the functions of data integrity verification, model parameter self-adaptive adjustment, multi-channel early warning notification, whole-process data tracing and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of aviation support, and specifically to a method based on an early warning model for the entire process of aircraft support tasks. Background Art

[0002] Aviation support is a series of support activities to ensure the safe and efficient operation of aircraft.

[0003] In the existing aviation support system, aircraft maintenance is a key link. Maintenance personnel conduct regular inspections, maintenance, and fault repairs on the fuselage, engines, avionics equipment, etc. of the aircraft according to strict maintenance manuals and procedures. For example, in engine maintenance, advanced detection instruments are used to monitor the performance parameters of the engine to ensure its normal operation.

[0004] Aviation material supply support is essential. The support department needs to establish a large aviation material inventory to ensure that suitable spare parts can be provided in a timely manner for replacement when components of the aircraft are damaged or malfunction. At the same time, there are strict requirements for the storage and management of aviation materials to ensure their quality and performance.

[0005] Ground support services include operations such as aircraft towing, refueling, cargo and passenger loading and unloading. The refueling vehicle will refuel the aircraft with standard fuel within the specified time, and the ground crew will methodically complete the luggage loading and unloading and passenger guidance work to ensure the punctuality of the flight.

[0006] Flight support also involves meteorological services. Meteorological personnel analyze meteorological data to provide accurate route weather forecasts for flight crews, helping the crew make reasonable flight decisions, avoid bad weather areas, and ensure flight safety. These links cooperate with each other to jointly form the existing aviation support system.

[0007] In aircraft support operations, there are numerous support tasks and complex processes, involving the collaborative work of multiple links and departments. Currently, the management and early warning of aircraft support tasks mostly rely on manual experience and simple schedule arrangements, lacking scientific and systematic early warning models and methods. This traditional method is prone to problems such as task delays, unreasonable resource allocation, and safety hazards. Therefore, it is of great practical significance to develop a model and method that can accurately predict the entire process of aircraft support tasks. Summary of the Invention

[0009] (I) Technical Problems to be Solved In view of the deficiencies of the prior art, the present invention provides a method based on an early warning model for the entire process of aircraft support, which has the advantages of realizing early warning of aircraft support tasks by establishing a scientific model and process, improving the execution efficiency and quality of support tasks, and ensuring the safe operation of aircraft. It solves the problems that the management and early warning of aircraft support tasks mostly rely on manual experience and simple schedule arrangements, lacking scientific and systematic early warning models and methods. This traditional method easily leads to task delays, unreasonable resource allocation, and potential safety hazards.

[0010] (2) Technical Solution To achieve the above objectives of realizing early warning of aircraft support tasks by establishing a scientific model and process, improving the execution efficiency and quality of support tasks, and ensuring the safe operation of aircraft, the present invention provides the following technical solutions: A method based on an early warning model for the entire process of aircraft support, comprising the following steps: Data collection and preprocessing: Comprehensively collect data related to the entire process of aircraft support, including but not limited to task information, resource information, historical task execution data, aircraft status data, aircraft maintenance records, and flight schedule information, etc., and clean and preprocess the collected data. Among them, numerical data is subjected to standardization and normalization processing, and text data is subjected to word segmentation, stemming, and vectorization processing; Task relationship analysis and modeling: Deeply analyze the logical relationships between tasks in the entire process of aircraft support, represent the task relationships using a directed acyclic graph, and quantitatively analyze the dependencies between tasks. Calculate the dependency strength index based on factors such as the sequence of tasks and the influence degree of pre-tasks on post-tasks; Construction of an early warning index system: Determine early warning indicators such as task progress deviation, resource utilization rate, and task risk level. Among them, the calculation formula for task progress deviation is: Among them, represents the progress deviation of task , represents the actual progress of task , represents the planned progress of task .

[0011] The calculation formula for resource utilization rate is: Among them, represents the utilization rate of resource , represents the used amount of resource , represents the total amount of resource .

[0012] The task risk level assessment comprehensively considers the task importance, complexity, historical execution situation, and task execution environment factors such as weather conditions and airport busyness, and calculates by establishing an analytic hierarchy structure model to determine the weights of various factors. Early warning model construction: Based on the task relationship model and the early warning index system, a machine learning algorithm is used to construct an early warning model. Then, an ensemble learning method is used to fuse multiple single models to improve stability and prediction accuracy. The ensemble learning methods used include but are not limited to Bagging, Boosting, and random forest ensemble, etc. And during the training process, data augmentation techniques are used to expand the training set data, and more training samples are generated through operations such as rotation, scaling, and adding noise. Early warning triggering and handling: Set an early warning threshold. When the task early warning index exceeds the threshold, an early warning is triggered. Different early warning thresholds are set for different types of tasks and are dynamically adjusted according to the task priority and the key degree to the guarantee process. Different handling measures are taken according to the early warning level, including establishing an emergency resource reserve mechanism. When a high-level early warning is triggered, emergency resources such as spare equipment, urgently deployed manpower, and special materials are preferentially allocated.

[0013] Preferably, in the data collection and preprocessing step, the data sources are diverse, data is obtained through multiple system interfaces, and the data cleaning process strictly ensures data quality, removing various data anomalies that may affect the model accuracy.

[0014] Preferably, in the task relationship analysis and modeling step, a directed acyclic graph accurately presents the task process and dependency relationship, and the dependency strength index can accurately quantify the closeness between tasks, providing an accurate relationship basis for the subsequent early warning model.

[0015] Preferably, in the early warning index system construction step, the determined early warning indicators comprehensively cover the key performance areas of aircraft guarantee tasks, the calculation formulas are scientific and reasonable, and the task risk level assessment fully considers the actual operating environment to ensure the accuracy and comprehensiveness of risk assessment.

[0016] Preferably, in the early warning model construction step, the selected machine learning algorithm and ensemble learning method are adapted to the data characteristics of aircraft guarantee tasks, and the data augmentation technology effectively improves the adaptability and generalization ability of the model to different task scenarios.

[0017] Preferably, in the early warning triggering and handling step, the early warning threshold is set flexibly and reasonably, which can accurately capture the abnormal trend of the task, and the emergency resource reserve mechanism ensures the smooth progress of the guarantee task in high-risk situations.

[0018] Preferably, the method has a data integrity verification mechanism during the data collection process to ensure that the collected data is not lost or damaged during transmission and storage, thus guaranteeing the reliability of the data.

[0019] Preferably, during the construction of the early warning model, the machine learning algorithms and ensemble learning methods adopted have the function of adaptive adjustment of model parameters, and can automatically optimize the model parameters according to the characteristics of the input data and the actual early warning effect.

[0020] Preferably, in the early warning trigger and processing steps, the early warning notification has the function of multi-channel sending. In addition to text messages and emails, it also includes in-site messages and other methods to ensure that relevant personnel can receive the early warning information in a timely manner.

[0021] Preferably, the method has the function of full-process data traceability, and can record and trace every operation and data change from data collection to early warning processing, which is convenient for fault troubleshooting and model optimization.

[0022] (III) Beneficial effects Compared with the prior art, the present invention provides a method based on an early warning model for the full-process tasks of aircraft support, and has the following beneficial effects: 1. The method based on the early warning model for the full-process tasks of aircraft support, through systematic and comprehensive data collection and preprocessing, as well as scientific and reasonable task relationship analysis and modeling, can accurately identify the complex relationships and potential risks between aircraft support tasks, greatly improving the accuracy and timeliness of early warning, effectively avoiding task delays, and ensuring the normal operation of flights.

[0023] 2. The method based on the early warning model for the full-process tasks of aircraft support, by constructing an early warning model with the help of a scientific early warning index system and advanced machine learning algorithms, and optimizing the model performance using ensemble learning and data augmentation techniques, enables the model to adapt to diverse support scenarios and task requirements, has strong versatility and adaptability, and helps to allocate resources reasonably and improve support efficiency.

[0024] 3. The method based on the early warning model for the full-process tasks of aircraft support, through a flexible early warning trigger mechanism and multi-channel early warning notification function, combined with an emergency resource reserve mechanism, ensures that relevant personnel can respond quickly and take measures in a timely manner when facing risks, effectively reducing potential safety hazards and ensuring the safe operation of aircraft. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is the architecture diagram of the early warning model for the full-process tasks of aircraft support as a whole of the present invention; Figure 2 is the data processing flow chart as a whole of the present invention DETAILED DESCRIPTION OF THE INVENTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] Please refer to Figure 1-2 , a method for an aircraft support full - process task early warning model of the present invention includes the following steps: (I) Data collection and pre - processing Use a specially developed data acquisition interface program to obtain comprehensive aircraft support - related data from multiple data sources such as the aircraft support task management system, resource management system, aircraft status monitoring system, aircraft maintenance record system, and flight schedule system. During the data transmission process, implement a data integrity verification mechanism, such as using algorithms like cyclic redundancy check (CRC), to ensure that the data is not lost or damaged during transmission and storage, and to guarantee the reliability of the data.

[0028] Use a data cleaning tool (such as the Pandas library in Python) to deeply clean the collected data, remove duplicate records, fill in missing values, and identify and correct outliers. For numerical data, use the standardization and normalization functions in the Scikit - learn library for processing; for text data, use a natural language processing library (such as NLTK) for word segmentation and stemming operations, and then use a word vector model (such as Word2Vec or GloVe) to convert the text into a vector representation so that the subsequent model can process text information.

[0029] (II) Task relationship analysis and modeling Organize experienced aircraft support task experts and professional process analysts to carefully sort out and confirm the logical relationships between tasks together to ensure the accuracy and integrity of the task relationships.

[0030] Use a professional graph database (such as Neo4j) to store and manage the task relationship model, and utilize its powerful graph - processing capabilities to efficiently query and analyze task relationships. By writing customized algorithm programs, calculate the dependence strength index based on factors such as the sequence of tasks and the influence degree of pre - tasks on post - tasks, and store it in the graph database to provide an accurate task relationship data basis for the early warning model.

[0031] (III) Early warning index system construction According to the characteristics of aircraft support tasks and the actual operation requirements, in-depth discussions with industry experts are carried out to determine the specific calculation methods and parameters of early warning indicators such as task progress deviation, resource utilization rate, and task risk level. For example, in the calculation of task progress deviation, clarify the acquisition methods and time nodes of actual progress and planned progress; in the calculation of resource utilization rate, accurately define the statistical scope of resource usage and total amount.

[0032] Regularly collect actual task execution data, use statistical analysis methods to evaluate the weight changes of task risk level assessment factors, and update the weight parameters in the analytic hierarchy process model in a timely manner to adapt to the dynamic changes in the actual operation environment of aircraft support tasks and ensure the accuracy and timeliness of risk assessment. When calculating the task risk level, use a professional analytic hierarchy process software (such as yaahp) to assist in determining the weights of various factors to ensure the scientificity and standardization of the assessment process.

[0033] (4) Early warning model construction Based on the characteristics of aircraft support task data and early warning requirements, select the algorithm with the best performance from various machine learning algorithms (such as random forest algorithm), and use the Scikit - learn library in Python to construct the model. During the model construction process, write an adaptive parameter adjustment function for the selected machine learning algorithm and ensemble learning methods (such as Bagging, Boosting, etc.). This function can automatically optimize the model parameters according to the feature distribution of the input data and the actual early warning effect of the model on the validation set to improve the model performance.

[0034] During the training process, use data augmentation techniques to expand the training set data. Different data augmentation strategies are adopted for different types of data. For example, for numerical data, new samples can be generated through random perturbation, interpolation, etc.; for text data, methods such as synonym replacement and sentence random shuffling can be used to increase data diversity. Use the ImageDataGenerator in the Keras library (if image - related data processing is involved) or custom functions to implement data augmentation operations to ensure that the model has good generalization ability.

[0035] (5) Early warning triggering and handling Set an initial early warning threshold in the early warning system. The threshold setting is based on historical data statistical analysis and industry experience standards. At the same time, develop a dynamic threshold adjustment algorithm to adjust the early warning threshold in real time according to the task type, priority, and key degree to the aircraft support process. For example, set a lower threshold for tasks related to key flight systems to ensure that potential risks can be detected in a timely manner; appropriately relax the threshold for non - key tasks to avoid excessive false alarms.

[0036] When the task warning index exceeds the threshold, the system automatically triggers a warning. The warning notification is sent through multiple channels. In addition to regular SMS and email notifications, it is also integrated with the internal communication system of the airport to achieve in-station message pushing, ensuring that relevant personnel can receive warning information in a timely manner. At the same time, an emergency resource management system is established to monitor and manage standby equipment, manpower, and special materials in real time. When a high-level warning is triggered, the emergency resource management system quickly allocates resources according to the preset allocation rules to ensure the smooth progress of the task. Moreover, the system regularly collects actual task execution data and compares it with the warning results for analysis, and uses data analysis tools (such as Tableau) to generate detailed reports, providing data support for model parameter adjustment and warning index system optimization, ensuring that the warning system can continuously adapt to the dynamic changes of aircraft support tasks and continuously improve the accuracy and effectiveness of warnings.

[0037] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method based on the aircraft support full-process mission warning model, characterized in that: The following steps are involved: Data collection and preprocessing: Comprehensively collect data related to the entire process of aircraft support, including but not limited to mission information, resource information, historical mission execution data, aircraft status data, aircraft maintenance records, and flight scheduling information, and clean and preprocess the collected data, including standardization and normalization of numerical data, and word segmentation, stem extraction, and vectorization of text data; Task relationship analysis and modeling: In-depth analysis of the logical relationship between tasks in the entire aircraft support process, using directed acyclic graphs to represent task relationships, and quantitative analysis of the dependencies between tasks, calculating dependency intensity indicators based on factors such as the order of tasks and the degree of influence of predecessor tasks on successor tasks; Construction of early warning indicator system: Determine early warning indicators such as task progress deviation, resource utilization rate, and task risk level. The calculation formula for task progress deviation is: The resource utilization calculation formula is: The mission risk level assessment comprehensively considers the mission importance, complexity, historical execution status, and mission execution environment factors such as weather conditions and airport busyness, and is calculated by establishing a hierarchical analysis structure model to determine the weight of each factor; Early warning model construction: Based on the task relationship model and early warning indicator system, the early warning model is constructed using machine learning algorithms. Then, multiple single models are integrated using ensemble learning methods to improve stability and prediction accuracy. The ensemble learning methods used include but are not limited to bagging, boosting, and random forest integration. In addition, data enhancement technology is used to expand the training set data during the training process, and more training samples are generated through operations such as rotation, scaling, and adding noise. Warning triggering and processing: Set warning thresholds. When the task warning indicators exceed the thresholds, the warning is triggered. Different warning thresholds are set for different types of tasks, and dynamically adjusted according to the task priority and the criticality of the guarantee process. Different processing measures are taken according to the warning level, including the establishment of an emergency resource reserve mechanism. When a high-level warning is triggered, emergency resources such as spare equipment, emergency manpower and special materials are allocated first.

2. The method according to claim 1, characterized in that: In the data collection and preprocessing steps, the data sources are diverse, and data is obtained through multiple system interfaces. The data cleaning process strictly ensures data quality and removes various data anomalies that may affect the accuracy of the model.

3. The method according to claim 1, characterized in that: In the task relationship analysis and modeling steps, the directed acyclic graph accurately presents the task process and dependency relationship, and the dependency strength index can accurately quantify the closeness between tasks, providing an accurate relationship basis for the subsequent early warning model.

4. The method according to claim 1, characterized in that: In the step of constructing the early warning indicator system, the determined early warning indicators comprehensively cover the key performance areas of the aircraft support mission, the calculation formula is scientific and reasonable, and the mission risk level assessment fully considers the actual operating environment to ensure the accuracy and comprehensiveness of the risk assessment.

5. The method according to claim 1, characterized in that: In the early warning model construction step, the selected machine learning algorithm and integrated learning method are adapted to the characteristics of aircraft support mission data, and the data enhancement technology effectively improves the adaptability and generalization ability of the model to different mission scenarios.

6. The method according to claim 1, characterized in that: In the warning triggering and processing steps, the warning threshold is set flexibly and reasonably, which can accurately capture abnormal trends in tasks, and the emergency resource reserve mechanism ensures the smooth progress of tasks under high-risk situations.

7. The method according to claim 1, characterized in that: This method has a data integrity verification mechanism during the data collection process to ensure that the collected data is not lost or damaged during transmission and storage, thereby ensuring the reliability of the data.

8. The method according to claim 1, characterized in that: During the construction of the early warning model, the machine learning algorithm and integrated learning method used have the function of adaptive adjustment of model parameters, and can automatically optimize the model parameters according to the characteristics of the input data and the actual early warning effect.

9. The method according to claim 1, characterized in that: In the early warning triggering and processing steps, the early warning notification has the function of sending through multiple channels. In addition to SMS and email, it also includes in-site messages to ensure that relevant personnel receive early warning information in a timely manner.

10. The method based on the aircraft support full process task warning model according to claim 1 is characterized in that: This method has the function of full-process data traceability, which can record and trace every step of operation and data changes from data collection to early warning processing, facilitating troubleshooting and model optimization.

Images