Long-term and short-term memory neural network based flight delay grading early warning method

Abstract

The invention discloses a long-term and short-term memory neural network based flight delay grading early warning method. The method includes analyzing an aviation meteorological message so that aviation meteorological data required by flight delay prediction can be obtained; performing multi-source data fusion to form an initial flight delay data set; converting non-numerical data into numericaldata by using semantic transformation, performing grading prediction on delay characteristics, and performing discrete partition on type characteristics and weather characteristics; performing data cleaning, missing value complementing and normalized processing to form a flight delay grading prediction standard data set, and performing partition; training long-term and short-term memory neural network based flight delay grading prediction models in batches on a training set; obtaining a long-term and short-term memory neural network model having the optimal hyperparameter on a verification set; performing verification on the performance of the optimal flight delay grading prediction model on a test set; and determining a delay early warning grade according to obtained flight delay grades through prediction. The method can effectively enhance the accuracy and reliability of flight delay early warning.

Description

technical field [0001] The invention belongs to the technical field of early warning methods for flight delays at airports, in particular to a hierarchical early warning method for flight delays based on long-short-term memory neural networks. Background technique [0002] With the sustained, rapid and healthy development of the national economy, the demand for air transport is also increasing. However, in recent years, the phenomenon of large-scale flight delays has become increasingly prominent, and has become a worldwide problem that plagues civil aviation departments and passengers. Vicious incidents such as passengers refusing to board the plane, bullying the plane, attacking the airport, and beating staff due to flight delays are common occurrences, which have damaged the image of civil aviation's high-quality service and seriously affected the safe operation order of the airport. In order to reduce the delays caused by the airlines themselves, especially the improper...

AI Technical Summary

Problems solved by technology

However, there are many reasons for flight delays, the main reason is insufficient capacity of the airport and airspace, other reasons, such as weather, airport scheduling, company plans, passengers, luggage, etc. may also cause flight delays

In addition, flight delays also have a chain reaction problem: when a flight is delayed, if the schedule is tight, it will affect the punctual arrival or departure of the next flight, which will indirectly affect more downstream flights and airports

[0005] The analysis of the actual flight delay data collected shows that these data contain a large amount of impurity data, and without reasonable and standardized preprocessing will directly affect the prediction accuracy of the prediction model

The formation of dirty data may be the loss of various data records or storage processes in reality and the distortion caused by wrong records or storage. If such data is directly used to train the model, the trained model parameters will be wrong. Misleading data, so that the predicted value after each data input has a relatively large change and the prediction accuracy is not high. Therefore, it is very necessary to perform strict preprocessing on the real-time running data, which can make the data set Applicable to various intelligent prediction algorithms

[0006] 2. Existing flight delay prediction models ignore the timing of flight delays

[0007] When traditional machine learning models process data, each input corresponds to an output, and each change of weight and bias item is based on the influence of each feature in a single piece of data, so these models cannot be used on the basis of Consider the interaction between multiple pieces of data

Images