WO2024114250A1 - 一种基于XGBoost的航班延误原因预判方法 - Google Patents

一种基于XGBoost的航班延误原因预判方法 Download PDF

Info

Publication number
WO2024114250A1
WO2024114250A1 PCT/CN2023/128387 CN2023128387W WO2024114250A1 WO 2024114250 A1 WO2024114250 A1 WO 2024114250A1 CN 2023128387 W CN2023128387 W CN 2023128387W WO 2024114250 A1 WO2024114250 A1 WO 2024114250A1
Authority
WO
WIPO (PCT)
Prior art keywords
flight
data
flight delay
model
categories
Prior art date
Application number
PCT/CN2023/128387
Other languages
English (en)
French (fr)
Inventor
李印凤
王秋玲
许亚男
阮昌
傅子涛
米雪玉
高志远
张�浩
Original Assignee
华北理工大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华北理工大学 filed Critical 华北理工大学
Publication of WO2024114250A1 publication Critical patent/WO2024114250A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/40—Business processes related to the transportation industry
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Definitions

  • China's civil aviation transport industry has entered a stage of rapid development. With the rapid growth of the number of flights, the regularity of flights is facing unprecedented challenges under the condition of limited overall support resources of various civil aviation units. China's civil aviation has significantly improved the regularity of flights by actively tapping potential and increasing technical investment to address the problem of flight delays.
  • the flight operation process involves multiple entities such as air traffic control, airlines, and airports, and the overall operation process is complex, with many links and large external disturbances, it is difficult to identify the responsible party and predict the cause of flight delays, which has a great impact on the management of flight regularity. Accurately predicting the cause of flight delays plays an important role in improving the level of flight management.
  • the purpose of the present invention is to overcome the shortcomings of the prior art and provide a method for predicting the cause of flight delays based on XGBoost.
  • a method for predicting flight delay causes based on XGBoost, comprising the following steps:
  • S1 Collect flight operation data, which includes flight plan data, flow control data and meteorological data;
  • S2 classify the flight delay reasons according to the flow control data, and associate the classified flight delay reasons with the flight plan data and weather data to obtain the associated sample data;
  • step S3 According to the categories of flight delay reasons, the number of delayed flights associated with each type of flight delay reason is counted, and the sample data obtained in step S2 is balanced using the ADASYN algorithm to obtain balanced sample data;
  • step S4 Construct a flight delay cause prediction model based on XGBoost, use the characteristic factors of the flight plan and the meteorological characteristic factors related to the flight in the balanced sample data obtained in step S3 as the input characteristic variables of the model, and use the flight delay cause category as the target variable of the model prediction output; and divide the balanced sample data in step S3 into a training set and a validation set, train and validate the flight delay cause prediction model, and obtain a trained flight delay cause prediction model;
  • step S5 The characteristic factors of the flight plan of the flight to be predicted and the meteorological characteristic factors related to the flight are used as the input of the model and input into the flight delay cause prediction model trained in step S4.
  • the target variable output by the model is the category of flight delay cause.
  • the XGBoost algorithm of the present invention has a good effect on predicting the causes of multi-classification delays and has a strong identification ability.
  • a relatively accurate prediction can be made, which can effectively improve the early warning and management capabilities of flight delays.
  • FIG1 is a schematic diagram showing the overall architecture of the flight delay cause prediction method based on XGBoost of the present invention.
  • a method for predicting flight delay causes based on XGBoost, as shown in FIG1 includes the following steps:
  • Flight operation data includes flight plan data, flow control data and meteorological data.
  • the flight plan data includes: flight number, registration number, aircraft model, flight date, departure airport, landing airport, planned take-off time, actual take-off time, planned landing time, actual landing time
  • the flow control data includes: flight number, registration number, flow control reason, flow control content and other information
  • the weather data includes: observation time, weather type, visibility, cloud base height, wind direction, wind speed and other information, among which the weather types include thunderstorms, rain, snowfall and fog.
  • S2 Classify the reasons for flight delays according to the flow control data, and associate the classified categories of flight delay reasons with the flight plan data and weather data of the flight to obtain the associated sample data.
  • the flight delay reasons are divided into five categories.
  • the flight delay reasons are divided into five categories: weather reasons, airline reasons, air traffic control reasons, airport reasons and other reasons, and these five categories of flight delay reasons are associated with the flight plan data and weather data of the flight. That is, the flight plan data information of the flight can be associated with the divided flight delay reason categories according to the flight number and registration number, and the flight plan data of the flight can be associated with the corresponding weather data according to the time information, thereby realizing the association of the flight delay reason categories with the flight plan data and weather data of the flight.
  • step S3 Analyze the number of flight delays according to the categories of flight delay reasons. Since the number of flight delays corresponding to each category of flight delay reasons may be very different, the sample data is unbalanced, which affects the training effect of the subsequent prediction model.
  • the adaptive synthetic algorithm (AdaptiveSynthetic) performs data balancing processing on the sample data obtained in step S2 to obtain balanced sample data.
  • the ADASYN algorithm is used to deal with the data imbalance problem, that is, different minority class samples are given corresponding weights, and then different numbers of samples are obtained to make the model data relatively balanced.
  • the specific steps include:
  • G is the number of delayed flights associated with each type of flight delay reason
  • m l is the number of data in the larger number category
  • m s is the number of data in the smaller number category
  • ⁇ is a random number in [0,1].
  • ri is the proportion of the larger number of categories in the neighbors
  • xi is the i-th sample data in the smaller number of category data
  • xi has K neighbors.
  • the number of new data is calculated as follows:
  • gi is the number of synthetic samples that need to be generated for each sample xi in the minority class.
  • S3.5 Calculate the number of smaller number categories of data generated one by one based on the number of new data to be generated for the smaller number categories.
  • the method of generating new data is as follows:
  • step S4 Construct a flight delay cause prediction model based on XGBoost, use the flight plan feature factors and flight-related meteorological feature factors in the balanced sample data in step S3 as the model's input feature variables, and use the flight delay cause category as the model's predicted output target variable.
  • the balanced sample data in step S3 is divided into a training set and a validation set to train and validate the flight delay cause prediction model.
  • the characteristic factors of a flight plan include: flight number, aircraft model, flight date, and planned take-off time; the flight-related meteorological characteristic factors include: visibility, cloud base height, wind direction, wind speed, thunderstorms, rainfall, snowfall, and fog.
  • the method of building a flight delay cause prediction model based on XGBoost is as follows:
  • Obj (t) is the objective function of the model
  • t is the t-th tree
  • N is the number of samples
  • the error with the true value y n , ⁇ ( ft ) is the complexity function of the tree.
  • ⁇ is the complexity parameter
  • T is the number of leaf nodes in the tree
  • ⁇ is a fixed coefficient
  • wj is the score of the jth leaf node.
  • f t (x n ) is the prediction result obtained by the tth tree
  • C is a constant
  • f(x) is a first-order function and ⁇ x is the change in the variable.
  • Ij is the sample set on node j
  • dn is the gradient value of each flight delay sample
  • hn is the second-order derivative of each flight delay sample.
  • the training set and validation set are put into the model for classification prediction and to determine the cause of the delay.
  • the prediction effect of the model is verified using precision, recall rate and F1-score value. Finally, a flight delay cause prediction model with accuracy that meets the requirements is obtained.
  • step S5 The characteristic factors of the flight plan of the flight to be predicted and the meteorological characteristic factors related to the flight are used as the input of the model and input into the flight delay cause prediction model trained in step S4.
  • the target variable output by the model is the category of flight delay cause.

Landscapes

  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明公开了一种基于XGBoost的航班延误原因预判方法，涉及空中交通管理技术领域，所述方法根据流控数据划分出航班延误原因类别，将各类航班延误原因类别与航班的飞行计划数据和天气数据相关联，并对样本数据进行数据平衡处理；然后构建基于XGBoost的航班延误原因预判模型，将平衡后的样本数据中的航班飞行计划的特征因素以及和航班相关的气象特征因素作为模型的输入特征变量，将航班延误原因类别作为模型预测输出的目标变量。最后将待预测的航班的飞行计划的特征因素以及和该航班相关的气象特征因素输入至训练好的航班延误原因预判模型中，模型输出航班延误原因类别。本发明能够有效提高航班延误的预警和管理能力。

Description

一种基于XGBoost的航班延误原因预判方法
本申请要求于2022年11月29日提交中国专利局、申请号为202211509271.0、发明名称为“一种基于XGBoost的航班延误原因预判方法”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及空中交通管理技术领域，特别是涉及一种基于XGBoost的航班延误原因预判方法。
背景技术
目前中国民用航空运输产业进入了飞速发展阶段，随着航班量飞跃式增长，在民航各单位总体保障资源有限的情况下，航班正常性面临着前所未有的挑战。中国民航通过积极挖掘潜力、加大技术投入等手段治理航班延误问题，使航班正常性有了明显的提升。但由于航班运行过程涉及空管、航司、机场多个主体，且整体运行流程复杂、环节多、外界扰动大，导致航班延误难以明确责任主体及预判航班延误原因，对航班正常性管理工作造成了很大的影响。准确预判航班延误原因对提升航班管理水平具有重要作用。
发明内容
本发明的目的在于克服现有技术的不足，提供一种基于XGBoost的航班延误原因预判方法。
本发明是通过以下技术方案实现的：
一种基于XGBoost的航班延误原因预判方法，包括以下步骤：
S1：采集航班运行数据，航班运行数据包括飞行计划数据、流控数据和气象数据；
S2：根据流控数据划分出航班延误原因类别，将划分的各类航班延误原因类别与航班的飞行计划数据和天气数据相关联，得到关联后的样本数据；
S3：根据划分的航班延误原因类别，统计各类航班延误原因关联延误航班的数量，采用ADASYN算法对步骤S2得到的样本数据进行数据平衡处理，得到平衡后的样本数据；
S4：构建基于XGBoost的航班延误原因预判模型，将步骤S3得到的平衡后的样本数据中的航班飞行计划的特征因素以及和航班相关的气象特征因素作为模型的输入特征变量，将航班延误原因类别作为模型预测输出的目标变量；并将步骤S3中平衡后的样本数据分为训练集和验证集，对航班延误原因预判模型进行训练和验证，得到训练好的航班延误原因预判模型；
S5：将待预测的航班的飞行计划的特征因素以及和该航班相关的气象特征因素作为模型的输入，输入至步骤S4训练好的航班延误原因预判模型中，模型输出的目标变量则为航班延误原因类别。
本发明的优点和有益效果为：
本发明的XGBoost算法对于多分类的延误原因预判有较好效果，有较强的鉴别能力，对具体每一类航班延误原因进行判定时，可以做到较为精准的预测，能够有效提高航班延误的预警和管理能力。
说明书附图
图1所示为本发明的基于XGBoost的航班延误原因预判方法的总体架构示意图。
具体实施方式
以下结合具体实施例对本发明作进一步详细说明。应当理解，此处所描 述的具体实施例仅用以解释本发明，并不用于限定本发明。
一种基于XGBoost的航班延误原因预判方法，参见图1，方法的步骤包括：
S1：采集航班运行数据。
航班运行数据包括飞行计划数据、流控数据和气象数据。
其中，飞行计划数据包括：航班号、注册号、机型、航班日期、起飞机场、降落机场、计划起飞时间、实际起飞时间、计划降落时间、实际降落时间；流控数据包括：航班号、注册号、流控原因、流控内容等信息；天气数据包括：观测时间、天气类型、能见度、云底高、风向、风速等信息，其中，天气类型包括雷暴、降雨、降雪和雾。
S2：根据流控数据对航班延误原因进行分类，并将划分的各类航班延误原因类别与航班的飞行计划数据和天气数据相关联，得到关联后的样本数据。
具体的讲，根据流控数据中流控原因和流控内容，划分出航班延误原因类别，本发明中，将航班延误原因分为天气原因、航空公司原因、空管原因、机场原因和其他原因五类，并将这五类航班延误原因类别与航班的飞行计划数据和天气数据相关联。即，根据航班号、注册号可以将航班的飞行计划数据信息和划分出的航班延误原因类别相关联，根据时间信息可以将航班的飞行计划数据和相应的天气数据相关联，进而实现航班延误原因类别与航班的飞行计划数据和天气数据相关联。
S3：根据所划分的航班延误原因类别对航班延误数量进行分析，由于各类航班延误原因类别所对应的航班延误数量差别可能会很大，使得样本数据不平衡，从而影响后续预测模型的训练效果。因此本发明采用ADASYN算 法(AdaptiveSynthetic，自适应合成算法)对步骤S2得到的样本数据进行数据平衡处理，得到平衡后的样本数据。
采用ADASYN算法处理数据不平衡问题，即随不同的少数类样本赋予相应权重，然后得到不同数量的样本，使模型数据变得相对平衡。具体包括以下步骤：
S3.1：根据航班延误原因，统计各类航班延误原因关联延误航班的数量：
各类航班延误原因关联延误航班的数量G计算公式如下：
G＝(ml-ms)×β
其中，G为各类航班延误原因关联延误航班的数量，ml为较多数量类别的数据数量，ms为较少数量类别的数据数量，β为[0,1]的随机数。
S3.2：根据近邻中多数类样本数，计算近邻中较多数量类别的比例。
比例的计算公式如下：
ri＝Δi/K
其中，ri为近邻中较多数量类别的比例，Δi为xi的K个近邻中多数类样本数，i＝1,...,ms，xi为较小数目类别数据中第i个样本数据，xi存在K个近邻。
S3.3：对近邻中较多数量类别的比例ri进行标准化。
标准化的公式如下：
其中，为标准化的近邻中较多数量类别的比例。
S3.4：根据较小数量类别的权重，逐个计算较小数量类别所要生成新数据的数目。
新数据的数目计算公式如下：
其中，gi为少数类别中每个样本xi需要生成的合成样本的数目。
S3.5：根据较小数量类别所要生成新数据的数目，逐个计算较小数量类别数据生成的数目。
新数据生成的方法如下：
其中，为合成数据，u＝1,...,gi，xzi为xi的K近邻中随机选取一个较小数目类别样本数据，λ为[0,1]的随机数；合成数据与原有数据组合形成平衡后的样本数据。
S4：构建基于XGBoost的航班延误原因预判模型，将步骤S3中平衡后的样本数据中的航班飞行计划的特征因素以及和航班相关的气象特征因素作为模型的输入特征变量，将航班延误原因类别作为模型预测输出的目标变量。并将步骤S3中平衡后的样本数据分为训练集和验证集，对航班延误原因预判模型进行训练和验证。
具体的讲，航班飞行计划的特征因素包括：航班号、机型、航班日期、计划起飞时间；航班相关的气象特征因素包括：能见度、云底高、风向、风速、雷暴、降雨、降雪和雾。
构建基于XGBoost的航班延误原因预判模型的方法如下：
首先定义初始的模型的目标函数：
其中，Obj(t)为模型的目标函数，t为第t棵树，N为样本的数量，为第n个航班延误原因数据样本中目标变量预测值与真实值yn的误差，Ω(ft)为树的复杂度函数。
树的复杂度函数表达式如下：
其中，γ为复杂度参数，T为树的叶子节点数，η为固定系数，wj为第j个叶子节点的得分数。
将偏差和方差函数进行结合，建立下一棵树时参考模型上次预测与实际值的误差作为输入的加法策略，得出目标函数，该目标函数表达式如下：
其中，为yn与的误差；ft(xn)为第t棵树得到的预测结果；C为常数；
选择二阶泰勒展开式的方法，实现梯度下降法，使得模型误差减小，具体公式如下：
其中，f(x)为一阶函数，Δx为变量改变量。
令 将其代入上述目标函数表达式可得：
式中，Ij为节点j上的样本集，dn为每个航班延误样本的梯度值，hn为每个航班延误样本的二阶导数。
航班延误原因预判模型构建完成后，采用训练集和验证集放入模型中进行分类预测，并判断延误原因，利用精确度、召回率和F1-score值对模型的预测效果进行校验，最后得到精确度满足要求的航班延误原因预判模型。
S5：将待预测的航班的飞行计划的特征因素以及和该航班相关的气象特征因素作为模型的输入，输入至步骤S4训练好的航班延误原因预判模型中，模型输出的目标变量则为航班延误原因类别。
以上对本发明做了示例性的描述，应该说明的是，在不脱离本发明的核心的情况下，任何简单的变形、修改或者其他本领域技术人员能够不花费创造性劳动的等同替换均落入本发明的保护范围。

Claims (4)

  1. 一种基于XGBoost的航班延误原因预判方法，其特征在于，包括以下步骤：
    S1：采集航班运行数据，航班运行数据包括飞行计划数据、流控数据和气象数据；
    S2：根据流控数据划分出航班延误原因类别，将划分的各类航班延误原因类别与航班的飞行计划数据和天气数据相关联，得到关联后的样本数据；
    S3：根据划分的航班延误原因类别，统计各类航班延误原因关联延误航班的数量，采用ADASYN算法对步骤S2得到的样本数据进行数据平衡处理，得到平衡后的样本数据；
    S4：构建基于XGBoost的航班延误原因预判模型，将步骤S3得到的平衡后的样本数据中的航班飞行计划的特征因素以及和航班相关的气象特征因素作为模型的输入特征变量，将航班延误原因类别作为模型预测输出的目标变量；并将步骤S3中平衡后的样本数据分为训练集和验证集，对航班延误原因预判模型进行训练和验证，得到训练好的航班延误原因预判模型；
    S5：将待预测的航班的飞行计划的特征因素以及和该航班相关的气象特征因素作为模型的输入，输入至步骤S4训练好的航班延误原因预判模型中，模型输出的目标变量则为航班延误原因类别。
  2. 根据权利要求1所述的基于XGBoost的航班延误原因预判方法，其特征在于：步骤S2，根据航班号、注册号将航班的飞行计划数据信息和划分出的航班延误原因类别相关联，根据时间信息将航班的飞行计划数据和相应的天气数据相关联，进而实现航班延误原因类别与航班的飞行计划数据和天气数据相关联。
  3. 根据权利要求1所述的基于XGBoost的航班延误原因预判方法，其特征在于：步骤S3包括以下步骤：
    S3.1：根据航班延误原因，统计各类航班延误原因关联延误航班的数量：
    各类航班延误原因关联延误航班的数量G计算公式如下：
    G＝(ml-ms)×β
    其中，G为各类航班延误原因关联延误航班的数量，ml为较多数量类别的数据数量，ms为较少数量类别的数据数量，β为[0,1]的随机数；
    S3.2：根据近邻中多数类样本数，计算近邻中较多数量类别的比例；比例的计算公式如下：
    ri＝Δi/K
    其中，ri为近邻中较多数量类别的比例，Δi为xi的K个近邻中多数类样本数，i＝1,...,ms，xi为较小数目类别数据中第i个样本数据，xi存在K个近邻；
    S3.3：对近邻中较多数量类别的比例ri进行标准化，标准化的公式如下：
    其中，为标准化的近邻中较多数量类别的比例；
    S3.4：根据较小数量类别的权重，逐个计算较小数量类别所要生成新数据的数目，计算公式如下：
    其中，gi为少数类别中每个样本xi需要生成的合成样本的数目；
    S3.5：根据较小数量类别所要生成新数据的数目，逐个计算较小数量类 别数据生成的数目；
    所述新数据生成的方法如下：
    其中，为合成数据，u＝1,...,gi，xzi为xi的K近邻中随机选取一个较小数目类别样本数据，λ为[0,1]的随机数；所述合成数据与原有数据组合形成平衡后的样本数据。
  4. 根据权利要求1所述的基于XGBoost的航班延误原因预判方法，其特征在于：构建基于XGBoost的航班延误原因预判模型的方法如下：
    首先定义初始的模型的目标函数：
    其中，Obj(t)为模型的目标函数，t为第t棵树，N为样本的数量，为第n个航班延误原因数据样本中目标变量预测值与真实值的误差，Ω(ft)为树的复杂度函数；
    树的复杂度函数表达式如下：
    其中，γ为复杂度参数，T为树的叶子节点数，η为固定系数，wj为第j个叶子节点的得分数；
    将偏差和方差函数进行结合，建立下一棵树时参考模型上次预测与实际值的误差作为输入的加法策略，得出目标函数，该目标函数表达式如下：
    选择二阶泰勒展开式的方法，实现梯度下降法，使得模型误差减小，具体公式如下：
    令 将其代入上述目标函数表达式可得：
    式中，dn为每个航班延误样本的梯度值，hn为每个航班延误样本的二阶导数。
PCT/CN2023/128387 2022-11-29 2023-10-31 一种基于XGBoost的航班延误原因预判方法 WO2024114250A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211509271.0A CN116150667A (zh) 2022-11-29 2022-11-29 一种基于XGBoost的航班延误原因预判方法
CN202211509271.0 2022-11-29

Publications (1)

Publication Number Publication Date
WO2024114250A1 true WO2024114250A1 (zh) 2024-06-06

Family

ID=86355221

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/128387 WO2024114250A1 (zh) 2022-11-29 2023-10-31 一种基于XGBoost的航班延误原因预判方法

Country Status (2)

Country Link
CN (1) CN116150667A (zh)
WO (1) WO2024114250A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116150667A (zh) * 2022-11-29 2023-05-23 华北理工大学 一种基于XGBoost的航班延误原因预判方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160189066A1 (en) * 2014-12-24 2016-06-30 General Electric Company Method and system for automatic evaluation of robustness and disruption management for commercial airline flight operations
CN105956621A (zh) * 2016-04-29 2016-09-21 南京航空航天大学 一种基于进化欠抽样集成学习的航班延误预警方法
US20220215760A1 (en) * 2019-05-28 2022-07-07 Sita Information Networking Computing Uk Limited System and method for flight arrival time prediction
CN116150667A (zh) * 2022-11-29 2023-05-23 华北理工大学 一种基于XGBoost的航班延误原因预判方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160189066A1 (en) * 2014-12-24 2016-06-30 General Electric Company Method and system for automatic evaluation of robustness and disruption management for commercial airline flight operations
CN105956621A (zh) * 2016-04-29 2016-09-21 南京航空航天大学 一种基于进化欠抽样集成学习的航班延误预警方法
US20220215760A1 (en) * 2019-05-28 2022-07-07 Sita Information Networking Computing Uk Limited System and method for flight arrival time prediction
CN116150667A (zh) * 2022-11-29 2023-05-23 华北理工大学 一种基于XGBoost的航班延误原因预判方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
TANG HONG, WANG DONG, SONG BO, CHU WENKUI, HE LINYUAN: "Classification of Flight Delay Based on Nonlinear Weighted XGBoost", JOURNAL OF SYSTEM SIMULATION, vol. 33, no. 9, 1 September 2021 (2021-09-01), pages 2261 - 2269, XP093176261, DOI: 10.16182/j.issn1004731x.joss.20-0372 *
ZHANG MINGLIANG; HOU XIA: "An integrated prediction method for departure flight delay based on XGBoost and MLP", JOURNAL OF BEIJING INFORMATION SCIENCE & TECHNOLOGY UNIVERSITY, BEIJING INFORMATION SCIENCE & TECHNOLOGY UNIVERSITY, CN, vol. 37, no. 3, 30 June 2022 (2022-06-30), CN , pages 41 - 45, XP009555037, ISSN: 1674-6864 *

Also Published As

Publication number Publication date
CN116150667A (zh) 2023-05-23

Similar Documents

Publication Publication Date Title
WO2024114250A1 (zh) 一种基于XGBoost的航班延误原因预判方法
CN111652427B (zh) 一种基于数据挖掘分析的航班到达时刻预测方法及系统
WO2021082393A1 (zh) 基于大数据深度学习的机场场面可变滑出时间预测方法
WO2021082394A1 (zh) 基于大数据深度学习的场面可变滑出时间预测系统
Yin et al. Machine learning techniques for taxi-out time prediction with a macroscopic network topology
CN108805337B (zh) 航空运行控制风险管控方法及系统
CN110675007A (zh) 航班延误预测方法、装置及计算机存储介质
TWI794516B (zh) 透過空中交通管理電子系統的用於空中管理的訓練和/或輔助平台及方法
CN115204466A (zh) 一种带通行限制的国际航线航路规划方法
CN111160612A (zh) 一种基于天气影响的离场航班延误分析及预测方法
CN110210648A (zh) 基于灰色长短期记忆网络的管制空域战略流量预测方法
CN114781704B (zh) 一种基于过站航班保障流程的航班延误预测方法
CN114647985A (zh) 锋面过境预测模型的训练方法及锋面过境的预测方法
Nan et al. Prediction of Departure Aircraft Taxi Time Based on Deep Learning.
CN112365037A (zh) 基于长短期数据预测模型的机场空域流量预测方法
Schultz et al. Classification of weather impacts on airport operations
Chou et al. A machine learning application for predicting and alerting missed approaches for airport management
CN110009939A (zh) 基于asm的航班延误预测及波及分析方法
CN117995027A (zh) 一种飞行计划备降场推荐方法、系统、设备及介质
CN116109212B (zh) 一种机场运行效率评价指标设计及监测方法
CN115759386B (zh) 一种民航航班执飞结果预测方法、装置及电子设备
CN115222150B (zh) 一种基于Stacking的航班延误预测方法及系统
CN110766064A (zh) 一种基于局部加权支持向量回归的离港滑行时间预测方法
Bongo et al. A hybrid fuzzy multi-criteria decision-making approach for mitigating air traffic congestion
Priyanka Prediction of airline delays using K-nearest neighbor algorithm

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23896404

Country of ref document: EP

Kind code of ref document: A1