Flight extra sliding time estimation method based on scene traffic situation feature prediction

Abstract

The invention belongs to the technical field of flight taxiing time estimation of scene operation optimization, and particularly relates to a flight additional taxiing time estimation method based on scene traffic situation feature prediction. The method comprises the following steps: calculating undisturbed slip-out time; preprocessing airport flight data, and constructing complex multi-runway scene traffic situation features and flight basic information features; constructing traffic situation features based on prediction; constructing a traffic situation and flight basic information data set based on prediction; an NOA-XGBOOST combination model is constructed; and the accuracy of the NOA-XGBOOST combination model is verified. According to the method, multiple factors are introduced to serve as input features, an NOA model, an XGBOOST model and an NOA-XGBOOST combined model are constructed respectively, and the NOA model, the XGBOOST model and the NOA-XGBOOST combined model are compared with the XGBOOST model, the RF model and the SVR model; and the XGBOOST combined model is verified to be high in estimation precision, estimation errors MAE and RMSE are reduced, and + / -3 min precision, + / -5 min precision and + 5-10 min precision are improved to some extent. The overall operation efficiency of the airport is improved, and the congestion phenomenon in the ground sliding period is reduced.

Description

Technical Field

[0001] The invention belongs to the technical field of taxi-out time estimation for airport surface operation management, and in particular relates to a method for estimating additional taxi-out time of a flight based on prediction of surface traffic situation characteristics. Background Art

[0002] As flight delays continue to increase, it is particularly important to accurately estimate the taxi time of aircraft at the airport. Flight taxi time estimation not only helps improve airport operational efficiency and reduce ground congestion, but also optimizes flight scheduling, improves punctuality and reduces fuel consumption. Through advanced machine learning and data analysis technology, airlines and airports can estimate taxi time more accurately, thereby optimizing resource allocation, reducing operating costs, improving passenger experience, and reducing carbon emissions, promoting the development of green airports.

[0003] The existing technologies for taxi-out time estimation first focus on the overall taxi-out time. However, taxi-out time is divided into two parts: undisturbed taxi-out time and additional taxi-out time. The undisturbed taxi-out time refers to the taxi-out time of the flight when some delay factors are not significant. The additional taxi-out time is the difference between the overall taxi-out time and the undisturbed taxi-out time. It is a key indicator of on-site operation management and reflects the operating efficiency of the on-site. When optimizing the taxi-out time, the additional taxi-out time is the focus of optimization. Secondly, in terms of feature input, the focus is on the overall flow of the on-site, while the micro-structure related flow, such as corridor flow, is ignored. In addition, the traffic situation characteristics of the on-site are based on historical flow information, which is not suitable for actual application. Finally, in terms of estimation methods, only machine learning algorithms are simply used, or combined, and research on optimizing and improving the parameters of existing machine learning methods has not been effectively carried out. Summary of the invention

[0004] The purpose of the present invention is to provide a method for estimating the additional taxiing time of a flight based on the prediction of the characteristics of the field traffic situation.

[0005] In order to solve the above technical problems, the present invention provides a flight additional taxiing time estimation method based on the prediction of scene traffic situation characteristics, including: calculating the undisturbed taxi-out time; preprocessing the airport flight data to construct complex multi-runway scene traffic situation characteristics and flight basic information characteristics; constructing traffic situation characteristics based on prediction; constructing a traffic situation and flight basic information data set based on prediction; constructing a NOA-XGBOOST combined model; and verifying the accuracy of the NOA-XGBOOST combined model.

[0006] Beneficial effects of the present invention:

[0007] (1) Focus on the estimation of additional taxi-out time rather than studying taxi-out time as a whole;

[0008] (2) In terms of the input factors for the extra taxiway time estimation, the influence of the corridor flow was considered, three new indicators were established, namely the corridor departure flow, corridor arrival flow and departure flow, and the traffic situation characteristics of the complex multi-runway scene were established;

[0009] (3) In terms of features, according to the calculation time nodes in the ACDM system, a traffic situation feature based on prediction is constructed and combined with the basic information features of the flight;

[0010] (4) In terms of estimation models, the NOA model, XGBOOST model, and NOA-XGBOOST combined model were constructed and compared with the XGBOOST, RF, and SVR models. It was also verified that the NOA-XGBOOST model had higher estimation accuracy than other models, with lower MAE and RMSE values, and improved R2, ±3-minute accuracy, ±5-minute accuracy, and +5-10-minute accuracy.

[0011] (5) In terms of example verification, taking the flight operation data of Shanghai Pudong International Airport as an example, it is verified that the estimation accuracy of the NOA-XGBOOST combined model is higher than that of the single XGBOOST model, the estimation error MAE and RMSE values ​​are reduced, and the ±3 minute accuracy, ±5 minute accuracy and +5-10 minute accuracy are improved;

[0012] (6) The method for estimating the additional taxiing time of a flight based on the prediction of the characteristics of the surface traffic situation of the present invention is beneficial to improving the overall operational efficiency of the airport and reducing congestion during ground taxiing. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 A flow chart of a method for estimating additional taxiing time of a flight based on prediction of scene traffic situation characteristics according to an embodiment of the present invention;

[0015] Figure 2 A flow chart of calculating the undisturbed slide-out time according to an embodiment of the present invention;

[0016] Figure 3 A data set construction process for an additional taxiing time estimation method according to an embodiment of the present invention;

[0017] Figure 4A comparison diagram of estimation results of different methods according to an embodiment of the present invention;

[0018] Figure 5 3 is a comparison diagram of estimation errors before and after removing structure-related features according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. 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.

[0020] like Figure 1 As shown, this embodiment provides a method for estimating additional taxiing time of a flight based on the prediction of scene traffic situation characteristics, including: calculating the undisturbed taxi-out time; preprocessing the airport flight data to construct complex multi-runway scene traffic situation characteristics and flight basic information characteristics; constructing traffic situation characteristics based on prediction; constructing a traffic situation and flight basic information data set based on prediction; constructing a NOA-XGBOOST combined model; and verifying the accuracy of the NOA-XGBOOST combined model.

[0021] In this embodiment, calculating the undisturbed taxiing time includes: grouping flights, screening arrival and departure flow indicators using the Pearson correlation coefficient, and establishing a linear regression model for calculation.

[0022] In an application scenario, specifically, Figure 2 As shown in the figure, 22 stand groups were first divided, and the stand groups and departure runways were paired one by one to form 88 sets of data. Then, the arrival and departure flow indicators that had temporal interactions with the studied flights were screened, and the one with the largest Pearson correlation coefficient with the taxi-out time was the screened arrival and departure flow indicator. Finally, the selected indicators were used to establish a regression model, so that the departure variable coefficient in the regression model was 1 and the arrival variable coefficient was 0, and the undisturbed taxiing time was calculated.

[0023] Subtract the calculated undisturbed taxi-out time from the overall taxi-out time for each flight to obtain the additional taxi-out time, and use this value as the basis for the estimated comparison.

[0024] In this embodiment, the airport flight data is preprocessed to construct the complex multi-runway scene traffic situation characteristics and flight basic information characteristics. The complex multi-runway scene traffic situation characteristics include departure flow, arrival flow, departure flow ratio, corridor departure flow and corridor arrival flow. Flight basic information characteristics include airline, aircraft type, restricted status, and time. Airline characteristics include domestic airlines and foreign airlines; aircraft type characteristics include Class C flights, Class D flights, Class E flights and Class F flights; restricted status characteristics include restricted flights and non-restricted flights; time characteristics include hours. Among them, departure flow ratio, corridor departure flow and corridor arrival flow are new features proposed in this embodiment.

[0025] The proportion of departure traffic refers to the ratio of departure traffic to the sum of arrival traffic and departure traffic. The corridor departure traffic refers to the number of other departing flights that taxi from the East (West) control area stand group to the West (East) control area departure runway through corridor TW3 (TW4) during the taxiing period. Correspondingly, the corridor arrival traffic refers to the number of other arrival flights.

[0026] In this embodiment, the method for constructing traffic situation characteristics based on prediction includes: using the calculation time point CTOT, COBT, CLDT and CIBT values ​​in the ACDM system to predict the scene traffic situation characteristics, and these values ​​are usually determined within 30 minutes or 1 hour before the aircraft takes off.

[0027] In this embodiment, the method for constructing a traffic situation and flight basic information dataset based on prediction includes:

[0028] According to the established traffic situation characteristics based on prediction and basic flight information characteristics, relevant data are selected from the original arrival and departure data, and the original data is preprocessed. For missing values, since the number of missing values ​​accounts for a small ratio of the total number, the corresponding data rows are directly deleted. For outliers, the standard deviation method is used, that is, the data with departure taxi time values ​​exceeding the mean plus or minus three times the standard deviation is deleted. Finally, the preprocessed data is used for calculation to obtain a data set based on the predicted traffic situation and basic flight information. The data set construction process is as follows: Figure 3 shown.

[0029] In this embodiment, the method for constructing an extra slide-out time estimation model composed of NOA and XGBOOST includes: constructing a NOA model; constructing an XGBOOST model; and constructing a NOA-XGBOOST combination model.

[0030] The constructed NOA model includes:

[0031] In the global search phase of the foraging and storage strategy phase, the position update formula is:

[0032]

[0033] Where: is the new position of the i-th nutcracker in the current generation; is the jth position of the i-th nutcracker in the current generation; U j and L j is a vector, including the upper and lower bounds of the j-th dimension in the optimization problem; γ is a random number generated according to the levy row; is the optimal solution of the jth dimension obtained so far; A, C and B are three different indicators randomly selected from the population to facilitate the exploration of high-quality food sources; τ1, τ2, r, r1 are random real numbers in the range of [0, 1]; is the j-th dimension mean of all solutions of the current population in the t-th iteration; μ is a number randomly generated between 0 and 1 (τ3) based on normal distribution (τ4), levy-flight (τ5);

[0034] In the local development phase of the foraging and storage strategy phase, the update formula for the position is:

[0035]

[0036] Where: λ is a number generated based on levy flight, τ3 is a random number between 0 and 1; l is a factor that linearly decreases from 1 to 0 in NOA development behavior;

[0037] In the global search phase of the cache search and recovery strategy, the position update formula is:

[0038]

[0039] Where: is the first reference point of the current position / cache of the ith nutcracker in the current iteration t; r1, r2, τ3, τ4, τ5, τ6 are random numbers between 0 and 1, and C is the index of a randomly selected solution from the population;

[0040] During the local development phase of the cache search and recovery strategy, the position update formula is:

[0041]

[0042] Where: is the first reference point currently cached by the ith Nutcracker at iteration t, is the second reference point currently cached by the i-th Nutcracker at iteration t;

[0043] All of the above updates require adjustments, namely:

[0044]

[0045] The construction of the XGBOOST model includes:

[0046] Initialize the base learner, that is

[0047]

[0048] In the formula is the initial estimate of the ith sample, y i is the actual value, n is the number of samples;

[0049] In each round of iteration m, the residual is calculated That is, the difference between the true value and the current estimate:

[0050]

[0051] In the formula, is the estimated value of the i-th sample in the m-1-th iteration;

[0052] Training base learner h m (x) Fitting residual Update the base learner:

[0053]

[0054] Where η is the learning rate, which controls the contribution of each base learner to the final model;

[0055] Add the new base learner to the model and update the model's estimate:

[0056]

[0057] Where M is the total number of iterations.

[0058] The construction of the NOA-XGBOOST combined model includes:

[0059] The objective function of the NOA model is set to the RMSE value of the estimation result, the optimization variable is set to the hyperparameter of XGBOOST, and the optimized hyperparameter value is input into the XGBOOST model for estimation, and finally the estimated value of the additional slip time is obtained.

[0060] The steps to implement the NOA-XGBOOST combined model include:

[0061] Step 1: Use 80% of the data based on the predicted traffic situation and basic flight information dataset as training samples and 20% of the data as test samples.

[0062] Step 2: Set the optimization range of XGBOOST hyperparameters and set the objective function of NOA to the RMSE value.

[0063] Step 3: Initialize the NOA and XGBOOST model parameters on the training set, and iteratively update them using the NOA algorithm to obtain the optimal XGBOOST hyperparameters.

[0064] Step 4: Use the trained model and optimal hyperparameters to estimate the additional slide time on the test set.

[0065] In this embodiment, the maximum number of model iterations is set to 100, the population size is set to 50, and the hyperparameters that need to be adjusted are the number of base learners, the maximum depth, the learning rate, the minimum split loss, the subsample ratio, the column sampling ratio, the L1 regularization coefficient, and the L2 regularization coefficient.

[0066] In this embodiment, the verification of the accuracy of the NOA-XGBOOST combination model includes: using mean absolute error (MAE), root mean square error (RMSE), ±3 minute accuracy, ±5 minute accuracy, and +5-10 minute accuracy as evaluation indicators; comparing the NOA-XGBOOST combination model with the evaluation indicators estimated by different models.

[0067] Optionally, the different models include: XGBOOST, RF, and SVR.

[0068] In this embodiment, optionally, a visualization method and actual operation data are used to verify the accuracy of the proposed additional slip-out time estimation model based on the NOA-XGBOOST model. The specific implementation process of the rationality includes:

[0069] The mean absolute error (MAE), root mean square error (RMSE), ±3 minute accuracy, ±5 minute accuracy, and +5-10 minute accuracy are used as evaluation indicators and calculated using the following formulas:

[0070]

[0071] Among them, y i is the true value, is an estimate, is the average of the true values.

[0072] 80% of the flight operation data set is selected as training samples, and 20% of the data is selected as test samples. Taking Shanghai Pudong International Airport as an example, the error terms estimated by different models are shown in Table 1. Figure 4As shown in the figure, the additional taxi-out time of the first 200 flights is shown. It can be found that in terms of additional taxi-out time estimation, the estimation effect of the NOA-XGBOOST combined model is significantly better than the XGBOOST model, SVR model and RF model, which improves the estimation accuracy.

[0073] Table 1 Estimation errors of NOA-XGBOOST, XGBOOST, RF, and SVR models

[0074]

[0075] Compared with XGBOOST, RF and SVR models, the NOA-XGBOOST model has lower MAE and RMSE indicators, but has improved ±3min accuracy, ±5min accuracy and +5-10min accuracy.

[0076] The results of the feature ablation experiment using the NOA-XGBOOST algorithm are shown in Table 2.

[0077] Table 2 NOA-XGBOOST algorithm feature ablation experiment

[0078]

[0079] It can be found that conventional traffic features have the greatest impact on the estimation results, followed by the three new features proposed. After adding the proposed features, the model has reduced MAE and RMSE indicators, while improving ±3min accuracy, ±5min accuracy and +5-10min accuracy. The error comparison of the first 200 flights before and after adding structural related features is shown in the figure below. Figure 5 shown.

[0080] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A method for estimating flight extra taxiing time based on the prediction of traffic situation characteristics, characterized in that: include: Calculate the undisturbed slide-out time; Pre-process airport flight data to construct complex multi-runway traffic situation characteristics and flight basic information characteristics; Construct prediction-based traffic situation features; Construct a data set of traffic situation and basic flight information based on prediction; Construct NOA-XGBOOST combination model; Verify the accuracy of the NOA-XGBOOST combined model.

2. The method for estimating the additional taxiing time of a flight according to claim 1, characterized in that: The calculation of the undisturbed taxi-out time includes: grouping flights, screening inbound and outbound flow indicators using the Pearson correlation coefficient, and establishing a linear regression model for calculation.

3. The method for estimating the additional taxiing time of a flight according to claim 2, characterized in that: The construction of complex multi-runway traffic situation characteristics and flight basic information characteristics includes: The traffic situation characteristics of the complex multi-runway scene include departure flow, arrival flow, departure flow ratio, corridor departure flow and corridor arrival flow; The basic flight information features include airline, aircraft type, restricted status, and time; The airline characteristics include domestic airlines and foreign airlines; The aircraft type characteristics include Class C flights, Class D flights, Class E flights, and Class F flights; The restricted status characteristics include restricted flights and non-restricted flights; The time characteristic includes hours.

4. The method for estimating the additional taxiing time of a flight according to claim 3, characterized in that: The construction of the traffic situation features based on prediction includes: The CTOT, COBT, CLDT and CIBT values ​​at the calculation time points in the ACDM system are used to predict the characteristics of the surface traffic situation. These values ​​are usually determined within 30 minutes or 1 hour before the aircraft takes off.

5. The method for estimating the additional taxiing time of a flight according to claim 4, characterized in that: The method for constructing a traffic situation and flight basic information data set based on prediction includes: According to the constructed traffic situation characteristics based on prediction and basic flight information characteristics, relevant data are selected from the original arrival and departure data, the original data is preprocessed, and the preprocessed data is used for calculation to obtain the traffic situation and basic flight information data set based on prediction.

6. The method for estimating the additional taxiing time of a flight according to claim 5, characterized in that: The estimation model constructed by combining NOA and XGBOOST includes: Construct NOA model; Build the XGBOOST model; Construct the NOA-XGBOOST combination model.

7. The method for estimating the additional taxiing time of a flight according to claim 6, characterized in that: The construction of the NOA model includes two stages: foraging and storage strategy and cache search and recovery strategy, each of which includes global search and local development; In the global search phase of the foraging and storage strategy phase, the position update formula is: Where: is the new position of the i-th nutcracker in the current generation; is the jth position of the i-th nutcracker in the current generation; U j and L j is a vector, including the upper and lower bounds of the j-th dimension in the optimization problem; γ is a random number generated according to the levy row; is the optimal solution of the jth dimension obtained so far; A, C and B are three different indicators randomly selected from the population to facilitate the exploration of high-quality food sources; τ1, τ2, r, r1 are random real numbers in the range of [0, 1]; is the j-th dimension mean of all solutions of the current population in the t-th iteration; μ is a number randomly generated between 0 and 1 (τ3) based on normal distribution (τ4), levy-flight (τ5); In the local development phase of the foraging and storage strategy phase, the update formula for the position is: Where: λ is a number generated based on levy flight, τ3 is a random number between 0 and 1; l is a factor that linearly decreases from 1 to 0 in NOA development behavior; In the global search phase of the cache search and recovery strategy, the position update formula is: Where: is the first reference point of the current position / cache of the ith nutcracker in the current iteration t; r1, r2, τ3, τ4, τ5, τ6 are random numbers between 0 and 1, and C is the index of a randomly selected solution from the population; During the local development phase of the cache search and recovery strategy, the position update formula is: Where: is the first reference point currently cached by the ith Nutcracker at iteration t, is the second reference point currently cached by the ith Nutcracker at iteration t; All of the above updates require adjustments, namely:

8. The method for estimating the additional taxiing time of a flight according to claim 7, characterized in that: The construction of the XGBOOST model includes: Initialize the base learner, that is In the formula is the initial estimate of the ith sample, y i is the actual value, n is the number of samples; In each round of iteration m, the residual is calculated That is, the difference between the true value and the current value: In the formula, is the estimated value of the i-th sample in the m-1-th iteration; Training base learner h m (x) Fitting residual Update the base learner: Where η is the learning rate, which controls the contribution of each base learner to the final model; Add the new base learner to the model and update the model's estimate: Where M is the total number of iterations.

9. The method for estimating the additional taxiing time of a flight according to claim 8, characterized in that: The constructed NOA-XGBOOST combined model includes: The objective function of the NOA model is set to the RMSE value of the estimation result, the optimization variable is set to the hyperparameter of XGBOOST, and the optimized hyperparameter value is input into the XGBOOST model for estimation, and finally the estimated value of the additional slip time is obtained.

10. The method for estimating the additional taxiing time of a flight according to claim 9, characterized in that: The steps to implement the NOA-XGBOOST combined model include: Step 1: Use 80% of the data in the dataset as training samples and 20% of the data as test samples; Step 2: Set the optimization range of XGBOOST hyperparameters and set the objective function of NOA to the RMSE value; Step 3: Initialize the NOA and XGBOOST model parameters on the training set, and iteratively update them using the NOA algorithm to obtain the optimal XGBOOST hyperparameters; Step 4: Use the trained model and optimal hyperparameters to estimate the additional slide time on the test set.

Images