A method for evaluating port entry sequencing

Abstract

The application discloses a kind of methods for evaluating approach sequencing, the evaluation method is based on the approach sequencing system of air traffic control terminal, steps include: collecting sequencing evaluation data, the sequencing evaluation data includes sequencing sequence data, time suggestion data and air route suggestion data;Based on the sequencing evaluation data, multidimensional approach sequencing evaluation is carried out, and the multidimensional approach sequencing evaluation includes sequencing sequence evaluation, time suggestion evaluation and air route suggestion evaluation;The multidimensional approach sequencing evaluation result is output.The application carries out sequencing sequence evaluation, time suggestion evaluation and air route suggestion evaluation based on the collected sequencing evaluation data respectively, effectively evaluates the accuracy of approach sequencing system in each key stage sequencing, based on operation data, objectively evaluates the performance of approach sequencing system in the process of operation, is favorable to the efficiency and reliability of approach sequencing system for controller to make decision analysis according to data index, meets the requirement of controller to each stage of approach flight.

Description

Technical Field

[0001] This invention relates to the field of flight sorting, and more specifically to an arrival sorting evaluation method. Background Technology

[0002] With rapid economic development, my country's civil aviation industry has also ushered in a golden age of development, with an increasing number of arriving flights. Relying solely on the individual abilities and experience of air traffic controllers is no longer sufficient to meet the demands of this era, hence the emergence of arrival sorting systems. Currently, academia and industry have conducted extensive research and exploration into the mechanisms of arrival sorting, resulting in a vibrant and diverse landscape of ideas. However, in the actual operation of civil aviation air traffic control departments, there is a lack of a reliable evaluation system for arrival sorting systems, making it difficult to quantify and evaluate the predictions and suggestions provided by these systems.

[0003] Existing arrival sequencing schemes research various mechanisms and strategies to effectively improve airport arrival efficiency, resolve arrival flight conflicts, and reduce the probability of flight delays, providing arrival sequencing suggestions for controllers. The invention patent, "An Arrival Sequencing Method Based on Machine Learning and Airport Capacity," uses machine learning and airport capacity as its core, effectively improving the accuracy of flight durations in approach airspace. It also employs machine learning to analyze real flight trajectories in approach airspace, providing route suggestions to resolve arrival flight slot conflicts. Furthermore, by incorporating airport capacity parameters, it balances the needs of both area control and terminal control, reducing the difficulty of implementing new arrival sequencing technologies.

[0004] However, existing arrival sorting systems still have some shortcomings. Taking the invention patent <An Arrival Sorting Method Based on Machine Learning and Airport Capacity> as an example: 1) Although time delay suggestions are calculated based on airport capacity and the number of expected arriving flights, there is a lack of analysis on whether the time delay suggestions are correct, and controllers cannot determine the accuracy of the time suggestions; 2) Although a method for calculating approach airspace route suggestions is given, there is a lack of evaluation on the accuracy of the route suggestions, and controllers cannot determine the accuracy of the route suggestions.

[0005] The reasons for the above-mentioned shortcomings include: The invention patent <An Arrival Ranking Method Based on Machine Learning and Airport Capacity> uses airport capacity as the core, and determines whether to give control time delay suggestions by calculating the number of expected arriving flights per unit time. However, in actual operation, time delays may be inaccurate due to various reasons. That is, the system should generate delay suggestions for flights but does not actually generate suggestions, or the system should not generate delay suggestions but generates delay suggestions. The patent lacks an effective method to detect and evaluate possible erroneous suggestions in operation.

[0006] The invention patent "An Arrival Ranking Method Based on Machine Learning and Airport Capacity" uses a machine learning-based method to provide route suggestions for flights approaching the airspace. However, the patent lacks a method to evaluate the accuracy of the route suggestions, and cannot give air traffic controllers sufficient confidence to trust the route suggestions made by the system. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides an arrival sorting evaluation method. The evaluation method is based on the arrival sorting system of an air traffic control terminal, and includes the following steps:

[0008] Obtain port arrival sorting data, which is generated by the port arrival sorting system;

[0009] Based on the port entry sorting data, sorting evaluation data is collected, which includes sorting sequence data, time suggestion data, and route suggestion data.

[0010] Based on the ranking evaluation data, a multi-dimensional port arrival ranking evaluation is performed, which includes ranking sequence evaluation, time suggestion evaluation, and route suggestion evaluation.

[0011] Output the multidimensional port entry ranking evaluation results.

[0012] Furthermore, the collected ranking and evaluation data includes collected ranking sequence data, specifically:

[0013] Obtain the timing for collecting the pre-set expected sorting sequence from the air traffic control terminal;

[0014] When any arriving flight meets the expected sorting sequence collection timing, the expected arrival sequence generated by the arrival sorting system at this time is collected according to the arrival sorting data, and the expected sorting sequence is saved to the expected sorting sequence data pool;

[0015] When there are actual landing flights, obtain the list of actual landing flights based on the arrival sorting data, and add the flight information of the actual landing flights to the list of actual landing flights.

[0016] Furthermore, when any arriving flight meets the expected sorting sequence collection timing, the expected arrival sequence generated by the arrival sorting system at this time is collected according to the arrival sorting data, and the expected sorting sequence is saved to the expected sorting sequence data pool, specifically as follows:

[0017] When any inbound flight meets the expected sorting sequence collection timing, the inbound flight that meets the expected sorting sequence collection timing will be determined as the center flight;

[0018] Obtain the number of sampled flights pre-set in the air traffic control terminal;

[0019] Based on the arrival sorting data and the number of sampled flights, the expected sorting sequence generated by the arrival sorting system based on the central flight is collected, and the expected sorting sequence is saved to the expected sorting sequence pool;

[0020] The expected sorting sequence includes the central flight and the flights before and after the central flight. The estimated number of inbound flights is Q, where Q represents the number of sampled flights.

[0021] Furthermore, the multi-dimensional arrival ranking evaluation based on the ranking evaluation data includes ranking sequence evaluation based on the ranking sequence data, specifically as follows:

[0022] All expected sorting sequences in the expected sorting sequence pool are grouped according to their corresponding collection timing, and all groups are traversed.

[0023] Calculate the average accuracy of the sorted sequences for each group;

[0024] The ranking of each group is evaluated based on the average accuracy of the ranking sequence until the ranking of all groups is evaluated.

[0025] Further, the calculation of the average accuracy of the sorted sequence for each group is specifically as follows:

[0026] Iterate through all sorted sequences within each group;

[0027] Obtain the expected sorting sequence of a single flight within the group from the expected sorting sequence pool to obtain the sorting sequence to be evaluated;

[0028] Obtain the relative position of the sorted sequence to be evaluated in the list of actual landing flights to obtain the actual landing sequence;

[0029] Obtain a preset character conversion list, and convert the sorting sequence to be evaluated and the actual landing sequence according to the character conversion list to obtain the sorting sequence string and the landing sequence string respectively;

[0030] Based on the sorted sequence string and the landing sequence string, calculate the Levinstein similarity between a single flight within the group and the actual landing sequence;

[0031] The average accuracy of the sorted sequence for each group is calculated based on the Levinstein similarity of all individual flights within each group.

[0032] Furthermore, the collected ranking evaluation data includes suggested data on collection time, specifically:

[0033] When the arrival sorting system prompts for a delay time suggestion, a flight delay time suggestion list is obtained based on the arrival sorting data, and the delay time suggestion is added to the flight delay time suggestion list;

[0034] When there are actual landing flights, obtain the list of actual landing flights based on the arrival sorting data, and add the flight information of the actual landing flights to the list of actual landing flights.

[0035] Furthermore, the entry ranking evaluation based on the ranking evaluation data includes the error time suggestion evaluation based on the time suggestion data, specifically as follows:

[0036] Iterate through the list of suggested flight delay times;

[0037] If there is a flight delay suggestion, obtain the relative position of the delayed flight in the list of actual landing flights;

[0038] Based on the arrival sorting data, the airport capacity parameters at the time of landing of the delayed flight are obtained. The airport capacity parameters include the maximum number of flights that the airport is allowed to land within the capacity time.

[0039] Starting from the relative position of the delayed flight in the list of actual landing flights, look backward in the list of actual landing flights to see the number of actual landing flights within the capacity time.

[0040] Determine whether the actual number of landing flights within the specified capacity time exceeds the maximum number of landing flights. If it does, the evaluation result is that the delay time suggestion is correct; otherwise, the evaluation result is that the delay time suggestion is incorrect.

[0041] Furthermore, the arrival ranking evaluation based on the ranking evaluation data also includes a missing time suggestion evaluation based on the time suggestion data, specifically:

[0042] Iterate through the list of actual landing flights;

[0043] Obtain the landing time and airport capacity parameters of a single flight in the real landing flight list. The airport capacity parameters include the maximum number of landing flights allowed by the airport within the capacity time.

[0044] Starting from the landing time of the individual flight, the number of actual landing flights within the specified capacity time is viewed backward from the list of actual landing flights.

[0045] Based on the airport capacity parameters and the actual number of landing flights within the capacity time, determine whether there are overcapacity flights. If so, obtain the flight information of the overcapacity flights from the list of actual landing flights.

[0046] Iterate through the list of suggested flight delay times to check if there are any suggested flight times for overcapacity flights. If so, the evaluation result is that the suggested delay time is correct; otherwise, the evaluation result is that the suggested delay time is missing.

[0047] This continues until the evaluation of missing time recommendations for all individual flights in the actual landing flight list is completed.

[0048] Furthermore, the collected ranking and evaluation data includes collected route suggestion data, specifically:

[0049] Based on the arrival sorting data, route suggestion data is collected, which includes the number of suggested waypoints and the single projected distance of each suggested waypoint to the actual track.

[0050] Furthermore, the arrival ranking evaluation based on the ranking evaluation data includes route suggestion evaluation based on the route suggestion data, specifically as follows:

[0051] Based on the number of suggested waypoints and the individual projected distance, calculate the average deviation distance of all suggested waypoints from the actual track;

[0052] Calculate the root mean square of the deviation distance based on the average deviation distance;

[0053] The route similarity is calculated based on the average deviation distance and the root mean square deviation distance, and the route suggestion evaluation is performed based on the route similarity.

[0054] The beneficial effects of this invention are reflected in the following aspects: based on the collected sorting evaluation data, sorting sequence evaluation, time suggestion evaluation, and route suggestion evaluation are carried out respectively, which effectively evaluates the accuracy of the arrival sorting system at each key stage. Based on the operational data, the performance of the arrival sorting system during operation is objectively evaluated, which is conducive to the decision-making analysis of the effectiveness and reliability of the arrival sorting system by air traffic controllers based on data indicators, and meets the requirements of air traffic controllers for arrival flights at each stage. Attached Figure Description

[0055] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0056] Figure 1 A flowchart of an arrival sorting and evaluation method provided in an embodiment of the present invention;

[0057] Figure 2A schematic diagram illustrating the data collection timing for an inbound sorting and evaluation method provided in an embodiment of the present invention;

[0058] Figure 3 This invention provides a flowchart of the sorting sequence data acquisition process for an inbound sorting evaluation method.

[0059] Figure 4 A flowchart of the sorting sequence evaluation method for port entry sorting provided in an embodiment of the present invention;

[0060] Figure 5 This is a flowchart of the time-suggested data collection process for an inbound sorting evaluation method provided in an embodiment of the present invention;

[0061] Figure 6 This invention provides a flowchart of an error time suggestion evaluation method for an inbound sorting evaluation method.

[0062] Figure 7 A flowchart for the missing time suggestion evaluation method of the port arrival ranking evaluation method provided in this embodiment of the invention;

[0063] Figure 8 This is a schematic diagram of the deviation distance difference in an inbound sorting evaluation method provided by an embodiment of the present invention;

[0064] Figure 9 This is a schematic diagram comparing the track stability of an arrival sorting evaluation method provided in an embodiment of the present invention. Detailed Implementation

[0065] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. These embodiments are merely illustrative of the technical solution of the present invention and are therefore intended to limit the scope of protection of the present invention.

[0066] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application should have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0067] like Figure 1 As shown, an arrival sorting evaluation method is provided. The evaluation method is based on the arrival sorting system of the air traffic control terminal, and the steps include:

[0068] S1: Obtain port arrival sorting data, which is generated by the port arrival sorting system;

[0069] S2: Collect sorting evaluation data based on the port entry sorting data. The sorting evaluation data includes sorting sequence data, time suggestion data, and route suggestion data.

[0070] S3: Perform a multi-dimensional port arrival ranking evaluation based on the ranking evaluation data. The multi-dimensional port arrival ranking evaluation includes ranking sequence evaluation, time suggestion evaluation, and route suggestion evaluation.

[0071] S4: Output the multi-dimensional port entry ranking evaluation results.

[0072] This embodiment evaluates arrival sequencing based on three dimensions: sequencing sequence, timing recommendations, and route recommendations. It should be noted that the order in which sequencing sequence, timing recommendations, and route recommendations are evaluated can be set by air traffic controllers according to timing requirements and is not limited here.

[0073] Firstly, the sorting sequence evaluation refers to evaluating the accuracy of the inbound sorting sequence generated by the inbound sorting system, specifically including:

[0074] S201: Collect sorted sequence data;

[0075] Specifically, the timing of the expected sorting sequence collection is obtained from the pre-set data on the air traffic control terminal. The specific time points and quantities of the expected sorting sequence collection can be set by air traffic controllers based on actual evaluation criteria. Figure 2 As shown, this includes, but is not limited to, the timing of the initial sorting of flights, the timing of flights entering the corridor entrance point X kilometers in advance, the timing of flights entering the approach airspace or being controlled by the approach sector, the timing of flights entering the trilateral area, and the timing of flights aligning with the runway.

[0076] After the port arrival sorting system is started, sorting sequence data is continuously collected, such as... Figure 3 As shown, the process includes: traversing all arriving flights; when any arriving flight meets the set collection timing for any predicted sorting sequence, the arriving flight meeting the predicted sorting sequence collection timing is identified as the center flight, and the number of sampled flights Q preset by the air traffic control terminal is obtained. Using this center flight as the sampling center, the predicted sorting sequence generated by the arriving flight system based on the center flight is collected according to the arriving flight sorting data and the number of sampled flights Q, and the predicted sorting sequence is saved to the predicted sorting sequence pool. Preferably, the predicted sorting sequence includes the center flight and the flights preceding the center flight. The sections are expected to include inbound flights and flights following the central terminal. The system predicts arrival flights for each segment. For example, when flight 4 meets the timing requirement for any predicted sorting sequence, flight 4 is designated as the center flight. In this case, the predicted landing order generated by the arrival sorting system based on all arriving flights is as follows:

[0077] (flight3→flight1→flight2→flight4→flight6→flight5→flight7→flight10→flight9)

[0078] If the number of sampled flights Q preset in the air traffic control terminal is 5, then the expected inbound flights before and after flight 4 at the data collection center, i.e., the expected sorting sequence, is as follows:

[0079] (flight1→flight2→flight4→flight6→flight5)

[0080] Furthermore, in the sorting sequence evaluation stage, the collection of sorting sequence data also includes: when there are actual landing flights, obtaining a list L of actual landing flights based on the arrival sorting data. A And add the actual landing information of the flights to the list of actual landing flights L. A This is to facilitate subsequent evaluation of the sorted sequence.

[0081] It should be noted that the collection cycle of sorted sequence data can be set by the controller, for example, it can be set to collect once per minute or every half minute. When all inbound flights have been traversed, the next collection cycle will begin.

[0082] S301: Evaluate the sorted sequence based on the sorted sequence data;

[0083] Specifically, such as Figure 4 As shown, data is retrieved from the expected sorting sequence pool, and all expected sorting sequences in the pool are grouped according to their corresponding collection times. For example, all arriving flights collected at the time of "flight entering corridor point X kilometers" are grouped into one group, or all arriving flights collected at the time of "flight entering approach airspace or being controlled by approach sector" are grouped into another group, and so on.

[0084] Iterate through all groups and all sorted sequences within each group. Retrieve the expected sorted sequence of a single flight within the group from the expected sorted sequence pool to obtain the sorted sequence to be evaluated. Then, obtain the actual landing flight list L containing the sorted sequence to be evaluated. A The relative positions in the sequence are used to obtain the actual landing sequence.

[0085] Furthermore, a preset character conversion list is obtained, and the sorted sequence to be evaluated and the actual landing sequence are converted according to the character conversion list to obtain the sorted sequence string and the landing sequence string.

[0086] For ease of understanding, this embodiment uses a single flight, flight4, as an example. The expected ranking sequence for flight4 within the group is obtained from the expected ranking sequence pool, resulting in the ranking sequence to be evaluated as follows:

[0087] (flight1→flight2→flight4→flight6→flight5)

[0088] Assuming the actual list of flights landing at this airport L A for:

[0089] (flight1→flight2→flight3→flight4→flight5→flight6→flight7→flight9→flight10)

[0090] Obtain the sorted sequence to be evaluated from the list of actual landing flights L A The relative positions in the sequence yield the actual landing sequence as follows:

[0091] (flight1→flight2→flight4→flight5→flight6)

[0092] Get the preset character conversion list:

[0093] flight1→A

[0094] flight2→B

[0095] flight3→C

[0096] flight4→D

[0097] flight5→E

[0098] flight6→F

[0099] Based on the character conversion list, the sorted sequence to be evaluated and the actual landing sequence are converted into characters respectively, resulting in the sorted sequence string s1 = ABDFE and the landing sequence string s2 = ABDEF.

[0100] Furthermore, after obtaining the sorted sequence string s1 and the final sequence string s2, the Levenstein distance is used to evaluate the accuracy of the sorted sequence. Preferably, the Levenstein distance represents the minimum number of edits required to transform string s1 into string s2, including operations such as replacement, insertion, and deletion. The formula for calculating the Levenstein distance is:

[0101]

[0102] Based on the sorted sequence string s1 and the landing sequence string s2, calculate the Levinstein similarity between a single flight within the group and the actual landing sequence. The calculation formula is as follows:

[0103]

[0104] In the formula, lev s1,s2 Let be the Levenstein distance between the strings, and len(s1) and len(s2) be the lengths of the sorted sequence string s1 and the unsorted sequence string s2, respectively.

[0105] Furthermore, the Lewinstein similarity of each flight within each group is calculated sequentially. Based on the Lewinstein similarity of all individual flights within each group, the average accuracy of the sorted sequence for each group is calculated using the following formula:

[0106]

[0107] In the formula, p is the average accuracy rate, and N is the mean accuracy rate. p p represents the total number of flights within this group. i Let be the Levinstein similarity of the i-th flight. A higher average accuracy of the sorted sequence indicates higher accuracy of the arrival sorting sequence generated by the arrival sorting system.

[0108] Based on the calculated average accuracy of the sorted sequence for each group, the sorting evaluation is performed on each group separately until the sorting evaluation of all groups is completed.

[0109] Secondly, the time suggestion evaluation refers to evaluating the accuracy of the time delay suggestions generated by the port arrival sequencing system, that is, evaluating whether a delay suggestion was not generated when it should have been generated, or whether a delay suggestion was generated when it should not have been generated. Specifically, this includes:

[0110] S202: Suggested data collection time;

[0111] Specifically, after the arrival sorting system is started, it continuously collects suggested time data, such as... Figure 5 As shown, this includes: when the arrival sorting system prompts for a delay time suggestion, obtaining a list of flight delay time suggestions L based on the arrival sorting data. s And add the delay time suggestion to the flight delay time suggestion list L s middle.

[0112] Furthermore, in the time suggestion evaluation phase, the collection of time suggestion data also includes: when there are actual landing flights, obtaining a list of actual landing flights L based on the arrival sorting data. A Simultaneously, it records the current airport capacity parameters and adds the actual landing flight information to the list of real landing flights. A middle.

[0113] S302: Evaluate the time recommendation based on the time recommendation data;

[0114] Specifically, the time recommendations generated by the arrival sequencing system are periodically evaluated. The specific evaluation cycle can be set by air traffic controllers based on actual evaluation needs, generally once daily at midnight. Time recommendation evaluation includes erroneous time recommendation evaluation and missing time recommendation evaluation. Erroneous time recommendation evaluation refers to situations where a delay time recommendation should not have been generated but was generated by the arrival sequencing system in the air traffic control terminal. Missing time recommendation evaluation refers to situations where a delay time recommendation should have been generated but was not generated by the arrival sequencing system.

[0115] Furthermore, such as Figure 6 As shown, during the error time suggestion evaluation phase, the flight delay time suggestion list L is traversed. s If there is a flight delay, please refer to the suggested list L. s If the list is not empty and contains flight delay suggestions, then query the list of actual arriving flights (L). A Get the list of delayed flights that actually landed. A The relative position of the delayed flight is determined, and the airport capacity parameters at the time of landing are obtained based on the arrival sorting data. The airport capacity parameters include the maximum number of flights that the airport is allowed to land within the capacity time.

[0116] If the airport capacity parameter is N, which is the maximum number of landing flights allowed at this airport within a T-minute time period, then the delayed flights are listed in the actual landing flight list L. A Starting from the relative position in the data, in the actual flight list L A The system checks the number of actual landing flights within T minutes. If the number of actual landing flights within T minutes exceeds the maximum number of landing flights N, the evaluation result is that the delay time suggestion is correct; otherwise, the evaluation result is that the delay time suggestion is incorrect. This continues until the flight delay time suggestion list L is reached. s Traversal complete, end error time suggestion evaluation.

[0117] Furthermore, such as Figure 7 As shown, during the missing time suggestion evaluation phase, the actual landing flight list L is traversed. A Get the actual list of landing flights L A The landing time of a single flight and the airport capacity parameters at the time of landing. If the airport capacity parameter is N, which is the maximum number of flights allowed to land at this airport within a T-minute time period, then the landing time of a single flight is taken as the starting point, and the actual number of landing flights in the list L is calculated. ALooking backwards through the list of actual landing flights within T minutes, if the number of actual landing flights does not exceed the maximum number of landing flights N, then there are no overcapacity flights, and the evaluation of the missing time for this single flight ends, moving on to the evaluation of the missing time for the next single flight. If the number of actual landing flights exceeds the maximum number of landing flights N, then there are overcapacity flights, and the evaluation should proceed from the list of actual landing flights L. A Obtain flight information for overcapacity flights.

[0118] Furthermore, iterate through the suggested list of flight delay times L s View the list of suggested flight delay times. s If there are flight time suggestions for flights exceeding airport capacity, the evaluation result is that the delay time suggestion is correct; otherwise, the evaluation result is that the delay time suggestion is missing, indicating that the arrival sorting system is missing delay time suggestions for flights exceeding airport capacity.

[0119] For the actual list of landing flights L A The above evaluation process is performed on each flight until the evaluation of missing time recommendations for all individual flights is completed.

[0120] Thirdly, the route suggestion evaluation uses a projection method to calculate the deviation distance difference, whereby the deviation distance difference refers to the shortest distance from the projected track point of the suggested route to the actual track. Figure 8 As shown. The deviation distance between the suggested routes generated by the arrival sequencing system and the actual flight paths of the flights is used as the evaluation criterion for similarity, thereby assessing the accuracy of the route suggestions from the arrival sequencing system. Specifically, this includes:

[0121] S203: Collect route suggestion data;

[0122] Specifically, route suggestion data is collected based on the port arrival sorting data. The route suggestion data includes the number of suggested waypoints and the single projected distance from each suggested waypoint to the actual track.

[0123] S303: Evaluate the route suggestion based on the route suggestion data;

[0124] Specifically, based on the number of suggested waypoints and the individual projected distance of each suggested waypoint to the actual track, the average deviation distance of all suggested waypoints from the actual track is calculated using the following formula:

[0125]

[0126] In the formula, N represents the average deviation distance. d To suggest the number of waypoints, N dThe purpose of -1 is to suggest that the last point of both the flight path and the actual flight path must be the runway touchdown point, so there is no offset. i Let be the single projected distance from the i-th suggested waypoint to the actual track.

[0127] Furthermore, the average deviation distance is insufficient to fully characterize the similarity between the suggested and actual tracks, such as... Figure 9 As shown, although the average projected distance of track L1 is smaller, its deviation fluctuation is greater and its stability is worse than that of track L2. In actual use at air traffic control sites, track L2 is generally considered to have better prediction performance. Therefore, this embodiment introduces the root mean square distance to evaluate the stability of the distance error, that is, the root mean square distance of the deviation is calculated based on the average deviation distance. The calculation formula is as follows:

[0128]

[0129] Based on the calculated average deviation distance The route similarity is calculated using the root mean square deviation (RMS) σ, to jointly assess the similarity between the proposed and actual routes. Assuming a weighting coefficient α, which can be set by air traffic controllers according to actual needs, typically 0.5, the route similarity calculation formula is as follows:

[0130]

[0131] As can be seen from the above formula, the larger the average deviation distance, the worse the similarity and the lower the quality of the route suggestion; conversely, the better the similarity, the better the quality of the route suggestion, thus realizing the evaluation of route suggestions based on route similarity.

[0132] Finally, the ranking sequence evaluation results, time suggestion evaluation results, and route suggestion evaluation results are output to the air traffic control terminal to provide reference data for air traffic controllers.

[0133] This invention obtains sorting sequence data from arrival sorting data generated by an arrival sorting system, calculates the average accuracy of the sorting sequence using Levenstein similarity, and then evaluates the accuracy of the arrival sorting sequence generated by the arrival sorting system. It also obtains time suggestion data from the arrival sorting data, analyzes whether there are erroneous or missing time suggestions based on the time suggestion data and airport capacity parameters, and evaluates the accuracy of the time suggestions generated by the arrival sorting system. Furthermore, it obtains route suggestion data from the sorting data, calculates the deviation distance between the suggested route and the actual route based on the route suggestion data, and introduces the root mean square deviation to characterize the stability of the deviation distance. Combining the deviation distance and the root mean square deviation, it calculates the similarity between the suggested route and the actual flight path, and evaluates the accuracy of the route suggestions generated by the arrival sorting system.

[0134] This invention evaluates the sorting sequence, time suggestion, and route suggestion based on the collected sorting evaluation data, effectively assessing the accuracy of the arrival sorting system at each key stage. Based on operational data, it objectively evaluates the performance of the arrival sorting system during operation, which helps air traffic controllers make decisions based on data indicators to analyze the effectiveness and reliability of the arrival sorting system, thus meeting the requirements of air traffic controllers for arrival flights at each stage.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for evaluating port arrival order, characterized in that, The evaluation method is based on the arrival sequencing system of the air traffic control terminal, and the steps include: Obtain port arrival sorting data, which is generated by the port arrival sorting system; Based on the port entry sorting data, sorting evaluation data is collected, which includes sorting sequence data, time suggestion data, and route suggestion data. Based on the ranking evaluation data, a multi-dimensional arrival ranking evaluation is performed, which includes ranking sequence evaluation, time suggestion evaluation, and route suggestion evaluation. Output the multi-dimensional port entry ranking evaluation results; The collected sorting and evaluation data includes collected sorting sequence data, specifically: Obtain the timing for collecting the pre-set expected sorting sequence from the air traffic control terminal; When any arriving flight meets the expected sorting sequence collection timing, the expected arrival sequence generated by the arrival sorting system at this time is collected according to the arrival sorting data, and the expected sorting sequence is saved to the expected sorting sequence data pool; Multi-dimensional arrival ranking evaluation is performed based on the ranking evaluation data, including ranking sequence evaluation based on the ranking sequence data, specifically as follows: All expected sorting sequences in the expected sorting sequence data pool are grouped according to their corresponding collection timing, and all groups are traversed. Iterate through all sorted sequences within each group; Obtain the expected sorting sequence of a single flight within the group from the expected sorting sequence pool to obtain the sorting sequence to be evaluated; Obtain the relative position of the sorted sequence to be evaluated in the list of actual landing flights to obtain the actual landing sequence; Obtain a preset character conversion list, and convert the sorting sequence to be evaluated and the actual landing sequence according to the character conversion list to obtain the sorting sequence string and the landing sequence string respectively; Based on the sorted sequence string and the landing sequence string, calculate the Levinstein similarity between a single flight within the group and the actual landing sequence; The average accuracy of the sorted sequence for each group is calculated based on the Levinstein similarity of all individual flights within each group. The ranking of each group is evaluated based on the average accuracy of the ranking sequence until the ranking evaluation of all groups is completed. The port arrival ranking evaluation is performed based on the ranking evaluation data, including the error time suggestion evaluation based on the time suggestion data, specifically as follows: Iterate through the list of suggested flight delay times; If there is a flight delay suggestion, obtain the relative position of the delayed flight in the list of actual landing flights; Based on the arrival sorting data, the airport capacity parameters at the time of landing of the delayed flight are obtained. The airport capacity parameters include the maximum number of flights that the airport is allowed to land within the capacity time. Starting from the relative position of the delayed flight in the list of actual landing flights, look backward in the list of actual landing flights to see the number of actual landing flights within the capacity time. Determine whether the actual number of landing flights within the specified capacity time exceeds the maximum number of landing flights. If it does, the evaluation result is that the delay time suggestion is correct; otherwise, the evaluation result is that the delay time suggestion is incorrect. The evaluation of port arrival ranking based on the aforementioned ranking evaluation data also includes the evaluation of missing time suggestions based on the aforementioned time suggestion data, specifically as follows: Iterate through the list of actual landing flights; Obtain the landing time and airport capacity parameters of a single flight in the real landing flight list. The airport capacity parameters include the maximum number of landing flights allowed by the airport within the capacity time. Starting from the landing time of the individual flight, the number of actual landing flights within the specified capacity time is viewed backward from the list of actual landing flights. Based on the airport capacity parameters and the actual number of landing flights within the capacity time, it is determined whether there are overcapacity flights. If it exists, then obtain the flight information of the overcapacity flight from the list of actual landing flights; Iterate through the list of suggested flight delay times to check if there are any suggested flight times for overcapacity flights. If so, the evaluation result is that the suggested delay time is correct; otherwise, the evaluation result is that the suggested delay time is missing. This continues until the evaluation of missing time recommendations for all individual flights in the actual landing flight list is completed.

2. The port arrival ranking evaluation method according to claim 1, characterized in that, When there are actual landing flights, obtain the list of actual landing flights based on the arrival sorting data, and add the flight information of the actual landing flights to the list of actual landing flights.

3. The port arrival ranking evaluation method according to claim 2, characterized in that, When any arriving flight meets the expected sorting sequence collection timing, the expected arrival sequence generated by the arrival sorting system at this time is collected based on the arrival sorting data, and the expected sorting sequence is saved to the expected sorting sequence data pool. Specifically: When any inbound flight meets the expected sorting sequence collection timing, the inbound flight that meets the expected sorting sequence collection timing will be determined as the center flight; Obtain the number of sampled flights pre-set in the air traffic control terminal; Based on the arrival sorting data and the number of sampled flights, the expected sorting sequence generated by the arrival sorting system based on the central flight is collected, and the expected sorting sequence is saved to the expected sorting sequence pool; The expected sorting sequence includes the central flight and (Q-1) / 2 expected inbound flights before and after the central flight, where Q is the number of sampled flights.

4. The port arrival ranking evaluation method according to claim 1, characterized in that, The collected sorting and evaluation data includes suggested data on collection time, specifically: When the arrival sorting system prompts for a delay time suggestion, a flight delay time suggestion list is obtained based on the arrival sorting data, and the delay time suggestion is added to the flight delay time suggestion list; When there are actual landing flights, obtain the list of actual landing flights based on the arrival sorting data, and add the flight information of the actual landing flights to the list of actual landing flights.

5. The port arrival ranking evaluation method according to claim 1, characterized in that, The collected, sorted, and evaluated data includes route suggestion data, specifically: Based on the arrival sorting data, route suggestion data is collected, which includes the number of suggested waypoints and the single projected distance of each suggested waypoint to the actual track.

6. The port arrival ranking evaluation method according to claim 5, characterized in that, The arrival ranking evaluation based on the ranking evaluation data includes route suggestion evaluation based on the route suggestion data, specifically as follows: Based on the number of suggested waypoints and the individual projected distance, calculate the average deviation distance of all suggested waypoints from the actual track; Calculate the root mean square of the deviation distance based on the average deviation distance; The route similarity is calculated based on the average deviation distance and the root mean square deviation distance, and the route suggestion evaluation is performed based on the route similarity.

Images