Airport delay level evaluation method based on space-time correlation

Abstract

The invention discloses an airport delay level evaluation method based on space-time correlation. The method comprises the following steps: constructing an associated airport set according to associated airport index analysis; analyzing the time relationship of the delay influence of the associated airport on the target airport by using the grey correlation degree; establishing an airport delay evaluation index system; an entropy weight method is adopted to carry out weight assignment on each airport delay evaluation index weight; and carrying out quantitative grading on the airport delay index based on a one-dimensional kernel density clustering algorithm so as to realize airport delay level evaluation. According to the method, the airport flight delay level can be objectively measured according to the operation characteristics of the target airport and the associated airport.

Description

Technical Field

[0001] This invention belongs to the technical field of civil airport flight delay assessment, specifically involving an airport delay level assessment method based on spatiotemporal correlation. Background Technology

[0002] With the rapid development of the civil aviation transportation industry, flight delays at airports have become increasingly serious, severely impacting airport operations and passenger travel. In 2018, over 20% of flights in the United States experienced delays, and the proportion of delayed domestic flights in China was close to 20%. Accurately assessing airport delay levels helps airports and air traffic management departments to scientifically adopt corresponding strategies to mitigate the negative impacts of flight delays.

[0003] Currently, research on airport delay assessment mainly focuses on two aspects: the construction of delay assessment indicator systems and the calculation methods for indicator weights. Regarding the construction of delay assessment indicator systems, most studies are based on the operational status of the target airport to extract indicators and construct a delay assessment indicator system based on the airport itself, neglecting the temporal and spatial impact of related airports. As for the calculation methods for indicator weights, there are subjective weighting, objective weighting, and a combination of subjective and objective weighting methods. However, in the process of shifting from quantitative indicators to graded assessments, most studies adopt subjective grading methods, with very few studies utilizing objective methods to calculate the boundaries of airport delay grades. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, the present invention aims to provide an airport delay level assessment method based on spatiotemporal correlation, thereby solving the problems in the construction of existing airport delay level assessment methods that do not consider the operational status of associated airports and that assessment indicators rely on subjective determination; the method of the present invention can objectively measure the airport flight delay level based on the operational characteristics of the target airport and associated airports.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] The present invention provides a method for assessing airport delay levels based on spatiotemporal correlation, comprising the following steps:

[0007] 1) Construct a set of associated airports based on the analysis of associated airport indices;

[0008] 2) Use grey relational analysis to analyze the temporal relationship between the impact of related airports on the delays of the target airport;

[0009] 3) Establish an airport delay assessment indicator system;

[0010] 4) The entropy weight method is used to assign weights to the delay assessment indicators for each airport;

[0011] 5) The airport delay index is quantitatively graded based on a one-dimensional kernel density clustering algorithm to achieve airport delay level assessment.

[0012] Further, step 1) specifically includes:

[0013] 11) Select airports that have direct flights to the target airport within the analysis period, calculate the route distances, and count the number of flights.

[0014] 12) Perform MAX-MIN standardization on route distance and number of flights. For the associated airport index matrix T composed of route distance and number of flights, for the m-th index and the n-th airport, the standardized index x' nm for:

[0015]

[0016] Where max(T) m ) represents the maximum value of the m-th indicator in the indicator matrix T, min(T) m Let x be the minimum value of the m-th indicator in the indicator matrix T. nm The value is located at row n and column m of the index matrix T;

[0017] 13) Based on the standardized index x' nm Calculate the correlation index R of related airports p R p =N p -D p , where N p To determine the number of flights operating at the p-th airport after standardization, D p This represents the standardized flight path distance to the p-th airport.

[0018] 14) Select the top N airports as the associated airport set of the target airport based on the associated airport index, where N is determined by the number of takeoffs and landings of the target airport and is an integer between 4 and 10.

[0019] Furthermore, step 2) specifically includes:

[0020] 21) Treat the target airport, related airports and related data as a gray system, determine the system attributes and internal factors of the gray system. The system attributes refer to the delay sequence of the target airport, that is, the number of delayed flights at the current airport; the internal factors refer to the number of delayed flights at the related airports in the preceding statistical time period.

[0021] 22) Construct the correlation matrix; the delay parent sequence is the target airport flight delay number matrix D for each unit time period, as follows:

[0022] D = {y1, y2, ..., y} i,...,y q}

[0023] The matrix P of the number of flight delays at associated airports in the preceding statistical time period is as follows:

[0024] P = {p1, p2, ..., p} j ,...,p k}

[0025] Among them, y i p represents the number of flight delays within the i-th preorder time period. j The matrix representing the number of flight delays at associated airports in the j-th statistical time period is as follows:

[0026] p j ={p j1 ,p j2 ,...,p ji ,...,p q}

[0027] Where, p ji This represents the number of delayed flights at the associated airport within the ij-th time unit.

[0028] 23) The correlation matrix is ​​processed to obtain the delay parent sequence matrix D' and the associated airport time feature sequence matrix p. j 'They are respectively:

[0029]

[0030] Where y1≠0,p j1 ≠0, 1≤i≤q, 1≤j≤k, where i and j are integers, q is the number of statistical samples, and k is the maximum number of preceding statistical time periods;

[0031] 24) Calculate the correlation coefficient between corresponding values; calculate the delay parent sequence matrix D' and the associated airport time feature sequence matrix p. j 'Difference between corresponding values ​​Δ j i, as follows:

[0032]

[0033] Then calculate the correlation coefficient γ between each corresponding value. ji ,as follows:

[0034]

[0035] Where ρ is the resolution coefficient, usually taken as 0.5; Δ m ax is Δ j The maximum value among the values ​​of i, Δ m in is Δ ji takes the minimum value among the values;

[0036] 25) Select the number of flight delays corresponding to the preceding statistical period with a large correlation coefficient as the impact indicator of the associated airport on the target airport;

[0037] 26) Repeat steps 22)-25) to determine the preceding statistical time period in which each associated airport has the greatest impact on the target airport, thereby determining the impact indicators of all associated airports on the target airport.

[0038] Furthermore, step 3) specifically includes:

[0039] Airport delay assessment is affected by airport capacity factors, related airport factors, and target airport operation factors. The airport delay assessment index system includes airport capacity indicators, related airport indicators, and target airport operation indicators.

[0040] Airport capacity indicators refer to indicators that represent the airport's carrying capacity and traffic levels, including the number of delayed arriving flights, the number of delayed departing flights, the number of scheduled flights, and the airport capacity ratio;

[0041] The associated airport index refers to the indicators that represent the flight delay situation of associated airports, including the number of flight delays and the flight delay time of each associated airport in the corresponding time period.

[0042] Target airport operational indicators refer to indicators that represent the operational status of the target airport, including the number of delayed passengers, the number of waiting passengers, and the security check delay time.

[0043] Furthermore, step 4) specifically includes:

[0044] The impact of each indicator on airport delays is differentiated, and the entropy weight method is used to assign weights to each airport delay assessment indicator.

[0045] 41) Calculate the normalized nonnegative matrix;

[0046] For M airport delay assessment indicators and R sample data, the normalized result is a non-negative matrix as follows:

[0047]

[0048] Where X is a non-negative matrix, x RM In this context, R represents the sample number, and M represents the airport number.

[0049] 42) Calculate the probability matrix based on the nonnegative matrix. For the element x in the g-th row and r-th column of the nonnegative matrix... gr Its probability p gr for:

[0050]

[0051] Among them, N k Summing for K sample data;

[0052] 43) Calculate the information entropy e of the m-th index. m and information utility value d m And finally, the weight ω of the computer field delay assessment index. m ,in:

[0053]

[0054] dm = 1 - e m

[0055]

[0056] Furthermore, step 5) specifically includes:

[0057] 51) Collect historical operational data of the target airport, including different delay levels during the airport's operating hours;

[0058] 52) Based on the weights of the airport delay assessment indicators obtained in step 4), calculate the airport delay assessment index. For T sets of sample data, there are T airport delay assessment indices, resulting in a one-dimensional array with a sample size of T.

[0059] 53) Calculate the kernel density estimation function for the t-th airport delay assessment index, as follows:

[0060]

[0061] In the formula, k(x) is the delay index, x is a point of the kernel density, and x(t) is the kernel density point of the t-th airport;

[0062]

[0063] Where K(x) is the Gaussian kernel function, α is the sensitivity factor (0 ≤ α ≤ 1), ω represents the bandwidth parameter, S is the scaling parameter of the Gaussian kernel function, and f(x) (k) f(x) is the kernel density estimate for the k-th sample point. (t) ) represents the kernel density estimate for the t-th sample point;

[0064] 54) The elements in array T are superimposed according to the kernel density estimation function, and the graded thresholds are used to achieve the airport delay level assessment.

[0065] This invention targets airport flight operations, constructs an airport delay assessment index system based on spatiotemporal relationships and airport operational characteristics, and uses the entropy weight method to assign weights to each assessment index to realize the calculation of airport delay indicators; based on historical data, it uses a one-dimensional kernel density clustering algorithm to classify the quantitative assessment results of airport delays, calculates the airport delay level threshold, and completes the airport delay level assessment.

[0066] The beneficial effects of this invention are:

[0067] This invention accurately establishes an airport delay assessment system and achieves precise evaluation. By constructing a set of associated airports and utilizing grey relational analysis to examine spatiotemporal relationships, it improves the assessment indicator system. It employs the entropy weight method to objectively determine indicator weights, avoiding subjective interference. Using a one-dimensional kernel density algorithm, it quantifies and grades airport delay indices based on historical data, obtaining scientific grading thresholds. This invention's method makes the assessment system comprehensive and accurate, providing strong support for airport and air traffic control decision-making and effectively mitigating the negative impact of flight delays. Attached Figure Description

[0068] Figure 1 This is a flowchart illustrating the principle of the method of the present invention. Detailed Implementation

[0069] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present invention.

[0070] Reference Figure 1 As shown, the present invention provides a method for assessing airport delay levels based on spatiotemporal correlation, comprising the following steps:

[0071] 1) Construct a set of associated airports based on the analysis of associated airport indices; specifically including:

[0072] 11) Select airports that have direct flights to the target airport within the analysis period, calculate the route distances, and count the number of flights.

[0073] 12) Perform MAX-MIN standardization on route distance and number of flights. For the associated airport index matrix T composed of route distance and number of flights, for the m-th index and the n-th airport, the standardized index x' nm for:

[0074]

[0075] Where max(T) m ) represents the maximum value of the m-th indicator in the indicator matrix T, min(T) m Let x be the minimum value of the m-th indicator in the indicator matrix T. nm The value is located at row n and column m of the index matrix T;

[0076] 13) Based on the standardized index x' nm Calculate the correlation index R of related airports p R p =N p -D p , where N p To determine the number of flights operating at the p-th airport after standardization, D p This represents the standardized flight path distance to the p-th airport.

[0077] 14) Select the top N airports as the associated airport set of the target airport based on the associated airport index, where N is determined by the number of takeoffs and landings of the target airport and is an integer between 4 and 10.

[0078] 2) Utilize grey relational analysis to analyze the temporal relationship between the impact of related airports on the target airport's delays; specifically including:

[0079] 21) Treat the target airport, related airports and related data as a gray system, determine the system attributes and internal factors of the gray system. The system attributes refer to the delay sequence of the target airport, that is, the number of delayed flights at the current airport; the internal factors refer to the number of delayed flights at the related airports in the preceding statistical time period.

[0080] 22) Construct the correlation matrix; the delay parent sequence is the target airport flight delay number matrix D for each unit time period, as follows:

[0081] D = {y1, y2, ..., y} i ,...,y q}

[0082] The matrix P of the number of flight delays at associated airports in the preceding statistical time period is as follows:

[0083] P = {p1, p2, ..., p} j ,...,p k}

[0084] Among them, y i p represents the number of flight delays within the i-th preorder time period. j The matrix representing the number of flight delays at associated airports in the j-th statistical time period is as follows:

[0085] p j ={p j1 ,p j2 ,...,p ji ,...,p q}

[0086] Where, p jiThis represents the number of delayed flights at the associated airport within the ij-th time unit.

[0087] 23) The correlation matrix is ​​processed to obtain the delay parent sequence matrix D' and the associated airport time feature sequence matrix p. j 'They are respectively:

[0088]

[0089] Where y1≠0,p j1 ≠0, 1≤i≤q, 1≤j≤k, where i and j are integers, q is the number of statistical samples, and k is the maximum number of preceding statistical time periods;

[0090] 24) Calculate the correlation coefficient between corresponding values; calculate the delay parent sequence matrix D' and the associated airport time feature sequence matrix p. j 'Difference between corresponding values ​​Δ j i, as follows:

[0091]

[0092] Then calculate the correlation coefficient γ between each corresponding value. ji ,as follows:

[0093]

[0094] Where ρ is the resolution coefficient, usually taken as 0.5; Δ m ax is Δ j The maximum value among the values ​​of i, Δ m in is Δ j i takes the minimum value among the values;

[0095] 25) Select the number of flight delays corresponding to the preceding statistical period with a large correlation coefficient as the impact indicator of the associated airport on the target airport;

[0096] 26) Repeat steps 22)-25) to determine the preceding statistical time period in which each associated airport has the greatest impact on the target airport, thereby determining the impact indicators of all associated airports on the target airport.

[0097] 3) Establish an airport delay assessment indicator system; specifically including:

[0098] Airport delay assessment is affected by airport capacity factors, related airport factors, and target airport operation factors. The airport delay assessment index system includes airport capacity indicators, related airport indicators, and target airport operation indicators.

[0099] Airport capacity indicators refer to indicators that represent the airport's carrying capacity and traffic levels, including the number of delayed arriving flights, the number of delayed departing flights, the number of scheduled flights, and the airport capacity ratio;

[0100] The associated airport index refers to the indicators that represent the flight delay situation of associated airports, including the number of flight delays and the flight delay time of each associated airport in the corresponding time period.

[0101] Target airport operational indicators refer to indicators that represent the operational status of the target airport, including the number of delayed passengers, the number of waiting passengers, and the security check delay time.

[0102] 4) The entropy weight method is used to assign weights to the delay assessment indicators for each airport; specifically including:

[0103] The impact of each indicator on airport delays is differentiated, and the entropy weight method is used to assign weights to each airport delay assessment indicator.

[0104] 41) Calculate the normalized nonnegative matrix;

[0105] For M airport delay assessment indicators and R sample data, the normalized result is a non-negative matrix as follows:

[0106]

[0107] Where X is a non-negative matrix, x RM In this context, R represents the sample number, and M represents the airport number.

[0108] 42) Calculate the probability matrix based on the nonnegative matrix. For the element x in the g-th row and r-th column of the nonnegative matrix... gr Its probability p gr for:

[0109]

[0110] Among them, N k Summing for K sample data;

[0111] 43) Calculate the information entropy e of the m-th index. m and information utility value d m And finally, the weight ω of the computer field delay assessment index. m ,in:

[0112]

[0113] dm = 1 - e m

[0114]

[0115] 5) Quantitatively classify airport delay indices based on a one-dimensional kernel density clustering algorithm to achieve airport delay level assessment; specifically including:

[0116] 51) Collect historical operational data of the target airport, including different delay levels during the airport's operating hours;

[0117] 52) Based on the weights of the airport delay assessment indicators obtained in step 4), calculate the airport delay assessment index. For T sets of sample data, there are T airport delay assessment indices, resulting in a one-dimensional array with a sample size of T.

[0118] 53) Calculate the kernel density estimation function for the t-th airport delay assessment index, as follows:

[0119]

[0120] In the formula, k(x) is the delay index, x is a point of the kernel density, and x(t) is the kernel density point of the t-th airport;

[0121]

[0122] Where K(x) is the Gaussian kernel function, α is the sensitivity factor (0 ≤ α ≤ 1), ω represents the bandwidth parameter, S is the scaling parameter of the Gaussian kernel function, and f(x) (k) f(x) is the kernel density estimate for the k-th sample point. (t) ) represents the kernel density estimate for the t-th sample point;

[0123] 54) The elements in array T are superimposed according to the kernel density estimation function, and the graded thresholds are used to achieve the airport delay level assessment.

[0124] This invention has many specific applications. The above description is only a preferred embodiment of this invention. It should be noted that for those skilled in the art, several improvements can be made without departing from the principle of this invention, and these improvements should also be considered within the scope of protection of this invention.

Claims

1. A method for assessing airport delay levels based on spatiotemporal correlation, characterized in that, The steps are as follows: 1) Construct a set of associated airports based on the analysis of associated airport indices; 2) Use grey relational analysis to analyze the temporal relationship between the impact of related airports on the delays of the target airport; 3) Establish an airport delay assessment indicator system; 4) The entropy weight method is used to assign weights to the delay assessment indicators for each airport; 5) The airport delay index is quantitatively graded based on a one-dimensional kernel density clustering algorithm to achieve airport delay level assessment.

2. The airport delay level assessment method based on spatiotemporal correlation according to claim 1, characterized in that, Step 1) specifically includes: 11) Select airports that have direct flights to the target airport within the analysis period, calculate the route distances, and count the number of flights. 12) Perform MAX-MIN standardization on route distance and number of flights. For the associated airport index matrix T composed of route distance and number of flights, for the m-th index and the n-th airport, the standardized index x' nm for: Where max(T) m ) represents the maximum value of the m-th indicator in the indicator matrix T, min(T) m Let x be the minimum value of the m-th indicator in the indicator matrix T. nm The value is located at row n and column m of the index matrix T; 13) Based on the standardized index x' nm Calculate the correlation index R of related airports p R p =N p -D p , where N p To determine the number of flights operating at the p-th airport after standardization, D p This represents the standardized flight path distance to the p-th airport. 14) Select the top N airports as the associated airport set of the target airport based on the associated airport index, where N is determined by the number of takeoffs and landings of the target airport.

3. The airport delay level assessment method based on spatiotemporal correlation according to claim 2, characterized in that, Step 2) specifically includes: 21) Treat the target airport, related airports and related data as a gray system, determine the system attributes and internal factors of the gray system. The system attributes refer to the delay sequence of the target airport, that is, the number of delayed flights at the current airport; the internal factors refer to the number of delayed flights at the related airports in the preceding statistical time period. 22) Construct the correlation matrix; the delay parent sequence is the target airport flight delay number matrix D for each unit time period, as follows: D={y1,y2,...,y i ,...,y q } The matrix P of the number of flight delays at associated airports in the preceding statistical time period is as follows: P={p1,p2,...,p j ,...,p k } Among them, y i p represents the number of flight delays within the i-th preorder time period. j The matrix representing the number of flight delays at associated airports in the j-th statistical time period is as follows: p j ={p j1 ,p j2 ,...,p ji ,...,p q } Where, p ji This represents the number of delayed flights at the associated airport within the ij-th time unit. 23) The correlation matrix is ​​processed to obtain the delay parent sequence matrix D' and the associated airport time feature sequence matrix p. j 'They are respectively: Where y1≠0,p j1 ≠0, 1≤i≤q, 1≤j≤k, where i and j are integers, q is the number of statistical samples, and k is the maximum number of preceding statistical time periods; 24) Calculate the correlation coefficient between corresponding values; calculate the delay parent sequence matrix D' and the associated airport time feature sequence matrix p. j 'Difference between corresponding values ​​Δ j i, as follows: Then calculate the correlation coefficient γ between each corresponding value. ji ,as follows: Where ρ is the resolution coefficient; Δ m ax is Δ j The maximum value among the values ​​of i, Δ m in is Δ j i takes the minimum value among the values; 25) Select the number of flight delays corresponding to the preceding statistical period with a large correlation coefficient as the impact indicator of the associated airport on the target airport; 26) Repeat steps 22)-25) to determine the preceding statistical time period in which each associated airport has the greatest impact on the target airport, thereby determining the impact indicators of all associated airports on the target airport.

4. The airport delay level assessment method based on spatiotemporal correlation according to claim 1, characterized in that, Step 3) specifically includes: Airport delay assessment is affected by airport capacity factors, related airport factors, and target airport operation factors. The airport delay assessment index system includes airport capacity indicators, related airport indicators, and target airport operation indicators. Airport capacity indicators refer to indicators that represent the airport's carrying capacity and traffic levels, including the number of delayed arriving flights, the number of delayed departing flights, the number of scheduled flights, and the airport capacity ratio; The associated airport index refers to the indicators that represent the flight delay situation of associated airports, including the number of flight delays and the flight delay time of each associated airport in the corresponding time period. Target airport operational indicators refer to indicators that represent the operational status of the target airport, including the number of delayed passengers, the number of waiting passengers, and the security check delay time.

5. The airport delay level assessment method based on spatiotemporal correlation according to claim 1, characterized in that, Step 4) specifically includes: The impact of each indicator on airport delays is differentiated, and the entropy weight method is used to assign weights to each airport delay assessment indicator. 41) Calculate the normalized nonnegative matrix; For M airport delay assessment indicators and R sample data, the normalized result is a non-negative matrix as follows: Where X is a non-negative matrix, x RM In this context, R represents the sample number, and M represents the airport number. 42) Calculate the probability matrix based on the nonnegative matrix. For the element x in the g-th row and r-th column of the nonnegative matrix... gr Its probability p gr for: Where, N k Summing for K sample data; 43) Calculate the information entropy e of the m-th index. m and information utility value d m And finally, the weight ω of the computer field delay assessment index. m ,in: dm = 1-e m 6. The airport delay level assessment method based on spatiotemporal correlation according to claim 5, characterized in that, Step 5) specifically includes: 51) Collect historical operational data of the target airport, including different delay levels during the airport's operating hours; 52) Based on the weights of the airport delay assessment indicators obtained in step 4), calculate the airport delay assessment index. For T sets of sample data, there are T airport delay assessment indices, resulting in a one-dimensional array with a sample size of T. 53) Calculate the kernel density estimation function for the t-th airport delay assessment index, as follows: In the formula, k(x) is the delay index, x is a point of the kernel density, and x(t) is the kernel density point of the t-th airport; Where K(x) is the Gaussian kernel function, α is the sensitivity factor (0 ≤ α ≤ 1), ω represents the bandwidth parameter, S is the scaling parameter of the Gaussian kernel function, and f(x) (k) f(x) is the kernel density estimate for the k-th sample point. (t) ) represents the kernel density estimate for the t-th sample point; 54) The elements in array T are superimposed according to the kernel density estimation function, and the graded thresholds are used to achieve the airport delay level assessment.

Images