A method and computer device for predicting characteristic values ​​of target airports

Abstract

A method and computer device for predicting characteristic values ​​of a target airport. The scheme obtains historical data of reference airports of the same type as the target airport and basic data of cities in the area where the reference airports are located, trains X prediction models based on the historical data of the reference airports and the basic data of cities in the area where the reference airports are located, and the X prediction models are used to predict airport passenger and cargo throughputs based on the basic data of cities. A comprehensive model of the X prediction models is constructed, and basic data of target cities in the area where the target airports are located are obtained. The airport passenger and cargo throughputs of the target airport are predicted based on the basic data of the target cities using the X prediction models. The prediction results of the airport passenger and cargo throughputs of the target airport using the X prediction models are then substituted into the comprehensive model, and the output results of the comprehensive model are used as the predicted values ​​of the airport passenger and cargo throughputs of the target airport, thereby realizing the prediction method of the characteristic values ​​of the target airport.

Description

Technical Field

[0001] The present invention relates to the technical field of big data processing, and in particular to an airport traffic prediction method and computer equipment. Background Art

[0002] The characteristic values ​​of civil airports mainly include the number of passengers during peak hours, the number of passenger aircraft take-offs and landings, and the number of cargo aircraft take-offs and landings.

[0003] The civil aviation industry is an important strategic industry of the country, and airports are important public infrastructure and a new source of power for economic and social development. In recent years, the number, scale and service scope of my country's civil transport airports have continued to expand, and the types of services have continued to enrich. The construction of civil aviation infrastructure with appropriate scale, strong guarantee and people's satisfaction is an important manifestation of deepening the structural reform of the civil aviation supply side, an important support for realizing the strategy of building a strong civil aviation country, and an inherent requirement for promoting the high-quality development of transport airports in the new era and better meeting the people's needs for better air travel.

[0004] Prediction is the basis of decision-making. Only with correct prediction can scientific decision-making be made. The prediction of civil airport characteristic values ​​is based on the history and current situation of airport development, investigation and statistical data, scientific qualitative analysis and quantitative calculation, and the development law of air transportation is analyzed to make scientific inferences on the development trend of air transportation. It is the basis of airport planning and construction, and the basis for determining the scale of airport construction and the flight schedule. If the characteristic value prediction result is too large, it may lead to a large construction scale, increase construction investment, premature occupation of land, and waste of resources. At the same time, the airport will bear too much depreciation cost and unit passenger operating cost; if the characteristic value prediction is too small, it may lead to a small construction scale, mismatch of the scale of key facilities such as parking spaces, runways, and terminals, which will affect the normal flight, operation safety and passenger experience, so that the airport often faces the situation of "starting renovation and expansion when it is completed and put into operation", which will disrupt the normal construction and development rhythm of the airport. In order to make the planning and construction of airports better adapt to the development needs of the air transportation market, the industry's requirements for the scientificity and accuracy of airport characteristic value prediction are also constantly improving.

[0005] At present, the prediction of airport aviation characteristic values ​​is generally based on manual methods. In the manual prediction process, the relevant data on civil aviation development such as economy, population, tourism, import and export in the area where the airport is located are generally collected and counted first. For the characteristic value prediction made under the background of the reconstruction and expansion of existing airports, the historical data of the airport are also collected; on this basis, the trend extrapolation method, regression analysis method, exponential smoothing method, market share method, BP neural network method, OD route segmentation method, etc. are applied to realize the business volume prediction process. Among them, the trend extrapolation method, regression analysis method, exponential smoothing method, and market share method are more common mathematical methods. The BP neural network method is a basic method of machine learning for nonlinear fitting, which can also be classified as a common mathematical method in essence. The OD route segmentation method is to use the data trend of each route for prediction. The prediction process is relatively complicated and greatly affected by human factors. At the same time, manual prediction generally cannot consider the impact of other modes of transportation such as high-speed rail on civil aviation travel demand. Under the combined influence, the objectivity of the prediction results of civil airport characteristic values ​​is not strong. Summary of the invention

[0006] In view of this, an embodiment of the present invention provides a target airport characteristic value prediction method and a computer device to achieve the prediction of the characteristic value of the target airport.

[0007] To achieve the above objectives, the embodiments of the present invention provide the following technical solutions:

[0008] A method for predicting characteristic values ​​of a target airport, comprising:

[0009] Obtaining airport historical data and basic data of cities in the area where the airport is located, wherein the airport historical data is the airport passenger and cargo throughput;

[0010] Based on the airport historical data and the city basic data, create X prediction models, wherein the prediction models are used to predict the airport passenger and cargo throughput based on the city basic data;

[0011] Obtain historical data of a reference airport and basic data of cities in the area where the airport is located, wherein the reference airport is an airport of the same category as the target airport;

[0012] Constructing a comprehensive model of the X prediction models;

[0013] Training the comprehensive model based on the prediction results of the airport passenger and cargo throughput of the reference airport based on the basic city data of the reference airport and the airport passenger and cargo throughput;

[0014] Obtaining basic data of a target city in the area where the target airport is located, and using the X prediction models to predict the airport passenger and cargo throughput of the target airport based on the basic data of the target city;

[0015] Substitute the airport passenger and cargo throughputs of the target airport into the comprehensive model, and use the output result of the comprehensive model as the predicted airport passenger and cargo throughputs of the target airport.

[0016] Optionally, in the above target airport characteristic value prediction method, the X prediction models include:

[0017] Time series prediction model, regression analysis prediction model, market share prediction model, BP neural network prediction model and Bopuga formula prediction model.

[0018] Optionally, in the above target airport characteristic value prediction method, when creating a time series method prediction model, the method includes:

[0019] Based on the airport passenger and cargo throughput and basic city data, create linear functions, power functions, quadratic functions and cubic functions for predicting the airport passenger and cargo throughput based on basic city data;

[0020] The relative error and the sum of squared errors between the predicted values ​​and the true values ​​of the linear function, power function, quadratic function and cubic function are calculated, and the function whose relative error and the sum of squared errors meet the preset difference requirements is selected as the time series prediction model.

[0021] Optionally, in the above target airport characteristic value prediction method, the comprehensive model is trained based on the prediction results of the X prediction models for the airport passenger and cargo throughput of the reference airport, including:

[0022] The basic city data and airport passenger and cargo throughput corresponding to a first preset number of reference airports are selected as training data, and the basic city data and airport passenger and cargo throughput corresponding to a second preset number of reference airports are selected as verification data to train the comprehensive model.

[0023] Optionally, the above target airport characteristic value prediction method further includes:

[0024] Determine whether there is a high-speed rail plan in the city where the target airport is located;

[0025] When there is a high-speed rail plan, obtain the distribution law of passenger and cargo throughput of the target airport on the high-speed rail strong competition line, medium competition line and weak competition line;

[0026] The passenger and cargo throughputs of the target airport are corrected based on the distribution law of the passenger and cargo throughputs on the high-speed rail strong competition lines, medium competition lines and weak competition lines and the competition parameters of the target airport on the strong competition lines, medium competition lines and weak competition lines.

[0027] Optionally, the above target airport characteristic value prediction method further includes:

[0028] Based on the revised passenger and cargo throughput of the target airport, the target airport's predicted annual peak hour passenger numbers, passenger aircraft takeoffs and landings, and all-cargo aircraft takeoffs and landings are calculated.

[0029] Optionally, in the above target airport characteristic value prediction method, the predicted annual peak hour passenger number of the target airport is calculated based on the revised passenger and cargo throughput of the target airport, specifically including:

[0030] Obtain the peak hour passenger throughput percentage that matches the corrected passenger throughput of the target airport;

[0031] The number of peak hour passengers at the target airport is calculated based on the peak hour passenger throughput rate and the corrected passenger throughput of the target airport.

[0032] Optionally, in the above target airport characteristic value prediction method, the predicted number of passenger aircraft take-offs and landings at the target airport is calculated based on the corrected passenger and cargo throughputs of the target airport, specifically including:

[0033] Obtaining the annual weighted average number of seats at the target airport, wherein the annual weighted average number of seats matches the airport category and the proportion of each aircraft type at the target airport;

[0034] Obtaining the annual passenger load factor of the target airport, where the annual passenger load factor is the annual passenger load factor of the reference airport;

[0035] The number of passenger aircraft take-offs and landings at the target airport is calculated based on the revised passenger throughput of the target airport, the annual weighted average number of seats and the annual passenger load factor.

[0036] Optionally, in the above target airport characteristic value prediction method, the target airport predicted all-cargo aircraft take-off and landing number is calculated based on the corrected passenger and cargo throughput of the target airport, specifically including:

[0037] Obtain the corrected cargo throughput calculation for the target airport;

[0038] Obtaining the proportion of cargo and mail transportation volume by all-cargo aircraft, where the proportion of cargo and mail transportation volume by all-cargo aircraft is the proportion of cargo and mail transportation volume by all-cargo aircraft at the reference airport;

[0039] Obtaining an average cargo load of all-cargo aircraft, wherein the average cargo load of all-cargo aircraft is the average cargo load of all-cargo aircraft at the reference airport;

[0040] The predicted take-off and landing number of all-cargo aircraft at the target airport is calculated based on the revised cargo throughput calculation of the target airport, the proportion of cargo and mail transportation volume of all-cargo aircraft, and the average cargo capacity of all-cargo aircraft.

[0041] A computer device comprising a memory and a processor;

[0042] The memory is used to store programs;

[0043] The processor is used to execute the program to implement each step of the target airport characteristic value prediction method described in any one of the above items.

[0044] Based on the above technical solution, the above method provided by the embodiment of the present invention obtains historical data of reference airports of the same type as the target airport and basic data of cities in the area where the reference airport is located, trains X prediction models based on the historical data of the reference airport and the basic data of cities in the area where the reference airport is located, and the X prediction models are used to predict the airport passenger and cargo throughput based on the basic data of the cities, constructs a comprehensive model of the X prediction models, obtains basic data of the target city in the area where the target airport is located, uses the X prediction models to predict the airport passenger and cargo throughput of the target airport based on the basic data of the target city, and then substitutes the prediction results of the airport passenger and cargo throughput of the target airport by the X prediction models into the comprehensive model, and uses the output results of the comprehensive model as the predicted value of the airport passenger and cargo throughput of the target airport, thereby realizing the prediction method of the characteristic value of the target airport (the characteristic value may include the airport passenger and cargo throughput) BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0046] Figure 1 A schematic diagram of a flow chart of a target airport characteristic value prediction method disclosed in an embodiment of the present application;

[0047] Figure 2 A schematic diagram of the structure of a computer device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0049] See also Figure 1, the target airport characteristic value prediction method disclosed in the embodiment of the present application may include:

[0050] Step S101: Acquire airport historical data and basic city data where the airport is located;

[0051] In this step, the airport referred to may be any of a large number of airports, and the airport historical data includes the airport passenger and cargo throughput, and the airport passenger and cargo throughput may refer to the annual throughput.

[0052] For example, the passenger throughput y of an airport can be collected by year t , freight throughput t The basic data of the city may refer to the gross urban product G of the area where the airport is located. t , population P t , tourism consumption output value T t 、 Industrial and trade output value I t 、The urban passenger throughput Ry in the region t and freight throughput Rz t 、National passenger throughput Ny t and freight throughput Nz t , these data can be counted by the data statistics department and then entered into the computing system used by this method. After the data collection is completed, the collected data is preprocessed, and the preprocessing refers to the application of the median to complete a single missing item and a single noise item, and the application of N quantiles to complete continuous missing items, where N is the number of missing items.

[0053] Step S102: creating X prediction models based on the airport historical data and the city basic data, wherein the prediction models are used to predict the airport passenger and cargo throughput based on the city basic data;

[0054] In this scheme, when predicting the airport passenger and cargo throughput, it is first necessary to use a variety of different prediction models to ensure the reliability of the prediction results.

[0055] The prediction models may include: a time series prediction model, a regression analysis prediction model, a market share prediction model, a BP neural network prediction model and a Boburg-Garman formula prediction model.

[0056] Time series prediction model: Use the time series method to predict airport passenger and cargo throughput.

[0057] In this step, after obtaining the airport historical data and city basic data, the time series method is used to predict the airport passenger and cargo throughput. t, study the temporal correlation variation of the airport's annual passenger throughput values, and establish a quantitative model to predict the airport's passenger throughput y.

[0058] In order to ensure the reliability of the prediction results, multiple functions can be used for prediction. The function with the highest prediction reliability is selected from the multiple functions as the function used in the time series method. This method uses four functions to simulate the changing law of time series. The four functions are linear function, power function, quadratic function and cubic function. Suppose the predicted year is t, and the passenger throughput in year t is y1, then:

[0059] The prediction equation for the linear function is: 11 =a+b×t;

[0060] The prediction equation for the power function is: 12 =a×t b ;

[0061] The prediction equation for the quadratic function is: 13 =a+b×t+c×t 2 ;

[0062] The prediction equation for the cubic function is y 14 =a+b×t+c×t 2 +d×t 3 ;

[0063] Then we can calculate the parameters a, b, c, and d of the four functions by using the least square method, that is, we can find the parameters so that the sum of squares of the errors after they are brought in is the smallest.

[0064] Then calculate the error and determine the function. Substitute the verification year into the prediction equation with the obtained parameters to calculate the verification year prediction value of each method, and calculate the relative error and the sum of squares of the error between the prediction value and the true value. And select the function whose relative error and sum of squares of the error meet the preset difference requirements as the method of the time series method, and its prediction value is recorded as y1.

[0065] Similarly, the predicted value of airport cargo and mail throughput based on the time series forecasting method is denoted as z1.

[0066] Regression analysis prediction model: Regression analysis method is used to predict the airport passenger and cargo throughput.

[0067] This method considers the role of factors related to air passenger throughput, mathematically processes relevant data from previous years, studies and determines the correlation between air passenger throughput and related influencing factors, establishes a regression equation with good correlation, determines the applicability of the regression model through the regression coefficient, and predicts the air passenger throughput for a period of time afterwards through extrapolation.

[0068] First, we build a model. We select the gross domestic product, population, tourism consumption output, and industrial trade output of the city where the airport is located as independent variables reflecting the foundation of aviation development, and select the air passenger throughput y as the dependent variable. We use the multiple linear regression method to obtain the relevant model: y2 = a × G t +b×P t +c× T t +d×I t +e.

[0069] Then we calculate the parameter. t , population P t , tourism consumption output value T t 、Industrial and trade output value I t and air passenger throughput y t Substitute into the model and calculate parameters a, b, c, d, and e.

[0070] Finally, the degree of fit is judged and the significance test is performed. Calculation:

[0071]

[0072]

[0073]

[0074]

[0075] Where y2(t) represents the predicted value of air passenger throughput in verification year t, y t represents the actual value of air passenger throughput in verification year t, n is the number of verification years, Represents the average of the true values. 2 When F is greater than the target value and F is greater than the F boundary value, it means that the fitting model is effective and reasonable. According to the obtained regression model, the aviation passenger throughput y2 based on the regression analysis method is calculated.

[0076] Similarly, the predicted value of the airport cargo and mail throughput based on regression analysis is denoted as z2.

[0077] Market share method prediction model: The market share method is used to predict airport passenger and cargo throughput.

[0078] In this method, the development of regional and national business volume is used as the overall support, and the business volume is predicted by calculating the market share of the airport in the regional and national aviation business volume.

[0079] First, calculate the historical airport market share. t , passenger throughput Ry in the area t、National passenger throughput Ny t Calculate the market share MSR of the airport's air passenger throughput in the region in year t t With a national market share of MSN t .

[0080]

[0081]

[0082] Then the market share and regional and national aviation business volume are predicted. The least squares curve fitting method is used to calculate the MSR. t and MSN t Fitting is performed to obtain the functional relationship between market share and year, and the MSR and MSN of the forecast year are calculated. There are many ways to predict regional and national air passenger throughput Ry and Ny in a year, which can be done by fitting and predicting through the least squares curve fitting method, or by consulting the relevant plans of regional and national planning departments.

[0083] Finally, the predicted value of the airport's air passenger throughput is calculated. The expression of the predicted value of air passenger throughput y3 based on the market share method is y3 = α × Ry × MSR + (1-α) × Ny × MSN.

[0084] Among them, α represents the degree of emphasis on regional share. The larger α is, the more emphasis the forecast results place on the impact of regional development and changes in the airport's regional share. Otherwise, it means that more emphasis is placed on the impact of national civil aviation development and changes in the airport's national share. It can be adjusted as needed.

[0085] Similarly, the predicted value z3 of the airport cargo and mail throughput based on the market share method can be calculated.

[0086] BP neural network prediction model: BP neural network method is used to predict the airport passenger and cargo throughput.

[0087] This method relies on the ideal analytical accuracy of BP neural network in processing multi-factor nonlinear systems, analyzes the internal relationship of multiple factors affecting airport aviation business volume, and predicts aviation passenger and cargo throughput.

[0088] First, the data is normalized. Due to the characteristics of the BP neural network and its requirements for input data, the gross product G of the city where the airport is located is t , population P t , tourism consumption output value T t 、Industrial and trade output value I t The historical data is normalized to improve the generalization ability of the neural network.

[0089] Then, a BP neural network is constructed. The BP neural network topology consists of an input layer, a hidden layer, and an output layer. This method mainly analyzes the impact of the gross domestic product, population, tourism consumption output value, and industrial trade output value of the city where the airport is located on the airport's air passenger throughput. Therefore, the input layer node is set to 4, the output layer node is set to 1, and LOGSIG is selected as the hidden layer activation function. The learning rate is fixed to 0.05, and the maximum number of iterations is set to 2000. The number of hidden layer nodes, initial weights and other parameters can be adjusted according to the training situation.

[0090] Then, the model is trained and verified. The collected data is sorted by year, and one sample is taken out of every four samples as a verification sample, and the rest are used as training samples to train the BP neural network. The predicted value of the verification sample is compared with the true value, and R is calculated. 2 Verify the model's fit.

[0091] Finally, the BP neural network prediction value is calculated. The predicted values ​​of GDP, population, tourism consumption output value, and industrial trade output value are brought into the verified BP neural network to calculate the air passenger throughput y4 based on the BP neural network method.

[0092] Similarly, the predicted value of the airport cargo and mail throughput based on the BP neural network method is denoted as z4.

[0093] The prediction model of the Bob-Gammon formula method: The Bob-Gammon formula method is used to predict the airport passenger and cargo throughput.

[0094] According to the following Bob Gamon formula, predict the airport's air passenger throughput y5.

[0095]

[0096] Among them, y t T t Annual passenger throughput, T t is the starting year for calculation, T0 is the year when the normal operation frequency begins to be appropriate, usually the year after commissioning, and n is a calculation constant, which is taken as 2 in the short-term forecast.

[0097] Similarly, the predicted value z5 of the airport cargo and mail throughput based on the Boburg-Gammon formula method can be calculated.

[0098] Step S103: Obtain historical data of the reference airport and basic data of the city where the airport is located.

[0099] The reference airport is an airport of the same category as the target airport. When determining the reference airport, cluster analysis can be used to analyze the characteristics of the target airport (the target airport is the airport to be newly built) and search for airports of the same type as the target airport as reference airports. Specifically, based on the collected airport information, the basic attributes of the target airport are analyzed and judged, mainly including: ① Whether the airport is a civil airport or a joint military and civil airport, ② Whether it is a port airport, and ③ Whether the city where it is located is a tourist city. The reference airport is then determined based on these three basic attributes. The reference airport can be pre-determined by the staff before implementing this plan.

[0100] In this step, in this solution, the reference airport refers to the reference airport of a city with the same urban scale and economic level as the airport planning location, and then the airport passenger throughput y is collected by year. t , freight throughput t , GDP of the city in the region t , population P t , tourism consumption output value T t 、Industrial and trade output value I t 、The urban passenger throughput Ry in the region t and freight throughput Rz t 、National passenger throughput Ny t and freight throughput Nz t , these data can be counted by the data statistics department and then entered into the computing system used by this method. After the data collection is completed, the collected data is preprocessed, and the preprocessing refers to the application of the median to complete a single missing item and a single noise item, and the application of N quantiles to complete continuous missing items, where N is the number of missing items.

[0101] Step S104: constructing a comprehensive model of the X prediction models.

[0102] In this step, after the X prediction models are determined, a comprehensive model of the X prediction models is constructed. At this time, the constructed comprehensive model is an initial model and needs to be trained and corrected.

[0103] In this step, the comprehensive model can be Among them, α, β, γ, δ, and ε are initial values.

[0104] Step S105: training the comprehensive model based on the prediction results of the airport passenger and cargo throughput of the reference airport based on the basic city data of the reference airport and the airport passenger and cargo throughput;

[0105] The comprehensive model established is The data corresponding to the first preset number of reference airports (basic city data and airport passenger and cargo throughput) are selected as verification samples, and the data corresponding to the second preset number of reference airports (basic city data and airport passenger and cargo throughput) are selected as training samples. The airport passenger and cargo throughput of the reference airport is taken as the true value. The coefficients (α, β, γ, δ, ε) of the comprehensive model are constantly modified to minimize the sum of squares of the deviations between the predicted value and the true value of the comprehensive model, that is, Minimize, solve α, β, γ, δ, ε.

[0106] After the training is completed, verification and solution are required. At this time, the verification sample is substituted into the comprehensive model to calculate the determination coefficient R 2 Verify the comprehensive prediction effect.

[0107]

[0108]

[0109]

[0110] where y * (t) represents the predicted value of air passenger throughput of the reference airport in verification year t, y t represents the actual value of the air passenger throughput of the reference airport in the verification year t, n is the number of verification years, It represents the average of the actual values ​​of the air passenger throughput of the reference airport in n verification years. If the determination coefficient is less than the preset target value, α, β, γ, δ, ε are re-selected and calculated; if the determination coefficient is greater than the preset target value, the calculated y1, y2, y3, y4, y5 are substituted into the comprehensive prediction formula to calculate the comprehensive prediction value y of air passenger throughput based on civil aviation development data. * .

[0111] Similarly, the comprehensive forecast value of air cargo and mail throughput z based on civil aviation development data can be calculated * .

[0112] Step S106: acquiring basic data of a target city in the area where the target airport is located, and using the X prediction models to predict the airport passenger and cargo throughput of the target airport based on the basic data of the target city.

[0113] In this step, basic data of the target city in the area where the target airport is located is obtained, and the X prediction models determined above are used to predict based on the basic data of the target city to obtain the airport passenger and cargo throughput of the target airport predicted by each prediction model.

[0114] Step S107: Substitute the airport passenger and cargo throughput of the target airport into the comprehensive model, and use the output result of the comprehensive model as the predicted airport passenger and cargo throughput of the target airport.

[0115] In this step, the airport passenger and cargo throughput of the target airport predicted by each prediction model is substituted into the trained comprehensive model. The comprehensive model will output a comprehensive prediction result based on the airport passenger and cargo throughput of the target airport predicted by each prediction model, and use the prediction result as the predicted passenger and cargo throughput of the target airport.

[0116] In the technical solution disclosed in another embodiment of the present application, taking into account the impact of high-speed rail planned in the future on the diversion of air passengers, in this solution, the predicted value of the comprehensive model can also be corrected based on the high-speed rail lines planned in the future.

[0117] Therefore, the technical solution disclosed in this embodiment may also include:

[0118] Determine whether there is a high-speed rail plan in the city where the target airport is located. When there is no high-speed rail plan, end the process. When there is a high-speed rail plan, obtain the distribution law of the passenger and cargo throughput of the target airport on the high-speed rail strong competition line, medium competition line and weak competition line, and correct the passenger and cargo throughput of the target airport based on the distribution law of the passenger and cargo throughput on the high-speed rail strong competition line, medium competition line and weak competition line and the competition parameters of the target airport on the strong competition line, medium competition line and weak competition line.

[0119] When obtaining the distribution law of the target airport's passenger and cargo throughput on the high-speed rail strong competition lines, medium competition lines and weak competition lines, generally speaking, there is competition between high-speed rail and flights. The applicant has found through research that the lines that meet the requirements of high-speed rail operation time ≤150min and departure interval ≤2h are set as strong competition lines, and high-speed rail has absolute advantages on these lines; the lines that meet the requirements of high-speed rail operation time within 150min but departure interval ≥2h, and the lines with high-speed rail operation time between 150min-240min are set as medium competition lines, and high-speed rail has a significant diversion effect on civil aviation on these lines; the lines with high-speed rail operation time between 240min-300min are set as weak competition lines, and civil aviation has relative advantages on these lines; the lines with high-speed rail operation time ≥300min are set as non-competitive lines, and civil aviation has absolute advantages on these lines.

[0120] Based on the above principles, the flights of the target airport are divided into strong competition routes, medium competition routes and weak competition routes. The probability (competition parameter) that passengers choose high-speed rail instead of air travel on strong competition routes is set as P1, the probability that passengers choose high-speed rail instead of air travel on medium competition routes is set as P2, and the probability that passengers choose high-speed rail instead of air travel on weak competition routes is set as P3. where i=1, 2, 3, U i V represents the utility of high-speed rail on the preset corresponding route, which is related to factors such as high-speed rail travel cost (fare), travel time, frequency, punctuality, convenience (network development level) and reliability; i It is the utility of civil aviation on the preset corresponding routes, which is related to factors such as civil aviation travel costs (fares), travel time, frequency, punctuality, convenience (network development level) and reliability.

[0121] By analyzing the composition of air passenger sources and destinations, we statistically calculated the proportion of passengers in the newly planned high-speed rail lines with strong competition, medium competition and weak competition (distribution law), and corrected the aviation business volume. The correction value is Y = y * ×(1-R1×P1-R2×P2-R3×P3).

[0122] Among them, R1 is the ratio of passengers on the planned new high-speed rail lines with strong competition to the total number of passengers, R2 is the ratio of passengers on the planned new high-speed rail lines with medium competition to the total number of passengers, and R3 is the ratio of passengers on the planned new high-speed rail lines with weak competition to the total number of passengers.

[0123] Similarly, the correction value Z of air cargo and mail throughput considering the diversion effect of high-speed rail can be calculated.

[0124] In addition to predicting the passenger and cargo throughput of the target airport, in this application, the predicted annual peak hour passenger number, passenger aircraft takeoffs and landings, and all-cargo aircraft takeoffs and landings of the target airport can also be calculated based on the revised passenger and cargo throughput of the target airport.

[0125] The target airport's predicted annual peak hour passenger numbers are calculated based on the revised target airport's passenger and cargo throughput, including:

[0126] Obtain the peak hour passenger throughput percentage θ that matches the corrected passenger throughput of the target airport;

[0127] The peak hour passenger number Y of the target airport is calculated based on the peak hour passenger throughput rate and the corrected passenger throughput of the target airport. peak , where, specifically, Y peak =Y×θ, where Y is the corrected passenger throughput of the target airport.

[0128] Where θ is the percentage of peak hour passenger throughput, which can be obtained based on the ratio of peak hour passenger flow to annual passenger volume recommended by the Federal Aviation Administration (FAA). The specific relationship is as follows:

[0129]

[0130] The target airport's predicted passenger aircraft take-offs and landings are calculated based on the revised target airport's passenger and cargo throughput, including:

[0131] The annual weighted average number of seats at the target airport is obtained, and the annual weighted average number of seats matches the airport category and the aircraft types of the target airport; the annual passenger load factor of the target airport is obtained, and the annual passenger load factor is the annual passenger load factor of the reference airport; the number of passenger aircraft take-offs and landings C of the target airport is calculated based on the revised passenger throughput of the target airport, the annual weighted average number of seats and the annual passenger load factor. p . C p =Y÷(SNum×PLF), where SNum is the predicted annual weighted average number of seats at the target airport, which is related to the airport category and the proportion of each aircraft type. The weighted average number of seats for Class B aircraft is 60, the weighted average number of seats for Class C aircraft is 160, the weighted average number of seats for Class D aircraft is 250, the weighted average number of seats for Class E aircraft is 360, and the weighted average number of seats for Class F aircraft is 550. Multiplying by the weight of each aircraft type in the predicted year, the weighted average number of seats for the target airport is the predicted annual weighted average number of seats; PLF is the predicted annual passenger load factor of the target airport, which can be calculated based on the historical passenger load factor of the reference airport.

[0132] The predicted take-off and landing times of all-cargo aircraft at the target airport are calculated based on the revised passenger and cargo throughput of the target airport, including:

[0133] Obtain the revised cargo throughput calculation of the target airport; obtain the proportion of cargo and mail transportation volume of all-cargo aircraft, which is the proportion of cargo and mail transportation volume of all-cargo aircraft at the reference airport; obtain the average cargo load of all-cargo aircraft, which is the average cargo load of all-cargo aircraft at the reference airport; based on the revised cargo throughput calculation of the target airport, the proportion of cargo and mail transportation volume of all-cargo aircraft and the average cargo load calculation of all-cargo aircraft, obtain the predicted take-offs and landings of all-cargo aircraft at the target airport.

[0134] Specifically, when calculating the number of all-cargo aircraft take-offs and landings, the target airport's predicted annual all-cargo aircraft take-offs and landings C can be calculated based on the target airport's cargo throughput correction value, the proportion of all-cargo aircraft cargo and mail transportation volume, and the average cargo capacity of all-cargo aircraft. g , the formula is: C g =Z×GRat÷Vol, where GRat is the predicted annual cargo and mail transport volume ratio of all-cargo aircraft, and Vol is the average cargo capacity of all-cargo aircraft.

[0135] In this embodiment, corresponding to the above method, a computer device is also disclosed, see Figure 2 , the device may include:

[0136] It may include: at least one processor 100, at least one communication interface 200, at least one memory 300 and at least one communication bus 400;

[0137] In the embodiment of the present invention, the number of the processor 100, the communication interface 200, the memory 300, and the communication bus 400 is at least one, and the processor 100, the communication interface 200, and the memory 300 communicate with each other through the communication bus 400; obviously, see Figure 2 The communication connections shown for the processor 100, the communication interface 200, the memory 300, and the communication bus 400 are merely optional;

[0138] Optionally, the communication interface 200 may be an interface of a communication module, such as an interface of a GSM module;

[0139] The processor 100 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0140] The memory 300 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0141] The processor 100 is specifically used to execute each step of the target airport characteristic value prediction method described in any one of the above embodiments.

[0142] For the convenience of description, the above system is described by dividing it into various modules according to its functions. Of course, when implementing the present invention, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0143] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0144] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0145] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0146] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0147] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting characteristic values ​​of a target airport, characterized in that: include: Obtaining airport historical data and basic data of cities in the area where the airport is located, wherein the airport historical data is the airport passenger and cargo throughput; Based on the airport historical data and the city basic data, create X prediction models, wherein the prediction models are used to predict the airport passenger and cargo throughput based on the city basic data; Acquire historical data of reference airports and basic data of cities in the areas where the airports are located, wherein the reference airports are airports of the same category as the target airport; wherein cluster analysis is used to analyze the characteristics of the target airports and determine airports of the same category as the target airport as reference airports; Constructing a comprehensive model of the X prediction models; Selecting basic city data and airport passenger and cargo throughput corresponding to a first preset number of reference airports as training data, selecting basic city data and airport passenger and cargo throughput corresponding to a second preset number of reference airports as verification data, and training the comprehensive model, wherein the airport passenger and cargo throughput in the training data is used as the true value, and the coefficients of the comprehensive model are continuously corrected so that the sum of squares of deviations between the predicted value and the true value of the comprehensive model is minimized, and the coefficients of the comprehensive model are solved; Obtaining basic data of a target city in the area where the target airport is located, and using the X prediction models to predict the airport passenger and cargo throughput of the target airport based on the basic data of the target city; Substitute the airport passenger and cargo throughputs of the target airport into the comprehensive model, and use the output result of the comprehensive model as the predicted airport passenger and cargo throughputs of the target airport.

2. The target airport characteristic value prediction method according to claim 1, characterized in that: The X prediction models include: Time series prediction model, regression analysis prediction model, market share prediction model, BP neural network prediction model and Bopuga formula prediction model.

3. The target airport characteristic value prediction method according to claim 2, characterized in that: When creating a time series forecasting model, the methods include: Based on the airport passenger and cargo throughput and basic city data, create linear functions, power functions, quadratic functions and cubic functions for predicting the airport passenger and cargo throughput based on basic city data; The relative error and the sum of squared errors between the predicted values ​​and the true values ​​of the linear function, power function, quadratic function and cubic function are calculated, and the function whose relative error and the sum of squared errors meet the preset difference requirements is selected as the time series prediction model.

4. The target airport characteristic value prediction method according to claim 1, characterized in that: Also includes: Determine whether there is a high-speed rail plan in the city where the target airport is located; When there is a high-speed rail plan, obtain the distribution law of passenger and cargo throughput of the target airport on the high-speed rail strong competition line, medium competition line and weak competition line; The passenger and cargo throughputs of the target airport are corrected based on the distribution law of the passenger and cargo throughputs on the high-speed rail strong competition lines, medium competition lines and weak competition lines and the competition parameters of the target airport on the strong competition lines, medium competition lines and weak competition lines.

5. The target airport characteristic value prediction method according to claim 4, characterized in that: Also includes: Based on the revised passenger and cargo throughput of the target airport, the target airport's predicted annual peak hour passenger numbers, passenger aircraft takeoffs and landings, and all-cargo aircraft takeoffs and landings are calculated.

6. The target airport characteristic value prediction method according to claim 5, characterized in that: The target airport's predicted annual peak hour passenger numbers are calculated based on the revised target airport's passenger and cargo throughput, including: Obtain the peak hour passenger throughput percentage that matches the corrected passenger throughput of the target airport; The number of peak hour passengers at the target airport is calculated based on the peak hour passenger throughput rate and the corrected passenger throughput of the target airport.

7. The target airport characteristic value prediction method according to claim 5, characterized in that: The predicted number of passenger aircraft take-offs and landings at the target airport is calculated based on the revised passenger and cargo throughput of the target airport, including: Obtaining an annual weighted average number of seats at the target airport, wherein the annual weighted average number of seats matches the airport category and the proportion of each aircraft type at the target airport; Obtaining the annual passenger load factor of the target airport, where the annual passenger load factor is the annual passenger load factor of the reference airport; The number of passenger aircraft take-offs and landings at the target airport is calculated based on the revised passenger throughput of the target airport, the annual weighted average number of seats and the annual passenger load factor.

8. The target airport characteristic value prediction method according to claim 5, characterized in that: The target airport's predicted all-cargo aircraft take-offs and landings are calculated based on the revised target airport's passenger and cargo throughput, including: Obtain the corrected cargo throughput calculation for the target airport; Obtaining the proportion of cargo and mail transportation volume by all-cargo aircraft, where the proportion of cargo and mail transportation volume by all-cargo aircraft is the proportion of cargo and mail transportation volume by all-cargo aircraft at the reference airport; Obtaining an average cargo load of all-cargo aircraft, wherein the average cargo load of all-cargo aircraft is the average cargo load of all-cargo aircraft at the reference airport; The predicted take-off and landing number of all-cargo aircraft at the target airport is obtained based on the cargo throughput calculation of the revised target airport, the proportion of cargo and mail transportation volume of all-cargo aircraft, and the average cargo capacity of all-cargo aircraft.

9. A computer device, characterized in that: including memory and processor; The memory is used to store programs; The processor is used to execute the program to implement each step of the target airport characteristic value prediction method as described in any one of claims 1-8.

Images