CN114943388A

CN114943388A - Airport real-time parking space distribution method

Info

Publication number: CN114943388A
Application number: CN202210722776.9A
Authority: CN
Inventors: 赵宁宁; 冯嘉蓬
Original assignee: Civil Aviation University of China
Current assignee: Civil Aviation University of China
Priority date: 2022-06-24
Filing date: 2022-06-24
Publication date: 2022-08-26

Abstract

The invention provides a method for allocating parking spaces in real time in an airport, which constructs an objective function by taking the shortest sliding distance of an aircraft and the highest occupancy rate of the parking spaces as an optimization target and simultaneously considers the safety constraint of scene operation to establish a real-time parking space allocation model, so that the real-time allocation model ensures the smooth production on the premise of scene safe operation; the model for real-time distribution of the stand of aircraft constructed by the invention not only can pre-distribute the flights of the airport on the same day, but also can distribute the flights under the condition of small-amplitude delay in real time, and is a model for real-time distribution of the stand of aircraft with the function of real-time distribution. The method has certain effect on reducing the aircraft congestion of airport channels and reducing the cost of fuel oil for the taxiing of airliners, and provides reference for the research of the problem of real-time distribution of parking positions.

Description

Airport real-time parking space distribution method

Technical Field

The invention belongs to the technical field of airports, and particularly relates to a method for allocating parking spaces in real time in an airport.

Background

In recent years, more and more foreign and domestic scholars study the problem of the stand by modeling and using different optimization algorithms. Although the research on the distribution of the stand at home and abroad is increasing in these years, the following disadvantages exist:

the current airplane parking space allocation research mostly stays on a pre-allocation level, and the real-time airplane parking space allocation research is less. Under the condition that flight delay occurs after flow control, weather change or sudden situations, the result of pre-distribution often fails and airplane stop conflicts are generated, most airports in China currently carry out real-time airplane position adjustment manually, and the scientificity and rationality need to be improved.

In the aspect of the algorithm selected by research, because the precise algorithm appears earlier, a plurality of scholars still use the precise algorithm to perform optimization calculation on the problem of the aircraft stop allocation at present, and in consideration of the construction of intelligent civil aviation in China and the continuous development of artificial intelligence, heuristic algorithms such as genetic algorithm, simulated annealing algorithm, ant colony algorithm and the like should be used for performing optimization analysis on the problem of the aircraft stop.

Although the conventional parking space allocation research has the main research targets of passengers, airports and other people with different interests, most students choose to research the passenger with the shortest walking distance. The walking distance has been selected in the past in order to improve the boarding comfort of passengers, and with the popularization of convenience facilities such as horizontal transportation elevators in airports, passengers are more convenient to transfer or board, and research on the boarding distance of passengers should not be performed.

Disclosure of Invention

In view of the above, the present invention is directed to a method for allocating parking spaces in real time in an airport, so as to solve the above-mentioned problems in the background art.

In order to achieve the purpose, the technical scheme of the invention is realized as follows:

an airport real-time parking space allocation method is characterized by comprising the following steps: comprises that

Establishing a parking space real-time distribution model by taking the shortest sliding distance of the aircraft and the highest parking space occupancy rate as optimization targets and taking scene operation safety as a constraint condition;

and performing optimization calculation on the parking space allocation model by using a genetic algorithm to obtain a parking space pre-allocation result.

Further, the function corresponding to the optimization objective is

f＝f ₁ +f ₂ (1)

f ₁ ＝min[∑ _i∈N ∑ _k∈M x _ik (DA _k +DD _K )] (2)

Where N represents a flight set (i, j. e.N; i 1,2,3 … 20; j 2,3,4 … 20)

M represents a set of parking positions (k is the same as M; k is 1,2,3 … 28)

DA denotes flight sliding distance set (DA) _k ∈DA；k∈M；k＝1,2,3…28)

DD represents flight slide-out distance set (DD) _k ∈DD；k∈M；k＝1,2,3…28)

Wherein f1 represents an optimization target with the shortest sliding distance, and the objective function is to solve the minimum sum of the sliding distance and the sliding-in distance of each flight, namely the total sliding distance is shortest;

f2 denotes maximizing stand use efficiency and reducing stand occupancy.

Further, the constraint condition is specifically

Wherein the content of the first and second substances,

；

E _ik indicating the moment when the flight i enters the wheel gear of the stand k;

L _ik indicating the wheel gear withdrawal moment when the flight i leaves the stand k;

E _jk a catch time indicating that a flight j using the same stand k after flight i enters a stand;

wherein, the formula (4) restricts that a flight can be arranged at only one stand; the formula (5) shows that one stand can only stop for one flight at most in the same time period; equation (6) indicates that two adjacent flights require a safe time interval of 35 minutes; equation (7) represents the constraint that each flight stops at the stand for at least one hour, which is the minimum time to pass; equations (8) and (9) indicate that the airport is operating at 5 to 24 hours per day, and flights are only allowed to land and land during this time frame.

Further, the calculation in step 3 is performed by using a genetic algorithm, which specifically comprises the following steps:

step 1: using the sliding distance of each parking space for gene coding;

step 2: generating an initial population in a random mode;

and step 3: setting an individual fitness function in the population;

and 4, step 4: and obtaining an optimal solution through selection, intersection and variation calculation, namely a final result of pre-allocation of the parking spaces.

The invention also provides an airport real-time parking space distribution device, which comprises

The model establishing device is used for establishing a parking space real-time distribution model by taking the shortest sliding distance of the aircraft and the highest parking space occupancy rate as optimization targets and taking scene operation safety as a constraint condition;

and the model solving device is used for carrying out optimization calculation on the parking space allocation model by utilizing a genetic algorithm to obtain a parking space pre-allocation result.

The invention also provides an electronic device comprising

At least one processor, and

at least one memory communicatively coupled to the processor, wherein:

the memory stores program instructions executable by the processor, the processor being capable of performing the method as described in any one of the preceding claims when called by the processor.

The present invention also provides a non-transitory computer-readable storage medium, which when executed by one or more processors, causes the processors to perform any of the methods described above.

Compared with the prior art, the airport real-time parking space allocation method has the following advantages:

(1) the invention constructs an objective function by taking the shortest sliding distance of an aircraft and the highest occupancy rate of the parking space as an optimization target, and simultaneously considers the safety constraint of scene operation to establish a real-time parking space distribution model, so that the real-time distribution model ensures the smooth production on the premise of scene safe operation;

(2) the parking space real-time distribution model constructed by the invention not only can pre-distribute the flights on the same day in an airport, but also can distribute the flights in real time under the condition of small-amplitude delay, and is a parking space distribution model with a real-time distribution function. The method has certain effect on reducing the aircraft congestion of airport channels and reducing the cost of fuel oil for the taxiing of airliners, and provides reference for the research of the problem of real-time distribution of parking positions.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate an embodiment of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:

FIG. 1 is a flow chart of initial population generation of the present invention;

FIG. 2 is a schematic illustration of a roulette selection according to the present invention;

FIG. 3 is a schematic cross-calculation of the present invention;

FIG. 4 is a schematic diagram of a variant calculation process according to the present invention;

FIG. 5 is a Tianjin coastal airport stop bitmap ZBTJ-2 of the present invention;

FIG. 6 is a schematic view of a first glide distance measurement of the present invention;

FIG. 7 is a second schematic diagram of the glide distance measurement of the present invention;

FIG. 8 is a sliding path diagram of the present invention

FIG. 9 is an iterative variation diagram of the present invention;

FIG. 10 is a Gantt diagram of an optimized pre-gate position of the present invention;

FIG. 11 is a Gantt diagram of an optimized stand (pre-allocated stand) of the present invention;

FIG. 12 is a delay generation diagram of the present invention.

Detailed Description

It should be noted that the embodiments and features of the embodiments of the present invention may be combined with each other without conflict.

In the description of the present invention, it is to be understood that the terms "central," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like are used in the orientations and positional relationships indicated in the drawings, which are based on the orientations and positional relationships indicated in the drawings, and are used for convenience in describing the present invention and for simplicity in description, but do not indicate or imply that the device or element so referred to must have a particular orientation, be constructed in a particular orientation, and be operated, and thus should not be construed as limiting the present invention. Furthermore, the terms "first", "second", etc. are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first," "second," etc. may explicitly or implicitly include one or more of that feature. In the description of the present invention, "a plurality" means two or more unless otherwise specified.

In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "connected" are to be construed broadly and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meaning of the above terms in the present invention can be understood by those of ordinary skill in the art through specific situations.

The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.

The invention provides a method for allocating airport real-time parking positions, which comprises the following steps of firstly explaining symbols used in modeling, and defining the following variables based on data used in the invention:

(1) n denotes a set of flights (i, j e N; i 1,2,3 … 20; j 2,3,4 … 20)

(2) M represents a set of parking positions (k is the same as M; k is 1,2,3 … 28)

(3) DA denotes the set of flight slide-in Distances (DA) _k ∈DA；k∈M；k＝1,2,3… 28)

(4) DD represents flight slide-out distance set (DD) _k ∈DD；k∈M；k＝1,2,3… 28)

(7)E _ik Indicating the moment when flight i enters stand k

(8)L _ik Indicating the departure of flight i from stand k at the time of the gear-off

(9)E _jk A catch time indicating that a flight j using the same stand k after a flight i enters a stand

The function corresponding to the optimization objective of the invention is:

f＝f ₁ +f ₂ (1)

f ₁ ＝min[∑ _i∈N ∑ _k∈M x _ik (DA _k +DD _K )] (2)

where f1 represents an optimization objective for the shortest taxi distance, and the objective function is to find the shortest sum of the taxi distance and the slide-in distance for each flight, i.e. the shortest total taxi distance. On one hand, the shortest sliding distance represents the minimum sliding oil quantity, so that the cost can be saved for the navigation driver; on the other hand, the method also shows that the taxiing time is minimum, the number of airplanes on the airport runway in a certain time period can be reduced, the runway capacity is increased, the speed of the aircraft entering the parking space and leaving the taxiway is increased, and the use efficiency of the parking space is improved.

And f2 represents maximizing stand use efficiency and reducing stand occupancy. Because the flight sliding distances corresponding to each stand are different, under the goal of the shortest total sliding distance, the selection of the stands is more inclined to certain stands with relatively shorter sliding distances on the basis of meeting the constraint condition, which shows that the number of used stands is reduced and the occupation of the stands is reduced.

The constraints used by the present invention are:

wherein, the formula (4) restricts that one flight can be arranged at only one stand; the formula (5) shows that one stand can only stop for one flight at most in the same time period; equation (6) indicates that two adjacent flights require a safe time interval of 35 minutes; equation (7) represents the constraint that each flight stops at the stand for at least one hour, which is the minimum time to pass; equations (8) and (9) indicate that the airport is operating at 5 to 24 hours per day, and flights are only allowed to land and land during this time frame.

The method selects a genetic algorithm to carry out optimization calculation on the parking space allocation model. The genetic algorithm has the advantages that due to the evolutionary characteristics of the genetic algorithm, the inherent properties of the problem do not need to be considered in calculation, and the calculation can be carried out on an objective function in any form. The concept of the genetic algorithm is derived from the Darwinian theory of evolution, the detailed problem to be researched by the genetic algorithm is used as required, data in the problem can be coded, codes are regarded as genes, coded results are regarded as individuals, codes of one group are different groups, calculation is carried out by the concept of superior and inferior selection in the theory of evolution on the basis of the groups, a man-made function setting is used as a natural rule, good results are selected from multiple possible results through multiple generations of heredity, bad results are removed, and therefore the optimal solution is obtained.

The Python implementation process using the genetic algorithm of the invention is as follows

(1) Gene coding

The gene coding is the root of the genetic algorithm, and the genetic algorithm must be used for coding firstly, because the coding condition determines the storage and expression form of the gene in the algorithm, and the subsequent overall calculation process is directly influenced. The genetic algorithm gene coding has various forms, can be binary coding or real number coding, if the binary coding is used, the process of crossing and variation can be influenced due to the high randomness of the binary coding, and the optimal solution can be possibly kept away when the operation is finished, so that the stability cannot be achieved, therefore, the invention adopts a self-created real number coding mode, and the sliding distance of each stand is directly used for coding, namely, the sliding distance data of each stand is the code of the stand, so that the subsequent data selection, crossing, variation and other calculations are facilitated, the optimization result is more conveniently obtained, and the condition that no feasible solution exists is avoided, as shown in table 1.

For example, the present invention describes an algorithm for allocating stops to 20 flights and minimizing the taxi distance. Because the present invention only focuses on the sliding distance, and the meaning of the chromosome is simple, one chromosome represents one individual, and the genetic composition of each individual is the total sliding distance corresponding to each of the 20 machine positions.

TABLE 1 Gene coding Table (parts)

(2) Generation of an initial population

Because genetic algorithms are inspired from natural reproduction, the first step in computational solution requires the determination of an initial population as the start of reproduction. The initial population of this study was generated in a completely random fashion. Some scholars will select better genes with some constraints when generating the initial population, and taking this study as an example, select some stands with closer sliding distances to add to the initial population when generating, so that the diversity of the population is reduced and premature may occur, and the above situation can be avoided by completely random means. The flow of initial population generation is shown in fig. 1.

The generation process is explained, in the example operation process, the initial population is performed on the basis of obtaining flight data, so after 20 flights are allocated with the stop positions in the first step, the second step of screening of constraint conditions is needed to meet the basic constraint conditions. And the constraints here do not include the constraint on the time taken by the stands, i.e. the time of use of the initial stands is allowed to overlap, which occurs because although no two flights enter the same stand simultaneously by the constraints, overlap in the time taken by the stands may occur due to the different times of passage of each flight. However, the initial population is only used as a precondition for starting operation, and the overlapped part participates in calculation, so that the subsequent calculation result is not influenced.

And after confirming that the stand meets the requirements, executing the third step, repeating the first step and the second step for 50 times to obtain an initial population containing 50 individuals, and spreading subsequent calculation and iteration around the initial population.

(3) Fitness function

In the genetic algorithm, how to evaluate the quality of an individual depends on a fitness function, the size of the fitness function represents the capability of the individual to adapt to the environment and determines the survival or death of the individual, and the higher the fitness is, the easier the individual can survive through natural selection, and vice versa. In order to obtain a final result and accelerate the solving speed, a fitness function needs to be reasonably set. Because the optimization goal of the invention is to find the minimum value of the sliding distance, the fitness function is also the minimum value, namely, the shorter the total sliding distance of the individual is, the stronger the fitness is, the more easily the individual can survive in the natural selection process, and the higher the fitness is reflected in the code.

(4) Selection calculation

The process of selecting and eliminating the advantages and the disadvantages in the nature is to screen through the natural environment, the selection and calculation in the genetic algorithm is used for simulating the natural selection process, and the scale in the selection process is the fitness function. If an individual is more adaptive, he is more likely to pass the gene to the next generation by natural selection.

However, the Selection in nature is probabilistic, and in order to realize the Selection process in the algorithm, the probability relation needs to be established, in the present research, the Selection calculation mode adopts Roulette Wheel Selection (Roulette Wheel Selection), which is a playback type random sampling method, the operation is simple, and the probability of each individual entering the next generation is equal to the ratio of the fitness value of the individual to the fitness value sum of the whole population. For example, five individuals need to be selected to determine whether to pass the gene to the next generation, and the fitness of the five individuals is p1, p2... p5 through fitness function calculation, so that the probability of each individual being selected is:

the probability of five individuals being selected makes up a wheel as shown in figure 2. It can be seen that if the fitness of an individual is larger, the advantage of the individual to be selected is larger, and the individual is easier to select in the natural selection process.

(5) Cross computation

The cross calculation refers to the mutual exchange of partial genes of two paired individuals of the previous generation in a certain way so as to form a new individual, in the genetic algorithm, the cross calculation belongs to the most important algorithm part, and the cross calculation endows the genetic algorithm with the capability of global search, and the generation and the variation calculation of the offspring individuals are required to be carried out after the genes are crossed. There are many ways of crossing, and the present study selects single point crossing as the crossing way, i.e., randomly selects 1 gene from 20 genes of the previous generation individuals to participate in the crossing calculation. The crossover probability of the present invention is set to 70% and the specific crossover operation is shown.

(6) Variance calculation

The mutation calculation is to change the codes of some filial individuals with certain probability to realize the effect of gene mutation. The variation calculation has two functions, the variation of organisms in the first and nature during reproduction is the same, the 'super individuals' in the population can be prevented from appearing earlier through the variation, the diversity of the population is increased, and the algorithm is prevented from being premature; secondly, the local searching capability is added to the genetic algorithm: when the previous generation is subjected to cross calculation, the generated offspring may be close to the optimal solution, and the speed of obtaining the optimal solution by the offspring can be accelerated by changing genes at certain positions of the offspring through mutation.

The general mutation calculation is divided into two steps, firstly, whether the offspring individuals generated in the crossing process of the previous generation have mutation or not depends on the mutation probability, and secondly, if the mutation occurs, the codes of the offspring are changed according to the mutation form set in the algorithm. In the present invention, the mutation probability is set to 3%, and an individual satisfying the mutation probability has a gene mutated into an allele, and the process of mutation calculation is shown in fig. 4.

The parameter settings of the genetic algorithm are shown in table 2. The reason why the number of iterations is selected to be 800 generations is that the iteration results are analyzed and determined after a plurality of trial calculations are performed, and the number of iterations is set to be 800 generations in order to make the results as close to the optimal solution as possible in consideration of the early-maturing situation that may occur in the genetic algorithm at the time of calculation.

TABLE 2 genetic Algorithm parameter settings

The application analysis is carried out by taking the Tianjin coastal international airport as an example, the requirements of each airport on the operation condition are not completely the same, and the operation condition can be revised by combining the conditions of the airports when the application is specifically applied to other airports.

(1) Flight landing runways have been determined by regulatory agencies. The invention uses the 16R runway of the Tianjin coastal airport to take off and land.

(2) The time information, model information and stand information of the flight are known. In the research, 20 flights and 28 stand positions are set, and all information of the flights, including time, model and the like, are assumed to be known; it is assumed that flights in this study are all airplanes in the A320 series or B737 series; it is assumed that the number and location of available stands are also known. In this study, a randomly generated flight sequence is used, but actual flight information may be introduced.

(3) The flight keeps constant speed during taxiing on the lanes. Assuming that the fuel consumption in this study is the average fuel consumption, the taxi fuel is only related to the distance of taxi, regardless of the difference in fuel consumption of the aircraft due to operations such as turning, acceleration, and deceleration.

After the assumption conditions are determined, the constraint conditions are explained. The constraints determined in this study are as follows:

(1) a flight can only stop at one stand.

(2) One stand can only be occupied by one flight at a time.

(3) If the same stand is used by two flights consecutively, there is a 35 minute safety interval between the two flights.

(4) The time of day jin coastal airport opening is 5: 00 to 24: 00.

(5) the flight takes the hold of the stand longer than the minimum station-passing time (60 minutes).

(6) Each flight represents a flight pair comprising two tasks, an inbound flight and an outbound flight, and the time that a stand is occupied is after the flight arrives and before the flight departs.

(7) The invention takes an Tianjin coastal airport as an experimental object, and assumes that flights all take off and land from a 16R runway. Because the sliding distance is selected as a main research target during modeling, sliding distance data corresponding to different parking positions of a flight during taking off and landing from a 16R runway needs to be obtained before optimization calculation is performed by using a genetic algorithm.

The flight glide distance data is obtained by taking an Tianjin coastal airport parking bitmap ZBTJ-2 (a satellite map in a Gaode map and a figure 5) in EAIP (2020-10.v1.18 version) as a template and using a ranging tool carried by the Gaode map to perform equal-scale measurement on the flight glide distance (figures 6 and 7).

The invention selects 28 stands in total, all located at Tianjin coastal airport T2 terminal building, wherein, because the fixed 16R runway is selected for taking off and landing, the stands 201 to 218 are closer relative to the runway; the 219 to 228 gate positions are remote relative to the runway.

When the sliding distance is measured, for inbound flights, the starting point of the measurement is the tail end of a 16R runway grounding zone, and the end point is the gallery bridge of the parking space; for an outbound flight, the starting point of the measurement is at the bridge of the aircraft stand and the end point is at the runway head of 16R. When the taxi path is selected, if the inbound flight stops at the station 201 and 211, the taxi path is composed of B4-C2-T3; if the vehicle stops at the 212-219 station, the sliding path consists of B4-C2-N3-Q; if the parking lot is stopped at the station 220 and 228, the sliding path is composed of B4-C2-N3-Q-N5-T5. The selection of the taxi path of the departure flight is consistent with the taxi path of the arrival flight in direction, and considering that taxi conflicts are reduced, the situation that the departure flight is located in the same taxi path is avoided as much as possible during the selection, and another route beside the taxi path of the arrival flight is generally selected. The reference when selecting the glide path diagram is shown in fig. 8.

After all the stand in and out distance measurements were made, the total glide distance data was summarized as shown in table 3. At present, an aircraft stop allocation model is established, key sliding distance data is obtained, and the next section starts to realize the optimization calculation process of the genetic algorithm.

TABLE 3 glide distance summary

The parameter settings of the genetic algorithm are shown in table 2.

The invention simulates the pre-allocation situation of the parking spaces of the Tianjin coastal airport, and supposing that the parking spaces 201 to 228 can be used for the airplane to park, 20 flight pairs which operate in the range of 5 to 24 are randomly generated, the randomly generated flight information of the case is shown in table 4, and the initial parking space information randomly allocated to the flights is shown in table 5.

TABLE 4 flight information Table

TABLE 4 flight information table

TABLE 5 initial stop level table

Pre-distribution and result analysis:

and according to the parameter setting of the table 2, calculating a genetic algorithm on the basis of the flight information of the table 4, and operating a Python program to obtain a result.

The iteration condition of the algorithm is firstly analyzed, and the iteration change of the genetic algorithm is shown in figure 9.

FIG. 9 is a graph of glide distance, ranging from 108 kilometers to 124 kilometers, in increments of 2 kilometers; the horizontal axis represents the range of iteration times from 0 generation to 800 generation, and the iteration times are increased at intervals of 100 generation.

It can be seen from the figure that the algorithm tends to be smooth and has basically converged around 350 generations, and when 800 iterations of the setting are completed, the change of the sliding distance can be ignored, so that the algorithm can be considered to have converged, and an approximate optimal solution is obtained.

The approximately optimal solution obtained through the simulation calculation of the genetic algorithm is the optimized parking space information, and the optimized result is the final result of the pre-allocation of the parking spaces, which is shown in table 6.

TABLE 6 parking space Pre-Allocation results

In order to visually display the optimization calculation process and the result of the genetic algorithm, the data in tables 4, 5 and 6 are plotted into a gate position Gantt chart before optimization and a gate position Gantt chart after optimization for comparing the gate position distribution conditions before and after optimization, as shown in FIGS. 10 and 11.

Firstly, comparing the two graphs, it can be seen that the flights before optimization only meet the requirements of constraint conditions, including exclusivity constraint, safety interval constraint and the like, and the flights are distributed on 14 stand positions, although the constraint conditions are met, the flights are scattered and occupy a plurality of stand positions; and the flights optimized according to the target with the shortest sliding distance are only distributed on 8 stands on the basis of meeting the constraint condition, which shows that the waste of stand resources is reduced, the use efficiency of the stands is increased, the number of the flights borne by each stand is uniform, the configuration of the stand resources of the airport is optimized, and the optimization target of reducing the occupation of the stands is met.

Secondly, the optimization objective of the present study is also based on airport and airline perspectives, considering that the assignment of stops minimizes the taxi distance (equivalent to minimizing taxi time and minimizing taxi fuel), reduces the possibility of the aircraft getting crowded on the road, and saves taxi fuel costs due to taxiing. In order to verify the validity of the genetic algorithm and whether the pre-allocation result is reasonable, the sliding distance data before and after optimization need to be compared, and the data obtained after algorithm calculation is shown in table 7:

TABLE 7 Pre-distribution optimization results

As can be seen from the table, the glide distance corresponding to the initial stand randomly allocated before optimization is 122472 meters; the sliding distance after optimization through calculation of a genetic algorithm is 108268 meters, the total optimized sliding distance is 14204 meters and is about 14 kilometers, and the optimization effect is obvious from the sliding distance.

By referring to assumed data in the course of the flight plan of China civil aviation university, assuming that the aircraft sliding speed is constant at 18 kilometers per hour, the sliding distance is converted into more intuitive sliding time and sliding oil amount, and the sliding time is saved by 47 minutes and the sliding fuel oil is saved by 852 kilograms. The operation conditions of 20 flight pairs of the Tianjin coastal airport are only simulated in the calculation example, in 10 months in 2021, the flight quantity of the Tianjin coastal airport per week can reach 3000, on the basis of the huge flight quantity, the fuel oil capable of being saved is more obvious, and under the background of the rising of the current international fuel oil price, more fuel oil cost for the aviation company is saved.

The method takes the sliding distance as an optimization target, uses a genetic algorithm to carry out optimization calculation, and the obtained optimization result proves that the consumption of sliding fuel oil can be reduced by pre-distributing the parking positions, so that the method has certain practical significance.

Real-time allocation and analysis of results thereof

To simulate a flight delay, a delay condition needs to be generated on the basis of the original flight information. On the basis of flight information for pre-allocation of the parking space, 4 flights are randomly selected from 20 flights, and delay of 10 minutes to 30 minutes is generated respectively. As for how to define the flight delay, the definition of the normal flight in the document of the 'civil aviation flight normal statistical approach' issued by the national civil aviation administration is that the normal flight belongs to the normal flight within the airport ground sliding time specified after the planned closing time of the cabin door, and the normal flight does not have abnormal conditions such as return, standby landing and the like or lands within 10 minutes after the planned opening time.

Although there is a definition of normal flights, in the documentation, the civil aviation administration of China does not specify in detail how long the actual and planned/off-gear times of a flight differ from each other will be considered a flight delay. For the judgment of flight delay, the current civil aviation industry in China combines the actual working condition, and generally thinks that if the actual time of a flight differs from the planned time by more than 30 minutes, the flight delay is regarded as the flight delay. Therefore, in the invention, the time of flight delay is considered to be added on the basis that the original scheduled time of the flight is already 30 minutes, namely, the invention generates 10-30 minutes delay, and actually delays the scheduled gear-withdrawing time and the scheduled gear-shifting time of the flight for 40-60 minutes. The actual flight information after the random generation delays is shown in fig. 12, using Python to write the code.

And sorting the generated flight delay information, wherein: flight 3 is delayed by 16 points; flight 7 delayed by 14 points; flight 11 is delayed by 21 points; flight 17 is delayed by 25 minutes, and details of flights that are delayed are summarized, and the flight delay is shown in table 8.

TABLE 8 flight delay situation table

Thus, the actual flight information after the flight delay is obtained. The delayed flight information is used as an example, the stall allocation model proposed by the research is used, the basic genetic algorithm written by Python is continuously used for calculating the example, and the parameter setting of the genetic algorithm is shown in Table 2. Meanwhile, since the delay condition is generated randomly, the final optimization result obtained by the allocation example in real time does not represent the optimization condition of the research and is only shown as the example. Results of the real-time dispensing are shown in table 9.

TABLE 9 real-time distribution results table

Comparing the pre-allocation results table 7 with the real-time allocation results table 9, wherein only 3 flights have changed, comprises:

the gate for flight 1 changes from 213 to 211;

flight 3's stand changed from 208 to 213;

the stand for flight 7 changes from 213 to 212.

On the basis that small delays occur to 4 flights, only 3 flights need to adjust the stop, and it can be seen that under the condition of small delays, real-time distribution is carried out by using the stop distribution model constructed by the research, only small changes of the stop are generated, and the real-time distribution of the model under the condition of small delays is proved to be effective.

In addition, since real-time allocation and pre-allocation use the same model, the optimization objectives are the same. From the optimization goal of reducing the occupation of the stand, the pre-distributed flights are distributed on 8 stands, and the flights distributed in real time are also distributed on 8 stands, so that the model provided by the research can reduce the occupation of the stands and increase the use efficiency of the stands no matter the model is applied to the pre-distribution or real-time distribution.

From the optimization goal of shortest sliding distance, the data (table 10) of the sliding distances distributed in real time are analyzed, and it can be seen that the sliding distance obtained by pre-distribution is 108268 meters, the sliding distance obtained by real-time distribution is 107737 meters, and the difference between the results of real-time distribution and pre-distribution is small, so that the model can still be proved to reduce the sliding distance by distributing parking spaces, reduce the sliding time of aircrafts in airports and reduce the possibility of lane congestion; and certain taxi fuel oil cost can be saved for the airline company. The parking space allocation model provided by the invention has certain practical significance.

TABLE 10 real-time distribution optimization results

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims

1. An airport real-time parking space allocation method is characterized by comprising the following steps: comprises that

2. The method for allocating the airport real-time parking spaces according to claim 1, wherein: the function corresponding to the optimization target is

f＝f ₁ +f ₂ (1)

f ₁ ＝min[∑ _i∈N ∑ _k∈M x _ik (DA _k +DD _K )] (2)

Wherein the content of the first and second substances,

n denotes flight set (i, j belongs to N; i is 1,2,3 … 20; j is 2,3,4 … 20)

M represents a set of parking positions (k is the same as M; k is 1,2,3 … 28)

DA denotes the set of flight slide-in Distances (DA) _k ∈DA；k∈M；k＝1,2,3…28)

DD representationFlight sliding-out distance set (DD) _k ∈DD；k∈M；k＝1,2,3…28)

f2 denotes maximizing stand use efficiency and reducing stand occupancy.

3. The allocation method of the real-time parking spaces of the airport according to claim 2, characterized in that: the constraint condition is specifically

Wherein the content of the first and second substances,

；

E _jk a catch time indicating that a flight j using the same stand k after flight i enters the stand;

4. The allocation method of the real-time parking spaces of the airport according to claim 1, characterized in that: the step 3 adopts a genetic algorithm for calculation, and specifically comprises the following steps:

step 1: using the sliding distance of each parking space for gene coding;

step 2: generating an initial population in a random mode;

and 3, step 3: setting an individual fitness function in a population;

and 4, step 4: and obtaining an optimal solution through selection, intersection and variation calculation, namely a final result of the pre-allocation of the parking space.

5. The utility model provides an airport real-time parking stall distributor which characterized in that: comprises that

6. An electronic device, characterized in that: comprises that

At least one processor, and

at least one memory communicatively coupled to the processor, wherein:

the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1-6.

7. A non-transitory computer-readable storage medium that, when executed by one or more processors, causes the processors to perform the method of any one of claims 1-6.