JP4355833B2 - Air traffic control business support system, aircraft position prediction method and computer program - Google Patents

Description

  The present invention relates to an air traffic control business support system, a method for predicting the position of an aircraft, and a computer program. More specifically, the present invention relates to an air traffic control business support system for monitoring and controlling the operation of an aircraft during flight from the ground, and the position of the aircraft. The present invention relates to a prediction method and a computer program.

  In conventional air traffic control, aircraft position information obtained from radar is output on the display screen of the control console, and the controller controls the flight plan (flight plan) output from the flight plan information processing system (FDP). The information is used for air traffic control. This method requires the knowledge, experience, and effort of the air traffic controller to operate the air traffic control.

  Therefore, in order to smoothly manage air traffic control, an aircraft route prediction technology using a Kalman filter has been developed as a technology for predicting the position of an aircraft (see, for example, Patent Document 1).

Japanese National Patent Publication No. 10-501059 Japanese Patent Laid-Open No. 9-189761 Yoshinori Uesaka, “Mathematical Foundations of Neurocomputing”, Modern Science, November 1993

  However, in the prior art, the controller predicts the future position of the airplane based on empirical knowledge from only the current position of the aircraft and limited information such as the flight plan. For this reason, when there is a lot of traffic, there has been a problem that the burden on the operation of the controller is great.

  In addition, the aircraft route prediction technology considered so far is based on a rule base, that is, linguistic knowledge, and its effect is questioned from the viewpoint of the effectiveness of the system. In the conventional route prediction technology, there is a problem that the predictable range is limited, for example, it is difficult to predict an aircraft having a complicated movement around an airport.

  Moreover, although the technique which tracks an aircraft using a dynamic model is known (for example, refer patent document 2), since there was no learning function, there existed a problem that an error could not be corrected immediately.

  The present invention has been made in view of such problems, and the object of the present invention is to effectively predict the route of an aircraft in air traffic control and to warn when abnormal operation is observed. An object of the present invention is to provide an air traffic control business support system, a method for predicting the position of an aircraft, and a computer program.

  A system that supports air traffic control work is truly useful because it has a structure that allows non-verbal learning as well as the controller gains experience, ie non-verbal knowledge. It is thought that it can build something.

  The present invention constructs a predictive model having a structure capable of acquiring non-verbal knowledge in order to support the air traffic control work of the controller. By using this prediction model, a system having a function of predicting the position of the aircraft in the controlled airspace and notifying the abnormality when the operation of the aircraft is abnormal is provided.

  In the present invention, in the controlled airspace, route prediction is performed using a prediction model constructed from observation records of radar information. Predict the future position of the aircraft in service by route prediction. If a discrepancy exceeding the threshold is found between the predicted value and the observed value, it is determined that the aircraft's trajectory is different from the normal one, and the controller is notified. The threshold value used in the individual route prediction can be changed according to the judgment of the controller, and can also be changed depending on the aircraft density in the airspace.

According to the first aspect of the present invention, an air traffic control business support system according to the present invention includes a storage unit that stores aircraft position information, and a neural circuit that inputs the position information stored in the storage unit as learning data. Construction means for constructing a prediction model having a learning function for predicting the future position of the aircraft using a network, acquisition means for acquiring new position information of the aircraft, and new information acquired by the acquisition means Prediction means for predicting the future position of the aircraft based on the position information and the prediction model constructed by the construction means, the new position information obtained by the obtaining means, and the prediction by the prediction means A distance calculating means for calculating a distance between the position of the aircraft at the time of obtaining new position information, and a distance calculating means A notification means for notifying information of an aircraft whose distance exceeds a predetermined threshold; an aircraft density calculation means for calculating the density of the aircraft in the airspace to be controlled based on the new position information acquired by the acquisition means; and the aircraft Threshold control means for changing the threshold based on the density of the aircraft calculated by the density calculation means .

  Here, display control means for displaying the future position of the aircraft predicted by the prediction means may be further provided.

Here, the information processing apparatus further includes an input unit that receives an input of information related to the evaluation of the comparison result notified by the notification unit, and the notification unit is based on the information related to the evaluation received from the input unit in response to a previous notification. The aircraft information may be notified.

Further, the threshold value control means may further change the threshold value in accordance with information relating to evaluation input by the input means.

According to a second aspect of the present invention, a method for predicting an aircraft position according to the present invention is a method for predicting an aircraft position, wherein a computer processing device that stores aircraft position information is stored. Constructing a prediction model having a learning function for predicting the future position of the aircraft using a neural network that inputs the position information as learning data; and acquiring new position information of the aircraft; Predicting the future position of the aircraft based on the acquired new position information and the constructed prediction model , the acquired new position information, and the predicted new position information. A step of calculating a distance between the aircraft position at the time of acquisition and information of the aircraft in which the calculated distance exceeds a predetermined threshold Comprising that a step, calculating the density of the aircraft in the airspace control target based on the acquired new position information, and a step of changing the threshold value based on the density of the calculated the aircraft.

According to a third aspect of the present invention, there is provided a computer program according to the present invention, comprising: a computer storing storage means for storing aircraft position information; and a neural network for inputting the position information stored in the storage means as learning data. Construction means for constructing a prediction model having a learning function for predicting the future position of the aircraft using the acquisition means , acquisition means for acquiring new position information of the aircraft, new position information acquired by the acquisition means Prediction means for predicting the future position of the aircraft based on the prediction model constructed by the construction means , the new position information acquired by the acquisition means, and the new position information predicted by the prediction means Distance calculating means for calculating the distance to the aircraft position at the time of obtaining the position, calculated by the distance calculating means Notification means for notifying aircraft information whose distance exceeds a predetermined threshold, aircraft density calculation means for calculating the density of the aircraft in the airspace to be controlled based on the new position information acquired by the acquisition means, and It functions as threshold control means for changing the threshold based on the density of the aircraft calculated by the aircraft density calculation means .

  According to the present invention, it is possible to visually grasp the time change of the air traffic control airspace, and it is possible to reduce the burden on the air traffic controller for airspace monitoring.

  Therefore, it is possible to ensure the safety interval of the aircraft by appropriate and effective control instructions, and support safe and smooth operation of the aircraft.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration of an entire system to which an air traffic control business support system according to the present invention is applied. The overall system 100 is configured by adding an air traffic control business support system 116 to a conventionally known radar information processing system 106, flight plan information processing system 114, and control console 108.

  The radar information processing system 106 is a computer system that processes radar information including flight number and position information of an aircraft. Here, the position information includes the altitude, latitude, longitude, speed vector, and the like of the aircraft.

  The radar information processing system 106 includes radar information RDP (Radar Data Processing System), ARTS (Automated Radar Terminal System), TRAD (Terminal Radar Alphanumeric Display System). Display system), and ODP (Oceanic Air Traffic Data Processing System).

  RDP tracks the aircraft by collating radar information and flight plan information from FDP (Flight Data Processing System), and the flight number of the aircraft is displayed on the radar display screen of the control console 108. This is a system that displays the flight information required for flight altitude control in alphanumeric characters.

  ARTS uses an electronic computer to process aircraft operation information included in radar information of ASR (Airport Surveillance Radar: Airport Surveillance Radar) / SSR (Secondary Surveillance Radar: Secondary Surveillance Radar). This system displays flight information necessary for control such as flight number, flight altitude, and ground speed of aircraft on the radar display screen by collating flight plan information from the FDP with radar information.

  TRAD uses a computer to process radar information input from airport surveillance radar, and displays information such as flight number, current altitude and control seat symbol of each aircraft based on flight plan information on the conventional radar display. This is a system for superimposing and displaying.

  The ODP calculates the position of the aircraft from the information obtained by data link communication and HF communication between the flight plan and the aircraft, and the current position, altitude, It is a system that displays interval, operation information, etc. on a display, and is used for air traffic control at sea. In addition, it displays the operational status of the aircraft based on ADS (Airport Dependent Surveillance) data and CPDLC (Controller Pilot Datalink Communications) information obtained via artificial satellites to improve accuracy, and supports control instructions and judgments. .

  The flight plan information processing system 114 is a computer system that processes flight plan information of an aircraft (IFR aircraft) performing instrument flight submitted in advance. Here, the flight plan information includes the flight number, departure airport and arrival airport (hereinafter referred to as “departure / arrival airport”), call sign, cruise altitude, and the like.

  The flight plan information processing system 114 includes CADIN (Common Aeronautical Data Interchange Network), FDP, and flight plan information database.

  CADIN consists of DTAX (Domestic Telecommunication Automatic Exchange and Aeronautical Data Processing System), AFTAX (Aeronautical Fixed Telecommunication Automatic Exchange and Aeronautical Data Processing System), data terminals installed at each airport, etc. It is a collective term, and is connected to systems such as AFTN (Aeronautical Fixed Telecommunications Network), control information processing system, Japan Meteorological Agency, Defense Agency, and airline company, etc., and relays various information necessary for aircraft operation.

  FDP automatically prints flight plan information for distribution to controllers, and provides flight plan information to other systems such as RDP and ARTS.

  The flight plan information database stores flight plan information.

  In the conventional system, the flight plan information created based on the flight slip (flight plan) submitted before the flight with respect to the aircraft 102 is input to the flight plan information processing system 114. The flight plan information processing system 114 passes the input flight plan information to the radar information processing system 106. The radar information processing system 106 processes the input radar information and the flight plan information received from the flight plan information processing system 114, and displays the above-described various information on the control console 108, thereby allowing the controller to Information is provided.

  The air traffic control business support system 116 receives outputs from the radar information processing system 106 and the flight plan information processing system 114, performs the following processing based on this information, and outputs the result to the control console 108.

  FIG. 2 is a block diagram illustrating an example of a functional configuration of the air traffic control business support system 116.

  The air traffic control business support system 116 includes an input device 202, a learning data storage unit 204, a position information extraction unit 206, a prediction model construction unit 210, a prediction model storage unit 212, a route prediction unit 213, a comparison unit 214, a notification unit 218, A display control unit 222 and a radar information acquisition unit 224 are included.

  The input device 202 includes a pointing device such as a keyboard and a mouse, and receives input from the user.

  The learning data storage unit 204 stores operation records. The flight record is a time-series record of aircraft position information included in radar information.

  The position information extraction unit 206 extracts the position information stored in the learning data storage unit 204.

  The prediction model construction unit 210 constructs a prediction model of the future position of the aircraft based on the position information extracted by the position information extraction unit 206.

  The prediction model storage unit 212 stores a prediction model of the future position of the aircraft constructed by the prediction model construction unit 210.

  The route prediction unit 213 predicts a future position based on the prediction model and the latest position information of the aircraft to be predicted.

  The comparison unit 214 compares the future predicted position of the aircraft with the actual aircraft track, and whether there is a divergence exceeding a predetermined threshold between the aircraft position indicated by the prediction model and the actual aircraft position. Is judged. The comparison unit 214 holds information related to the evaluation of the notification result of the flight plan information support system 116.

  The display control unit 222 displays the position of the aircraft indicated by the prediction model constructed by the prediction model construction unit 210 on the screen of the control console 108.

  The notification unit 218 notifies the controller of the comparison result by the comparison unit 214 via the display control unit 222. For example, as a result of the comparison, the notifying unit 218 determines that an aircraft having a deviation exceeding a predetermined threshold between the position indicated by the prediction model and the position indicated by the flight plan information as an abnormal aircraft is sent to the controller via the display control unit 220. Notice.

  The present invention can be realized by providing the computer program to a computer system from a storage medium storing a computer program that realizes each function of the above-described functional blocks. A processing device such as a CPU of a computer system reads and executes a computer program stored in the recording medium. In this case, the function of the present invention is also realized by the computer program code itself read from the storage medium. Further, the above functions may be realized by an operating system or the like running on the computer system performing part or all of the actual processing based on the instruction of the program code read by the computer system.

  Examples of the recording medium include a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a CD-R / W, a DVD, a nonvolatile memory card, and a ROM.

  Next, with reference to the flowchart of FIG. 3, the procedure of the process performed by the air traffic control business support system according to the present embodiment will be described.

  Radar information of the aircraft in operation is input to the radar information processing system 106 at 10-second intervals (S302). Radar information includes aircraft position information.

  The air traffic control business support system 116 extracts radar information accumulated in the radar information processing system 106 every minute. Next, the flight plan database is referred to based on the flight name of the radar information, and the corresponding flight plan information is extracted.

  Next, the radar information is sorted for each departure and arrival airport included in the extracted flight plan information and stored in the learning data storage unit 204 (S304).

  The position information extraction unit 206 extracts the position information stored in the learning data storage unit 204 at a predetermined timing (S306). Here, for each prediction model to be constructed, learning data in the target time zone of the departure / arrival airport pair is extracted. The prediction model construction unit 210 constructs a prediction model using this information (S308), and stores it in the prediction model storage unit 212 (S310). In this way, a prediction model having a learning function is constructed by using the learning data.

  The prediction model is constructed for each airport and each coordinate component. The reason why the memory and prediction model is constructed for each airport is because the route of the aircraft is generally determined for each airport pair (for example, between Haneda and Fukuoka).

  The route prediction unit 213 acquires position information of an aircraft whose future operation route is to be predicted at the timing of extracting position information from the radar information (S312). Further, a prediction model of the same departure and arrival airport as the aircraft whose flight route is to be predicted is obtained from the prediction model storage unit 212 (S314). As will be described later, a plurality of prediction models are read out from the prediction model storage unit 212. And the future position of the aircraft in operation is predicted using the acquired latest position information and prediction model (S316). In step S316, since the route prediction is performed using a plurality of prediction models, the predicted position is given as a plurality of points. A sphere including the prediction result is set as a prediction range. The predicted range is maintained for a certain period of time for comparison with the actual track.

  In step S318, the prediction range derived from the prediction result of the prediction model is output from the display control unit 222 to the control console 108.

  On the other hand, the comparison unit 214 acquires position information of an aircraft for which a future operation route is to be predicted at the timing of extracting position information from radar information. Moreover, the information of the prediction range of the aircraft at the time of acquiring the latest position information is acquired from the prediction range information of the aircraft predicted by the route prediction unit 213 in the past. Then, the acquired prediction range is compared with the actual aircraft position information (S322). If the difference between the actual aircraft position and the predicted range is equal to or greater than a predetermined value (hereinafter referred to as “deviation allowable range”) as a result of the comparison, it is determined that the aircraft is operating abnormally, and the display controller 222 controls the control console. The controller is warned by outputting the fact to 108 (S324).

  Next, a method for constructing a prediction model will be specifically described.

  The air traffic control business support system 116 creates a prediction model using neural network theory in order to predict the position of a future aircraft using the position information in the radar information as learning data. A prediction model is created for each airport.

  FIG. 4 shows an example of the timing and time interval for creating the prediction model. The air traffic control business support system creates a 12-hour model every 12 hours, creates a 72-hour model every 24 hours, and stores the model in the prediction model storage unit 212.

  In the 12-hour model, the latest 12 hours of data are used as learning data at 0:00 and 12:00 every day. In the 72-hour model, the data for the latest three days (72 hours) is set as learning data at 0:00 every day.

  An example of the structure of a prediction model using a neural network is shown in FIG. In order to predict position information, one model is prepared for each of the three components of three-dimensional coordinates (x, y, z). Future three-dimensional coordinates are predicted by the prediction model, but a model is prepared for each of the x, y, and z components of the coordinates.

Next, a method for predicting the position of the aircraft after t = 1 minute using a neural network will be described with reference to FIG. In the example shown in FIG. 5, the data set includes components of three-dimensional coordinates (x, y, z). As the learning data, the current data set t-1 (x, y, z), the data set t-2 (x, y, z) one minute before the current time, and the data set t-3 two minutes before the current time (x, y, z), four data sets of data set t-4 (x, y, z) three minutes before the current time are extracted. This data set consists of data for each component (d (t−1, x), d (t−1, y), d (t−1, z),..., D (t−4, z)) Is broken down into Here, d (t−n, i) indicates the data of the i component n−1 minutes before the present time. In this way, the position information of the same aircraft for the past four time points at the present time (that is, data from three minutes before the current zero minutes) is input to the input layer of the neural network model. Next, in the hidden layer (1), data h ₁ (1), h ₁ (2),..., H ₁ (k) are created using the data of the input layer. Next, in the hidden layer (2), data h ₂ (1), h ₂ (2),..., H ₂ (l) are created using the data of the input layer. In this way, the position data d (t, x), d (t, y), and d (t, z) after one minute are output to the output layer as prediction targets. The arithmetic processing in such a neural network model is described in Non-Patent Document 1, for example. For example, as shown in FIG. 6, when focusing on h ₁ of the input layer (d layer) and the hidden layer (1), h ₁ (1) is the parameter w (1), w (2), w (3), ... expressed by linear combination of input layer data weighted by w (n). Therefore, h ₁ (i) can be expressed by equation (1).

When h ₁ (1) exceeds a certain value, h ₁ (1) sends a signal to the _second hidden layer h ₂ . h ₁ and h ₂ are also linked by parameters.

In the initial stage of the model, the parameters and the threshold values of the cells h ₁ and h ₂ are set at random. Even if it is input in this state, only a good number is output to the output layer. Therefore, using the learning data set (the past data set whose output results are known in advance), the error between the output from the output layer and the actual results is reduced in all learning data sets. Adjust parameters and thresholds between cells.

  The learning data stored in step S304 in FIG. 3 is a set of “correct answer when an input data is input”, that is, a question and an answer. For example, when predicting the position of 10:04 using four data of 10:00, 10: 1, 10:02, 10:03, the problem is the data from 10:00 to 10:03 The answer will be 10:04.

  Since the problem and answer of this learning data may vary depending on time and weather, create different models using different learning data and select the one with the smallest error as shown in FIG. .

  In the example shown in FIG. 5, an example of a neural network structure having no loop is shown, but a neural network model having a loop may be used in order to improve learning efficiency.

Next, a method for determining the future prediction range of the aircraft will be described in more detail. The route prediction unit 213 selects the following four prediction models corresponding to the departure airport and arrival airport information of the aircraft (prediction target aircraft) displayed on the control console.
(1) The last 12 hour model created. For example, if the prediction time is 15:00, the model is created using the position information from 00:00 to 12:00 on the same day as learning data.
(2) A 12-hour model created using the location information of the time zone including the same time of the previous day as learning data.
(3) A 12-hour model created using the location information of the time zone including the same time 7 days ago as learning data.
(4) A 72-hour model created using the location information of the same day in the previous year and two days before and after as learning data.

  Then, the smallest sphere that encloses four points of each prediction model is set as the prediction range.

  For example, if only the model (1) is used, the prediction depends only on the data of the same day, and it is possible to consider the case where a route with a different route depending on the day of the week matches, the difference in route between daytime and nighttime, etc. Disappear. Therefore, using the same time zone of the previous day and data of the same day of the week for the prediction time zone, the result is obtained from the greatest common divisor. As described above, by using a plurality of prediction models, it is possible to determine the prediction range in consideration of weather, time fluctuation, day of the week fluctuation, and seasonal fluctuation.

  FIG. 7 shows an example of a screen displaying the prediction range obtained in this way. The black square in the figure indicates the actual position of the aircraft with flight number XXX021, and the direction of travel is indicated by a short line extending from the square to the upper left. Further, the inner diameter of the circle drawn with a bold line indicates the predicted range, and the outer diameter indicates the allowable deviation range.

  Next, with reference to FIG. 8, a process for determining an abnormal operation will be described.

  One minute after the prediction, an abnormal operation can be determined. In order to determine an abnormal operating aircraft that operates differently from the normal operating mode, a range (deviation allowable range) that allows a deviation between the predicted range and the actual position is set. As a result, even if the actual aircraft position falls outside the predicted range, it is judged as normal operation if it is within the deviation allowance range, or it is not within the deviation allowance range even if the actual aircraft position is within the forecast range. Judged as abnormal operation.

  The distance between the sphere and the actual position is set as l, and the allowable deviation range δ is compared. As an initial value, the allowable deviation range δ is a radius of a sphere of the prediction range.

  As shown in FIG. 8A, when l ≦ δ, it is determined that the operation is normal. On the other hand, if l> δ as shown in FIG.

  FIG. 9 shows an example of a screen displaying a warning. In the figure, the white square indicates the actual position of the aircraft with flight number XXX021, and the direction of travel is indicated by a short line extending upward from the square. Further, the inner diameter of the circle drawn with a bold line indicates the predicted range, and the outer shape indicates the allowable deviation range. The □ mark changes from black to white, and the □ mark is surrounded by a thin line circle to indicate abnormal operation.

  In addition, you may perform the warning showing abnormal operation by outputting a warning sound from the audio | voice output apparatus (not shown) with which the control console was equipped.

  Further, in the process of displaying the comparison result between the prediction range and the actual position, information indicating an error between the prediction range and the actual position may be displayed.

  In addition, in the case of indicating an axial shift from the front to the back of the drawing, information indicating that may be displayed in proximity to the □ mark. For example, the letter “H” can be displayed when the actual position is higher than the allowable deviation range, and the letter “L” can be displayed when the actual position is lower than the allowable deviation range.

(Second Embodiment)
Next, with reference to FIGS. 2 and 3, an embodiment in which the threshold value is changed based on the determination of the controller will be described.

  When a warning is given by the process of step S324, the controller determines whether the warning is correct. If the controller who is notified that the control airspace is abnormal determines that the determination result is incorrect, he / she inputs information indicating that the determination is incorrect from the input device 202 to the air traffic control support system 116. be able to. By inputting this information, the comparison unit 214 increases the allowable deviation range δ and sets it to the value of the distance from the center of the prediction range sphere to the current position.

  In addition, even if the controller has not been notified of abnormal operation from the system, if the controller determines that the controlled airspace is abnormal, the air traffic control business support system will confirm that the controlled airspace is abnormally operated. Information to be shown can be input from the input device 202. By inputting this information, the comparison unit 214 lowers the value of the allowable deviation range δ from the original value and sets it to a value smaller than the distance from the center of the prediction range sphere to the current position. In this way, a further learning function is added to the prediction model.

(Third embodiment)
Next, another embodiment for feeding back the controller's determination will be described.

  When receiving an input from the controller as an unnecessary warning, the comparison unit 214 sets δ = 1. Further, the aircraft density ω in the controlled sector is acquired from the radar information at the time of the warning, and stored in association with the value of the allowable deviation range δ. The aircraft density can be obtained by dividing the number of aircraft detected by radar information by the area of the airspace to be controlled. This process is repeated, and when two or more pieces of controller judgment information are obtained, δ is obtained as a function of ω.

  FIG. 10 is a diagram illustrating an example of a method for setting the allowable deviation range δ. In the initial state, the threshold δ is constant as shown in FIG. When the function of ω is obtained as a function as shown in FIG. 10B by the deviation allowable range δ stored over a plurality of aircraft densities, the abnormal operation is subsequently determined using this function.

  That is, in step S322 in FIG. 3, the comparison unit 214 obtains the aircraft density ω from the radar information, and obtains a deviation allowable range δ corresponding to this ω from the above function. Then, the actual position of the aircraft and the allowable deviation range δ are compared. If the function shown in FIG. 10B is obtained, if the density of the aircraft in the controlled airspace is low, a large divergence between the actual position of the aircraft and the predicted position is allowed, and the density of the aircraft is high. In this case, it is possible to strictly determine the deviation between the actual position of the aircraft and the predicted position. In this way, a further learning function is added to the prediction model.

  According to the embodiment described above, in the air traffic control, the future position is predicted using the prediction model constructed as the learning data based on the position information of the aircraft in the controlled airspace and the judgment of the controller, and there is an abnormality in the operation of the aircraft. By notifying the controller when a problem is detected, a system that improves the reliability of safer and more efficient aircraft operation can be provided.

  Various modifications other than those described above are possible. However, as long as it is based on the technical idea described in the claims, the modifications are within the technical scope of the present invention.

It is a block diagram which shows the structure of the whole system to which the air traffic control work support system which concerns on this invention is applied. It is a block diagram which shows an example of a function structure of an air traffic control work support system. It is a flowchart which shows the procedure of the process implemented by the air traffic control work support system which concerns on this embodiment. It is a figure which shows an example of the timing and time interval which create a prediction model. It is a figure which shows the structure of the prediction model using the neural network which concerns on one Embodiment of this invention. It is a figure which shows the example of the arithmetic processing in a neural network model. It is a figure which shows the example which displays the prediction range which concerns on one Embodiment of this invention. It is a figure which shows the process which determines the abnormal operation which concerns on one Embodiment of this invention. It is a figure which shows the example which displays the warning which concerns on one Embodiment of this invention. It is a figure which shows the relationship between threshold value (delta) and aircraft density which concern on one Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Overall system 102 Aircraft 106 Radar information processing system 114 Flight plan information processing system 108 Control table 202 Input device 204 Learning data storage unit 206 Location information extraction unit 210 Prediction model construction unit 212 Prediction model storage unit 213 Route prediction unit 214 Comparison unit 218 Notification unit 222 Display control unit 224 Radar information acquisition unit

Claims (6)

Storage means for storing aircraft position information;
Construction means for constructing a prediction model having a learning function for predicting the future position of the aircraft using a neural network that inputs the position information accumulated in the accumulation means as learning data;
Obtaining means for obtaining new position information of the aircraft;
Prediction means for predicting the future position of the aircraft based on the new position information acquired by the acquisition means and the prediction model built by the construction means ;
Distance calculation means for calculating a distance between the new position information acquired by the acquisition means and the position of the aircraft at the time of acquiring the new position information predicted by the prediction means;
A notifying means for notifying information of an aircraft whose distance calculated by the distance calculating means exceeds a predetermined threshold;
Aircraft density calculation means for calculating the density of the aircraft in the airspace to be controlled based on the new position information acquired by the acquisition means;
An air traffic control business support system, comprising: threshold control means for changing the threshold based on the aircraft density calculated by the aircraft density calculation means .   The air traffic control business support system according to claim 1, further comprising display control means for displaying a future position of the aircraft predicted by the prediction means. The information processing apparatus further includes input means for receiving information related to the evaluation of the comparison result notified by the notification means, and the notification means is configured to receive the aircraft based on the information related to the evaluation received from the input means in response to a previous notification. The air traffic control business support system according to claim 1 or 2 , characterized in that: It said threshold control means, air traffic control service support system according to claim 3, characterized by further changing the threshold value in accordance with the information on the evaluation inputted by the input means. A method for predicting the position of an aircraft, wherein a computer processing device that stores aircraft position information includes:
Constructing a prediction model having a learning function for predicting the future position of the aircraft using a neural network that inputs the accumulated position information as learning data;
Obtaining new position information of the aircraft;
Predicting the future position of the aircraft based on the acquired new position information and the constructed prediction model ;
Calculating a distance between the acquired new position information and the position of the aircraft at the time of acquiring the predicted new position information;
Notifying aircraft information where the calculated distance exceeds a predetermined threshold;
Calculating the density of the aircraft in the controlled airspace based on the acquired new location information;
Changing the threshold based on the calculated density of the aircraft . Computer
Accumulation means for accumulating aircraft position information,
Construction means for constructing a prediction model having a learning function for predicting the future position of the aircraft using a neural network that inputs the position information accumulated in the accumulation means as learning data;
Acquisition means for acquiring new position information of the aircraft ;
Prediction means for predicting the future position of the aircraft based on the new position information acquired by the acquisition means and the prediction model established by the construction means ;
Distance calculation means for calculating a distance between the new position information acquired by the acquisition means and the position of the aircraft at the time of acquiring the new position information predicted by the prediction means;
A notification means for notifying information of an aircraft whose distance calculated by the distance calculation means exceeds a predetermined threshold;
Aircraft density calculation means for calculating the density of the aircraft in the airspace to be controlled based on the new position information acquired by the acquisition means; and
A computer program that functions as threshold control means for changing the threshold based on the density of the aircraft calculated by the aircraft density calculation means .

Images