Flight trajectory prediction method, system and device and medium

Abstract

The invention relates to the technical field of aircraft trajectory prediction, and particularly relates to a flight trajectory prediction method, system and device and a medium, and the method specifically comprises the steps: obtaining a data sequence, a control sequence and a state; the method comprises the following steps: constructing a FlightKooba model based on an HIPPO module and a Kooba module; according to the data sequence, a prediction coefficient is iteratively calculated by adopting an HIPPO module, and three matrixes are obtained after calculation by a Kooba module; splicing two matrixes to form a prediction matrix; according to the control sequence, a Kooba module is adopted to calculate a control coefficient through back propagation; obtaining a control sequence of a future moment according to the control coefficient and the control sequence; according to the prediction matrix, the state and the control sequence of the future moment, calculating to obtain the state of the future moment; predicting the flight path of the aircraft according to the state of the future moment and the coefficient matrix; the problem of low accuracy of flight path prediction in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft trajectory prediction, and in particular to a flight trajectory prediction method, system, device and medium. Background Art

[0002] A flight trajectory is the path of successive positions an aircraft traverses in three-dimensional space from takeoff to landing. It typically includes multi-dimensional information such as time, space, and dynamic maneuvers. For example, a civil airliner's trajectory includes phases such as takeoff, climb, cruise, descent, and landing; a fighter jet's trajectory may include complex maneuvers such as cobra maneuvers, tailspins, and high-angle-of-attack maneuvers, while helicopters also have specialized trajectories such as hovering and sideways flight. Flight trajectory information is crucial and plays an indispensable role in various fields.

[0003] In terms of civil aviation management: 1. Flight route optimization: Using four-dimensional trajectory (4DT) technology to predict flight arrival times, the system dynamically adjusts routes to avoid congested airspace, reducing fuel consumption and delays. For example, a Shenzhen-Kunming flight took a curved route due to mountainous terrain and weather conditions, and real-time trajectory synchronization technology was used to optimize the route. 2. Airspace resource allocation: The trajectory-based operation (TBO) mode optimizes traffic flow sequences by collaboratively planning conflict-free trajectories.

[0004] 3. Safety supervision and emergency response: Flight trajectory data is transmitted to the ground control center in real time. Radar and ADS-B technology are used to monitor deviations and assist in accident investigation and emergency decision-making.

[0005] 4. Flight Quality Analysis: By quantifying the difference between the actual and expected paths of an aircraft and combining it with multi-dimensional data models, this provides a scientific basis for flight control, safety assessment, and system improvements.

[0006] In addition, in terms of drone applications: 1. Agricultural plant protection: Drones spray pesticides based on horizontal trajectory, using path planning to cover farmland and reduce repetitive work. For example, multi-rotor drones use gridded paths to optimize spraying efficiency. 2. Logistics and delivery: Urban logistics drones use the A* algorithm to avoid buildings and adjust their routes based on real-time weather data to improve delivery efficiency. For example, JD.com drones use dynamic trajectory planning to deliver supplies in mountainous areas. 3. Aerial photography and mapping: Fixed-wing drones use serpentine or spiral trajectories to capture high-resolution images and combine them with SLAM technology to construct 3D maps. For example, DJI drones use pre-set trajectories to capture panoramic data during disaster assessments.

[0007] In terms of transportation and urban management, drone traffic control uses flight trajectory big data to analyze hotspots and optimize airspace allocation. For example, the FAA monitors low-altitude aircraft through the Unmanned Aerial Traffic Management (UTM) system.

[0008] In these applications, flight trajectory is the core data, which shows the core position of flight trajectory data in various aircraft applications.

[0009] Traditionally, traditional dynamic models or purely data-driven models are generally used to predict the flight trajectory of an aircraft. However, dynamic models rely too much on precise physical equations and have difficulty handling complex maneuvers (such as cobra maneuvers and high angle of attack maneuvers) and environmental disturbances. Purely data-driven models lack physical interpretability, have weak online update capabilities, and have difficulty integrating control inputs. As a result, the above models are unable to accurately predict the flight trajectory.

[0010] Therefore, we propose a method that can accurately predict the flight trajectory of an aircraft. Summary of the Invention

[0011] The present invention solves the problem of low accuracy of flight trajectory prediction in the prior art by providing a flight trajectory prediction method, system, device and medium, and achieves the technical effect of improving the accuracy of prediction results.

[0012] In a first aspect, the present invention provides a flight trajectory prediction method, specifically comprising: Get the data sequence of the aircraft at the current moment and the current control sequence of the aircraft ; Build a FlightKooba model based on the HIPPO (High-order Polynomial Projection Operators) module and the Kooba module; According to the data sequence of the aircraft at the current moment , use the HIPPO module to iteratively calculate the prediction coefficient ; According to the prediction coefficient , the coefficient matrix is ​​obtained after calculation by the Kooba module , state transfer matrix and the control matrix ; The state transfer matrix and control matrix Splicing to form a prediction matrix ; According to the current control sequence of the aircraft , the control coefficient is obtained by back propagation calculation using the Kooba module ; According to the control coefficient and the current control sequence of the aircraft , obtain the control sequence of the aircraft at the future time ; According to the prediction matrix and the control sequence of the aircraft at future times , and introduce state variables , calculate the state of the aircraft at the future time ; According to the state of the aircraft at the future time and coefficient matrix Predicting the flight trajectory of an aircraft .

[0013] Furthermore, the present invention provides a specific step of iteratively calculating a prediction coefficient using a HIPPO module based on a data sequence of an aircraft at a current moment, including: Construct including basis functions and the prediction coefficients The observation function ; The prediction coefficients are obtained by using Fourier series to find the coefficients Expressions of According to the discretized data sequence of the aircraft at the current moment , calculate the prediction coefficient Discrete expression of ; According to the prediction coefficient The discrete expression of .

[0014] Furthermore, the present invention provides a Kooba module, comprising an InBlock unit and a KBlock unit; The InBlock unit is used to obtain the control coefficient through back propagation calculation ; The KBlock unit is used to predict the , calculate the coefficient matrix , state transfer matrix and the control matrix , and is used to transform the state transfer matrix and control matrix Splicing to form a prediction matrix .

[0015] Furthermore, the present invention provides a method for obtaining a control sequence of an aircraft at a future time based on a control coefficient and a control sequence of the aircraft at a current time, specifically comprising: , Where, is the control sequence of the aircraft at the current moment.

[0016] Furthermore, the present invention provides a method for calculating the state of an aircraft at a future time based on a prediction matrix and a control sequence of an aircraft at a future time, and introducing state variables, which specifically includes: , Where, is a state variable.

[0017] Furthermore, the present invention provides a method for predicting the flight trajectory of an aircraft based on the aircraft's future state and coefficient matrix, which specifically includes: , Where, Predicted flight trajectory.

[0018] In a second aspect, the present invention provides a flight trajectory prediction system, comprising a data acquisition unit for obtaining a data sequence of the aircraft at the current moment. and the current control sequence of the aircraft ; Model building processing unit, used to build the FlightKooba model and the data sequence of the aircraft at the current moment and the current control sequence of the aircraft Calculate the state of the aircraft at the future time and coefficient matrix ; The flight trajectory prediction unit is used to predict the future state of the aircraft and coefficient matrix Predicting the flight trajectory of an aircraft .

[0019] Furthermore, the present invention provides a model building processing unit, including a HIPPO module, for iteratively calculating the prediction coefficients ; InBlock unit, used to obtain the control coefficient through back propagation calculation ; KBlock unit, used to calculate the prediction coefficients , calculate the coefficient matrix , state transfer matrix and the control matrix , and is used to transform the state transfer matrix and control matrix Splicing to form a prediction matrix ; Prediction unit, based on control coefficient and the current control sequence of the aircraft , obtain the control sequence of the aircraft at the future time ; According to the prediction matrix , the current status of the aircraft and the control sequence of the aircraft at future times , calculate the state of the aircraft at the future time .

[0020] In a third aspect, the present invention provides an electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to execute to implement a flight trajectory prediction method.

[0021] In a fourth aspect, the present invention provides a non-temporary computer-readable storage medium, which, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to execute a flight trajectory prediction method.

[0022] One or more technical solutions provided in the present invention have at least the following technical effects or advantages: 1. The FlightKooba (flight trajectory prediction) model, which combines the advantages of the HIPPO module (high-order polynomial projection operator module) and the Kooba module (prediction module), effectively solves the problem that traditional dynamics models have difficulty handling complex maneuvers (such as cobra maneuvers and high angle of attack maneuvers) and environmental disturbances by accurately calculating the state transition matrix A and the control matrix B and integrating them into the prediction matrix K. Furthermore, compared to purely data-driven models, the FlightKooba model not only enhances physical interpretability but also has stronger online update capabilities, enabling better integration of control inputs, thereby significantly improving the accuracy of flight trajectory predictions.

[0023] 2. Because the FlightKooba model can process and analyze aircraft data and control sequences in real time, it provides strong support for dynamic adjustments to flight missions. By identifying potential conflicts (such as route intersections and areas of severe weather) in advance and issuing timely warnings, the model allows the flight control system or pilot to react quickly, optimize the flight path, and reduce unnecessary delays and fuel consumption. This capability greatly improves the accuracy of flight mission planning and the efficiency of execution, making flight operations more efficient and economical. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 A schematic flow chart of a method of the present invention is shown; Figure 2 This is a schematic diagram of the FlightKooba model structure of the present invention. DETAILED DESCRIPTION

[0026] The embodiments of the present invention solve the technical problem in the prior art that traditional dynamic models are difficult to handle complex maneuvers (such as cobra maneuvers and high angle of attack maneuvers) and environmental disturbances by providing a flight trajectory prediction method, system, device and medium.

[0027] The technical solution of the present invention is to solve the above technical problems, and the overall idea is as follows: First, obtain the data sequence of the aircraft at the current moment and the current control sequence of the aircraft ; Then build the FlightKooba model based on the HIPPO module and the Kooba module; According to the data sequence of the aircraft at the current moment , use the HIPPO module to iteratively calculate the prediction coefficient ; According to the prediction coefficient , the coefficient matrix is ​​obtained after calculation by the Kooba module , state transfer matrix and the control matrix ; The state transfer matrix and control matrix Splicing to form a prediction matrix ; According to the current control sequence of the aircraft , the control coefficient is obtained by back propagation calculation using the Kooba module ; According to the control coefficient and the current control sequence of the aircraft , obtain the control sequence of the aircraft at the future time ; According to the prediction matrix and the control sequence of the aircraft at future times , and introduce state variables , calculate the state of the aircraft at the future time ; Finally, according to the state of the aircraft at the future moment and coefficient matrix Predicting the flight trajectory of an aircraft .

[0028] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0029] First, the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. Furthermore, the character " / " in this document generally indicates an "or" relationship between the associated objects.

[0030] Example 1 As attached Figure 1-Figure 2 As shown, first obtain the data sequence of the aircraft at the current moment and the control sequence of the aircraft at the current moment; wherein the data sequence of the aircraft at the current moment , represents various parameters that need to be predicted for the future state of the aircraft, such as average altitude, heading, longitude and latitude. These parameters can be obtained by sampling historical flight data from normal flight training, for example, a sampling sequence of the values ​​of various indicators of the flight system during a normal flight training; The current control sequence of the aircraft , represents various parameters related to the aircraft state that do not need to be predicted, such as flight speed and various engine parameters. These parameters can also come from the sampling of various parameters of the flight system during a normal flight. It is foreseeable that choosing more appropriate control features can achieve better prediction results; According to the data sequence of the aircraft at the current moment , use the HIPPO module to iteratively calculate the prediction coefficient ; This step is the training step, used to train the prediction coefficient ; The prediction coefficient These coefficients are primarily used to represent the coefficients preceding each basis function in the observation function. These coefficients approximate the actual observation function through polynomial approximation, thereby capturing the time-varying pattern of the aircraft's state. These coefficients are derived by analyzing input data sequences such as average altitude, heading, latitude, and longitude.

[0031] According to the prediction coefficient , the coefficient matrix is ​​obtained after calculation by the Kooba module , state transfer matrix and the control matrix ; The state transfer matrix and control matrix Splicing to form a prediction matrix ; According to the current control sequence of the aircraft , the control coefficient is obtained by back propagation calculation using the Kooba module ; This step is the training step, used to train the control coefficient , where the control coefficient Used to measure the impact of external control inputs on the system state. For example, during the operation of an aircraft, engine thrust, rudder angles, etc. are all adjustable control parameters that directly affect the aircraft's motion state. These coefficients are calculated by processing the control sequence (such as flight speed, various engine parameters, etc.) through the Kooba module and using the backpropagation algorithm.

[0032] In addition, the Kooba module includes an InBlock unit (control unit) for obtaining the control coefficient through back propagation calculation ; The KBlock unit (prediction unit) is used to calculate the prediction coefficients , calculate the coefficient matrix , state transfer matrix and the control matrix , and is used to transform the state transfer matrix and control matrix Splicing to form a prediction matrix .

[0033] According to the control coefficient and the current control sequence of the aircraft , obtain the control sequence of the aircraft at the future time ; , Where, is the control sequence of the aircraft at the current moment.

[0034] According to the prediction matrix and the control sequence of the aircraft at future times , and introduce state variables , calculate the state of the aircraft at the future time ; , Where, is a state variable.

[0035] Finally, according to the state of the aircraft at the future moment and coefficient matrix Predicting the flight trajectory of an aircraft ; , Where, is the predicted flight trajectory.

[0036] In addition, for a practical embodiment, if we want to predict the data sequence of a certain aircraft A data in, such as average altitude , because during the flight, It changes over time, but the data we get (which has been normalized) is sampled data , the selected control sequence There are four features: E1 FFlow, E1 OilT, E1 OilP, and E1 RPM. The corresponding sampling data values ​​are: ; ; ; .

[0037] Now we need to predict , that is, predict one step forward.

[0038] 1. First, according to Iterate out the coefficients ; 2. Then according to generate 、 、 ; 3. 、 Stitched together to form ; 4. Input the first 6 data of the control sequence into the Kooba module for training, and get ; 5. Then finally according to get ; 6. Then according to Predict the state for 126 steps ; 7. Finally, according to , calculate the average altitude , that is, the predicted result of the average altitude in step 126 is obtained.

[0039] It should be noted that for a data sequence that is continuous in time , predict each data value in the future time step, and then the connection is the flight trajectory of the aircraft.

[0040] The technical solutions in the above embodiments of the present invention have at least the following technical effects or advantages: The FlightKooba model, which combines the advantages of the HIPPO and Kooba modules, effectively addresses the difficulties traditional dynamics models face in handling complex maneuvers (such as the Cobra maneuver and high angle of attack maneuvers) and environmental disturbances by precisely calculating the state transition matrix A and the control matrix B and integrating them into the prediction matrix K. Furthermore, compared to purely data-driven models, the FlightKooba model not only enhances physical interpretability but also possesses stronger online update capabilities, enabling better integration of control inputs and significantly improving the accuracy of flight trajectory predictions.

[0041] Example 2 For the specific calculation process of iteratively calculating the prediction coefficient using the HIPPO module, we first choose to introduce the polynomial approximation method to approximate the observation function: , in, It's about time The basis functions of is the coefficient before each basis function. In this formula, we have no way of knowing the prediction coefficient The value of needs to be calculated. In order to facilitate the calculation, we stipulate the basis function It needs to satisfy the properties of the standard orthogonal basis (Legendre polynomials or Fourier basis in Hippo can be selected here). So far, we can refer to the method of Hippo theory to Solve it.

[0042] Refer to the Fourier series to find the coefficients We multiply the left and right sides of formula (1) by Conjugate to get the coefficients The calculation formula (assuming The domain of , the actual upper and lower limits will be determined based on (The specific situation may change accordingly): , In order to make the above formula fit more general time , find one arrive Mapping Then the coefficient calculation formula is obtained: , Then, taking the derivative of both sides of equation (3) and using the method of integration by parts, we can obtain: , The following calculations depend on and The specific form of the coefficient is suspended here. derivation of , because we now need to determine specific form.

[0043] Therefore, the SSM (State Space Model) is combined with the Koopman theory, so here We choose the Legendre polynomial with modified coefficients (whose domain is ,also is the boundary value of Legendre polynomial) that is: , in is a Legendre polynomial, because there is a recurrence relationship between the derivatives of Legendre polynomials: , This relationship will play a big role in combining SSM and Koopman theoretical methods later.

[0044] Now introduce the mapping ,Will Mapped to (meaning only focus on the length ), and then put it into formula (4) to get: , There is a recursive formula in Legendre polynomials: , Continuing to iterate, we can get: , Among them, when When it is an even number, otherwise .

[0045] Therefore, formula (7) can be further deduced: , From formula (1), we can get: , So that : , Substituting (12) into (10), we have: , in, This formula, due to It represents our The number of terms to be approximated by the polynomial, so we can choose is an even number, that is, it is approximated by an even number of terms. On this basis, and Must be one odd and one even, when When it is an even number, If is an even number, , ;when When is an odd number, If is an odd number, , .Right now Therefore, formula (13) can be simplified as: , By introducing the matrix, the above formula can be transformed into: , in, ,and: , , Introducing a scaling factor , making , the calculation result is: , , Now we want to solve equation (15). To facilitate the iterative solution by computer, we discretize the equation and first process the left side of the equation. Approximately and The slope between two points is: , Then we come to the discretization on the right side of the equation. Since the data we get is discrete, we think that between is a constant value, so we can Discrete ,and There are two ways to discretize , one is to think that between, , the other is that , that is, the two values ​​of the front and back endpoints of the interval, At this time, if you consider , we can get: , Rearranging the formula, we can get: , At this time, if you consider , we can get: , Rearranging the formula, we can get: , With the above two equations (22) and (24), the third estimate emerges: The value approach is to consider that: , This approximate estimation method is more reasonable than any of the previous two methods, which is the bilinear form often mentioned. Substituting equation (25) into equation (21), we have: , Rearranging the formula, we can get: , in It is a hyperparameter, which means the distance between discrete points, also known as the step size.

[0046] Therefore, we have: , , So that: , In this way, Equation (15) is successfully discretized.

[0047] Now, we will consider how to perform batch training on (28). Start writing when the initial state , then: , in, for , for Indicates input The observed value from forward discrete values ​​to discrete values ​​(a total of values), that is, a total of iterations step; However, training is not just an iteration, so it is necessary to derive the update formula for the subsequent state space. Following the above ideas, we can get: , Similarly, we can derive a formula similar to formula (31): , Because of which for , so in actual iteration, only one calculation is required Matrix The result of each power increase is saved and can be used until the end of the iteration.

[0048] Now, through continuous iteration, we have calculated formula (15), which is Approximate solution of . In the previous section, we will , after the iteration is completed, it means that we have obtained the observation function Approximate expression of: , which is formula (1). Because is a transformation of the Legendre polynomial we chose at the beginning, and It is the last iteration .

[0049] Now there is , and we chose ,in is a Legendre polynomial. In the Legendre polynomial, there is a derivative relationship between its adjacent terms (same as formula (6)): , That is: , The Legendre polynomial before Denoted as coefficient , then: , We refer to the construction of the state equation and introduce the state variable ,make: , So, there is a state vector: , and: , From equation (36), we can inversely solve ,have: , Expressed in state variables: , Substituting (41) into the recursive formula of state variables, we have: , This can be written in matrix form: , in: , , The output equation is: , in .

[0050] According to the control sequence , the control coefficient is obtained by back propagation calculation of the inblock unit ; Input control sequence , the left side of (36) can be written as: , Now integrate equations (36) and (47), we have: , According to the previous method, we can get a similar form to (41) from (48): , In this way, the final A matrix remains unchanged, but the B matrix becomes: , The control coefficient can be obtained .

[0051] The technical solutions in the above embodiments of the present invention have at least the following technical effects or advantages: The FlightKooba model processes and analyzes aircraft data and control sequences in real time, providing strong support for dynamic adjustments to flight missions. By identifying potential conflicts (such as route intersections and areas of severe weather) in advance and issuing timely warnings, the model allows the flight control system or pilot to react quickly, optimize the flight path, and reduce unnecessary delays and fuel consumption. This capability greatly improves the accuracy of flight mission planning and the efficiency of execution, making flight operations more efficient and economical.

[0052] Example 3 Based on the same inventive concept, another embodiment of the present invention provides a flight trajectory prediction system for implementing the embodiments of the present invention, including a data acquisition unit for obtaining a data sequence of an aircraft at a current moment, a control sequence of the aircraft at a current moment, and a state of the aircraft at a current moment; Model building processing unit, used to build the FlightKooba model and the data sequence of the aircraft at the current moment and the current control sequence of the aircraft Calculate the state of the aircraft at the future time and coefficient matrix ; The flight trajectory prediction unit is used to predict the future state of the aircraft and coefficient matrix Predicting the flight trajectory of an aircraft .

[0053] In addition, the model building processing unit includes a HIPPO module for iteratively calculating the prediction coefficients ; InBlock unit, used to obtain the control coefficient through back propagation calculation ; KBlock unit, used to calculate the prediction coefficients , calculate the coefficient matrix , state transfer matrix and the control matrix , and is used to transform the state transfer matrix and control matrix Splicing to form a prediction matrix ; Prediction unit, based on control coefficient and the current control sequence of the aircraft , obtain the control sequence of the aircraft at the future time ; According to the prediction matrix , the current status of the aircraft and the control sequence of the aircraft at future times , calculate the state of the aircraft at the future time .

[0054] Example 4 Based on the same inventive concept, another embodiment of the present invention provides an electronic device for implementing the flight trajectory prediction method described in the embodiment of the present invention.

[0055] An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is configured to execute to implement a flight trajectory prediction method.

[0056] Based on the same inventive concept, a non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to implement a flight trajectory prediction method.

[0057] Since the electronic device described in this embodiment is an electronic device used to implement the information processing method in the embodiment of the present invention, based on the information processing method described in the embodiment of the present invention, those skilled in the art will be able to understand the specific implementation of the electronic device of this embodiment and its various variations. Therefore, how the electronic device implements the method in the embodiment of the present invention will not be described in detail here. As long as the electronic device used by those skilled in the art to implement the information processing method in the embodiment of the present invention falls within the scope of protection of the present invention.

[0058] The technical solutions in the above embodiments of the present invention have at least the following technical effects or advantages: Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0059] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0060] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0061] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0062] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0063] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A flight trajectory prediction method, characterized in that: Specifically include: Obtain the current data sequence and control sequence of the aircraft; Build the FlightKooba model based on the HIPPO module and the Kooba module; According to the data sequence of the aircraft at the current moment, the prediction coefficient is iteratively calculated using the HIPPO module; According to the prediction coefficients, the coefficient matrix, state transfer matrix and control matrix are obtained after calculation by the Kooba module; Concatenate the state transfer matrix and the control matrix to form a prediction matrix; According to the control sequence of the aircraft at the current moment, the control coefficients are obtained by back propagation calculation using the Kooba module; Obtaining the control sequence of the aircraft at a future time according to the control coefficient and the control sequence of the aircraft at a current time; According to the prediction matrix and the control sequence of the aircraft at the future time, and by introducing the state variables, the state of the aircraft at the future time is calculated; The flight trajectory of the aircraft is predicted based on the aircraft's state and coefficient matrix at future moments.

2. The flight trajectory prediction method according to claim 1, wherein: The specific steps of iteratively calculating the prediction coefficient using the HIPPO module based on the current data sequence of the aircraft include: Construct including basis functions and the prediction coefficients The observation function ; The prediction coefficients are obtained by using Fourier series to find the coefficients Expressions of According to the discretized data sequence of the aircraft at the current moment , calculate the prediction coefficient Discrete expression of ; According to the prediction coefficient The discrete expression of .

3. The flight trajectory prediction method according to claim 2, wherein: The Kooba module includes an InBlock unit and a KBlock unit; The InBlock unit is used to obtain the control coefficient through back propagation calculation ; The KBlock unit is used to predict the , calculate the coefficient matrix , state transfer matrix and the control matrix , and is used to transform the state transfer matrix and control matrix Splicing to form a prediction matrix .

4. The flight trajectory prediction method according to claim 3, wherein: The method for obtaining the control sequence of the aircraft at a future time based on the control coefficient and the control sequence of the aircraft at a current time specifically includes: ， Where, is the control sequence of the aircraft at the current moment.

5. The flight trajectory prediction method according to claim 4, wherein: The method for calculating the state of the aircraft at a future time based on the prediction matrix and the control sequence of the aircraft at a future time, and introducing state variables, specifically includes: ， Where, is a state variable.

6. The flight trajectory prediction method according to claim 5, wherein: The method for predicting the flight trajectory of an aircraft based on the aircraft's state at a future moment and a coefficient matrix specifically includes: ， Where, is the predicted flight trajectory.

7. A flight trajectory prediction system, characterized in that: It includes a data acquisition unit for obtaining a data sequence of the aircraft at the current moment and a control sequence of the aircraft at the current moment; Model building processing unit, used to build the FlightKooba model and the data sequence of the aircraft at the current moment and the current control sequence of the aircraft Calculate the state of the aircraft at the future time and coefficient matrix ; The flight trajectory prediction unit is used to predict the future state of the aircraft and coefficient matrix Predicting the flight trajectory of an aircraft .

8. The flight trajectory prediction system according to claim 7, wherein: The model building processing unit includes a HIPPO module for iteratively calculating the prediction coefficient ; InBlock unit, used to obtain the control coefficient through back propagation calculation ; KBlock unit, used to calculate the prediction coefficients , calculate the coefficient matrix , state transfer matrix and the control matrix , and is used to transform the state transfer matrix and control matrix Splicing to form a prediction matrix ; Prediction unit, based on control coefficient and the current control sequence of the aircraft , obtain the control sequence of the aircraft at the future time ; According to the prediction matrix , the current status of the aircraft and the control sequence of the aircraft at future times , calculate the state of the aircraft at the future time .

9. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute to implement a flight trajectory prediction method according to any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium, characterized in that When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to implement a flight trajectory prediction method as described in any one of claims 1 to 6.

Images