CN-121596892-B - Aircraft track prediction method and system based on space gridding and heuristic search

CN121596892BCN 121596892 BCN121596892 BCN 121596892BCN-121596892-B

Abstract

The embodiment relates to the technical field of aerospace safety, and provides an aircraft track prediction method and system based on space meshing and heuristic search. By means of self-adaptive grid resolution adjustment of high-speed aircraft track prediction and rapid calculation of possible high-probability tracks according to aircraft types, the problems that in the existing high-speed aircraft track prediction, calculation resource coordination is inflexible, aircraft type perception is insensitive and calculation binding physical constraint is low are solved.

Inventors

QI JIANHUAI
LI YUXIN
HU JINHUA
CHENG YANG
WANG YINCHENG
WANG MINGSHUAI
ZHENG WEIFAN
XU GUOQIAN

Assignees

深圳市永达电子信息股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260123

Claims (9)

1. An aircraft trajectory prediction method based on spatial meshing and heuristic search, comprising: Discretizing continuous six-dimensional state space information of the aircraft into grid points with different resolutions on each grid level respectively, and determining prior probability distribution of each grid point; Acquiring a current state parameter of the aircraft, dynamically selecting an optimal grid level from a plurality of grid levels according to the state parameter, and mapping a state vector of the aircraft to the optimal grid level, wherein the state parameter comprises a speed and an acceleration; Determining maneuver modes from a pre-constructed maneuver mode library according to the type of the current aircraft, and generating candidate tracks based on a preset kinematic integration model; Calculating the state transition cost of the candidate track based on a state transition cost function, and estimating the time cost from the current state to the target area based on a heuristic function; Taking the sum of the state transition cost and the time cost as the total cost, and performing heuristic search on a discretized grid space based on the candidate track to find the optimal track of the aircraft reaching the target area; The method further comprises the steps of, before discretizing the continuous six-dimensional state space information of the aircraft into grid points with different resolutions on each grid level, respectively: Constructing L grid levels with different resolutions, wherein the resolution Deltal of the first-layer grid meets the geometric series relation, deltal=Delta 0 ·ρ (l-1); Δ 0 is the basic resolution, and rho is the scaling factor; Discretizing continuous six-dimensional state space information of the aircraft into grid points with different resolutions on each grid level respectively, wherein the six-dimensional state space information comprises: According to a preset three-dimensional space boundary and the resolution delta l of each grid level, uniformly dividing the three-dimensional space position information into a plurality of cube grid units, wherein the center point or the vertex of each cube grid unit represents a discrete position state, so that a position discrete grid comprising the position state is formed; performing discrete division on the three-dimensional speed component according to a preset speed range and the resolution Deltal of each grid level to form a speed discrete grid comprising a speed state; the position discrete grid and the speed discrete grid are combined through Cartesian products to form complete discrete state points, and each point represents a specific position and speed combination state.
2. The aircraft trajectory prediction method based on spatial meshing and heuristic search of claim 1, wherein determining a priori probability distribution for each grid point comprises: Wherein r is the position vector of the current grid point, sum traverses all known target regions of interest n, w n is the priority weight of the target region, f n (z) is a height preference function, and alpha is a distance attenuation coefficient; distance from the current grid point to the target region r n ; A priori probabilities for the i, j, k grid points; The height of the position of the grid point i, j, k is given, N is the number of all grid points, and i, j, k is the three-dimensional coordinate of the grid point.
3. The aircraft trajectory prediction method based on spatial meshing and heuristic search of claim 2, characterized in that dynamically selecting an optimal mesh level from a plurality of mesh levels according to the state parameters comprises: Based on the state parameters, dynamically calculating an optimal grid level index l optimal through an online learning decision function, wherein the decision function is as follows: l optimal = floor( L* (λ* (||a||/a max ) + (1-λ)* (||v||/v scale ) ) ) Wherein L is the total grid layer number, a max is the maximum acceleration reference value, v scale is the speed reference threshold value, lambda is the maneuvering strength weight factor, lambda is more than or equal to 0 and less than or equal to 1, and is used for adjusting the relative importance of acceleration and speed in decision, v is the speed, and a is the acceleration.
4. The aircraft trajectory prediction method based on spatial meshing and heuristic search of claim 3, wherein the pre-built maneuver pattern library comprises a plurality of maneuver patterns, wherein each maneuver pattern is associated with a trigger probability, wherein the trigger probability is determined by a base probability and an adjustment based on the aircraft type.
5. The method for predicting an aircraft trajectory based on spatial meshing and heuristic search of claim 4, wherein determining maneuver patterns from a library of pre-constructed maneuver patterns based on the current type of aircraft and generating candidate trajectories based on a pre-set kinematic integration model comprises: According to the current aircraft type, expanding a plurality of candidate maneuver instruction from the current state in a weighted random selection mode according to the maneuver instruction of each maneuver mode and the associated trigger probability thereof from the corresponding maneuver mode library; Determining a constraint acceleration for each candidate maneuver instruction, wherein the constraint acceleration is a function of a pneumatic constraint, a handling capacity constraint and a fuel consumption constraint; based on the constraint acceleration, a preset kinematic integral model is adopted to carry out recursive calculation on the current state, and a prediction state of the next time step is generated; and iteratively executing the maneuvering action expansion, physical constraint correction and constraint state recursion process to generate a continuous state point sequence which starts from the current state and is in a future period of time, and generating a plurality of candidate tracks.
6. The aircraft trajectory prediction method based on spatial meshing and heuristic search of any one of claims 1-5, wherein the state transition costs consist of weighted sums of physical feasibility costs, speed costs, altitude costs, and probability costs; the heuristic time cost from the current state to the target area comprises the following steps: h(s) = (κ type / v avg ) · d scaled (s, G) h(s) is heuristic time cost from the current state to the target area, kappa type is an adjustment factor corresponding to the current type of aircraft, v avg is average speed of the current type of aircraft in the current environment, and d scaled (s, G) is a scaling distance from the state s to the target area G.
7. The method of aircraft trajectory prediction based on spatial meshing and heuristic search of claim 6, wherein performing a heuristic search on a discretized mesh space based on candidate trajectories to find an optimal trajectory for an aircraft to reach a target area comprises: calculating the sum of the state transition cost and heuristic time cost of each segment as the total cost of the segment; selecting a candidate track segment with the minimum total cost for path expansion to obtain a new state node; And reversely backtracking all the selected candidate track fragments, and connecting to form the optimal track.
8. An aircraft trajectory prediction system based on spatial meshing and heuristic search, comprising: The construction module is used for discretizing continuous six-dimensional state space information of the aircraft into grid points with different resolutions on each grid level respectively and determining prior probability distribution of each grid point; The system comprises a selection and mapping module, a state vector mapping module, a speed calculation module and a speed calculation module, wherein the selection and mapping module is used for acquiring a current state parameter of the aircraft, dynamically selecting an optimal grid level from a plurality of grid levels according to the state parameter, and mapping a state vector of the aircraft to the optimal grid level; The generation module is used for determining maneuvering modes from a pre-constructed maneuvering mode library according to the type of the current aircraft and generating candidate tracks based on a pre-set kinematic integration model; The cost calculation module is used for calculating the state transition cost of the candidate track based on a state transition cost function and estimating the time cost from the current state to the target area based on a heuristic function; the searching module is used for performing heuristic searching on the discretized grid space according to the sum of the state transition cost and the time cost as the total cost so as to find the optimal track of the aircraft reaching the target area; the construction module is also used for constructing L grid levels with different resolutions, wherein the resolution Deltal of the first-layer grid meets the geometric series relation, deltal=Delta 0 ·ρ (l-1); Δ 0 is the basic resolution, and rho is the scaling factor; The method comprises the steps of uniformly dividing three-dimensional space position information into a plurality of cube grid units according to preset three-dimensional space boundaries and resolution delta l of each grid level, enabling the center point or the vertex of each cube grid unit to represent a discrete position state, forming a position discrete grid comprising the position state, carrying out discrete division on three-dimensional speed components according to a preset speed range and the resolution delta l of each grid level, forming a speed discrete grid comprising a speed state, combining the position discrete grid and the speed discrete grid through Cartesian products, and jointly forming a complete discrete state point, wherein each point represents a specific position and speed combination state.
9. An electronic device comprising one or more processors, a memory for storing one or more computer programs, characterized in that the computer programs are configured to be executed by the one or more processors, the programs comprising method steps for performing the spatial gridding and heuristic search based aircraft trajectory prediction method according to any one of claims 1-7.

Description

Aircraft track prediction method and system based on space gridding and heuristic search Technical Field The application relates to the technical field of aerospace safety, in particular to an aircraft track prediction method and system based on space meshing and heuristic search. Background Trajectory prediction of high-speed aircraft in near space is one of the key protection steps against the fight of enemy sky. The dynamic property, the accuracy and the real-time property of the flight trajectory prediction are important guarantees for continuously maintaining the space safety of the air defense anti-reflection system. With the continuous improvement of the maneuvering performance of modern military targets, particularly the appearance of novel threats such as hypersonic aircrafts, stealth and navigation high-speed aircrafts and the like, modern space combat rules show the trend of quick response, accurate striking and flexible burst prevention, and a serious challenge is formed for the existing air defense system and the traditional reconnaissance and early warning technology. In today's large environments where high speed aircraft in close proximity are rapidly developing, air defense and reverse guidance systems face increasingly severe safety challenges. The existing aircraft track prediction method adopts a grid with fixed resolution to discretize a state space and combines a kinematic model (such as a uniform speed and a cooperative turning model) to predict the track, and has strong universality but weak pertinence, so that prediction deviation caused by target type difference is difficult to effectively process. The traditional aircraft trajectory prediction evaluation mechanism mainly has the following limitations: (1) Fixed resolution grids lack dynamic adjustment mechanisms. Most of the existing grid dividing methods adopt fixed grid resolution and cannot be dynamically adjusted according to the target state. This results in difficult computational resource coordination and inefficient utilization, such as unnecessary fine mesh used when the target is flying steadily at low speed, wasted computational resources, too coarse mesh used when the target is maneuvering at high speed, and loss of critical detail information. (2) The cost function is simple, and the physical constraint condition is not deeply bound. The trajectory evaluation function (cost function) adopted in the existing prediction method is usually too simplified, is concentrated on the geometric or kinematic level, lacks modeling and environmental constraints on the physical feasibility of the aircraft, such as the influence of aerodynamic characteristics of non-deeply bound targets, environmental avoidance requirements and the like on the maneuverability, and causes that an algorithm may search out a trajectory which is mathematically 'optimal' but physically unrealizable, so that the prediction precision is reduced under complex conditions. (3) The target type difference perception capability is weak, and the prediction model lacks personalized self-adaptive adjustment. The existing system generally regards different aircrafts as homogeneous targets, adopts the same set of model parameters for prediction, cannot convert multi-source detection information such as radars, photoelectricity and the like into key priori knowledge such as a target maneuvering mode library and the like, so that the accurate identification and classification capability of the target type are weak, the internal parameters of the model cannot be predicted dynamically, and finally, the motion model is insensitive to specific behaviors of the aircrafts and has large prediction deviation. In view of the above, there is a need in the art for improvements. Disclosure of Invention The application aims to provide an aircraft track prediction method and system based on space meshing and heuristic search, which can be used for meeting the requirements of modern defense systems by adopting a track prediction method which is adaptive to target types, balance precision and efficiency and deeply binding physical constraint conditions. The accuracy and the flexibility of the flight trajectory prediction are improved by deeply fusing dynamic grid division and improving a heuristic search algorithm, and a reference is provided for improving the space-sky security defense. In a first aspect, the application provides an aircraft trajectory prediction method based on spatial meshing and heuristic search, which has the following technical scheme: Discretizing continuous six-dimensional state space information of the aircraft into grid points with different resolutions on each grid level respectively, and determining prior probability distribution of each grid point; Acquiring a current state parameter of the aircraft, dynamically selecting an optimal grid level from a plurality of grid levels according to the state parameter, and mapping a state vector of the aircraft to th