CN-121683049-B - Aircraft optimal trajectory screening method and system based on adjustable phase quantum search

CN121683049BCN 121683049 BCN121683049 BCN 121683049BCN-121683049-B

Abstract

The invention relates to the technical field of low-altitude safety, and provides an aircraft optimal track screening method and system based on adjustable phase quantum search. The method comprises the steps of generating candidate tracks according to flight task constraint, calculating cost value, distributing Boolean mark based on cost threshold to identify high-quality solutions, obtaining output indexes by utilizing Grover quantum search algorithm to conduct balanced quantum superposition state iterative search and measurement of coding track indexes based on adjustment search phase, mapping the indexes into tracks, obtaining optimized tracks through classical local optimization, adaptively adjusting phase parameters and cost threshold based on performance evaluation, and repeatedly executing until the end conditions are met, and determining the optimal tracks. The strategy is online reconfigurable by introducing an adjustable phase mechanism, a feedback closed loop is established to realize parameter self-adaptive optimization, and the real-time requirements of scenes such as urban air traffic control, unmanned aerial vehicle clusters and the like are met.

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: 20260209

Claims (10)

1. An aircraft optimal trajectory screening method based on adjustable phase quantum search is characterized by comprising the following steps: The track coding step comprises the steps of generating a plurality of candidate tracks according to flight task constraint and calculating the comprehensive cost value of the candidate tracks, distributing a Boolean mark for each track based on a preset cost threshold to identify a high-quality solution, and generating an Oracle lookup table for mapping track indexes of the candidate tracks to the Boolean marks; Constructing a quantum Oracle operator based on the Boolean mark, and determining an adjustable phase parameter according to task requirements to construct a generalized diffusion operator; carrying out iterative execution on quantum superposition states coded with track indexes, wherein the quantum operations comprise the Oracle operator and the generalized diffusion operator, so as to obtain an output index; The method comprises the steps of outputting an output index, mapping the output index into a corresponding candidate track, carrying out local fine tuning and smoothing by taking the candidate track as an initial demodulation classical track optimization algorithm to obtain a current round of optimization track, adjusting the adjustable phase parameter and the cost threshold value based on performance evaluation of the quantum search, and repeatedly executing the track coding step and the quantum search step based on the adjusted adjustable phase parameter and the cost threshold value, wherein when a preset ending condition is met, determining the current obtained optimization track as the optimal track of the flight task.
2. The aircraft optimal trajectory screening method based on adjustable phase quantum search of claim 1, wherein the flight mission constraints include a start position, an end position, airspace constraints, and an aircraft dynamics model, and wherein the airspace constraints include obstacles, no-fly zones, and terrain constraints.
3. The aircraft optimal trajectory screening method based on adjustable phase quantum search of claim 2, wherein the comprehensive cost value comprises a multi-dimensional cost vector C (τ i ) that calculates a candidate trajectory τi: ; wherein w L ,w R ,w F ,w T E [0,1] is a weight coefficient; Exposing costs for threat; time of flight costs; is the cost of the flight path; The consumption cost is for cruising; The method for assigning a Boolean mark to each track based on the preset cost threshold value to identify a high-quality solution comprises the steps of defining a Boolean mark function f (tau i ) for generating a judgment result of the Boolean mark, wherein: If C (τ i )≤C th ), f (τ i ) =1, the solution is marked as a good solution, otherwise f (τ i )=0;C th is a preset cost threshold.
4. The aircraft optimal trajectory screening method based on adjustable phase quantum search of claim 3, wherein the generating an Oracle lookup table mapping the trajectory index of the candidate trajectory to the boolean flag comprises: Based on the judging result of the Boolean mark, establishing a one-to-one mapping relation between the unique index of each candidate track and the candidate track mark value to form the Oracle lookup table; And packaging the Oracle lookup table and the adjustable phase parameters into a quantum instruction packet, and sending the quantum instruction packet to a quantum simulator for constructing the quantum Oracle operator.
5. The method for screening the optimal trajectory of the aircraft based on the adjustable phase quantum search according to claim 4, wherein constructing quantum Oracle operators based on the boolean flags and determining an adjustable phase parameter according to task requirements to construct generalized diffusion operators comprises: introducing auxiliary quantum bits, associating a track index state with the Boolean mark through a controlled gate network, applying phase inversion operation to a state marked as a high-quality solution, and executing inverse operation solution calculation to empty the auxiliary quantum bits, so as to realize negative phase marking of a base state corresponding to the high-quality solution, thereby obtaining the quantum Oracle operator; The construction of the generalized diffusion operator comprises the steps of constructing a reflection operator related to an average probability amplitude based on the adjustable phase parameter, wherein the reflection intensity of the reflection operator is precisely controlled by the adjustable phase parameter, and when the adjustable phase parameter is a flat angle, the reflection operator is degenerated into a standard Grover diffusion operator to serve as the generalized diffusion operator.
6. The method for screening an optimal trajectory of an aircraft based on adjustable phase quantum search according to claim 5, wherein performing iterative execution of quantum operations consisting of the Oracle operator and the generalized diffusion operator on the quantum stack state encoded with the trajectory index to obtain an output index comprises: distributing n qubits as main registers on a quantum simulator, wherein N is the number of candidate tracks; Invoking a unitary operator UN to prepare an equilibrium superposition state containing all N candidate track indexes; combining the generalized diffusion operator and the Oracle operator to form a single iteration operator; repeatedly executing the iterative operator Q times, so that the probability amplitude corresponding to the track index marked as the high-quality solution is coherently amplified; All qubits of the main register are measured and collapsed with high probability to obtain the output index.
7. The aircraft optimal trajectory screening method based on adjustable phase quantum search according to claim 1, wherein mapping the output index into a corresponding candidate trajectory, performing local fine tuning and smoothing by using the candidate trajectory as an initial demodulation classical trajectory optimization algorithm, and obtaining a current round of optimization trajectory, comprises: mapping the output index back to a classical track database, and searching track parameters uniquely corresponding to the index of the candidate track; And taking the verified candidate track as a high-quality initial solution, calling a sequence quadratic programming algorithm to perform local fine tuning and smoothing, and generating the current round of optimized track.
8. The method for screening an optimal trajectory of an aircraft based on an adjustable phase quantum search according to claim 7, wherein the adjusting the adjustable phase parameter and the cost threshold based on the performance evaluation of the current quantum search comprises: Establishing a performance evaluation model, and evaluating the searching performance according to the weighted scores of the time consumption and the solution quality of the searching, wherein the solution quality is determined according to the percentile of the comprehensive cost value of the current round of optimization tracks in the historical solution; If the solution quality is lower than a preset quality threshold and the current iteration number does not reach the upper limit, reducing the adjustable phase parameters according to a first proportion coefficient; if the search time exceeds a preset time threshold, increasing the adjustable phase parameters according to a second proportionality coefficient; And counting the proportion of the track marked as the high-quality solution in the historical multi-cycle searching to the total candidate track, if the proportion is continuously lower than a first proportion threshold value, increasing the cost threshold value, and if the proportion is continuously higher than a second proportion threshold value, reducing the cost threshold value.
9. An aircraft optimal trajectory screening system based on adjustable phase quantum search, comprising: The track coding module is used for generating a plurality of candidate tracks according to flight task constraint and calculating the comprehensive cost value of the candidate tracks, distributing a Boolean mark for each track based on a preset cost threshold to identify a high-quality solution, and generating an Oracle lookup table for mapping track indexes of the candidate tracks to the Boolean marks; The quantum search module is used for constructing a quantum Oracle operator based on the Boolean mark, and determining an adjustable phase parameter according to task requirements to construct a generalized diffusion operator; The optimization module is used for mapping the output index into a corresponding candidate track, carrying out local fine tuning and smoothing by taking the candidate track as an initial demodulation by using a classical track optimization algorithm to obtain a current round of optimization track, adjusting the adjustable phase parameter and the cost threshold value based on performance evaluation of the quantum search, and repeatedly executing track coding and quantum search based on the adjusted adjustable phase parameter and the cost threshold value, wherein when a preset ending condition is met, the current obtained optimization track is determined to be the optimal track of the flight task.
10. 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 an aircraft optimal trajectory screening method based on an adjustable phase quantum search according to any one of claims 1-8.

Description

Aircraft optimal trajectory screening method and system based on adjustable phase quantum search Technical Field The application relates to the technical field of low-altitude safety, in particular to an aircraft optimal track screening method and system based on adjustable phase quantum search. Background With the rapid development of low-altitude economy, the limitation of the traditional aircraft track planning method is gradually enlarged, the search time of the classical algorithm is exponentially exploded along with the increase of the degree of freedom, so that the complexity of solution is increased, and the real-time re-planning requirement in a dynamic environment cannot be met, for example, the number of feasible solutions in a high-dimensional continuous space is huge, so that the time required for planning a safe track in a dense obstacle or complex space is too long, meanwhile, the uncertainty of solution is one of important factors for limiting track generation, for example, the random sampling-based RRT algorithm can quickly find the feasible solution, but the global optimality of solution is difficult to ensure, so that the planned track may be far around, not smooth and has high energy consumption, which is unacceptable for unmanned aerial vehicles with long endurance or limited fuel. The computational force limitations of classical computing force computing systems to make a compromise between computation speed and optimality of the solution. Quantum computing is taken as the most potential computing force breakthrough direction at the present stage, and quantum parallelism provides theoretical possibility for realizing exponential acceleration search in huge combined space. The Grover quantum search algorithm proposed in 1996 is a typical representation of this potential, however, it has three inherent assumptions of database size of 2 power, stationary phase rotation, and prior knowledge of the number of target solutions, and these ideal laboratory conditions are severely disjointed from the real world unstructured, a priori information deficient, scale-random flight planning scenarios, resulting in its inability to be directly applied. In the large environment of rapid development of low-altitude economy, a low-altitude unmanned aerial vehicle track intelligent planning system faces increasingly severe computing pressure. The existing aircraft track planning method adopts a traditional classical calculation paradigm to carry out track planning, the core calculation of the method totally depends on the serial and parallel processing capacity of a classical computer, and a feasible solution or an optimized solution is searched in a solution space in a traversing way, so that the high-quality path is still difficult to quickly generate when the method is used for coping with low-altitude intensive and high-dynamic unmanned plane motion scenes, although the solution quality is higher. The traditional aircraft trajectory prediction planning mechanism mainly has the following limitations: 1. The solving efficiency is low, and the real-time requirement cannot be met. The computational complexity of classical algorithms is often polynomial or even exponential in relation to the problem size (e.g., discrete accuracy of state space, number of obstacles). For example, the search time of the a algorithm in the three-dimensional refinement grid expands sharply, and the RRT algorithm finds a feasible solution faster, but the number of sampling points needed to obtain a high quality solution is large, and the optimization process is slow. Under the scene that unmanned aerial vehicle suddenly avoids barrier or urban logistics multimachine collaborative planning and the like need "second level" or even "millisecond level" response, the calculation delay of classical algorithm becomes unacceptable bottleneck, and the system is forced to only make a compromise between "reducing planning precision" and "sacrificing response speed", so that low-altitude flight safety and efficiency are seriously affected. 2. The search flexibility is low. The traditional Grover algorithm uses fixed rotation phase, the search step length is fixed, the search strategy cannot be flexibly adjusted when the problem of different target solution ratios is faced, the target solution ratio is high, the search is easy to be excessive, and the convergence is slow when the target solution ratio is low. The user cannot balance and control the search speed and the precision according to the task emergency degree, and meanwhile, the scale of the solution space required by the algorithm must be a power of 2, so that the algorithm is difficult to adapt to diversified actual task demands. 3. Feedback learning and adaptation are weak. Existing classical planning systems are typically in a "one-time" mode. I.e. the planning process is terminated after generating the trajectory from the current input. The system lacks a closed loop feedback me