Aviation data analysis method based on quantum collaborative travel mouse optimization

Abstract

The invention discloses an aviation data analysis method based on quantum collaborative traveler optimization, and belongs to the technical field of data clustering and intelligent optimization. The method aims at the problems that when a traditional K-mode algorithm processes aviation operation state data, an initial center is sensitive, local optimum is prone to being achieved, and adaptability is poor. The method comprises the steps that firstly, a clustering center is initialized through a quantum tunneling strategy; a clustering center is optimized by adopting a quantum collaborative artificial travel mouse optimization algorithm (QCALA) fusing adaptive step size disturbance, multi-cave collaborative excavation, quantum chaos spiral foraging and a three-mode switching escape mechanism; and finally, clustering is completed in combination with a K-mode algorithm. According to the method, the stability, precision and convergence speed of aviation operation data clustering are remarkably improved, the aviation operation state can be accurately judged, and reliable data analysis support is provided for low-altitude economic aviation operation planning and risk management.

Description

Technical Field

[0001] This invention belongs to the field of cluster analysis technology, specifically relating to an aeronautical data analysis method based on quantum cooperative lemming optimization. Background Technology

[0002] With the rapid development of the low-altitude economy and urban air traffic (UAM), aviation operation data is experiencing explosive growth. This data (such as flight numbers, airlines, departure and arrival airports, times, and delay status) is mostly discrete (categorical), exhibiting high dimensionality, mixed types, and multimodal characteristics. Extracting operational patterns and identifying risks from this data is crucial for aviation safety management and efficiency improvement.

[0003] Cluster analysis is an effective method for processing unlabeled data. The K-mode algorithm, an extension of K-means for discrete data, uses the mode and Hamming distance and is a classic method for processing categorical data. However, the traditional K-mode algorithm is extremely sensitive to the selection of initial cluster centers, and its iterative update greedy strategy is prone to getting trapped in local optima. When faced with high-dimensional and complex aeronautical data, the stability and accuracy of the clustering results are difficult to guarantee.

[0004] To address these issues, researchers have attempted to combine swarm intelligence optimization algorithms (such as particle swarm optimization and genetic algorithms) with K-modes to optimize the cluster center search process. While some progress has been made, these optimization algorithms themselves suffer from premature convergence and an imbalance between global and local search capabilities. The recently proposed Artificial Lemming Algorithm (ALA) simulates various behaviors of lemmings and shows potential in optimization problems, but it still has shortcomings in initial population quality, adaptive search step size, and finesse of local development, limiting its effectiveness when directly applied to optimizing aeronautical data clustering. Summary of the Invention

[0005] To address the aforementioned problems, the purpose of this invention is to provide an aviation data analysis method based on quantum cooperative lemming optimization, which effectively overcomes the inherent defects of K-modes and integrates a novel clustering method with high-performance optimization algorithms to meet the practical needs of high-precision and high-stability clustering analysis of aviation operation data.

[0006] To achieve the above objectives, the technical solution adopted by the present invention includes: An aeronautical data analysis method based on quantum cooperative lemming optimization includes the following steps: S1. Obtain an aviation operation status dataset containing multiple discrete attributes, preprocess the aviation operation status dataset, and determine the number of clusters K for the aviation operation status. S2. Based on the number of clusters K, a quantum tunneling strategy is used to generate an initial lemming population containing N individuals in the solution space; where each individual is encoded as a set of candidate cluster centers, representing a candidate clustering scheme. S3. Starting with the initial lemming population, set the current iteration number t=0, and execute the QCALA iterative process of the quantum cooperative artificial lemming optimization algorithm. Each iteration outputs a final updated lemming population. The iterative process specifically includes: S3.1, Based on the current iteration number t and the preset maximum iteration number T max Given the initial energy value E0, calculate the current energy factor E(t); S3.2. Based on the value of the energy factor E(t), select a behavior mode for each individual in the population to obtain all individuals whose positions have been updated; whereby each individual whose position has been updated is encoded as a set of updated cluster centers; when E(t)>1, the exploration behavior is executed; the exploration behavior updates the individual's position through an adaptive step size perturbation mechanism and / or a multi-cavity collaborative digging mechanism; when E(t)≤1, the exploitation behavior is executed; the exploitation behavior updates the individual's position through a quantum chaotic spiral foraging mechanism and / or a three-mode switching escape mechanism. S3.3: For individuals whose positions have been updated, perform the K-modes clustering assignment step on the aviation operation status dataset, calculate the corresponding clustering objective function value W(C,Z), and calculate the individual's fitness value based on this value; S3.4: Select the individual with the best fitness value from the current population, perform a mode update step of K-mode clustering on the updated cluster center of its code to obtain a set of locally optimized cluster centers, and update the code of the best individual accordingly. S3.5: Let t = t + 1, and output the current updated lemming population and its flight operation status. Determine whether the flight operation status of the current updated lemming population meets the preset convergence condition. If not, take the current updated lemming population as input and return to step S3.1 for the next iteration. If yes, take the current updated lemming population as the final updated lemming population and proceed to step S4. S4: The final cluster center encoded by the updated lemming population obtained in S3, and the final clustering result obtained by dividing the aviation operation status dataset based on the final cluster center; S5: Determine the aviation operation status based on the cluster partitioning results obtained in S4.

[0007] Preferably, the data preprocessing in S1 includes: cleaning the original dataset of aviation operation status, processing missing and outlier values, and digitally encoding all discrete attributes; the number of clusters K is determined by prior business knowledge or clustering effectiveness indicators, including the silhouette coefficient.

[0008] Preferably, the specific implementation of the quantum tunneling strategy in S2 is achieved through the quantum tunneling probability P. t The initial lemming population is obtained by correcting the positions of randomly initialized individuals using the tunneling intensity factor δ. The mathematical model is as follows: Corrected position = Initial random position + P t ×δ× random perturbation term; Among them, the quantum tunneling probability P t =0.4, where δ is a coefficient that decreases with increasing iteration number. N represents the initial number of lemmings in the population.

[0009] Preferably, the formula for calculating the energy factor E(t) in S3.1 is as follows:

[0010] Here, rand is a random number in the interval [0,1].

[0011] Preferably, the adaptive step size perturbation mechanism in S3.2 specifically replaces the fixed Brownian motion step size in the original artificial lemming algorithm's migration behavior with a normal distribution vector BW that follows an adaptive standard deviation σ. The formula for calculating the standard deviation σ is:

[0012] Where σ0 is the initial standard deviation, β is the step size decay coefficient, t is the current iteration number, and Entropy is the entropy feedback term calculated based on the population fitness value, which is used to dynamically adjust the step size according to the population diversity. The mathematical model for the multi-cavity collaborative excavation mechanism is as follows:

[0013]

[0014]

[0015] in, Let i be the position of the i-th individual in the (t+1)-th iteration. Z represents the position of the i-th individual in the t-th iteration; ω is the linearly decaying inertia weight, decreasing linearly from 0.9 to 0.4; c1 and c2 are acceleration coefficients adaptively adjusted using a cosine function; r1 and r2 are random numbers in the interval [0,1]; Z best For the individual's historical optimal position, Zpbest This represents the group's historically optimal position.

[0016] Preferably, the position update formula for the quantum chaotic spiral foraging mechanism in S3.2 is:

[0017] in, Represents the Logistic mapping; This represents a random step size generated by the Lévy distribution.

[0018] The three-mode switching escape mechanism selects one of the following three modes based on random probability to update the execution location:

[0019] Where p_m represents the probability threshold for mode switching, which can be understood as a constant and can take any value from 0.3 to 0.5; d_best represents the distance of the current individual to the global optimum; Cauchy(0,1) represents the standard Cauchy distribution, which, due to its heavy-tailed characteristics, can generate a large perturbation force to help escape local optima; levy(β) represents the Levy flight random step size with exponent β, which gradually increases from 1.3 to 2.0; UB,LB represent the upper and lower bounds of the solution space; rand represents a random number in the interval [0,1].

[0020] Preferably, the clustering objective function value W(C,Z) in S3.3 is as follows:

[0021] Where W(C, Z) represents the total cost of clustering, which is the sum of the Hamming distances of all samples to the center of their respective clusters. Minimizing this value is the goal of clustering; k represents the preset number of clusters; Ci represents the i-th cluster, which contains the set of all samples assigned to that cluster; Zi represents the cluster center of the i-th cluster; and xi represents the i-th sample in the dataset.

[0022] Preferably, the preset convergence condition in S3.5 is one of the following two cases: the current iteration number t reaches the preset maximum iteration number T. max T max =500; The improvement in the optimal fitness value over 20 consecutive generations is less than the threshold ξ=10. 5 .

[0023] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the quantum cooperative lemming-optimized aeronautical data analysis method of this application.

[0024] A computer program product includes a computer program / instructions that, when executed by a processor, implement the quantum cooperative lemming-optimized aeronautical data analysis method of this application.

[0025] Compared with the prior art, the advantages of the present invention are: The aviation data analysis method based on quantum collaborative lemming optimization of this invention significantly improves the stability, accuracy and convergence speed of aviation operation data clustering, can accurately identify aviation operation status, and provides reliable data analysis support for low-altitude economic aviation operation planning and risk management. Attached Figure Description

[0026] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart illustrating the present invention; Figure 2 This invention relates to a flowchart of the multi-strategy collaborative optimization artificial lemming algorithm; Figure 3 These are the clustering results of different algorithms on the publicly available aviation dataset from the Whale Community. Detailed Implementation

[0027] The present invention will be further described in detail below with reference to specific embodiments. These descriptions are for explanation purposes only and are not intended to limit the scope of the invention.

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0030] Example (1) Implementation scenario This embodiment is applied to aviation operation status analysis in the context of the low-altitude economy. It addresses the massive amounts of aviation operation data generated by emerging business models such as urban air traffic (UAM) and drone logistics by using cluster analysis to identify different operation modes and delay risk levels, providing support for aviation operation planning and risk management. The experimental data uses publicly available aviation datasets from the Whale Community.

[0031] (2) Setting experimental parameters Maximum number of iterations T max =500; The initial energy value is E0 = 2.0; Tunneling probability P t =0.4; The initial standard deviation σ0 = 0.5; Inertia weight ω: Decreases linearly from 0.9 to 0.4; Acceleration coefficients c1 and c2: adaptively adjusted based on the cosine function, with values ​​ranging from [0.5 to 2.5]. The helical shape factor b = 1.5 Convergence threshold ξ=10 5 The number of clusters, K=4 (determined through silhouette coefficient verification, corresponding to four risk levels: high, medium-high, medium-low, and low). Initial lemming population size N=30 (3) Specific implementation steps Step S1: Data Acquisition and Preprocessing Obtain the publicly available aviation dataset from the Whale Community. This dataset contains 8 categorical attributes: flight identifier, airline, departure airport, arrival airport, scheduled time, day of the week, flight duration, and delay status.

[0032] Data cleaning: The mean imputation method was used to handle missing values ​​of flight duration, and the deletion method was used to handle outliers (data that deviate from the mean by 3 times the standard deviation) of key attributes such as delay status. Numerical encoding: Label encoding is applied to all categorical attributes. For example, "Airline A" is encoded as 1, "Airline B" is encoded as 2, "delay" is encoded as 1, and "on time" is encoded as 0.

[0033] Step S2: Initial lemming population generation (based on quantum tunneling strategy) Thirty initial individuals are randomly generated, and each individual is encoded as 4 cluster centers (corresponding to K=4). The dimension of each cluster center is consistent with the dimension of the attributes in the dataset (8 dimensions). The initial individual position is corrected using a quantum tunneling strategy, and the correction formula is as follows: Corrected individual position = random initial position × (1 + 0.4 × δ × e) t / 500 ) The initial iteration number t=0, the initial tunneling intensity factor δ=0.8, and it decreases by 0.1 for every 10 iterations, with a minimum decrease to 0.2.

[0034] S3. Starting with the initial lemming population, set the current iteration number t=0, and execute the QCALA iterative process of the quantum cooperative artificial lemming optimization algorithm. Each iteration outputs a final updated lemming population. The iterative process specifically includes: S3.1 Calculate the current energy factor E(t) based on the current iteration number t, the preset maximum iteration number Tmax, and the initial energy value E0; In this embodiment, let Tmax = 500, E0 = 2.0, t = 100, rand = 0.3, then E(100) = 2.0 * (1 - 100 / 500)^0.3 ≈ 1.66. S3.2. Based on the value of the energy factor E(t), select a behavior pattern for each individual in the population to obtain all individuals whose positions have been updated; wherein, each individual whose position has been updated is encoded as a set of updated cluster centers; When E(t) > 1, the individual performs exploratory behavior. For example, an individual initially encoded as an 'afternoon flight' might significantly adjust its 'time' and 'airport' attributes through 'adaptive step-size perturbation' to explore the possibility of changing to an 'early morning flight' or using a different hub airport. When E(t) ≤ 1, the individual performs exploitation behavior. For example, an individual currently encoded as close to the 'short-haul high-risk' pattern (cluster 3) might fine-tune its attribute values ​​(such as airline, departure airport) through a 'quantum chaotic spiral foraging' mechanism to precisely match the characteristics of this high-risk pattern. This exploratory behavior updates the individual's position through an adaptive step-size perturbation mechanism and / or a multi-cavity collaborative mining mechanism. The adaptive step size perturbation mechanism in this embodiment specifically replaces the fixed Brownian motion step size in the original artificial lemming algorithm's migration behavior with a normal distribution vector BW that follows an adaptive standard deviation σ. The formula for calculating the standard deviation σ is as follows:

[0035] Where σ0 is the initial standard deviation, β is the step size decay coefficient, t is the current iteration number, and Entropy is the entropy feedback term calculated based on the population fitness value, which is used to dynamically adjust the step size according to the population diversity. The mathematical model of the multi-cavity collaborative excavation mechanism in this embodiment is as follows:

[0036]

[0037]

[0038] in, Let i be the position of the i-th individual in the (t+1)-th iteration. Z represents the position of the i-th individual in the t-th iteration; ω is the linearly decaying inertia weight, decreasing linearly from 0.9 to 0.4; c1 and c2 are acceleration coefficients adaptively adjusted using a cosine function; r1 and r2 are random numbers in the interval [0,1]; Z best For the individual's historical optimal position, Z pbest This represents the group's historical optimal position. c1 and c2 are two parameters that change according to a cosine function over time (iteration number t), rather than being fixed values. The right-hand arrows in the formula represent a vector.

[0039] When E(t) ≤ 1, assuming an individual is currently near a relatively optimal solution, E(t) = 0.6, it enters the exploitation phase. If "Quantum Chaotic Spiral Foraging" is selected, its position will undergo a spiral-like fine search around the current optimal solution, with a chaotic perturbation (generated by chaos_i) superimposed to find a better solution within the local range. If "Three-Mode Switching Escape" is selected, the system randomly selects a mode: for example, if mode two (random escape) is selected, the individual will make a large-scale jump with a certain probability, attempting to escape the current local optimum.

[0040] The exploitation behavior is performed; the exploitation behavior updates the individual's position through a quantum chaotic spiral foraging mechanism and / or a three-mode switching escape mechanism; The position update formula for the quantum chaotic spiral foraging mechanism in this embodiment is:

[0041] in, The Logistic map is a classic chaotic system used to generate sequences with ergodicity and pseudo-randomness, helping algorithms to better cover the solution space and avoid getting trapped in loops during local searches. The chaotic sequences generated by the map are used to enhance the ergodicity and randomness of local searches. This represents a random step size generated by the Levy distribution.

[0042] The three-mode switching escape mechanism in this embodiment selects one of the following three modes for execution location update based on random probability:

[0043] Where p_m represents the probability threshold for mode switching, which can be understood as a constant and can be taken as 0.4; d_best represents the distance from the current individual to the global optimum; Cauchy(0,1) represents the standard Cauchy distribution, which, due to its heavy-tailed characteristics, can generate a large perturbation force to help escape local optima; levy(β) represents the Levy flight random step size with exponent β, which gradually increases from 1.3 to 2.0; UB,LB represent the upper and lower bounds of the solution space; rand represents a random number in the interval [0,1].

[0044] The three-mode switching escape mechanism in this embodiment is a method that allows an individual (lemming) to randomly select three different strategies to update its position, aiming to help the algorithm escape local optima.

[0045] S3.3: For the updated cluster centers corresponding to the individuals whose positions have been updated, perform the K-modes clustering assignment step on the aviation operation status dataset, calculate the corresponding clustering objective function value W(C,Z), and calculate the fitness value of the individual based on this value; For each individual after the update (with 4 cluster centers), perform the K-modes assignment steps: calculate the Hamming distance between each sample and the 4 cluster centers, and assign the sample to the nearest cluster; calculate the clustering objective function value W(C,Z) (the sum of the Hamming distances of all samples to their respective cluster centers); the fitness value is calculated as follows: For example, if an individual's W(C,Z)=120, then its fitness value is 1 / (1+120)≈0.00826.

[0046] S3.4: Select the individual with the best fitness value from the current population, perform a mode update step of K-mode clustering on the updated cluster center of its code to obtain a set of locally optimized cluster centers, and update the code of the best individual accordingly. In this embodiment, the individual with the highest fitness value (let's say the 12th individual) is selected from the current 30 individuals. The K-mode mode update step is then performed on the four cluster centers encoded by this individual: the mode of each attribute within each cluster is counted, the cluster centers are updated to the corresponding modes, the locally optimized cluster centers are obtained, and the encoding of the 12th individual is updated.

[0047] S3.5: Let t = t + 1, and output the current updated lemming population and its flight operation status. Determine whether the flight operation status of the current updated lemming population meets the preset convergence condition. If not, take the current updated lemming population as input and return to step S3.1 for the next iteration. If yes, take the current updated lemming population as the final updated lemming population and proceed to step S4. Determine if the convergence condition is met: If t=500, or the optimal fitness value increases by less than 10 over 20 consecutive generations. If the condition is 5, then stop iterating; otherwise, return to S3.1 to continue iterating. In this embodiment, the algorithm converges on the 6th iteration (the optimal fitness value has not significantly improved for 20 consecutive generations).

[0048] S4: The final cluster center encoded by the updated lemming population obtained in S3, and the final division of the aviation operation status dataset based on the final cluster center, are used to obtain the cluster division result.

[0049] This embodiment outputs the final cluster center matrix (four 8-dimensional vectors) and sample membership matrix (each sample corresponds to a cluster label: 1-4), as shown below (example): Cluster 1 (High Risk) Cluster Center Key Attributes: Planned Time = Morning (Code 1), Flight Duration = Short-haul (Code 1), Delay Status = 1 (Delay); Cluster 2 (medium-high risk) cluster center key attributes: planned time = afternoon (code 2), flight duration = medium-short distance (code 2), delay status = 1 (delay); Cluster 3 (low to medium risk) cluster center key attributes: planned time = evening (code 3), flight duration = medium to long distance (code 3), delay status = 0 (on time); Cluster 4 (low risk) cluster center key attributes: planned time = night (code 4), flight duration = long distance (code 4), delay status = 0 (on time).

[0050] S5: Determine the aviation operation status based on the cluster partitioning results obtained in S4.

[0051] Cluster 1 (High Risk): Includes 1,200 flights with a delay rate of 66.11%. It is characterized by short-haul routes in the morning and mainly involves small airlines and regional airports. Cluster 2 (Medium-high risk): Includes 1,800 flights with a delay rate of 49.51%. It is characterized by short- and medium-haul routes in the afternoon and involves small and medium-sized airlines. Cluster 3 (Low to Medium Risk): Includes 2,500 flights with a delay rate of 35.13%, typically characterized by medium- and long-haul routes during the evening hours, involving major airlines; Cluster 4 (Low Risk): Includes 3,000 flights with a delay rate of 30.67%. It is characterized by long-haul routes during nighttime hours and involves major airlines and hub airports.

[0052] Based on the above characteristics, an aviation operation status analysis report is generated to identify high-risk operation scenarios (short-haul feeder flights in the morning) and low-risk operation scenarios (long-haul hub flights at night), providing a basis for airspace resource scheduling and delay risk warning.

[0053] (4) Effect verification Experiments on the Hejing Community aviation dataset show that the QCALA-Kmodes implementation in this example converges in only 6 iterations, a 76.9% speed improvement compared to the traditional K-modes (26 iterations). The clustering results show significant pattern differences: high-risk clusters have a delay rate of 66.11%, mainly short-haul routes in the morning; low-risk clusters have a delay rate of 30.67%, mainly long-haul routes; and medium-risk clusters have delay rates ranging from 35.13% to 49.51%. The difference in delay risk between clusters reaches 35.44%, significantly better than the clustering effect of traditional methods.

[0054] In summary, this embodiment verifies the effectiveness of the method of the present invention in aviation data analysis, which can accurately identify operational status patterns and provide reliable support for low-altitude economic risk management.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

[0056] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0057] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0058] Furthermore, the various implementation methods disclosed in this solution can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content invented by this disclosure.

Claims

1. An aeronautical data analysis method based on quantum cooperative lemming optimization, characterized in that, Includes the following steps: S1. Obtain an aviation operation status dataset containing multiple discrete attributes, preprocess the aviation operation status dataset, and determine the number of clusters K for the aviation operation status. S2. Based on the number of clusters K, a quantum tunneling strategy is used to generate an initial lemming population containing N individuals in the solution space. Each individual is coded as a set of candidate cluster centers, representing a candidate clustering scheme; S3. Starting with the initial lemming population, set the current iteration number t=0, and execute the QCALA iteration process of the quantum cooperative artificial lemming optimization algorithm. Each iteration outputs a final updated lemming population. The iteration process specifically includes: S3.1, Based on the current iteration number t and the preset maximum iteration number T max Given the initial energy value E0, calculate the current energy factor E(t); S3.

2. Based on the value of the energy factor E(t), select a behavior pattern for each individual in the population to obtain all individuals whose positions have been updated; wherein, each individual whose position has been updated is encoded as a set of updated cluster centers; When E(t)>1, an exploration behavior is executed; the exploration behavior updates the individual's position through an adaptive step size perturbation mechanism and / or a multi-cavity collaborative excavation mechanism. When E(t)≤1, exploitation behavior is performed; the exploitation behavior updates the individual's position through a quantum chaotic spiral foraging mechanism and / or a three-mode switching escape mechanism; S3.3: For the updated cluster centers corresponding to the individuals whose positions have been updated, perform the K-modes clustering assignment step on the aviation operation status dataset, calculate the corresponding clustering objective function value W(C,Z), and calculate the fitness value of the individual based on this value; S3.4: Select the individual with the best fitness value from the current population, perform a mode update step of K-mode clustering on the updated cluster center of its code to obtain a set of locally optimized cluster centers, and update the code of the best individual accordingly. S3.5: Let t = t + 1, and output the current updated lemming population and its flight operation status. Determine whether the flight operation status of the current updated lemming population meets the preset convergence condition. If not, take the current updated lemming population as input and return to step S3.1 for the next iteration. If yes, take the current updated lemming population as the final updated lemming population and proceed to step S4. S4: The final clustering center encoded by the final updated lemming population obtained in S3, and the final division of the aviation operation status dataset based on the final clustering center, are used to obtain the clustering result; S5: Determine the aviation operation status based on the cluster partitioning results obtained in S4.

2. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in claim 1, characterized in that, The data preprocessing described in S1 includes: cleaning the original dataset of aviation operation status, processing missing and outlier values, and digitally encoding all discrete attributes; the number of clusters K is determined by prior business knowledge or clustering effectiveness indicators, including the silhouette coefficient.

3. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in claim 1, characterized in that, The specific implementation of the quantum tunneling strategy described in S2 is achieved through the quantum tunneling probability P. t The initial lemming population is obtained by correcting the positions of randomly initialized individuals using the tunneling intensity factor δ, and the mathematical model is as follows: Corrected position = Initial random position + P t ×δ× random perturbation term; Among them, the quantum tunneling probability P t =0.4, where δ is a coefficient that decreases with increasing iteration number. N represents the initial number of lemmings in the population.

4. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in claim 1, characterized in that, The formula for calculating the energy factor E(t) described in S3.1 is as follows: Here, rand is a random number in the interval [0,1].

5. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in claim 1, characterized in that, The adaptive step size perturbation mechanism described in S3.2 specifically replaces the fixed Brownian motion step size in the original artificial lemming algorithm's migration behavior with a normal distribution vector BW that follows an adaptive standard deviation σ. The formula for calculating the standard deviation σ is as follows: Where σ0 is the initial standard deviation, β is the step size decay coefficient, t is the current iteration number, and Entropy is the entropy feedback term calculated based on the population fitness value, which is used to dynamically adjust the step size according to the population diversity. The mathematical model for the multi-cavity collaborative excavation mechanism is as follows: in, Let i be the position of the i-th individual in the (t+1)-th iteration. Z represents the position of the i-th individual in the t-th iteration; ω is the linearly decaying inertia weight, decreasing linearly from 0.9 to 0.4; c1 and c2 are acceleration coefficients adaptively adjusted using a cosine function; r1 and r2 are random numbers in the interval [0,1]; Z best For the individual's historical optimal position, Z pbest This represents the group's historically optimal position.

6. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in claim 1, characterized in that, The position update formula for the quantum chaotic spiral foraging mechanism described in S3.2 is: in, Represents the Logistic mapping; Denotes a random step size generated by a Lévy distribution; The three-mode switching escape mechanism selects one of the following three modes based on random probability to update the execution location: Where p_m represents the probability threshold for mode switching, which can be understood as a constant and can take any value from 0.3 to 0.5; d_best represents the distance of the current individual to the global optimum; Cauchy(0,1) represents the standard Cauchy distribution, which, due to its heavy-tailed characteristics, can generate a large perturbation force to help escape local optima; levy(β) represents the Levy flight random step size with exponent β, which gradually increases from 1.3 to 2.0; UB,LB represent the upper and lower bounds of the solution space; rand represents a random number in the interval [0,1].

7. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in claim 1, characterized in that, The clustering objective function value W(C,Z) described in S3.3 is as follows: Where W(C, Z) represents the total cost of clustering, which is the sum of the Hamming distances of all samples to the center of their respective clusters. Minimizing this value is the goal of clustering; k represents the preset number of clusters; Ci represents the i-th cluster, which contains the set of all samples assigned to that cluster; Zi represents the cluster center of the i-th cluster; and xi represents the i-th sample in the dataset.

8. The aeronautical data analysis method based on quantum cooperative lemming optimization as described in any one of claims 1-7, characterized in that, The preset convergence condition in S3.5 is one of the following two cases; The current iteration number t has reached the preset maximum iteration number T. max T max =500; The improvement in the optimal fitness value over 20 consecutive generations is less than the threshold ξ=10. 5 .

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the aeronautical data analysis method based on quantum cooperative lemming optimization as described in any one of claims 1-8.

10. A computer program product, characterized in that, Includes a computer program / instruction, which, when executed by a processor, implements the aeronautical data analysis method based on quantum cooperative lemming optimization as described in any one of claims 1-8.

Images