A method for aircraft low-altitude penetration route planning based on improved ant colony algorithm

By introducing particle swarm algorithm optimization and adaptive pheromone strategy into the ant colony algorithm, the local optimization and slow convergence speed problems of standard ant colony algorithm in low-altitude penetration route planning are solved, and efficient three-dimensional terrain follow-up route planning is achieved. , improving the quality and efficiency of the aircraft's low-altitude penetration route.

CN115268500BActive Publication Date: 2024-11-08CHINESE PEOPLES LIBERATION ARMY UNIT 92728
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211040576.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-08-29
Publication Date
2024-11-08
Estimated Expiration
2042-08-29

AI Technical Summary

Technical Problem

Standard ant colony algorithms are prone to local optimality and optimization stagnation in aircraft low-altitude penetration route planning. The performance advantages and disadvantages are strongly correlated with key parameters, and the convergence speed is slow, resulting in poor planning results.

Method used

By reconstructing the three-dimensional map, setting height, angle and distance constraints, using the particle swarm algorithm to optimize the information and expected heuristics of the ant swarm algorithm, constructing a fitness function, combining roulette selection nodes and adaptive pheromone volatile factors, design Pheromone update strategy, optimize the ant colony algorithm parameters to improve the efficiency of route planning.

Benefits of technology

It has achieved rapid generation of aircraft penetration routes, improved the efficiency of three-dimensional terrain following route planning, formed a general solution for aircraft low-altitude penetration, and improved the quality and efficiency of route planning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115268500B_ABST
    Figure CN115268500B_ABST
Patent Text Reader

Abstract

The invention discloses a method for low-altitude aircraft penetration route planning based on an improved ant colony algorithm, comprising: reconstructing a three-dimensional map to establish a solution space; setting a height constraint condition, an angle constraint condition and a distance constraint condition; optimizing and solving the parameters of the ant colony algorithm by a particle swarm algorithm to obtain an optimal information heuristic factor and an optimal expected heuristic factor; constructing a fitness function based on pheromone concentration and heuristic value; selecting nodes by roulette; updating the pheromone corresponding to the node; judging whether the standard deviation variation range of the most recent 10 calculation results is within 10% of the median of the current iterative cycle number calculation result; if so, increasing the adaptive pheromone volatility factor by 5% within its definition domain, and jumping back to the step of updating the pheromone corresponding to the node; if not, judging whether the improved ant colony algorithm reaches a preset maximum number of iterative operations. The method is used in three-dimensional terrain following route planning and can quickly generate aircraft penetration routes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of aircraft route planning, and in particular to an aircraft low-altitude penetration route planning method based on an improved ant colony algorithm. Background Art

[0002] In modern air strikes, ground-based air defense early warning systems are the main threat to penetrating aircraft. Limited by the curvature of the earth and the topography, the downward-looking capability of ground-based air defense early warning systems is generally weaker than their upward-looking capability. Taking advantage of this feature, in modern air strikes, aircraft often use terrain such as valleys for cover, and mainly fly at low altitudes close to the ground. Faced with the complex low-altitude battlefield environment, it is difficult to meet combat needs by relying solely on pilots' improvisations. Therefore, it is necessary to plan several low-altitude penetration routes before the mission begins.

[0003] Route planning is an important part of aviation combat mission planning and is also the primary guarantee for the implementation of combat missions. Carrying out refined route planning can effectively improve mission effectiveness. At present, most aviation air strike route planning is two-dimensional plane route planning, and there are few three-dimensional plans that consider elevation attributes such as topography. For aircraft to penetrate low altitudes in valleys, elevation attributes such as topography are extremely important constraints. Therefore, three-dimensional route planning has greater practical significance.

[0004] The essence of route planning is path planning. In recent years, there are many ways to solve this problem, including traditional methods such as analytical method, artificial potential field method, A* algorithm, as well as emerging intelligent algorithms such as ant colony algorithm, genetic algorithm, and firefly algorithm, and some results have been achieved. In comparison, the ant foraging behavior of the ant colony algorithm is highly similar to path planning. Therefore, using the ant colony algorithm to solve the path planning problem is the most reasonable and the most adaptable solution. However, the standard ant colony algorithm also has the disadvantages of being prone to local optimality, optimization stagnation, strong correlation between performance and key parameters, prominent empirical and random key parameter settings, and slow convergence speed. It is not effective when used directly for route planning. Summary of the invention

[0005] In view of this, an embodiment of the present invention provides a method for planning an aircraft low-altitude penetration route based on an improved ant colony algorithm to solve the problem that the standard ant colony algorithm is used in the prior art to plan a low-altitude penetration route with poor results.

[0006] The embodiment of the present invention provides a method for low-altitude penetration route planning of an aircraft based on an improved ant colony algorithm, comprising:

[0007] Reconstruct the three-dimensional map to establish the solution space;

[0008] Set height constraints, angle constraints, and distance constraints;

[0009] The parameters (α, β) of the ant colony algorithm are optimized and solved by the particle swarm algorithm to obtain the best information heuristic factor and the best expected heuristic factor; where α is the information heuristic factor and β is the expected heuristic factor.

[0010] Based on the pheromone concentration and heuristic value, a fitness function is constructed;

[0011] Select nodes via roulette wheel;

[0012] Update the pheromone corresponding to the node;

[0013] Determine whether the standard deviation of the last 10 calculation results is within 10% of the median of the calculation results of the current iteration cycle;

[0014] If yes, then increase the adaptive pheromone volatility factor by 5% within its definition domain and jump back to the step of updating the pheromone corresponding to the node;

[0015] If not, it is determined whether the improved ant colony algorithm has reached a preset maximum number of iterative operations.

[0016] Optionally, reconstructing the three-dimensional map to establish a solution space includes:

[0017] A vertex of the map data is selected as the origin of the coordinate system, and a coordinate system is established with the increasing directions of longitude, latitude and altitude as the three coordinate axes and the extreme values ​​of the map data as the end points of the three coordinate axes. The planning space is discretized by dividing the space equally. The set of grid points required for route planning is extracted to form the solution space.

[0018] Optionally, the height constraint conditions include: the waypoint height is greater than the map terrain height; the line connecting the front and rear waypoints is higher than the height of the terrain between the two points; on the same plumb plane, the waypoint with the lowest height is preferentially selected as the transfer node;

[0019] The angle constraint conditions include: setting the turning angle corresponding to the minimum turning radius as the minimum turning angle;

[0020] Distance constraints include: setting the minimum available track length and the maximum available track length between two waypoints.

[0021] Optionally, the parameters (α, β) of the ant colony algorithm are optimized and solved by a particle swarm algorithm to obtain the best information heuristic factor and the best expected heuristic factor, including:

[0022] A two-dimensional search space is formed by the definition domain of information heuristic factor and the definition domain of expectation heuristic factor;

[0023] Get the population size and the position and velocity of each particle in the population in the two-dimensional search space;

[0024] The fitness function in the ant colony algorithm is used as the objective function;

[0025] In the iterative process, the mapping of the objective function on each particle is calculated to update the position and velocity of the corresponding particle;

[0026] Calculate the result of the objective function through the error function;

[0027] If the error remains unchanged within the preset number of times, the particle swarm algorithm is judged to have converged, and the corresponding optimal information heuristic factor and optimal expected heuristic factor are obtained.

[0028] Optionally, updating the pheromone corresponding to the node includes:

[0029] Real-time update link:

[0030] After the current ant selects the next node, it updates the pheromone of the next node;

[0031] Path update phase:

[0032] After all ants have traversed all paths, the optimal ant crawling path is selected; the pheromone of each node on the optimal ant crawling path is updated.

[0033] Optionally, the path updating step also includes:

[0034] Based on the pheromone intensity factor of the best ant, the worst pheromone intensity factor, the best path, the worst path, the number of ants choosing the best path, the number of ants choosing the worst path and the total number of ants, the pheromone change is obtained;

[0035] Based on the pheromone change, the pheromone concentration of the best ant path node is increased accordingly, and the pheromone concentration of the worst ant path node is decreased accordingly.

[0036] Optionally, the path updating step also includes:

[0037] Set the definition domain of the adaptive pheromone volatility factor;

[0038] Initialize the adaptive pheromone volatility factor to the minimum value;

[0039] During the iterative calculation process, if the standard deviation of the accumulated 10 calculation results is within 10% of the median of the calculation results of the current iterative cycle number, the adaptive pheromone volatility factor is increased by 5% within the definition domain of the adaptive pheromone volatility factor.

[0040] Optionally, the domain of the adaptive pheromone volatility factor is set to [0.3, 1).

[0041] Optionally, the heuristic value is characterized by the travel distance and the node height:

[0042] η i,j,k =S·σ1d i,j,k ·σ2h i,j,k

[0043] Among them, S is the judgment operator, which is used to judge whether the next node is reachable; d i,j,k h is the distance between the ant's current node and the next node and the distance between the ant's next node to move to and the end point; i,j,k is the height of the node where the ant is currently located; σ1 is the weight coefficient of the travel distance; σ2 is the weight coefficient of the node height; and σ2≥σ1.

[0044] Beneficial effects of the embodiments of the present invention:

[0045] 1. The embodiment of the present invention proposes an aircraft low-altitude penetration route planning method based on an improved ant colony algorithm, which can quickly generate an aircraft penetration route. By improving the standard ant colony algorithm, it is used in three-dimensional terrain following route planning, providing useful assistance to relevant practitioners in route planning. At the same time, this method can be promoted and applied to other similar fields.

[0046] 2. This invention proposes a method for generating a three-dimensional terrain following route based on an improved ant colony algorithm. This method focuses on strengthening the positive gain of the best ants in the ant colony and weakening the negative gain of the worst ants, designs a pheromone update strategy; comprehensively considers the influence of the distance, height, and turning angle of feasible waypoints, designs a node movement strategy; and uses a particle swarm algorithm to intelligently optimize and solve the core parameters of the ant colony algorithm. In view of this, a universal solution for route planning for low-altitude aircraft penetration problems is formed. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0048] Figure 1 A flow chart of a method for low-altitude penetration route planning of an aircraft based on an improved ant colony algorithm in an embodiment of the present invention is shown;

[0049] Figure 2 A schematic diagram of ant tree movement in an embodiment of the present invention is shown;

[0050] Figure 3 A flowchart of another method for planning a low-altitude penetration route for an aircraft based on an improved ant colony algorithm in an embodiment of the present invention is shown;

[0051] Figure 4A variation trend diagram of the (α, β) parameter combination error of the particle swarm algorithm optimizing the ant colony algorithm in an embodiment of the present invention is shown;

[0052] Figure 5 A schematic diagram of a low-altitude penetration route obtained by an aircraft low-altitude penetration route planning method based on an improved ant colony algorithm in an embodiment of the present invention is shown.

[0053] Figure 6 The comparison results of the calculation process between the improved ant colony algorithm provided by the embodiment of the present invention and the traditional ant colony algorithm are shown. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0055] The embodiment of the present invention provides a method for low-altitude penetration route planning of an aircraft based on an improved ant colony algorithm, such as Figure 1 As shown, including:

[0056] Step S10, reconstructing the three-dimensional map to establish a solution space.

[0057] Before planning a route, we must first reconstruct the three-dimensional map to establish the solution space. Here, a vertex of the map data is used as the coordinate origin, the increasing directions of longitude, latitude and altitude are used as the three coordinate axes, and the extreme values ​​of the map data are used as the end points of each coordinate axis to establish a three-dimensional coordinate system. The enclosed cubic area constitutes the route planning space. The space is discretized by dividing the space equally, extracting the grid points required for route planning, and constructing the solution space from the set of these points.

[0058] Step S20, setting height constraint conditions, angle constraint conditions and distance constraint conditions.

[0059] In this embodiment, the height constraint conditions include: the height of the waypoint is greater than the height of the map terrain; the line connecting the front and rear waypoints is higher than the height of the terrain between the two points; on the same plumb plane, the waypoint with the lowest height is preferentially selected as the transfer node. The angle constraint conditions include: setting the turning angle corresponding to the minimum turning radius as the minimum turning angle. The distance constraint conditions include: setting the minimum available track length and the maximum available track length between two waypoints.

[0060] In a specific embodiment, in order to achieve terrain following to the greatest extent, strict requirements should be imposed on the height of waypoints. First, the waypoints should be reachable, that is, the waypoint height should be greater than the map terrain height; second, the track should be able to cross obstacles, that is, the line connecting the front and rear waypoints should be higher than the height of the terrain between the two points; third, on the same vertical plane, the waypoint with the lowest height should be selected as the transfer node as much as possible, that is, is a set of optional waypoints that meet requirement 1, I,J,k is the vertical plane where the point (I,J) lies.

[0061] In actual flight, the waypoint setting is greatly affected by its turning radius. The smaller the turning radius, the smaller the corresponding turning angle, the higher the energy required for maneuvering, the greater the fuel consumption, and the more obvious the impact on the flight time. Therefore, within the scope of conditions, route planning should try to search for relatively smooth flight routes. Constrained by the minimum turning radius, the waypoint turning angle θ i,j,k ≥θ lim ,θ lim is the minimum turning angle corresponding to the minimum turning radius.

[0062] Due to the limitations of the navigation system, aircraft control system and maneuverability, the track distance between every two waypoints in route planning needs to be constrained, that is, d i,j,k d i+1,j+1,k+1 ∈[d min ,d max ],d min d max are the minimum and maximum available track lengths, respectively.

[0063] Step S30, optimizing and solving the parameters (α, β) of the ant colony algorithm by using the particle swarm algorithm to obtain the best information heuristic factor and the best expected heuristic factor.

[0064] In this embodiment, α is the information heuristic factor, which reflects the degree of guidance of the ant colony search by the accumulated information during the movement process. The larger the value, the greater the role played in the movement of the ants, and the more likely the ants are to choose the previous path; the smaller the α value, the weaker the ants' memory of the previous crawling path; when α = 0, it is a traditional random greedy algorithm, and the probability of the shortest point of the path being selected is the highest. β is the expectation heuristic factor, which reflects the degree of guidance of the prior knowledge on the ant colony search. The larger the value, the more the ants tend to choose the local shortest path; the smaller the β value, the less the ants care about the prior knowledge; when β = 0, the ants will completely determine the movement path according to the pheromone concentration.

[0065] (α, β) is proportional to the convergence speed of the algorithm and inversely proportional to the search randomness. In order to ensure the quality of the algorithm, it is necessary to select appropriate (α, β) values. According to relevant research conclusions, considering the algorithm convergence speed and global search efficiency, take α∈[1,3], β∈[1,5]. Since there is currently no complete theoretical system to support the determination of the optimal combination of the two, the conventional practice is to use an exhaustive method based on specific research problems through a large number of repetitive experiments to continuously try and get a satisfactory combination. Since this method is highly dependent on subjective experience, it is time-consuming and inefficient, which seriously restricts the optimal performance of the algorithm. In order to solve this problem, this embodiment uses a particle swarm algorithm to optimize and solve (α, β).

[0066] Step S40, constructing a fitness function based on the pheromone concentration and the heuristic value.

[0067] In this embodiment, the pheromone concentration and the heuristic value are comprehensively considered to construct a fitness function:

[0068]

[0069]

[0070] Among them, κ is the angle compensation coefficient, which represents the influence of the turning angle on the selection of waypoints. When it is less than θ lim When the turning radius of the aircraft at this waypoint exceeds its inherent minimum turning radius requirement, the waypoint is abandoned; conversely, the larger the angle, the stronger the positive feedback brought by the algorithm. α is the information heuristic factor, and β is the expectation heuristic factor. τ i,j,k is the pheromone concentration.

[0071] In a specific embodiment, the heuristic value η i,j,k Characterized by travel distance and node height:

[0072] η i,j,k =S·σ1d i,j,k ·σ2h i,j,k

[0073]

[0074]

[0075] Among them, S is the judgment operator, which is used to judge whether the next node is reachable; d i,j,k h is the distance between the ant's current node and the next node and the distance between the ant's next node to move to and the end point; i,j,kis the height of the node where the ant is currently located; σ1 is the weight coefficient of the travel distance; σ2 is the weight coefficient of the node height; and σ2≥σ1. The purpose is to try to choose a point with a lower flight altitude when the action distance is equal, so as to give priority to the requirement of terrain following.

[0076] Step S50, selecting a node by roulette.

[0077] In this embodiment, if Figure 2 As shown, it is assumed that the ant can only move in the increasing direction along the vertical axis, and the moving range is limited, and the movable step length is ε I,J,K , using the tree-like moving method, the next waypoint search space is formed on the plumb projection plane, that is, the point set x i+1,j+1,k+1 =x i,j,k +ε I,J,K The probability of ant movement at each point is: The next waypoint is determined based on the probability of movement through the roulette method.

[0078] Step S60: updating the pheromone corresponding to the node.

[0079] Determine whether the standard deviation of the most recent 10 calculation results is within 10% of the median of the calculation results of the current iteration cycle number.

[0080] If yes, then step S701 is executed to increase the adaptive pheromone volatility factor by 5% within its definition domain, and then the process returns to step S60 to enter the next cycle.

[0081] If not, step S702 is executed to determine whether the improved ant colony algorithm has reached a preset maximum number of iterative operations. If the maximum number of iterative operations has been reached, the operation is terminated.

[0082] The embodiment of the present invention proposes a method for low-altitude penetration route planning of an aircraft based on an improved ant colony algorithm. By improving the standard ant colony algorithm, it is used in three-dimensional terrain following route planning, providing useful assistance to relevant practitioners in route planning. At the same time, this method can be extended and applied to other similar fields.

[0083] As an optional implementation, step S30 includes:

[0084] Step S301: forming a two-dimensional search space with an information heuristic factor definition domain and an expected heuristic factor definition domain.

[0085] Step S302, obtaining the population size and the position and velocity of each particle in the population in the two-dimensional search space.

[0086] Step S303: Using the fitness function in the ant colony algorithm as the objective function.

[0087] Step S304: during the iteration process, the mapping of the objective function on each particle is calculated to update the position and velocity of the corresponding particle.

[0088] Step S305, calculating the result of the objective function by using the error function.

[0089] If the error remains unchanged within the preset number of times, it is determined that the particle swarm algorithm has converged, and the corresponding optimal information heuristic factor and optimal expected heuristic factor are obtained. In a specific embodiment, the preset number of times is set according to actual needs.

[0090] The particle swarm algorithm originates from the study of bird foraging behavior. According to the rules of "individual cognition" and "group collaboration", it searches for the area around the individual closest to the food, and updates its own moving speed and location, and finally gathers around the food. The present invention makes full use of the superiority of the particle swarm algorithm in parameter optimization and the convenience of algorithm implementation to seek the best combination of (α, β).

[0091] In this embodiment, first, a two-dimensional search space Dim is formed with a (α, β) domain, and the position of each particle in the search space in a population of size n is X p ={X p,q} p∈n,q∈Dim , speed is V p = {V p,q} p∈n,q∈Dim The fitness function in the ant colony algorithm is used as the objective function. In the rth iteration, the particle position and velocity are updated by calculating the mapping of the objective function on each particle, that is,

[0092]

[0093]

[0094] Among them, c1 and c2 are acceleration factors, RND1 and RND2 are random numbers in the [0,1] interval, ω r is the inertia weight, which indicates the particle's ability to inherit the previous velocity.

[0095]

[0096] Among them, ω start is the initial inertia weight, ω end is the inertia weight at the end of the iteration, and R is the total number of iterations.

[0097] Record the objective function value F(gen) obtained in each loop iteration, and establish the error function E(gen) = |F(gen)-F(gen-1)|. If the error remains unchanged within a finite number of times, the algorithm is considered to have converged, and the corresponding (α, β) is what is sought.

[0098] As an optional implementation, step S60 includes:

[0099] Real-time update link:

[0100] After the current ant selects the next node, the pheromone of the next node is updated.

[0101] In this embodiment, the real-time update indicates the influence of the current ant pheromone on the population, that is, after the current ant selects a node, the pheromone of the node is updated to guide the crawling route of the subsequent ants.

[0102] τ i,j,k =(1-ρ)τ i,j,k

[0103] Path update phase:

[0104] After all ants have traversed all paths, the optimal ant crawling path is selected; the pheromone of each node on the optimal ant crawling path is updated.

[0105] In this embodiment, path updating represents the impact of the best individual pheromone on the population. After all ants have traversed all paths, the best ant's crawling path is selected, and the pheromone of each node on the path is updated to ensure that the path guides the ant population after the next cycle begins.

[0106] τ i,j,k =(1-ρ)τ i,j,k +Δτ i,j,k

[0107]

[0108] Among them, ρ is the pheromone volatility factor, Q is the pheromone intensity factor, Δτ i,j,k is the increment of pheromone in the node (i, j, k) of this cycle. When the value of pheromone increment is positive, it means that the pheromone concentration increases, and when the value of pheromone increment is negative, it means that the pheromone concentration decreases.

[0109] As an optional implementation, the path updating step further includes:

[0110] Based on the pheromone intensity factor of the best ant, the worst pheromone intensity factor, the best path, the worst path, the number of ants choosing the best path, the number of ants choosing the worst path and the total number of ants, the pheromone change is obtained;

[0111] Based on the pheromone change, the pheromone concentration of the best ant path node is increased accordingly, and the pheromone concentration of the worst ant path node is decreased accordingly.

[0112] In this embodiment, the path pheromone update of the ant colony algorithm is improved:

[0113] Reinforcement learning strategy: adjust the pheromone concentration according to the crawling paths of good and bad ants to ensure the overall excellence of the population and improve the algorithm's optimization ability. The main operation is to increase the pheromone concentration of the best ant path node to strengthen the guiding role of the best ant on the ant colony; reduce the pheromone concentration of the worst ant path node to reduce the deviation effect of the worst ant crawling route on the ant colony. Based on this idea, the increment of pheromone Δτ i,j,k It can be expressed as:

[0114]

[0115] Among them, q1 and q2 are the pheromone intensity factors of the best and worst ants respectively, m1 and m2 represent the individuals in the best and worst ants respectively, and L min is the optimal path, M is the total number of ants in the population, M min is the number of ants that select the optimal path, L max is the worst path, M max is the number of ants that choose the worst path.

[0116] As an optional implementation, the path updating step further includes:

[0117] Set the definition domain of the adaptive pheromone volatility factor;

[0118] Initialize the adaptive pheromone volatility factor to the minimum value;

[0119] During the iterative calculation process, if the standard deviation of the accumulated 10 calculation results is within 10% of the median of the calculation results of the current iterative cycle number, the adaptive pheromone volatility factor is increased by 5% within the definition domain of the adaptive pheromone volatility factor.

[0120] In this embodiment, the path pheromone update of the ant colony algorithm is improved:

[0121] Adaptive pheromone volatility factor adjustment strategy: The pheromone volatility factor ρ reflects the degree of mutual influence between ant colony individuals. When ρ is too large, the probability of selecting the searched path again increases, the algorithm randomness is weakened, and it is easy to fall into the local optimum; when ρ is too small, the residual pheromone concentration of the path is high, the algorithm randomness is enhanced, and the convergence speed is slowed down. Therefore, the best solution is to dynamically adjust the ρ value according to the results of the loop iteration. According to the conclusions of relevant research, the value range of ρ is generally between [0.3,1). Here, consider taking an adaptive pheromone volatility factor, and dynamically adjust the ρ value by judging the degree of deviation from the optimal path under a certain iterative operation. Specifically: first set the initial value of ρ to 0.3. During the calculation process, if the standard deviation of the cumulative 10 calculation results is within 10% of the median of the calculation results of the current iteration cycle, it means that the ρ value is increased by 5% within its definition domain.

[0122] In a specific embodiment, Figure 3 As shown, the algorithm is mainly implemented through the following steps:

[0123] Step 1: Set up the planning space of the three-dimensional terrain in the valley area, and construct the solution space by discretizing the planning space through the space equal division method.

[0124] Step 2: Establish a small-scale population of individuals and execute the improved ant colony algorithm to find the optimal path.

[0125] Step 3: Taking the ant colony moving path in step 2 as the evaluation target, the particle swarm algorithm is used to optimize and solve the optimal parameter combination (α, β) in the ant colony algorithm.

[0126] Step 4: The change in the standard deviation of the optimal path is used as the basis for judging the end of the particle swarm algorithm. If the standard deviation of the 30 results does not change, it is considered that the particle swarm algorithm has converged, and (α, β) at this time is what is required;

[0127] Step 5: Establish a larger population of individuals, substitute the optimization parameters of step 3, and execute the improved ant colony algorithm to find the optimal path;

[0128] Step 6: Determine whether the standard deviation of the accumulated 10 calculation results is within 10% of the median of the calculation results of the current iteration cycle. If so, increase the ρ value by 5% within its definition domain and enter the next cycle; if not, proceed to the next step;

[0129] Step 7: Determine whether the algorithm has reached the maximum number of iterations and the calculation ends.

[0130] Taking a certain three-dimensional terrain as an example, the low-altitude penetration route planning method proposed above is verified. Assume that the volume of the terrain after dimensionless processing is 21×21×10, and the coordinates of the starting point and the end point are O(1,18,5) and O*(21,8,6) respectively. The aircraft needs to find a shortest flight route that fits the undulating terrain as closely as possible without contacting it.

[0131] In the first step, a small ant colony with only 5 individuals is initialized, the number of iterations is set to 10, the optimal path of ant colony movement is used as the objective function, and the (α, β) parameter combination in the ant colony algorithm is optimized by the particle swarm algorithm. The population size of the particle swarm algorithm is 15, and the acceleration factor is 1.5. Figure 4 As shown in the figure, after 41 steps of iterative calculation, the result has converged. At this time, the corresponding combination of (α, β) is (1.9667, 3.9272).

[0132] In the second step, a medium-sized ant colony with 20 individuals is regenerated. According to the method provided in the above embodiment, the (α, β) combination parameters obtained in the first step are substituted to optimize the low-altitude penetration route. The results are shown in Figure 5 From the perspective of route quality, the algorithm in this embodiment can still search for feasible routes well under many constraints. Compared with the individual fitness changes obtained by the traditional ant colony algorithm, Figure 6 The algorithm of this embodiment can obtain the best individual fitness of 117.3 after 105 iterations, while the standard ant colony algorithm falls into the local optimal solution many times and has not fully converged until the maximum number of iterations of 300 steps. At this time, the corresponding local optimal individual fitness is 126.2.

[0133] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for low-altitude aircraft penetration route planning based on an improved ant colony algorithm, characterized in that: include: Reconstruct the three-dimensional map to establish the solution space; Set height constraints, angle constraints, and distance constraints; The parameters (α, β) of the ant colony algorithm are optimized and solved by the particle swarm algorithm to obtain the best information heuristic factor and the best expected heuristic factor; where α is the information heuristic factor and β is the expected heuristic factor; Based on the pheromone concentration and heuristic value, a fitness function is constructed; Select nodes via roulette wheel; Updating the pheromone corresponding to the node; Determine whether the standard deviation of the last 10 calculation results is within 10% of the median of the calculation results of the current iteration cycle; If yes, then the adaptive pheromone volatility factor is increased by 5% within its definition domain, and the process returns to the step of updating the pheromone corresponding to the node; If not, then determine whether the improved ant colony algorithm has reached the preset maximum number of iterative operations; Reconstruct the 3D map to establish the solution space, including: A coordinate system is established by selecting a vertex of the map data as the origin of the coordinates, taking the increasing directions of longitude, latitude and altitude as the three coordinate axes, and taking the extreme values ​​of the map data as the end points of the three coordinate axes; discretizing the planning space by equally dividing the space; extracting the set of grid points required for route planning to form the solution space; The height constraint conditions include: the waypoint height is greater than the map terrain height; the line connecting the front and rear waypoints is higher than the terrain height between the two points; on the same plumb plane, the waypoint with the lowest height is preferentially selected as the transfer node; The angle constraint condition includes: setting the turning angle corresponding to the minimum turning radius as the minimum turning angle; The distance constraint condition includes: setting a minimum available track length and a maximum available track length between two waypoints; The heuristic value is characterized by the travel distance and the node height: or i,j,k =S·σ1d i,j,k ·σ2h i,j,k Among them, S is the judgment operator, which is used to judge whether the next node is reachable; d i,j,k h is the distance between the ant's current node and the next node and the distance between the ant's next node to move to and the end point; i,j,k is the height of the node where the ant is currently located; σ1 is the weight coefficient of the travel distance; σ2 is the weight coefficient of the node height; and σ2≥σ1; i, j, k represent nodes; Based on the pheromone concentration and heuristic value, the fitness function is constructed including: Among them, κ is the angle compensation coefficient, which represents the turning angle θ i,j,k The influence on the selection of waypoints is that when the turning angle θ i,j,k Less than θ lim When the turning radius of the aircraft at this waypoint exceeds its inherent minimum turning radius requirement, the waypoint is abandoned; on the contrary, the larger the angle, the stronger the positive feedback brought by the algorithm; α is the information inspiration factor, β is the expectation inspiration factor; τ i,j,k is the pheromone concentration.

2. The method for aircraft low-altitude penetration route planning based on improved ant colony algorithm according to claim 1 is characterized in that: The particle swarm algorithm is used to optimize the parameters (α, β) of the ant colony algorithm to obtain the best information heuristic factor and the best expected heuristic factor, including: A two-dimensional search space is formed by the definition domain of information heuristic factor and the definition domain of expectation heuristic factor; Obtaining a population size and a position and a velocity of each particle in the population in the two-dimensional search space; Taking the fitness function in the ant colony algorithm as the objective function; In the iterative process, by calculating the mapping of the objective function on each particle, the position and velocity of the corresponding particle are updated; Calculating the result of the objective function through an error function; If the error remains unchanged within a preset number of times, it is determined that the particle swarm algorithm has converged, and the corresponding optimal information heuristic factor and the optimal expected heuristic factor are obtained.

3. The method for low-altitude aircraft penetration route planning based on improved ant colony algorithm according to claim 1 is characterized in that: Updating the pheromone corresponding to the node includes: Real-time update link: After the current ant selects the next node, the pheromone of the next node is updated; Path update phase: After all ants have traversed all paths, an optimal ant crawling path is selected; and the pheromone of each node on the optimal ant crawling path is updated.

4. The method for aircraft low-altitude penetration route planning based on improved ant colony algorithm according to claim 3 is characterized in that: The path updating step also includes: Based on the pheromone intensity factor of the best ant, the worst pheromone intensity factor, the best path, the worst path, the number of ants choosing the best path, the number of ants choosing the worst path and the total number of ants, the pheromone change is obtained; Based on the pheromone change amount, the pheromone concentration of the best ant path node is correspondingly increased, and the pheromone concentration of the worst ant path node is correspondingly reduced.

5. The method for aircraft low-altitude penetration route planning based on improved ant colony algorithm according to claim 3 is characterized in that: The path update phase also includes: Setting the definition domain of the adaptive pheromone volatility factor; Initializing the adaptive pheromone volatility factor to a minimum value; During the iterative calculation process, if the standard deviation of the accumulated 10 calculation results is within 10% of the median of the calculation results of the current iterative cycle number, the adaptive pheromone volatility factor is increased by 5% within the definition domain of the adaptive pheromone volatility factor.

6. The method for low-altitude aircraft penetration route planning based on improved ant colony algorithm according to claim 5 is characterized in that: The definition domain of the adaptive pheromone volatility factor is set to [0.3, 1).