Multi-agent collaborative trajectory planning method and planning system considering radar threats in three-dimensional environment
By improving the ant colony optimization method to plan UAV trajectories in a 3D environment, the problem of multi-agent cooperative trajectory planning under complex terrain and radar threats was solved, achieving path optimization and enhanced security.
Patent Information
- Application Number
- CN202311766596.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-21
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2043-12-21
Smart Images

Figure CN117850471B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of path planning, and in particular relates to a multi-agent collaborative track planning method and a planning system thereof in a three-dimensional environment taking into account radar threats. Background Art
[0002] In recent years, the application of UAVs (Unmanned Aerial Vehicles) in military warfare and civilian fields has become increasingly widespread. With its strong adaptability, wide coverage, sophisticated structure and flexible maneuverability, UAVs have been successfully applied to many application scenarios that manned aircraft are unable to handle, and therefore have attracted much attention at home and abroad. In the civilian field, UAVs have also achieved rapid development in logistics and express delivery, earthquake relief, aerial photography and other fields, and their application scope and depth have been continuously expanded. However, with the complexity of the environment, a single UAV is often unable to perform specific tasks due to the limitations of sensor detection capabilities and load carrying. The complementary capabilities and coordinated actions of multiple UAVs can significantly improve the overall combat effectiveness of UAVs, so UAV swarm combat has become a new combat method that military powers around the world are competing to study.
[0003] Due to factors such as battlefield complexity, target uncertainty, and sensor uncertainty, drones need to obtain one or more feasible flight paths as references when performing missions. In complex terrain and highly confrontational environments, how to plan a suitable flight path to meet mission requirements is crucial to the success of the mission. In addition, the planned route needs to meet specific flight constraints and requirements, which increases the complexity of the problem. Summary of the invention
[0004] The present invention provides a multi-agent collaborative trajectory planning method in a three-dimensional environment taking radar threats into consideration, so as to solve the problem that when a UAV performs a mission in a complex terrain and highly confrontational environment, the planned route needs to meet specific flight constraints and requirements.
[0005] The present invention also provides a multi-agent collaborative trajectory planning system considering radar threats in a three-dimensional environment, which is used to implement a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment.
[0006] The present invention is achieved through the following technical solutions:
[0007] A multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment, the planning method specifically comprising the following steps:
[0008] Step 1: Load the mission environment terrain data and use the DEM data file with a resolution of 30 meters as the mission terrain;
[0009] Step 2: Establish the task environment;
[0010] Step 3: Define the feasible route structure;
[0011] Step 4: Define the constraints during the flight process, including dynamic constraints, multi-UAV coordination constraints, environmental safety constraints, and radar safety constraints;
[0012] Considering the limitations of the drone's propulsion system, its flyable speed range is also restricted.
[0013]
[0014] Among them, v min is the minimum flight speed of the drone. is the flight speed of any UAV at time t, v max is the maximum flight speed of the drone;
[0015] At the same time, in order to reduce the probability of being detected during flight, the flight altitude range relative to the terrain is limited.
[0016]
[0017] in, is the flight altitude of any UAV at time t, H min is the minimum flight altitude from the terrain, H max is the maximum flight altitude relative to the terrain;
[0018] Step 5: Based on the constraints in step 4, define the cost function during the flight process;
[0019] Step 6: Use the improved ant colony method to solve the cost function of step 5;
[0020] Step 7: Iterate and optimize to obtain the final planned route to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the mission.
[0021] Furthermore, the step 3 is specifically to use the following formula to represent the path structure:
[0022]
[0023] Among them, (x i ,y i , z i ) is the three-dimensional coordinate of the drone, t i is the flight time of the drone, v i is the flight speed of the drone, φ i and γ i Represent the angles of the track respectively.
[0024] Furthermore, the dynamic constraint is specifically:
[0025] The variable angles between adjacent waypoints are restricted to meet the performance constraints of the UAV, as follows:
[0026]
[0027] The multi-UAV coordination constraints are specifically:
[0028] The collaborative constraints of mission benefits between UAVs are as follows:
[0029] ||X i (t)-X j (t)||≥d safe ,i≠j (6)
[0030] Formula (6) At the same time, the distance between any UAVs is kept outside the safe distance to achieve spatial coordination;
[0031]
[0032] The UAV in equations (7) and (8) can achieve more precise time control through speed control.
[0033] Furthermore, the environmental safety constraints are specifically:
[0034]
[0035] The radar safety constraints are specifically:
[0036] The UAV is modeled using the ellipsoid RCS model, where c1 and c2 represent the relevant parameters of the radar.
[0037]
[0038] Substitute the relative angle between the UAV and the radar into the above formula (10) to calculate the probability of detection:
[0039]
[0040] The impact of multiple radars on the overall detection probability is calculated by the following formula,
[0041]
[0042] Furthermore, the step 5 defines the cost function during the flight process, specifically,
[0043]
[0044] J total =ω1Jlength +ω2J safe +ω3J energy +ω4J time (17)
[0045] Wherein, N represents the number of path points, and based on this, the sections on the path with a detection probability greater than 0.5 are counted to characterize the detection risk of the path, and ω1~ω4 are the corresponding indicator weights.
[0046] Furthermore, the step 6 is specifically as follows:
[0047] Pheromone initialization, where τ max and τ min Indicates the concentration range of pheromones;
[0048] Construct feasible solutions based on energy consumption, path length, safety and overlap;
[0049] Design state transition rules;
[0050] Pheromones updated and height corrected.
[0051] Furthermore, the pheromone initialization is specifically to use a pheromone initialization method based on height variation to guide individual ants to search towards feasible areas, and the ants will not select infeasible areas marked by other ants.
[0052]
[0053] The construction of the feasible solution is specifically as follows:
[0054] In the construction process, the energy consumption, path length, safety and overlap indicators are comprehensively considered. σ1, σ2, σ3 and σ4 represent the importance of the heuristic information of energy consumption, path length, safety and overlap, respectively.
[0055]
[0056] For path length heuristic information, the remaining distance is estimated by calculating the straight-line distance between each feasible point in the candidate feasible node set and the end point, so that the ants can reach the end point faster.
[0057]
[0058] η energy =η smooth ·η smooth1 (twenty three)
[0059] For path smoothing heuristic information, the angle information state changes of the nodes in the feasible candidate set are mainly considered. More consideration is given to path points with slow changes in order to achieve a smooth transition of the path.
[0060]
[0061] For path safety information, calculate the distance between each point in the candidate point set and the obstacles in the environment and select the feasible point with the largest distance as the candidate point.
[0062]
[0063] R(P j )=R(P j )+1,P j ∈alive ant (26)
[0064] For path repetition information, by processing the information of ants that successfully construct feasible solutions and applying corresponding processing to their path points, a path that is repeatedly explored reduces the possibility of subsequent exploration, thereby stimulating the algorithm to explore undeveloped areas.
[0065] Furthermore, the state transition rule is specifically designed as follows: the following pseudo-random strategy is used to select feasible path points from the candidate point set. The pseudo-random strategy can ensure that the heuristic information and uncertainty are comprehensively considered in the process of state transition, thereby avoiding the algorithm from falling into the local optimum too early.
[0066]
[0067] Furthermore, the present invention considers three updating modes of pheromones, namely, local and global updating of attraction pheromones and updating of warning pheromones.
[0068] τ(P j )=(1-ρ local )τ(P j )+ρ local (τ0+Δτ(P j )) (29)
[0069]
[0070] The above formula uses the membership function in fuzzy mathematics to estimate the importance of the state point, so as to guide the key points in the subsequent search process.
[0071] τ(P j )=(1-ρ global )τ(P j )+ρ global Δτ(Pj ) (31)
[0072]
[0073] The above formula uses the path point on the global optimal individual of the current iteration and applies pheromone increment to it to ensure that the ant colony can quickly move towards a feasible solution and accelerate the convergence speed.
[0074]
[0075]
[0076] The above formula marks the path points on the infeasible solution individuals in the construction process, so as to selectively avoid the infeasible path points in the calculation process of heuristic information, thereby preventing the algorithm from consuming too many resources in the infeasible area;
[0077] Specifically, the altitude correction is to impose a relative distance limit on the terrain altitude obtained by interpolation calculation in order to ensure that the flight altitude of the UAV always maintains a safe distance from the ground. The corrected waypoint coordinates can be expressed as P j (x j ,y j ,h(x j ,y j )+H min );h(x j ,y j ) is the corresponding interpolation function.
[0078] A multi-agent collaborative trajectory planning system considering radar threats in a three-dimensional environment, the planning system comprising a parameter definition module and a route planning module;
[0079] The parameter definition module is used to load the mission environment terrain data, using a DEM data file with a resolution of 30 meters as the mission terrain;
[0080] Establishing the task environment;
[0081] Define feasible route structures;
[0082] Define the constraints during the flight process, including dynamic constraints, multi-UAV coordination constraints, environmental safety constraints, and radar safety constraints;
[0083] Considering the limitations of the drone's propulsion system, its flyable speed range is also restricted.
[0084]
[0085] At the same time, in order to reduce the probability of being detected during flight, the flight altitude range relative to the terrain is limited.
[0086]
[0087] Based on the constraints, define the cost function during the flight process;
[0088] The route planning module is used to solve the cost function using an improved ant colony method;
[0089] The final planned route is obtained through iterative optimization to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the mission.
[0090] The beneficial effects of the present invention are:
[0091] The present invention provides a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment, which is used to solve the problem that when a UAV performs a mission in a complex terrain and highly confrontational environment, the planned route needs to meet specific flight constraints and requirements. The present invention first establishes an interaction model between UAVs and radar threats in a three-dimensional terrain environment and defines the corresponding mission objective function in combination with specific mission requirements. Further, an improved ant colony optimization method is designed in combination with the characteristics of the mission terrain to obtain a feasible trajectory that meets the mission requirements. The terrain data is randomly downloaded from the geospatial data cloud and smoothed preprocessed. Three UAVs were used for simulation experiments, and different mission requirements were set to verify the effectiveness of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0092] Figure 1 It is a flow chart of the method of the present invention.
[0093] Figure 2 It is the algorithm flow chart of the present invention.
[0094] Figure 3 It is a schematic diagram of the environment representation of the present invention.
[0095] Figure 4 It is a schematic diagram of the inter-visibility detection of the present invention.
[0096] Figure 5 It is a schematic diagram of the structure of the mixed pheromone of the present invention.
[0097] Figure 6 It is a comparison chart of the effects of different pheromone initial methods of the present invention.
[0098] Figure 7 It is a schematic diagram of the task environment of the present invention.
[0099] Figure 8 It is a box plot of cost value distribution of the present invention.
[0100] Fig. 9It is the cost convergence curve of the present invention.
[0101] Fig.10 It is a feasible track schematic diagram of the present invention. DETAILED DESCRIPTION
[0102] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.
[0103] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.
[0104] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.
[0105] The following is attached to this application specification Figure 1-10 , the technical solutions in the embodiments of the present application are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.
[0106] In the following description, many specific details are set forth to facilitate a full understanding of the present application, but the present application may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present application. Therefore, the present application is not limited to the specific embodiments disclosed below.
[0107] Embodiment 1
[0108] The present invention provides a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment, and the planning method specifically comprises the following steps:
[0109] Step 1: Load the mission environment terrain data and use the DEM data file with a resolution of 30 meters as the mission terrain;
[0110] Step 2: Establish the task environment;
[0111] Step 3: Define the feasible route structure;
[0112] Step 4: The present invention uses a 3-DOF model to model the UAV and establishes constraints based on it, wherein the constraints include dynamic constraints, multi-UAV coordination constraints, environmental safety constraints, and radar safety constraints;
[0113] Considering the limitations of the drone's propulsion system, its flyable speed range is also restricted.
[0114]
[0115] Among them, v min is the minimum flight speed of the drone. is the flight speed of any UAV at time t, v max is the maximum flight speed of the drone;
[0116] At the same time, in order to reduce the probability of being detected during flight, the flight altitude range relative to the terrain is limited.
[0117]
[0118] in, is the flight altitude of any UAV at time t, H min is the minimum flight altitude from the terrain, H max is the maximum flight altitude relative to the terrain;
[0119] Step 5: Based on the constraints in step 4, define the cost function during the flight process;
[0120] Step 6: Use the improved ant colony method to solve the cost function of step 5;
[0121] Step 7: Iterate and optimize to obtain the final planned route to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the mission.
[0122] Specifically, the step 1 is that the real three-dimensional grid terrain data can be downloaded from the geospatial data cloud (http: / / www.gscloud.cn / home). In the present invention, a DEM data file with a resolution of 30 meters is used as the task terrain. The elevation data at the grid intersection can be directly read and obtained, and the elevation data at the middle of the grid can be obtained by interpolation fitting. The grid map schematic diagram under the task environment is as follows: Figure 3 shown.
[0123] The step 2 specifically includes the establishment of the mission environment. In the present invention, {V, S, T, E, D, C} is used to represent the multi-UAV collaborative trajectory planning problem. Represents a collection of drones. and Respectively represent the starting point and end point of the task. U is the number of drones. E is the overall mission environment. is the threat in the environment, N T is the number of threats.
[0124] Furthermore, the step 3 specifically includes using Path to represent the path structure.
[0125]
[0126] Among them, (x i ,y i , z i ) is the three-dimensional coordinate of the drone, t i is the flight time of the drone, v i is the flight speed of the drone, φ i and γ i Represent the angles of the track respectively.
[0127] Furthermore, the dynamic constraint is specifically:
[0128] The variable angles (heading angle change and pitch angle change) between adjacent waypoints are restricted to meet the performance constraints of the UAV, as shown below:
[0129]
[0130] in, is the change in heading angle of the UAV at time t, and is the spatial position coordinate of the UAV at time t+1, is the pitch angle change between track points at adjacent moments, Δφ max and Δγ max Indicates the change threshold of the corresponding angle;
[0131] The multi-UAV coordination constraints are specifically:
[0132] In order to ensure the mission efficiency between UAVs, corresponding coordination constraints are also defined.
[0133] The collaborative constraints of mission benefits between UAVs are as follows:
[0134] ||X i (t)-X j (t)||≥d safe,i≠j (6)
[0135] Among them, X i (t) and X j (t) are the three-dimensional spatial positions of different UAVs at time t, d safe is the safety distance threshold between drones;
[0136] The constraint of equation (6) above ensures that the distance between any UAVs at the same time is kept outside the safe distance, in order to achieve spatial coordination;
[0137]
[0138] Among them, L i is the resulting path length, V min and V max Indicates the speed range of the drone, For the fastest arrival time, is the slowest arrival time, t i is the flight time of the drone, represents a collection of drones;
[0139] The constraints of equations (7) and (8) above ensure that the time when the drones arrive at the destination has an intersection, thereby ensuring temporal coordination, that is, the drones can achieve more precise time control through speed control.
[0140] Furthermore, the environmental safety constraints are specifically:
[0141] In order to avoid collisions between the drone and obstacles in the environment, environmental constraints are defined
[0142]
[0143] Among them, Obs is the set of environmental obstacles, Indicates that there is no intersection between the path and the obstacle;
[0144] The radar safety constraints are specifically:
[0145] In order to keep the UAV performing tasks with a low probability of being discovered as much as possible, the UAV is modeled using the ellipsoid RCS model, where c1 and c2 represent the relevant parameters of the radar.
[0146]
[0147] Among them, σ represents the radar reflection cross-sectional area of the UAV relative to the radar, R i Indicates the distance between the drone and the radar;
[0148] Substitute the relative angle between the UAV and the radar into the above formula (10) to calculate the probability of detection:
[0149]
[0150] Among them, a, b, c are the ellipsoid shape parameters, λ e ,μ e represents the angle of the UAV relative to the radar, φ, λ, μ are the azimuth, pitch and tilt angles of the UAV respectively;
[0151] The influence of multiple radars K on the overall detection probability is considered and calculated by the following formula:
[0152]
[0153] in, It is a switching function, which means that the calculation is performed only when the joint detection probability exceeds the specified threshold.
[0154] Furthermore, the step 5 defines the cost function during the flight process, specifically,
[0155] The purpose of this invention is to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the task. The corresponding calculation method is as follows:
[0156]
[0157]
[0158] J total =w1J length +w2J safe +w3J enery +w4J time (17)
[0159] Where N represents the number of path points, and based on this, the sections on the path with a detection probability greater than 0.5 are counted to characterize the detection risk of the path, ω1~ω4 are the corresponding indicator weights, and J length is the path length cost, J safe is the path safety cost, J energye is the path energy consumption cost, J time is the path coordination cost, J total is the total path cost.
[0160] Furthermore, the step 6 is specifically as follows:
[0161] Pheromone initialization, where τ max and τ min Indicates the concentration range of pheromones;
[0162] Construct feasible solutions based on energy consumption, path length, safety and overlap;
[0163] Design state transition rules;
[0164] Pheromones updated and height corrected.
[0165] Furthermore, the pheromone initialization is specifically as follows: considering that the global uniform initialization method of pheromone in the traditional ant colony algorithm is not applicable to the grid three-dimensional environment in the present invention, a pheromone initialization method based on height variation is proposed to guide the ant individuals to search towards a feasible area, where τ max and τ min Indicates the concentration range of pheromones. In addition, in order to avoid invalid searches of repeated areas, warning pheromones are introduced. By marking infeasible individuals, multiple searches can be avoided to improve search efficiency. It should be noted that in the present invention, an independent maintenance strategy is adopted for attraction pheromones, while a common maintenance strategy is adopted for warning pheromones, that is, ants will not select infeasible areas marked by other ants.
[0166]
[0167] Among them, e is the natural index, K is the artificial attenuation factor used to control the change of pheromone with the terrain height difference, H is the terrain height of the current path point, and H forbid The flight altitude threshold is prohibited;
[0168] The construction of the feasible solution is specifically as follows:
[0169] In the construction process, the energy consumption, path length, safety and overlap indicators are comprehensively considered. σ1, σ2, σ3 and σ4 represent the importance of the heuristic information of path length, energy consumption, safety and overlap, respectively.
[0170]
[0171]
[0172] Among them, S Pi Indicates the current waypoint P i The set of subsequent feasible nodes,
[0173] For the path length heuristic information η length , by calculating the straight-line distance between each feasible point in the candidate feasible node set and the end point, the remaining distance is estimated so that the ants can reach the end point faster.
[0174]
[0175] η energy =η smooth ·η smooth1 (twenty three)
[0176] Among them, φ Pi and slope(P i ) represent waypoints P i The heading angle and environmental obstacle gradient, where the environmental gradient can be obtained by prior differentiation of the mission terrain;
[0177] For path smoothing heuristic information η smooth , mainly considers the change of the angle information state of the nodes in the feasible candidate set, and gives more consideration to the path points with slow changes in order to achieve a smooth transition of the path.
[0178]
[0179] Among them, F(P j ,Obs) represents the calculation of waypoint P j The distance function to the environmental obstacles, in this case, is calculated using the Euclidean distance;
[0180] For the path security information η safety , calculate the distance between each point in the candidate point set and the obstacles in the environment and select the feasible point with the largest distance as the candidate point,
[0181]
[0182] R(P j )=R(P j )+1,P j ∈alive ant (26)
[0183] Among them, R(P j ) is used to record the number of times the current candidate track point appears in the feasible solution, that is, the more times a track point appears, the smaller the probability of being selected;
[0184] For the path repeatability information η repetition ,By processing the information of ants that successfully construct feasible solutions, and applying corresponding processing to their path points, a path that is repeatedly explored reduces the possibility of subsequent exploration, thus stimulating the algorithm to explore unexplored areas.
[0185] Furthermore, the state transition rule is specifically designed as follows: the following pseudo-random strategy is used to select feasible path points from the candidate point set. The pseudo-random strategy can ensure that the heuristic information and uncertainty are comprehensively considered in the process of state transition, thereby avoiding the algorithm from falling into the local optimum too early.
[0186]
[0187] Among them, ξ(P j ) represents the concentration of danger pheromone, q0∈(0,1) is the artificially specified judgment threshold, q is a randomly generated pseudo-random number, and the calculation formula of roulette is shown in (28).
[0188] Furthermore, the present invention considers three updating modes of pheromones, namely, local and global updating of attraction pheromones and updating of warning pheromones.
[0189] τ(P j )=(1-ρ local )τ(P j )+ρ local (τ0+Δτ(P j )) (29)
[0190]
[0191] The above formula uses the membership function in fuzzy mathematics to estimate the importance of the state point, so as to guide the key points in the subsequent search process.
[0192] τ(P j )=(1-ρ local )τ(P j )+ρ local (τ0+Δτ(P j )) (31)
[0193]
[0194] The above formula uses the path point on the global optimal individual of the current iteration and applies pheromone increment to it to ensure that the ant colony can quickly move towards a feasible solution and accelerate the convergence speed.
[0195]
[0196]
[0197] The above formula marks the path points on the infeasible solution individuals in the construction process, so as to selectively avoid the infeasible path points in the calculation process of heuristic information, thereby preventing the algorithm from consuming too many resources in the infeasible area;
[0198] Specifically, the altitude correction is to impose a relative distance limit on the terrain altitude obtained by interpolation calculation in order to ensure that the flight altitude of the UAV always maintains a safe distance from the ground. The corrected waypoint coordinates can be expressed as P j (x j ,yj ,h(x j ,y j )+H min );h(x j ,y j ) is the corresponding interpolation function.
[0199] The beneficial effects of the present invention are further illustrated below in conjunction with specific embodiments.
[0200] The terrain data was randomly downloaded from the geospatial data cloud and smoothed. Three drones were used for simulation experiments, and different mission requirements were set to verify the effectiveness of the present invention. The key parameters of the algorithm are summarized in Table 1.
[0201] Table 1 Key simulation parameters
[0202]
[0203] from Figure 8-Figure 9 It can be seen that the proposed method can quickly explore the solution space and avoid repeated searches for infeasible solutions in the early stages of iteration, so it can quickly obtain high-quality feasible solutions. It is superior to the comparison algorithm in terms of the number of iterative convergence, solution quality, and solution stability.
[0204] Table 2 Statistical data under different methods
[0205]
[0206]
[0207] Combination Fig.10 As can be seen from Table 2, the method can plan a reasonable flight path, that is, flying along the valley in the terrain while reducing the risk of being detected by using the radar detection blind area caused by terrain shielding. From the statistics of the relevant information of the path, it can be seen that the path obtained by the present invention has certain advantages over the comparison algorithm in terms of path length, dangerous proportion, arrival time and energy consumption, thereby proving the effectiveness of the present invention.
[0208] Finally, it should be noted that the above embodiments are only used to describe the technical solution of the present invention rather than to limit the technical method. The present invention can be extended to other modifications, changes, applications and embodiments in application, and therefore it is believed that all such modifications, changes, applications and embodiments are within the spirit and teaching scope of the present invention.
[0209] Embodiment 2
[0210] The present invention also provides a multi-agent collaborative trajectory planning system considering radar threats in a three-dimensional environment, wherein the planning system includes a parameter definition module and a route planning module;
[0211] The parameter definition module is used to load the mission environment terrain data, using a DEM data file with a resolution of 30 meters as the mission terrain;
[0212] Establishing the task environment;
[0213] Define feasible route structures;
[0214] Define the constraints during the flight process, including dynamic constraints, multi-UAV coordination constraints, environmental safety constraints, and radar safety constraints;
[0215] Considering the limitations of the drone's propulsion system, its flyable speed range is also restricted.
[0216]
[0217] Among them, v min is the minimum flight speed of the drone. is the flight speed of any UAV at time t, v max is the maximum flight speed of the drone;
[0218] At the same time, in order to reduce the probability of being detected during flight, the flight altitude range relative to the terrain is limited.
[0219]
[0220] in, is the flight altitude of any UAV at time t, H min is the minimum flight altitude from the terrain, H max is the maximum flight altitude relative to the terrain;
[0221] Based on the constraints, define the cost function during the flight process;
[0222] The route planning module is used to solve the cost function using an improved ant colony method;
[0223] The final planned route is obtained through iterative optimization to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the mission.
[0224] The real three-dimensional grid terrain data of the parameter definition module can be downloaded from the geospatial data cloud (http: / / www.gscloud.cn / home). In the present invention, a DEM data file with a resolution of 30 meters is used as the task terrain. The elevation data at the grid intersection can be directly read and obtained, and the elevation data at the middle of the grid can be obtained by interpolation fitting. The grid map schematic diagram under the task environment is shown in the figure below: Figure 3 shown.
[0225] The establishment of the task environment of the parameter definition module is specifically the establishment of the task environment. In the present invention, {V, S, T, E, D, C} is used to represent the multi-UAV collaborative trajectory planning problem. Represents a collection of drones. and Respectively represent the starting point and end point of the task. U is the number of drones. E is the overall mission environment. is the threat in the environment, N T is the number of threats.
[0226] Furthermore, the parameter definition module defines the feasible route structure specifically as follows: Path is used to represent the path structure.
[0227]
[0228] Among them, (x i ,y i , z i ) is the three-dimensional coordinate of the drone, t i is the flight time of the drone, v i is the flight speed of the drone, φ i and γ i Represent the angles of the track respectively.
[0229] Furthermore, the dynamic constraint is specifically:
[0230] The variable angles (heading angle change and pitch angle change) between adjacent waypoints are restricted to meet the performance constraints of the UAV, as shown below:
[0231]
[0232]
[0233] in, is the change in heading angle of the UAV at time t, and is the spatial position coordinate of the UAV at time t+1, is the pitch angle change between track points at adjacent moments, Δφmax and Δγ max Indicates the change threshold of the corresponding angle;
[0234] The multi-UAV coordination constraints are specifically:
[0235] In order to ensure the mission efficiency between UAVs, corresponding coordination constraints are also defined.
[0236] The collaborative constraints of mission benefits between UAVs are as follows:
[0237] ||X i (t)-X j (t)||≥d safe ,i≠j (6)
[0238] Among them, X i (t) and X j (t) are the three-dimensional spatial positions of different UAVs at time t, d safe is the safety distance threshold between drones;
[0239] The constraint of equation (6) above ensures that the distance between any UAVs at the same time is kept outside the safe distance, in order to achieve spatial coordination;
[0240]
[0241] Among them, L i is the resulting path length, V min and V max Indicates the speed range of the drone, For the fastest arrival time, is the slowest arrival time, t i is the flight time of the drone, represents a collection of drones;
[0242] The constraints of equations (7) and (8) above ensure that the time when the drones arrive at the destination has an intersection, thereby ensuring temporal coordination, that is, the drones can achieve more precise time control through speed control.
[0243] Furthermore, the environmental safety constraints are specifically:
[0244] In order to avoid collisions between the drone and obstacles in the environment, environmental constraints are defined
[0245]
[0246] Among them, Obs is the set of environmental obstacles, Indicates that there is no intersection between the path and the obstacle;
[0247] The radar safety constraints are specifically:
[0248] In order to keep the UAV performing tasks with a low probability of being discovered as much as possible, the UAV is modeled using the ellipsoid RCS model, where c1 and c2 represent the relevant parameters of the radar.
[0249]
[0250] Among them, σ represents the radar reflection cross-sectional area of the UAV relative to the radar, R i The distance between the UAV and the radar is expressed by the relative angle between the UAV and the radar. The detection probability is calculated by substituting the relative angle between the UAV and the radar into the above formula (10).
[0251]
[0252] Among them, a, b, c are the ellipsoid shape parameters, λ e ,μ e Represents the angle of the drone relative to the radar, φ, λ, μ are the azimuth, pitch and tilt angles of the drone respectively
[0253] The influence of multiple radars K on the overall detection probability is considered and calculated by the following formula:
[0254]
[0255] in, It is a switching function, which means that the calculation is performed only when the joint detection probability exceeds the specified threshold.
[0256] Furthermore, the cost function of the parameter definition module during the flight process is specifically defined as follows:
[0257] The purpose of this invention is to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the task. The corresponding calculation method is as follows:
[0258]
[0259] J total =ω1J lenght +ω2J safe +ω3J energy +ω4J time (17)
[0260] Where N represents the number of path points, and based on this, the sections on the path with a detection probability greater than 0.5 are counted to characterize the detection risk of the path, ω1~ω4 are the corresponding indicator weights, and J length is the path length cost, J safe is the path safety cost, J energye is the path energy consumption cost, Jtime is the path coordination cost, J total is the total path cost.
[0261] Furthermore, the route planning module uses an improved ant colony method to solve the cost function as follows:
[0262] Pheromone initialization, where τ max and τ min Indicates the concentration range of pheromones;
[0263] Construct feasible solutions based on energy consumption, path length, safety and overlap;
[0264] Design state transition rules;
[0265] Pheromones updated and height corrected.
[0266] Furthermore, the pheromone initialization is specifically as follows: considering that the global uniform initialization method of pheromone in the traditional ant colony algorithm is not applicable to the grid three-dimensional environment in the present invention, a pheromone initialization method based on height variation is proposed to guide the ant individuals to search towards a feasible area, where τ max and τ min Indicates the concentration range of pheromones. In addition, in order to avoid invalid searches of repeated areas, warning pheromones are introduced. By marking infeasible individuals, multiple searches can be avoided to improve search efficiency. It should be noted that in the present invention, an independent maintenance strategy is adopted for attraction pheromones, while a common maintenance strategy is adopted for warning pheromones, that is, ants will not select infeasible areas marked by other ants.
[0267]
[0268] Among them, e is the natural index, K is the artificial attenuation factor used to control the change of pheromone with the terrain height difference, H is the terrain height of the current path point, and H forbid The flight altitude threshold is prohibited
[0269] The construction of the feasible solution is specifically as follows:
[0270] In the construction process, the energy consumption, path length, safety and overlap indicators are comprehensively considered. σ1, σ2, σ3 and σ4 represent the importance of the heuristic information of path length, energy consumption, safety and overlap, respectively.
[0271]
[0272] in, Indicates the current waypoint P i The set of subsequent feasible nodes,
[0273] For the path length heuristic information η length , by calculating the straight-line distance between each feasible point in the candidate feasible node set and the end point, the remaining distance is estimated so that the ants can reach the end point faster.
[0274]
[0275] η energy =η smooth ·η smooth1 (twenty three)
[0276] Among them, φ Pi and slope(P i ) represent waypoints P i The heading angle and environmental obstacle gradient, where the environmental gradient can be obtained by prior differentiation of the mission terrain
[0277] For path smoothing heuristic information η smooth , mainly considers the change of the angle information state of the nodes in the feasible candidate set, and gives more consideration to the path points with slow changes in order to achieve a smooth transition of the path.
[0278]
[0279] Among them, F(P j ,Obs) represents the calculation of waypoint P j The distance function to the environmental obstacles, in this case, is calculated using the Euclidean distance
[0280] For the path security information η safety , calculate the distance between each point in the candidate point set and the obstacles in the environment and select the feasible point with the largest distance as the candidate point,
[0281]
[0282] R(P j )=R(R j )+1,P j ∈alive ant (26)
[0283] Among them, R(P j ) is used to record the number of times the current candidate track point appears in the feasible solution, that is, the more times a track point appears, the smaller the probability of being selected
[0284] For the path repeatability information η repetition,By processing the information of ants that successfully construct feasible solutions, and applying corresponding processing to their path points, a path that is repeatedly explored reduces the possibility of subsequent exploration, thus stimulating the algorithm to explore unexplored areas.
[0285] Furthermore, the state transition rule is specifically designed as follows: the following pseudo-random strategy is used to select feasible path points from the candidate point set. The pseudo-random strategy can ensure that the heuristic information and uncertainty are comprehensively considered in the process of state transition, thereby avoiding the algorithm from falling into the local optimum too early.
[0286]
[0287] Among them, ξ(P j ) represents the concentration of danger pheromone, q0∈(0,1) is the artificially specified judgment threshold, q is a randomly generated pseudo-random number, and the calculation formula of roulette is shown in (28).
[0288] Furthermore, the present invention considers three updating modes of pheromones, namely, local and global updating of attraction pheromones and updating of warning pheromones.
[0289] τ(P j )=(1-ρ local )τ(P j )+ρ local (τ0+Δτ(P j )) (29)
[0290]
[0291] The above formula uses the membership function in fuzzy mathematics to estimate the importance of the state point, so as to guide the key points in the subsequent search process.
[0292] τ(P j )=(1-ρ global )τ(P j )+ρ global Δτ(P j ) (31)
[0293]
[0294] The above formula uses the path point on the global optimal individual of the current iteration and applies pheromone increment to it to ensure that the ant colony can quickly move towards a feasible solution and accelerate the convergence speed.
[0295]
[0296]
[0297] The above formula marks the path points on the infeasible solution individuals in the construction process, so as to selectively avoid the infeasible path points in the calculation process of heuristic information, thereby preventing the algorithm from consuming too many resources in the infeasible area;
[0298] Specifically, the altitude correction is to impose a relative distance limit on the terrain altitude obtained by interpolation calculation in order to ensure that the flight altitude of the UAV always maintains a safe distance from the ground. The corrected waypoint coordinates can be expressed as P j (x j ,y j ,h(x j ,y j )+H min );h(x j ,y j ) is the corresponding interpolation function.
[0299] Specifically, it is used to implement a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment.
Claims
1. A multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment, characterized in that: The planning method specifically comprises the following steps: Step 1: Load the mission environment terrain data and use the DEM data file with a resolution of 30 meters as the mission terrain; Step 2: Establish the task environment; Step 3: Define the feasible route structure; Step 4: Define the constraints during the flight process, including dynamic constraints, multi-UAV coordination constraints, environmental safety constraints, and radar safety constraints; The speed range in which drones can fly is restricted. Among them, v min is the minimum flight speed of the drone. is the flight speed of any UAV at time t, v max is the maximum flight speed of the drone; Limit the range of altitudes that the drone can fly relative to the terrain. in, is the flight altitude of any UAV at time t, H min is the minimum flight altitude from the terrain, H max is the maximum flight altitude relative to the terrain; Step 5: Based on the constraints in step 4, define the cost function during the flight process; Step 6: Use the improved ant colony method to solve the cost function of step 5; The step 6 is specifically as follows: Pheromone initialization, where τ max and τ min Indicates the concentration range of pheromones; Construct feasible solutions based on energy consumption, path length, safety and overlap; Design state transition rules; Pheromone update and height correction; The pheromone initialization is specifically: using a pheromone initialization method based on height variation to guide individual ants to search towards feasible areas, and the ants will not choose infeasible areas marked by other ants. Among them, e is the natural index, K is the artificial attenuation factor used to control the change of pheromone with the terrain height difference, H is the terrain height of the current path point, and H forbid The flight altitude threshold is prohibited; The construction of the feasible solution is specifically as follows: In the construction process, energy consumption, path length, safety and overlap indicators are comprehensively considered. σ1, σ2, σ3 and σ4 represent the importance of the heuristic information of energy consumption, path length, safety and overlap, respectively. in, Indicates the current waypoint P i The set of subsequent feasible nodes, is the path length heuristic, is the energy consumption heuristic, is the security heuristic, is the iterative path repetition heuristic term, P target To set the target point coordinates, P j is the node coordinate in the extended node set, and slope(P i ) represent waypoints P i The heading angle and environmental obstacle gradient; For path length heuristic information, the remaining distance is estimated by calculating the straight-line distance between each feasible point in the candidate feasible node set and the end point, so that the ants can reach the end point faster. or energy =the smooth ·or smooth1 (23) For path smoothing heuristic information, the angle information state changes of the nodes in the feasible candidate set are mainly considered. More consideration is given to path points with slow changes in order to achieve a smooth transition of the path. For path safety information, calculate the distance between each point in the candidate point set and the obstacles in the environment and select the feasible point with the largest distance as the candidate point. R(P j )=R(P j )+1,P j ∈alive ant (26) For path repetition information, by processing the information of ants that successfully construct feasible solutions and applying corresponding processing to their path points, a path that is repeatedly explored reduces the possibility of subsequent exploration, thereby stimulating the algorithm to explore unexplored areas; Step 7: Iterate and optimize to obtain the final planned route to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the mission.
2. According to claim 1, a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment is characterized in that: The step 3 is specifically as follows: the path structure is represented by the following formula: Among them, (x i ,y i ,z i ) is the three-dimensional coordinate of the drone, t i is the flight time of the drone, v i is the flight speed of the drone, φ i and γ i Represent the angles of the track respectively.
3. According to claim 1, a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment is characterized in that: The dynamic constraints are specifically: The variable angles between adjacent waypoints are restricted to meet the performance constraints of the UAV, as follows: in, is the variable angle between adjacent waypoints, and is the spatial position coordinate of the UAV at time t+1, is the pitch angle change between track points at adjacent moments, Δφ max and Δγ max Indicates the change threshold of the corresponding angle, The multi-UAV coordination constraints are specifically: The collaborative constraints of mission benefits between UAVs are as follows: ||X i (t)-X j (t)||≥d safe ,i≠j (6) Among them, X i (t) and X j (t) are the three-dimensional spatial positions of different UAVs at time t, d safe is the safety distance threshold between drones, Formula (6) At the same time, the distance between any UAVs is kept outside the safe distance to achieve spatial coordination; Among them, L i is the resulting path length, V min and V max Indicates the speed range of the drone, For the fastest arrival time, is the slowest arrival time, t i is the flight time of the drone, represents a collection of drones; The UAV in equations (7) and (8) can achieve more precise time control through speed control.
4. According to claim 1, a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment is characterized in that: The environmental safety constraints are specifically: Among them, Obs is the set of environmental obstacles, Indicates that there is no intersection between the path and the obstacle; The radar safety constraints are specifically: The UAV is modeled using the ellipsoid RCS model, where c1 and c2 represent the relevant parameters of the radar. Among them, σ represents the radar reflection cross-sectional area of the UAV relative to the radar, R i Indicates the distance between the drone and the radar; Substitute the relative angle between the UAV and the radar into the above formula (10) to calculate the probability of detection: Among them, a, b, c are the ellipsoid shape parameters, λ e ,μ e represents the angle of the UAV relative to the radar, φ, λ, μ are the azimuth, pitch and tilt angles of the UAV respectively; The impact of multiple radars on the overall detection probability is calculated by the following formula, Among them, P d is the impact of multiple radars on the overall detection probability, is the switch function.
5. According to claim 1, a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment is characterized in that: The step 5 defines the cost function during the flight process, specifically: I total =ω1J length +ω2J safe +ω3J energy +ω4J time (17) Where N represents the number of path points, and based on this, the sections on the path with a detection probability greater than 0.5 are counted to characterize the detection risk of the path, ω1~ω4 are the corresponding indicator weights, and J length is the path length cost, J safe is the path safety cost, J energye is the path energy consumption cost, J time is the path coordination cost, J total is the total path cost, P d This is the impact of multiple radars on the overall detection probability.
6. According to claim 1, a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment is characterized in that: The state transfer rule is specifically designed as follows: the following pseudo-random strategy is used to select feasible path points from the candidate point set. The pseudo-random strategy can ensure that the heuristic information and uncertainty are comprehensively considered during the state transfer process, thereby avoiding the algorithm from falling into the local optimum too early. Among them, ξ(P j ) represents the concentration of dangerous pheromone, q0∈(0,1) is the artificially specified judgment threshold, q is a randomly generated pseudo-random number, 7. According to claim 1, a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment is characterized in that: The present invention considers three updating modes of pheromones, namely, local and global updating of attraction pheromones and updating of warning pheromones. τ(P j )=(1-ρ local )τ(P j )+r local (τ0+Δτ(P j )) (29) The above formula uses the membership function of fuzzy mathematics to estimate the importance of the state point, so as to guide the key points in the subsequent search process. τ(P j )=(1-ρ global )τ(P j )+r global Δτ(P j ) (31) The above formula uses the path point on the global optimal individual of the current iteration and applies pheromone increment to it to ensure that the ant colony can quickly move towards a feasible solution and accelerate the convergence speed. The above formula marks the path points on the infeasible solution individuals in the construction process, so as to selectively avoid the infeasible path points in the calculation process of heuristic information, thereby preventing the algorithm from consuming too many resources in the infeasible area; The height correction is specifically: in order to ensure that the flying height of the UAV always maintains a safe distance from the ground, a relative distance limit is imposed on the terrain height obtained by interpolation calculation. The corrected waypoint coordinates can be expressed as P j (x j ,y j ,h(x j ,y j )+H min ).
8. A multi-agent collaborative trajectory planning system considering radar threats in a three-dimensional environment, characterized in that: The system uses a multi-agent collaborative trajectory planning method considering radar threats in a three-dimensional environment as described in any one of claims 1-7, and the system includes: The parameter definition module is used to load the mission environment terrain data, using a DEM data file with a resolution of 30 meters as the mission terrain; Establishing the task environment; Define feasible route structures; Define the constraints during the flight process, including dynamic constraints, multi-UAV coordination constraints, environmental safety constraints, and radar safety constraints; Limit the speed range within which drones can fly. Among them, v min is the minimum flight speed of the drone. is the flight speed of any UAV at time t, v max is the maximum flight speed of the drone; Limit the range of altitudes that the drone can fly relative to the terrain. in, is the flight altitude of any UAV at time t, H min is the minimum flight altitude from the terrain, H max is the maximum flight altitude relative to the terrain; Based on the constraints, define the cost function during the flight process; The route planning module is used to solve the cost function using the improved ant colony method; The final planned route is obtained through iterative optimization to minimize the path length, energy consumption, arrival time difference and detection probability required for the UAV to perform the mission.
