Distributed collaborative target search method and device for unmanned aerial vehicles in dynamic environment
A distributed collaborative target search method optimized by rasterization and ant colony algorithm solves the problems of search efficiency and safety of UAV swarms in dynamic environments, and achieves efficient and safe target discovery and area coverage.
Patent Information
- Application Number
- CN202311804902.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-26
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2043-12-26
AI Technical Summary
In dynamic environments, existing technologies for collaborative search missions involving drone swarms suffer from problems such as stationary targets, inability to quickly detect targets, low area coverage, and algorithms that are prone to getting trapped in local optima, which can lead to mission failure.
A distributed collaborative target search method is adopted. By rasterizing the task environment, the overall probability density distribution, uncertainty metric, attraction and repulsion pheromones and partition markers are used for initialization. The search area is optimized by combining ant colony algorithm, a search benefit function is established, and the update of UAV status and environmental information is optimized to ensure search safety and efficiency.
It improves the search efficiency and safety of drone swarms in dynamic environments, enabling rapid target discovery, reducing resource waste in cross-regional searches, and enhancing the reliability and effectiveness of mission completion.
Smart Images

Figure CN117908560B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of multi-agent collaborative technology, and in particular to a distributed collaborative target search method and device for unmanned aerial vehicles in a dynamic environment. Background Art
[0002] In recent years, various unmanned aerial vehicles (UAVs) have been increasingly used in military warfare and the civilian sector. UAVs are widely sought after at home and abroad because of their strong adaptability, wide range of action, sophisticated structure, and flexible mobility, and they can achieve many application scenarios that are difficult for manned aircraft to achieve. In the military field, they are widely used in intelligence reconnaissance, battlefield assessment, and bomb-carrying attacks; in the civilian field, they are also developing rapidly in logistics and express delivery, earthquake relief, aerial photography, and other fields, and UAVs are currently expanding their breadth and depth of application. However, as the battlefield environment becomes increasingly complex, a single UAV is often unable to perform a given task due to limitations in sensor detection capabilities and weapon payloads, while the complementary capabilities and coordinated actions of multiple UAVs can significantly improve the overall combat effectiveness of UAVs in terms of mission completion, robustness, and reliability. Therefore, UAV swarm combat is becoming a new combat style that the world's military powers are competing to study.
[0003] In order to achieve the coordinated combat of large-scale drone swarms and enhance the overall efficiency of drone swarms in completing tasks, it is necessary to design a scientific and effective planning and decision-making method for drone swarms. Without appropriate planning and decision-making methods, drone swarms may have conflicts in time, space, and tasks, which will not only fail to play the scale advantage, but may even lead to mission failure. Therefore, drone swarm task planning is the core of achieving the advantages of swarm combat. Among them, drone swarm collaborative search task planning will be an important means of performing regional search in the future.
[0004] Due to the complexity of the mission area, unknown targets, and sensor uncertainty, the battlefield environment is dynamic, and real-time decisions can only be made through real-time communication between drones, and pre-planning of search missions is not possible. For the task planning problem, domestic and foreign literature has made a lot of contributions to the search aspect. However, although most of the literature has enabled drone clusters to have the ability of collaborative search, there are still problems such as static targets, inability to quickly discover targets, low regional coverage, and algorithms that are prone to falling into local optimality.
[0005] Therefore, how to provide a method for collaborative target search for drones in a dynamic environment has become a technical problem that needs to be solved urgently in this field. Summary of the invention
[0006] In order to solve the above technical problems, the present invention provides a distributed collaborative target search method and device for drones in a dynamic environment. The method first rasterizes the task environment and uses attributes of different dimensions to characterize the information in the environment in order to obtain efficient search results. Further, on the basis of defining the voronoi diagram, the uncertainty difference between the drone search areas is balanced with the minimum uncertainty between areas as an indicator to reduce the efficiency loss caused by cross-area search. In addition, the relevant search benefit function is defined and the ant colony algorithm is used to solve the problem to obtain the optimal control sequence and obtain the state of the drone at the next moment and the environmental perception and repeat this process until the task is completed. The method can improve the search efficiency on the basis of ensuring the search safety and has high practical value. .
[0007] Based on the same inventive concept, the present invention has four independent technical solutions:
[0008] 1. A distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment, comprising the following steps:
[0009] S1, obtain the task environment map and rasterize it;
[0010] S2, initializing the rasterized map based on the overall probability density distribution, uncertainty measurement index, attraction and repulsion pheromone, and partition mark;
[0011] S3, updating the probability of existence of a target within the search range based on the rasterized map;
[0012] S4. Calculate the total uncertainty of the mission area corresponding to each UAV according to the uncertainty measurement index and the partition mark, and optimize the regional division;
[0013] S5. Calculating task benefits and task constraints based on the attraction and repulsion pheromones;
[0014] S6. Establishing an optimization model based on the target existence probability, task benefit and task constraint, and optimizing the target through an ant colony algorithm, wherein the task benefit is maximized as the optimization target;
[0015] S7, updating the state information and environment information of the drone at the next moment;
[0016] S8. Repeat steps S3 to S7 until the search task is completed, and update the communication topology relationship between the drones at a preset communication interval and unify the environmental information between the drones.
[0017] Furthermore, the uncertainty metric is expressed as follows:
[0018]
[0019] Among them, k represents the current time, represents the uncertainty in the grid (x, y) at the current moment, p xy (k) represents the probability of the target existing in the grid (x, y) at the current moment.
[0020] Further, the attraction and repulsion pheromones are represented as follows:
[0021] S a (k+1)=(1-V a )[(1-G a )(S a (k)+d a (EV))+GP a (k)];
[0022] S r (k+1)=(1-V r )[(1-G r )(S r (k)+d r V)+GP r (k)];
[0023] Among them, V a V r G a G r is the pheromone transmission and volatility coefficient, GP a (k) and GP r (k) is the pheromone concentration transmitted from the adjacent grid, E is the unit matrix, V is the path matrix, S a (k+1) and S r (k+1) is the concentration of the attracting and repelling pheromones at the next moment of the grid. The pheromone concentration of the grid is expressed as
[0024] Furthermore, when the rasterized map is initialized, a voronoi marker map attribute is also included;
[0025] Step S4 includes:
[0026] S41. Calculate the total uncertainty of each UAV's corresponding mission area, expressed as follows:
[0027] The uncertainty differences between the search areas are expressed as follows:
[0028]
[0029] Where o(i) is the total uncertainty of the mission area corresponding to the i-th UAV, σ xy (k) is the partition mark;
[0030] S42, optimize and adjust the center point of the Voronoi diagram to obtain the area division result with the minimum uncertainty difference, when |Δ j+1 -Δ j |≤θ, where θ is any decimal greater than 0. P * is the center point of the Voronoi diagram.
[0031] Furthermore, the task benefit and task constraint are expressed as follows:
[0032]
[0033] Among them, ω1, ω2, ω3>0 are the corresponding weights and satisfy ω1+ω2+ω3=1, Indicates the target detection benefit brought by searching the specified area, represents the environmental exploration benefit of searching the specified area, Indicates the cumulative pheromone gain of searching the specified area. Represents a partition constraint, Represents a safety constraint.
[0034] Furthermore, target optimization through ant colony algorithm includes:
[0035] S61, pheromone initialization, the pheromone in the no-fly zone obstacle area is 0, the attraction pheromone in the initial environment is greater than the repulsion pheromone p att >p rep ;
[0036] S62. Solution construction: each ant individual represents a feasible solution, and the length of the solution is the length of the prediction time window {0,1,...,-1} 1×T , where 0, 1, and -1 represent the unit change in the direction of travel at the next moment relative to the current direction;
[0037] S63, calculation of heuristic information, taking task benefits, task constraints and repeated search items as heuristic information:
[0038]
[0039]
[0040]
[0041]
[0042] Among them, ζ, χ and Indicates the importance of the corresponding heuristic information, |N xy| represents the total number of available neighbors around the grid, Indicates the location of the obstacle. represents the location information of UAV i at time k, R detect Indicates the detection range of the drone. represents the size of repulsive pheromone in the environment, N search Indicates the number of neighbors that have been searched. represents the size of the reward pheromone in the environment, Indicates the heuristic information size of the current grid;
[0043] S64, state transition rules and pheromone update, the next state is selected using the following pseudo-random method and the pheromone of the corresponding grid is updated after successful update:
[0044]
[0045]
[0046]
[0047]
[0048] in, Indicates the probability of selecting the next node at the next moment k+1 in the current grid (x, y). N xy Represents a set of subsequent optional nodes, and Respectively represent the pheromone concentration of the grid at the current moment and the next moment, Indicates the size of the heuristic information of the current grid, α and β represent the weights of the two respectively. q0 is the artificially set judgment threshold, and q is a random number randomly generated between [0,1]. ρ local and ρ global Represent the local and global pheromone volatility coefficients respectively, τ0 is the artificially set fixed pheromone increment, is based on the predicted cost R(S i (k))Calculate the pheromone increment within the grid (x,y).
[0049] Furthermore, in step S8, the neighbor set N with which the k-drone can communicate and exchange information at any time is i (k) The following conditions are met:
[0050] N i (k) = {U j |||s i (k)-s j (k)||2≤R com ,U j ∈Nuav};
[0051] Among them, U j represents the neighbors within the communication range of UAV i, s i (k) represents the state of UAV i at time k, s j (k) represents the state of UAV j at time k, R com Represents the communication radius, N uav Represents the set of all drones.
[0052] 2. A distributed collaborative target search device for unmanned aerial vehicles in a dynamic environment, comprising:
[0053] Map acquisition module, used to obtain the task environment map and rasterize it;
[0054] Initialization module, used to initialize the rasterized map based on four attributes: overall probability density distribution, uncertainty metric, attraction and repulsion pheromone, and partition mark;
[0055] A first updating module, configured to update the existence probability of a target within a search range based on the rasterized map;
[0056] A region optimization module, used to calculate the total uncertainty of the mission area corresponding to each UAV according to the uncertainty measurement index and the partition mark, and optimize the region division;
[0057] A constraint module, for calculating task benefits and task constraints based on the attraction and repulsion pheromones;
[0058] An optimization module is used to establish an optimization model based on the target existence probability, task benefit and task constraints, and perform target optimization through an ant colony algorithm, wherein the task benefit is maximized as the optimization target;
[0059] The second updating module is used to update the state information and environment information of the drone at the next moment;
[0060] The loop module is used to repeatedly execute the first update module to the second update module until the search task is completed, and to update the communication topology relationship between the drones at a preset communication interval and unify the environmental information between the drones.
[0061] 3. A computer-readable storage medium storing a computer program, wherein the computer program implements the above method when executed by a processor.
[0062] 4. An electronic device, comprising a processor and a storage device, wherein the storage device stores a plurality of instructions, and the processor is used to read the plurality of instructions in the storage device and execute the above method.
[0063] The distributed collaborative target search method and device for unmanned aerial vehicles in a dynamic environment provided by the present invention have at least the following beneficial effects:
[0064] The uncertainty differences between different regions during the UAV search mission were comprehensively considered and an online solution method framework was designed, which improved the search efficiency by reducing cross-region searches. Through iterative division of the mission environment, the mission area with relatively balanced uncertainty was found. On this basis, a search benefit function was constructed and solved using an improved ant colony method. This method can improve the search efficiency while ensuring the search safety and has high practical value. BRIEF DESCRIPTION OF THE DRAWINGS
[0065] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0066] Figure 1 It is a flow chart of a distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment of the present invention;
[0067] Figure 2 It is a schematic diagram of a search path of a distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment of the present invention;
[0068] Figure 3 It is a target search probability schematic diagram of a distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment of the present invention;
[0069] Figure 4 It is a schematic diagram of the distance between the UAV and the dynamic obstacle in a distributed collaborative target search method for UAV in a dynamic environment of the present invention;
[0070] Figure 5 It is a schematic diagram of the minimum distance between drones in a distributed collaborative target search method for drones in a dynamic environment of the present invention;
[0071] Figure 6 It is a statistical diagram of the number of iterations of a distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment of the present invention;
[0072] Figure 7 It is a statistical relationship diagram between environmental uncertainty and search times of a distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment of the present invention. DETAILED DESCRIPTION
[0073] 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.
[0074] 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.
[0075] 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.
[0076] The following is a clear and complete description of the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. 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 this application.
[0077] 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.
[0078] Embodiment 1:
[0079] See also Figure 1 In some embodiments, a distributed collaborative target search method for a UAV in a dynamic environment is provided, comprising the following steps:
[0080] S1, obtain the task environment map and rasterize it;
[0081] S2, initializing the rasterized map based on the overall probability density distribution, uncertainty measurement index, attraction and repulsion pheromone, and partition mark;
[0082] S3, updating the probability of existence of a target within the search range based on the rasterized map;
[0083] S4. Calculate the total uncertainty of the mission area corresponding to each UAV according to the uncertainty measurement index and the partition mark, and optimize the regional division;
[0084] S5. Calculating task benefits and task constraints based on the attraction and repulsion pheromones;
[0085] S6. Establishing an optimization model based on the target existence probability, task benefit and task constraint, and optimizing the target through an ant colony algorithm, wherein the task benefit is maximized as the optimization target;
[0086] S7, updating the state information and environment information of the drone at the next moment;
[0087] S8. Repeat steps S3 to S7 until the search task is completed, and update the communication topology relationship between the drones at a preset communication interval and unify the environmental information between the drones.
[0088] Specifically, in step S1, each UAV loads the mission environment map and rasterizes it and defines the relevant detection distance threshold to ensure the success of the search. Load the mission environment map Ω and discretize it into L x ×L y Assume that N grids are randomly distributed in the environment. target The task target, whose coordinates are T i ,i=1,2,...,N target express. dispatchN uav A UAV performs a task, and its coordinates can be expressed as Each drone is equipped with corresponding sensors to collect environmental information, and its sensing range is represented by R sensor Indicates that the detection range is represented by R detect Indicated by R, the communication range is com To facilitate searching, we assume that each grid has certain attributes Represent the probability of target existence, the probability of environmental uncertainty, the pheromone map, and the voronoi marker map, which are used to distinguish the different states of the grid at different times. We assume that the drones fly at the same altitude, and all targets and obstacles within the detection range can be detected. To ensure safety, we define a safe distance R between drones. safe During the search process, the distance between drones and between drones and obstacles must always be greater than the safe distance.
[0089] In step S2, we can obtain prior information about the target position distribution before executing the task. Assuming that the probability distribution of the target in space follows a Gaussian distribution, its overall probability density distribution can be calculated as follows:
[0090]
[0091] In the formula, represents the approximate location of the target obtained through prior information, is the uncertainty variance, c n is the target existence intensity. It should be noted that the target existence probability of each grid must satisfy p xy (k)∈[0,1]. At the same time, due to the possible no-fly zones in the environment, the probability of the target existing in this area is set to p xy (k)=0.
[0092] Considering that the prior information of the drone may be inaccurate and there is still uncertainty in the environment, we define an uncertainty metric for each grid. The uncertainty metric is expressed as follows:
[0093]
[0094] Among them, k represents the current time, represents the uncertainty in the grid (x, y) at the current moment, p xy (k) represents the probability of the target existing in the grid (x, y) at the current moment.
[0095] We introduce attraction and repulsion pheromone maps to guide the drone to move towards unsearched areas while avoiding repeated searches of the same area. The attraction and repulsion pheromone maps are represented as follows:
[0096] S a (k+1)=(1-V a )[(1-G a )(S a (k)+d a (EV))+GP a (k)];
[0097] S r (k+1)=(1-V r )[(1-G r )(S r (k)+d r V)+GP r (k)];
[0098] Among them, V a V r G a G r is the pheromone transmission and volatility coefficient, GP a (k) and GP r (k) is the pheromone concentration transmitted from the adjacent grid, E is the unit matrix, V is the path matrix, Sa (k+1) and S r (k+1) is the concentration of the attracting and repelling pheromones at the next moment of the grid. The pheromone concentration of the grid is expressed as
[0099] For efficient collaborative search, we define a partition mark σ for each grid xy (k) When searching, the drone will prioritize searching within its own area to avoid wasting resources caused by frequent cross-region searches.
[0100] As a better implementation method, the relative position relationship between drones changes over time, which will cause the communication topology between drones to change over time. Therefore, we define that at any time, the set of neighbors with which k drones can communicate and exchange information needs to meet the following conditions:
[0101] N i (k) = {U j |||s i (k)-s j (k)||2≤R com ,U j ∈N uav};
[0102] Among them, U j represents the neighbors within the communication range of UAV i, s i (k) represents the state of UAV i at time k, s j (k) represents the state of UAV j at time k, R com Represents the communication radius, N uav Represents the set of all drones.
[0103] In each communication interval T com ,UAVs unify and update environmental information and various maps stored locally on each UAV through corresponding information fusion mechanisms.
[0104] In step S3, based on the environment map and communication topology established above, we update the target existence probability within the sensor range according to the following formula:
[0105]
[0106] Among them, b xy (k) = 1 means the target is detected, b xy (k) = 0 means no target is detected, p d represents the detection probability of the sensor, p f Represents the false alarm probability of the detector.
[0107] Step S4 includes the following steps:
[0108] S41. In order to ensure the balance between the search areas corresponding to the drones, the total uncertainty of the mission area corresponding to each drone is calculated, which is expressed as follows:
[0109] The uncertainty differences between the search areas are expressed as follows:
[0110]
[0111] Where o(i) is the total uncertainty of the mission area corresponding to the i-th UAV, σ xy (k) is the partition mark;
[0112] S42, optimize and adjust the center point of the Voronoi diagram to obtain the area division result with the minimum uncertainty difference, when |Δ j+1 -Δ j |≤θ, where θ is any decimal greater than 0. P * is the center point of the Voronoi diagram.
[0113] In step S5, we define the relevant mission benefits to guide the drone to search for a more efficient area. The mission benefits and mission constraints are expressed as follows:
[0114]
[0115] Among them, ω1, ω2, ω3>0 and satisfy ω1+ω2+ω3=1, Indicates the target detection benefit brought by searching the specified area, represents the environmental exploration benefit of searching the specified area, Indicates the cumulative pheromone gain of searching the specified area. Represents a partition constraint, Represents a safety constraint.
[0116] Specifically, step S5 includes the following sub-steps:
[0117] S51. Calculate the target detection benefit brought by searching the specified area, expressed as follows:
[0118]
[0119]
[0120] Among them, δ p is the target existence threshold;
[0121] S52, calculating the environmental exploration benefit of searching the specified area, which is expressed as follows;
[0122]
[0123] S53, calculate the cumulative pheromone benefits of searching the specified area, expressed as follows:
[0124]
[0125] S54. Calculate the safety constraints in the search process, expressed as follows:
[0126]
[0127]
[0128] The above constraints indicate that as the distance between drones decreases, the cost value will gradually increase, which ensures the safety of flight to a certain extent. The following constraints represent the cost calculation method between drones and obstacles:
[0129]
[0130]
[0131] S55. Calculate the partition constraints during the search process, as shown below:
[0132]
[0133]
[0134] Among them, C1 is a large constant used to distinguish infeasible solutions, γ is a decay constant used to characterize the urgency of the constraint, and T is the prediction window length.
[0135] Based on the above analysis, we established an overall optimization model to obtain the control input sequence at the next moment by maximizing the task benefit
[0136]
[0137] In step S6, the target optimization by using the ant colony algorithm includes:
[0138] S61, pheromone initialization, the pheromone in the no-fly zone obstacle area is 0, and this area will not attract the drone. In addition, in order to ensure the effectiveness of the initial search phase, the attraction pheromone in the initial environment is set to be greater than the repulsion pheromone p att >p rep ;
[0139] S62. Solution construction: each ant individual represents a feasible solution, and the length of the solution is the length of the prediction time window {0,1,...,-1} 1×T, where 0, 1, and -1 represent the unit change in the direction of travel at the next moment relative to the current direction;
[0140] S63, calculation of heuristic information, taking task benefits, task constraints and repeated search items as heuristic information, aims to balance search and obstacle avoidance;
[0141]
[0142]
[0143]
[0144]
[0145] Among them, ζ, χ and Indicates the importance of the corresponding heuristic information, |N xy | represents the total number of available neighbors around the grid, Indicates the location of the obstacle. represents the location information of UAV i at time k, R detect Indicates the detection range of the drone. represents the size of repulsive pheromone in the environment, N search Indicates the number of neighbors that have been searched. represents the size of the reward pheromone in the environment, Indicates the heuristic information size of the current grid;
[0146] S64, state transition rules and pheromone update, the next state is selected using the following pseudo-random method and the pheromone of the corresponding grid is updated after successful update:
[0147]
[0148]
[0149]
[0150]
[0151] in, Indicates the probability of selecting the next node at the next moment k+1 in the current grid (x, y). N xy Represents a set of subsequent optional nodes, and Respectively represent the pheromone concentration of the grid at the current moment and the next moment, Indicates the size of the heuristic information of the current grid, α and β represent the weights of the two respectively. q0 is the artificially set judgment threshold, and q is a random number randomly generated between [0,1]. ρlocal and ρ global Represent the local and global pheromone volatility coefficients respectively, τ0 is the artificially set fixed pheromone increment, is based on the predicted cost R(S i (k))Calculate the pheromone increment within the grid (x,y).
[0152] In step S7, after obtaining the optimal input sequence, the state information of the drone at the next moment is calculated. The state of the drone at adjacent moments is updated using linearity, which is expressed as follows:
[0153]
[0154] Among them, s i (k)=(x i (k),y i (k),φ i (k)) represents the position and heading of the drone
[0155] The beneficial effects of this embodiment are further illustrated below through specific simulation experiments.
[0156] We randomly downloaded terrain data from the geospatial data cloud and downsampled it into a 100*100 mission area, which contains impassable terrain obstacles and randomly moving dynamic obstacles. The positions of the drone and the target are randomly generated in the feasible space. Our mission goal is to maximize the search efficiency and the number of target detections while ensuring safety. The key parameters are summarized in Table 1.
[0157] Table 1
[0158] <![CDATA[Sensor detection probability p d > 0.9 <![CDATA[False alarm probability p of the sensor f > 0.1 <![CDATA[Iteration time upper limit Iter max > 1500 Prediction window length T 8 <![CDATA[Number N of unmanned aerial vehicles uav > 5 <![CDATA[Target number N target > 15 <![CDATA[Safe distance R safe > 5 <![CDATA[Detection distance R detect > 6 <![CDATA[Sensor radius R sensor > 2 Information decay factor γ 0.9 <![CDATA[Communication interval T com > 10
[0159] Figure 2 , Figure 3 The distribution information of drones and targets, the detection results of drones on targets, and the search path information between drones are displayed. It can be seen that this method can search for all predetermined targets based on prior information, and the detection probability reaches 100%. Figure 2 It can be seen that the search paths between the UAVs overlap less, which also shows that the regional search strategy we proposed can improve the search efficiency and reduce invalid repeated searches.
[0160] Figure 4 , Figure 5 The changes in the closest distance curves between the drone and the dynamic obstacle and the number of iterations are shown, which proves that Figure 3 The safety and flyability of the flight path. Figure 5It can be seen that our method can always maintain a safe distance between drones during the search process. The comparison method will cause the distance between drones to be too low, which reduces safety.
[0161] Figure 6 The box plot shows the minimum number of iterations required for 50 repeated experiments in the same environment. It can be seen that our method has the lowest mean number of iterations and a compact distribution, which is significantly superior to other methods and can meet the task requirements of quickly searching for targets in dynamic environments.
[0162] Figure 7 The curves of our method and other methods on environmental uncertainty are shown. It can be seen that our method can quickly reduce the uncertainty of the overall environment and eliminate the uncertainty first (reduced to 0). Other algorithms require more iterations to complete. The local enlarged figure shows the final number of iterations, which is consistent with Figure 6 The number of terminations of our method is consistent with that of
[0163] Embodiment 2:
[0164] In some embodiments, a distributed collaborative target search device for a drone in a dynamic environment is provided, comprising:
[0165] Map acquisition module, used to obtain the task environment map and rasterize it;
[0166] Initialization module, used to initialize the rasterized map based on four attributes: overall probability density distribution, uncertainty metric, attraction and repulsion pheromone, and partition mark;
[0167] A first updating module, configured to update the existence probability of a target within a search range based on the rasterized map;
[0168] A region optimization module, used to calculate the total uncertainty of the mission area corresponding to each UAV according to the uncertainty measurement index and the partition mark, and optimize the region division;
[0169] A constraint module, for calculating task benefits and task constraints based on the attraction and repulsion pheromones;
[0170] An optimization module is used to establish an optimization model based on the target existence probability, task benefit and task constraints, and perform target optimization through an ant colony algorithm, wherein the task benefit is maximized as the optimization target;
[0171] The second updating module is used to update the state information and environment information of the drone at the next moment;
[0172] The loop module is used to repeatedly execute the first update module to the second update module until the search task is completed, and to update the communication topology relationship between the drones at a preset communication interval and unify the environmental information between the drones.
[0173] As a preferred implementation, in the initialization module, the uncertainty metric is expressed as follows:
[0174]
[0175] Among them, k represents the current time, represents the uncertainty in the grid (x, y) at the current moment, p xy (k) represents the probability of the target existing in the grid (x, y) at the current moment.
[0176] The attraction and repulsion pheromones are represented as follows:
[0177] S a (k+1)=(1-V a )[(1-G a )(S a (k)+d a (EV))+GP a (k)];
[0178] S r (k+1)=(1-V r )[(1-G r )(S r (k)+d r V)+GP r (k)];
[0179] Among them, V a V r G a G r is the pheromone transmission and volatility coefficient, GP a (k) and GP r (k) is the pheromone concentration transmitted from the adjacent grid, E is the unit matrix, V is the path matrix, S a (k+1) and S r (k+1) is the concentration of the attracting and repelling pheromones at the next moment of the grid. The pheromone concentration of the grid is expressed as
[0180] When initializing the rasterized map, the voronoi marker map attribute is also included;
[0181] As a preferred implementation, the region optimization module is also used for:
[0182] Calculate the total uncertainty of each UAV's corresponding mission area, expressed as follows:
[0183] The uncertainty differences between the search areas are expressed as follows:
[0184]
[0185] Where o(i) is the total uncertainty of the mission area corresponding to the i-th UAV, σ xy (k) is the partition mark;
[0186] Optimize and adjust the center point of the Voronoi diagram to obtain the area division result with the minimum uncertainty difference. When |Δ j+1 -Δ j |≤θ, where θ is any decimal greater than 0. P * is the center point of the Voronoi diagram.
[0187] As a preferred implementation, in the optimization module, the task benefits and task constraints are expressed as follows:
[0188]
[0189] Among them, ω1, ω2, ω3>0 are the corresponding weights and satisfy ω1+ω2+ω3=1, Indicates the target detection benefit brought by searching the specified area, represents the environmental exploration benefit of searching the specified area, Indicates the cumulative pheromone gain of searching the specified area. Represents a partition constraint, Represents a safety constraint.
[0190] As a preferred implementation, the optimization module is also used for:
[0191] Pheromone initialization, the pheromone in the no-fly zone obstacle area is 0, and the attraction pheromone in the initial environment is greater than the repulsion pheromone p att >p rep ;
[0192] The solution is constructed by letting each ant individual represent a feasible solution, and the length of the solution is the length of the prediction time window {0,1,...,-1} 1×T , where 0, 1, and -1 represent the unit change in the direction of travel at the next moment relative to the current direction;
[0193] The calculation of heuristic information takes the task benefit, task constraints and repeated search items as heuristic information:
[0194]
[0195]
[0196]
[0197]
[0198] Among them, ζ, χ and Indicates the importance of the corresponding heuristic information, |N xy | represents the total number of available neighbors around the grid, Indicates the location of the obstacle. represents the location information of UAV i at time k, R detect Indicates the detection range of the drone. represents the size of repulsive pheromone in the environment, N search Indicates the number of neighbors that have been searched. represents the size of the reward pheromone in the environment, Indicates the heuristic information size of the current grid;
[0199] The state transition rule and pheromone update use the following pseudo-random method to select the next state and update the pheromone of the corresponding grid after successful update:
[0200]
[0201]
[0202]
[0203]
[0204] in, Indicates the probability of selecting the next node at the next moment k+1 in the current grid (x, y). N xy Represents a set of subsequent optional nodes, and Respectively represent the pheromone concentration of the grid at the current moment and the next moment, Indicates the size of the heuristic information of the current grid, α and β represent the weights of the two respectively. q0 is the artificially set judgment threshold, and q is a random number randomly generated between [0,1]. ρ local and ρ global Represent the local and global pheromone volatility coefficients respectively, τ0 is the artificially set fixed pheromone increment, is based on the predicted cost R(S i (k))Calculate the pheromone increment within the grid (x,y).
[0205] As a preferred implementation, in the loop module, the neighbor set N with which k drones can communicate and exchange information at any time is i (k) The following conditions are met:
[0206] N i (k) = {U j |||s i (k)-s j (k)||2≤R com ,U j ∈N uav};
[0207] Among them, U j represents the neighbors within the communication range of UAV i, s i (k) represents the state of UAV i at time k, s j (k) represents the state of UAV j at time k, R com Represents the communication radius, N uav Represents the set of all drones.
[0208] Embodiment three:
[0209] In some embodiments, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program implements the above method when executed by a processor.
[0210] Embodiment 4:
[0211] In some embodiments, an electronic device is provided, including a processor and a storage device, wherein the storage device stores a plurality of instructions, and the processor is used to read the plurality of instructions in the storage device and execute the above method.
[0212] It should be understood that in the embodiments of the present application, the processor referred to may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.
[0213] The memory may include a read-only memory, a flash memory, and a random access memory, and provides instructions and data to the processor. A part or all of the memory may also include a nonvolatile random access memory.
[0214] It should be understood that the above-mentioned integrated module / unit, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such an understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The above-mentioned computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the above-mentioned computer program includes computer program code, and the above-mentioned computer program code can be in source code form, object code form, executable file or some intermediate form. The above-mentioned computer-readable medium may include: any entity or device capable of carrying the above-mentioned computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the above-mentioned computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.
[0215] The distributed collaborative target search method and device for drones in a dynamic environment provided by the above embodiment comprehensively considers the uncertainty differences between regions during the drone's search mission and designs a set of online solution method frameworks, thereby improving search efficiency by reducing cross-region searches. By iteratively dividing the mission environment, a mission area with relatively balanced uncertainty is found, and on this basis, a search benefit function is constructed and solved using an improved ant colony method. This method can improve search efficiency while ensuring search security and has high practical value.
[0216] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
[0217] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0218] In the embodiments provided in the present application, it should be understood that the disclosed apparatus / terminal equipment and method can be implemented in other ways. For example, the apparatus / equipment embodiments described above are only schematic, for example, the division of the above modules or units is only a logical function division, and in actual implementation, other division methods can be used, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.
[0219] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.
Claims
1. A distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment, characterized in that: The steps include: S1, obtain the task environment map and rasterize it; S2, initializing the rasterized map based on the overall probability density distribution, uncertainty measurement index, attraction and repulsion pheromone, and partition mark; S3, updating the probability of existence of a target within the search range based on the rasterized map; S4. Calculate the total uncertainty of the mission area corresponding to each UAV according to the uncertainty measurement index and the partition mark, and optimize the regional division; S5. Calculating task benefits and task constraints based on the attraction and repulsion pheromones; S6. Establishing an optimization model based on the target existence probability, task benefit and task constraint, and optimizing the target through an ant colony algorithm, wherein the task benefit is maximized as the optimization target; S7, updating the state information and environment information of the drone at the next moment; S8, repeating steps S3 to S7 until the search task is completed, and updating the communication topology relationship between the drones at a preset communication interval and unifying the environmental information between the drones; The uncertainty metric is expressed as follows: ; in, Indicates the current moment, Represents the current time grid The uncertainty within Represents the current time grid The probability of the target existing within When initializing the rasterized map, the voronoi marker map attribute is also included; Step S4 includes: S41. Calculate the total uncertainty of each UAV's corresponding mission area, expressed as follows: The uncertainty differences between the search areas are expressed as follows: ; in, is the total uncertainty of the mission area corresponding to the i-th UAV, It is a partition mark; S42, optimize and adjust the center point of the voronoi diagram to obtain the area division result with the minimum uncertainty difference, when satisfying Stop when is any decimal greater than 0, where , is the center point of the voronoi diagram; The task benefits and task constraints are expressed as follows: ; in, is the corresponding weight and satisfies , Indicates the target detection benefit brought by searching the specified area, Represents the environmental exploration benefit of searching the specified area, Indicates the cumulative pheromone gain of searching the specified area. Represents a partition constraint, , Represents a safety constraint.
2. The method according to claim 1, characterized in that: The attraction and repulsion pheromones are represented as follows: ; ; in, is the pheromone transmission and volatility coefficient, and is the pheromone concentration transmitted from the adjacent grid, E is the unit matrix, V is the path matrix, and is the concentration of attraction and repulsion pheromone at the next moment of the grid. The pheromone concentration of the grid is expressed as .
3. The method according to claim 1, characterized in that Objective optimization through ant colony algorithm includes: S61, pheromone initialization, the pheromone in the no-fly zone obstacle area is 0, the attraction pheromone in the initial environment is greater than the repulsion pheromone ; S62, solution construction, let each ant individual represent a feasible solution, the length of the solution is the length of the prediction time window , where 0, 1, and -1 represent the unit change in the direction of travel at the next moment relative to the current direction; S63, calculation of heuristic information, taking task benefits, task constraints and repeated search items as heuristic information: S64, state transition rules and pheromone update, the next state is selected in a pseudo-random manner and the pheromone of the corresponding grid is updated after successful update.
4. The method according to any one of claims 1 to 3, characterized in that: In step S8, at any time The set of neighbors with which the drone can communicate and exchange information The following conditions must be met: ; in, Indicates that the drone The set of neighbors within the communication range, Indicates drone exist The state of the moment, Indicates drone exist The state of the moment, represents the communication radius, Represents the set of all drones.
5. A search device using the distributed collaborative target search method for unmanned aerial vehicles in a dynamic environment according to any one of claims 1 to 3, characterized in that: include: Map acquisition module, used to obtain the task environment map and rasterize it; Initialization module, used to initialize the rasterized map based on four attributes: overall probability density distribution, uncertainty metric, attraction and repulsion pheromone, and partition mark; A first updating module, configured to update the existence probability of a target within a search range based on the rasterized map; A region optimization module, used to calculate the total uncertainty of the mission area corresponding to each UAV according to the uncertainty measurement index and the partition mark, and optimize the region division; A constraint module, for calculating task benefits and task constraints based on the attraction and repulsion pheromones; An optimization module is used to establish an optimization model based on the target existence probability, task benefit and task constraints, and perform target optimization through an ant colony algorithm, wherein the task benefit is maximized as the optimization target; The second updating module is used to update the state information and environment information of the drone at the next moment; The loop module is used to repeatedly execute the first update module to the second update module until the search task is completed, and to update the communication topology relationship between the drones at a preset communication interval and unify the environmental information between the drones.
6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.
7. An electronic device comprising a processor and a storage device, characterized in that: The storage device stores a plurality of instructions, and the processor is used to read the plurality of instructions in the storage device and execute the method according to any one of claims 1 to 4.
Citation Information
Patent Citations
Multi-UAV cooperation target searching method based on multi-group ant colony algorithm
CN108829140A
Unmanned aerial vehicle cluster collaborative dynamic target searching method based on improved pigeon inspired optimization
CN114020031A
Multi-ant colony search submarine target collaborative path optimization method based on rule heuristic
CN114840016A
Multi-AUV (Autonomous Underwater Vehicle) collaborative target searching method based on ant colony algorithm
CN115061465A
Distributed collaborative target searching method for unmanned aerial vehicle cluster
CN115185297A
