Unmanned aerial vehicle multi-task point approaching reconnaissance online route planning method

Abstract

The invention discloses an unmanned aerial vehicle multi-task point approaching reconnaissance online route planning method, which comprises the following steps: firstly, determining all to-be-reconnaissance task points in a task area, merging the dense task points according to the distribution of the task points and the detection breadth of an airborne detection device of an unmanned aerial vehicle, and determining equivalent task points to obtain an equivalent task point set; secondly, determining a reconnaissance sequence of equivalent task points by adopting a greedy algorithm according to a position distribution relationship between the unmanned aerial vehicle and the equivalent task point set; and finally, according to the reconnaissance sequence of the equivalent task points, on the basis of a heuristic A * method, adding maneuverability constraints of the turning radius and the flight speed and direction of the unmanned aerial vehicle, and considering an on-way no-fly zone to perform obstacle avoidance so as to complete approaching reconnaissance route planning of all the equivalent task points in the task area. The method has the advantages of high reliability, strong real-time performance and strong engineering practicalness, can be used for scenes such as regional situation awareness, search and rescue and the like, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to the technical field of route planning for unmanned autonomous flying platforms, and in particular to an online route planning method for close-in reconnaissance of multiple mission points of unmanned aerial vehicles (UAVs), which can be used for fixed-wing UAVs, rotary-wing UAVs, ground unmanned vehicles, and the like. Background Art

[0002] With the rapid development of autonomous unmanned systems, drones (UAVs), as long-endurance unmanned vehicles, are playing a vital role in reconnaissance and search, replacing manned platforms in reconnaissance and exploration of dangerous areas. Reconnaissance UAVs typically need to receive multiple mission points and conduct autonomous reconnaissance of each point. Path planning technology, which guides the UAV to quickly traverse multiple points and acquire regional information, is crucial for completing reconnaissance missions. Efficient online path planning for close-in reconnaissance of multiple mission points is crucial for UAVs to leverage their reconnaissance advantages. When performing online path planning for close-in reconnaissance of multiple mission points, constraints such as the distribution of mission points, no-fly zones, UAV detection performance, and UAV maneuverability must be comprehensively considered. Currently, planning is typically performed on a high-performance ground-based computing platform and then uploaded to an onboard computer via wireless communication devices for execution. This relies on a reliable communication link, but the actual mission area may be subject to various electromagnetic interference conditions. Therefore, it is necessary to design an online path planning method for close-in reconnaissance of multiple mission points that can be deployed and run on an onboard low-resource computing platform to enhance the autonomous reconnaissance capabilities of UAVs. Summary of the Invention

[0003] The purpose of the present invention is to provide an online route planning method for close-in reconnaissance of multiple mission points by unmanned aerial vehicles (UAVs), so as to realize efficient close-in reconnaissance of multiple scattered and dense points in a mission area by a single UAV. During planning, the regional distribution of mission points, the distribution of no-fly zones, the detection performance of the UAV, and the maneuverability of the UAV need to be comprehensively considered, and the real-time requirements of the online planning of the UAV need to be met.

[0004] In order to achieve the above tasks, the present invention adopts the following technical solutions:

[0005] An online route planning method for UAV multi-mission point close reconnaissance, comprising:

[0006] First, all the task points to be surveyed in the mission area are determined. Based on the distribution of the task points and the detection width of the UAV's onboard detection device, the dense task points are merged and equivalent task points are determined to obtain an equivalent task point set.

[0007] Secondly, according to the position distribution relationship between the UAV and the equivalent task point set, a greedy algorithm is used to determine the reconnaissance order of the equivalent task points;

[0008] Finally, according to the reconnaissance order of equivalent mission points, the heuristic A* method is used to increase the maneuverability constraints of the UAV's turning radius and flight speed direction, and obstacle avoidance is performed considering the no-fly zones along the way, so as to complete the close reconnaissance route planning for all equivalent mission points in the mission area.

[0009] Furthermore, based on the distribution of task points and the detection width of the UAV-mounted detection device, the dense task points are merged and equivalent task points are determined to obtain an equivalent task point set, including:

[0010] Calculate the relative distance dist_t2t(i,j) between every two task points i and j in the task area;

[0011] Based on the relative distance dist_t2t(i,j) between task points i and j, if dist_t2t(i,j)≤ωW sm , then task points i, j are dense task points, and task point j is regarded as the dense point of task point i; where ω is the preset proportional coefficient, D sm The detection width of the UAV's onboard detection device;

[0012] Calculate the number of dense points corresponding to each task point in turn and construct a dense point information matrix;

[0013] Sort the dense point information matrix in descending order to obtain the order of task point numbers from the largest to the smallest number of dense points. t ;

[0014] Sort by task point number t , by order t (1)To order t (n) The corresponding task point number is calculated in sequence to calculate the equivalent task point; where order t (1) order t (n) represents the order of the task points with the most and least dense points, respectively, and n is the total number of task points;

[0015] For task points with dense points, the mean of the north and east positions of all task points is calculated as the coordinates of the equivalent task point; for task points without dense points, the equivalent task point of the task point is itself.

[0016] Construct an equivalent task point set from all equivalent task points.

[0017] Furthermore, in the process of calculating equivalent task points, if the dense point sequence corresponding to task point A contains dense point C, and the dense point sequence corresponding to task point B also contains dense point C, and the number of dense points corresponding to A is greater than the number of dense points corresponding to B, then when calculating the equivalent task points of task point A and all its dense points, dense point C is considered to participate in the calculation, while when calculating the equivalent task points of task point B and its dense point sequence, dense point C is not considered.

[0018] Furthermore, based on the position distribution relationship between the UAV and the equivalent mission point set, a greedy algorithm is used to determine the reconnaissance order of the equivalent mission points, including:

[0019] Construct a relative distance matrix, which includes the distance between the current position of the UAV and each equivalent mission point, as well as the distance between every two equivalent mission points;

[0020] Use a greedy algorithm to find the optimal solution based on the relative distance matrix:

[0021] At the beginning, the equivalent mission point closest to the UAV is selected as the first approach point;

[0022] When the first approach point is determined, the next equivalent task point is selected as the approach point in sequence based on the principle that the current equivalent task point is closest to other equivalent task points, until all equivalent task points are traversed, and finally the equivalent task point set after order decision is obtained.

[0023] Furthermore, the heuristic A* method sets the heuristic cost function, node expansion direction, and node expansion step size, and expands the nodes from the starting point to the end point under the constraints, and finally backtracks to obtain the optimal route; the starting point is the previous equivalent mission point or the starting point of the UAV, and the end point is the current equivalent mission point;

[0024] At each step of node expansion, the node with the smallest cost function is used as the valid node, and the sum of the cumulative distance from the starting point to the current node and the Euclidean distance between the current node and the end point is used; when expanding the node at each step, the expansion direction is selected, and the expansion step size is used as the distance increment for each expansion, and then the path cost of the node after expansion in each direction is calculated, and then the optimal path is determined; when the node is expanded to the end point, according to the recorded correspondence between each node and its "parent node", the node sequence from the starting point to the end point is backtracked to form the reconnaissance waypoint between the current equivalent task point and the previous equivalent task point.

[0025] Furthermore, with a turning radius R zw As the node expansion step, it is calculated as follows:

[0026]

[0027] step=R zw

[0028] Among them, V ref represents the expected flight speed of the drone, g is the acceleration due to gravity, and φ is the expected roll angle when turning.

[0029] Furthermore, when performing node expansion, the direction strategy is as follows:

[0030]

[0031] Among them, α wpt (n) represents the expandable direction of node n; [1,0]; [0,1]; [-1,0]; [0,-1]; [1,1]; [-1,-1]; [1,-1]; [-1,1] represent the north, east, south, west, northeast, southwest, northwest, and southeast directions respectively in the North-Sky-East coordinate system.

[0032] Furthermore, a safe distance is reserved at the edge of the no-fly zone. Each time a node is expanded, the distance between the node and the center of the no-fly zone is determined. The node's safety is determined based on the distance. Only when the node is safe is it considered a valid node:

[0033]

[0034] Where d(n,k) represents the distance between node n and the kth no-fly zone, obs(k,1), obs(k,2), and obs(k,3) represent the north position, east position, and radius of the kth no-fly zone, respectively. safe Indicates that a safe distance is reserved, flag safe (n)=1 indicates that node n is safe.

[0035] Furthermore, after completing the close reconnaissance route planning for the last equivalent mission point, considering the terminal flight speed direction θ v , with 2R zw Plan a rectangular circling waiting route for the side length, θ v The calculation is as follows:

[0036]

[0037] Where wpt(t,1) and wpt(t,2) represent the northing and easting positions of the last reconnaissance waypoint to the last equivalent mission point, respectively; wpt(t-1,1) and wpt(t-1,2) represent the northing and easting positions of the second-to-last reconnaissance waypoint to the last equivalent mission point, respectively;

[0038] The circling route is calculated as follows:

[0039] wpt(t+1,:)=wpt(t,:)+2R zw [cos(θv ),sin(θ v )]

[0040] wpt(t+2,:)=wpt(t+1,:)+2R zw [cos(θ v +90),sin(θ v +90)]

[0041] wpt(t+3,:)=wpt(t+2,:)+2R zw [cos(θ v +180),sin(θ v +180)]

[0042] wpt(t+4,:)=wpt(t+3,:)+2R zw [cos(θ v +270),sin(θ v +270)]

[0043] Where wpt(t,:) represents the North Celestial East coordinate of the t-th reconnaissance waypoint.

[0044] A terminal device includes a processor, a memory, and a computer program stored in the memory; when the processor executes the computer program, the online route planning method for close-in reconnaissance of multiple mission points of an unmanned aerial vehicle is implemented.

[0045] A computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the online route planning method for close-in reconnaissance of multiple mission points of an unmanned aerial vehicle is implemented.

[0046] Compared with the prior art, the present invention has the following technical features:

[0047] The method of the present invention reduces the planned flight distance and improves the reconnaissance efficiency by preprocessing dense points and deciding the approach order of mission points. According to the performance of the aircraft, the node step size and expansion direction of the A* algorithm are improved in a targeted manner to improve the real-time performance of the algorithm. Finally, the UAV's approach reconnaissance route planning function for multiple mission points is realized, and the real-time requirements of the deployment of airborne weak resource platforms are met, which has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Flow chart of online route planning for UAV multi-mission point close reconnaissance;

[0049] Figure 2 Route planning result 1 (dense point preprocessing + approach order decision + improved A* + circling circle);

[0050] Figure 3Route planning result 2 (dense point preprocessing + approach order decision + A*);

[0051] Figure 4 Route planning result 3 (dense point preprocessing + A*);

[0052] Figure 5 Route planning result 4 (A*). DETAILED DESCRIPTION

[0053] See attached Figure 1 The present invention provides an online route planning method for UAV multi-mission point close reconnaissance, comprising the following steps:

[0054] First, all the task points to be surveyed in the mission area are determined. Based on the distribution of the task points and the detection width of the UAV's onboard detection device, the dense task points are merged and equivalent task points are determined to obtain an equivalent task point set, thus avoiding repeated surveys of dense task points.

[0055] Secondly, according to the position distribution relationship between the UAV and the equivalent task point set, a greedy algorithm is used to determine the reconnaissance order of the equivalent task points, thereby reducing the invalid range and improving the traversal efficiency of the equivalent task points;

[0056] Finally, according to the reconnaissance order of equivalent mission points, the heuristic A* method is used to increase the maneuverability constraints of the UAV's turning radius and flight speed direction, and obstacle avoidance is performed considering the no-fly zones along the way, so as to complete the close reconnaissance route planning for all equivalent mission points in the mission area.

[0057] The UAV multi-mission point close-in reconnaissance route planning method provided by the present invention has high reliability and strong real-time performance, strong engineering practicality, and a wide range of application scenarios. It can be expanded to various military and civilian unmanned autonomous flying platforms for use in scenarios such as situation construction, search and rescue, and has broad application prospects.

[0058] The specific design process of the present invention is described in detail below with reference to the accompanying drawings.

[0059] 1. Determine dense task points based on the distribution of task points and construct equivalent task points.

[0060] For the situation where there are multiple task points to be reconnaissanced in the task area, the locations are relatively concentrated, that is, multiple task points are concentrated around ωD sm If the task points are within the circle with a radius of , then these task points are considered to be dense task points; where ω is the preset proportional coefficient, D sm It is the detection width of the UAV’s onboard detection device.

[0061] To improve reconnaissance efficiency, the geometric center position of the dense task points, task_point, is used as the equivalent task point for subsequent close-in reconnaissance route planning, as follows:

[0062] (1.1) Calculate the relative distance dist_t2t(i,j) between each two task points i and j in the task area as follows:

[0063]

[0064] Among them, Taskpt(i,1) and Taskpt(i,2) represent the north position and east position of the i-th task point in the North Sky East coordinate system.

[0065] (1.2) Based on the relative distance dist_t2t(i,j) between task points i and j, if dist_t2t(i,j)≤ωW sm , then task points i and j are dense task points, and task point j is regarded as the dense point of task point i.

[0066] Calculate the number of dense points corresponding to each task point in turn, construct the dense point information matrix Task_mj_num, Task_mj_num(i) represents the number of dense points of the i-th task point; and obtain the dense point sequence matrix Task_mj_index corresponding to each task point, where Task_mj_index(i) contains the number information of the dense points of the i-th task point.

[0067] (1.3) Sort the dense point information matrix Task_mj_num in descending order to obtain the order of the task point numbers from the largest to the smallest number of dense points. t ; among them order t (1) order t (n) represents the serial numbers of the task points with the most and least dense points respectively, and n is the total number of task points.

[0068] (1.4) Sort by task point number t , by order t (1)To order t (n) The corresponding task point number is calculated in sequence:

[0069] For task points with dense points (Task_mj_num(i)>0), the mean of the north and east positions of all task points is calculated as the coordinates of the equivalent task point; for task points without dense points (Task_mj_num(i)=0), that is, scattered task points, the equivalent task point of the task point is itself.

[0070] Through the above steps, all equivalent task points are used to construct an equivalent task point set task_point, where task_point(i,1) and task_point(i,2) represent the north position and east position of the i-th equivalent task point in the North Sky East coordinate system, respectively.

[0071] In the process of calculating equivalent task points, the calculated dense point serial numbers are marked, and the calculation is not repeated when calculating subsequent equivalent task points; for example, the dense point sequence corresponding to task point A contains dense point C, and the dense point sequence corresponding to task point B also contains dense point C, and the number of dense points corresponding to A is greater than the number of dense points corresponding to B, then when calculating the equivalent task points of task point A and all its dense points, dense point C is considered to participate in the calculation, while when calculating the equivalent task points of task point B and its dense point sequence, dense point C is not considered.

[0072] 2. Decision on the order of approaching reconnaissance at equivalent mission points.

[0073] After determining the equivalent task point set task_point, in order to reduce the planned route distance and improve reconnaissance efficiency, a greedy algorithm is used to make decisions on the order of approaching the equivalent task points to determine the optimal approach sequence. Based on this idea, the distance between the UAV and the equivalent task point and the distance between each equivalent task point are used as input. The specific steps are as follows:

[0074] Construct the relative distance matrix dist; dist(1,j) in the matrix dist represents the distance between the current position of the drone and the j-th equivalent task point, and dist(i,j) represents the distance between the i-th equivalent task point and the j-th equivalent task point.

[0075] Considering the limited flight time of the UAV and the limited number of equivalent mission points, a greedy algorithm is used to find the optimal solution based on the relative distance matrix dist:

[0076] At the beginning, the equivalent mission point closest to the drone is selected as the first approach point:

[0077] min_dist_index(i)=min(dist(i,:))

[0078] task_new(i,1)=task_point(min_dist_index(i),1)

[0079] task_new(i,2)=task_point(min_dist_index(i),2)

[0080] Among them, min_dist_index(i) is the serial number of the i-th equivalent task point to be scouted, min(dist(i,:)) means taking the serial number corresponding to the minimum value of the i-th row in dist, task_new(i,1) and task_new(i,2) respectively represent the north position and east position of the equivalent task point selected after the order decision.

[0081] When the first approach point is determined, the next equivalent task point is selected as the approach point according to the principle of the closest distance between the current equivalent task point and other equivalent task points, until all equivalent task points are traversed:

[0082] min_dist_index(i)=min(dist(min_dist_index(i-1),:))

[0083] task_new(i,1)=task_point(min_dist_index(i),1)

[0084] task_new(i,2)=task_point(min_dist_index(i),2)

[0085] Finally, the equivalent task point set task_new after the order decision is obtained.

[0086] 3. Planning of close reconnaissance routes for equivalent mission points.

[0087] The approach reconnaissance route planning for equivalent task points is carried out according to the approach reconnaissance order of equivalent task points. In order to ensure the flyability and real-time planning of the planned route, the heuristic planning method is used to plan the route between adjacent equivalent task points.

[0088] Since the traditional heuristic A* method does not consider the maneuverability of the UAV when planning the route, the turning radius R of the UAV is used. zw Improve the node expansion step size, constrain the node expansion direction with the UAV flight direction, avoid sharp angles between the two routes, and meet the UAV flight requirements; in addition, plan a circling standby route after completing the reconnaissance of the last equivalent mission point.

[0089] The A* method sets the heuristic cost function, node expansion direction, and node expansion step size. Under the constraints, the nodes are expanded from the starting point to the end point, and the optimal route is finally obtained by backtracking. The specific process is as follows:

[0090] The cost function f(n) is used as the basis for node expansion. At each step of node expansion, the node with the minimum cost function is used as the valid node. The setting of the cost function takes into account the actual path cost and the expected cost. The cost function is the sum of the cumulative distance g(n) from the starting point (previous equivalent task point) to the current node and the Euclidean distance h(n) between the current node and the end point (current equivalent task point), as follows:

[0091] f(n)=g(n)+h(n)

[0092]

[0093] Among them, x(n) and y(n) represent the north position and east position of the nth node, task_new(i-1,1) and task_new(i-1,2) represent the north position and east position of the i-1th equivalent task point to be reconnaissanced.

[0094] The traditional A* algorithm selects eight directions as expansion directions at each step when expanding nodes. The expansion step size step is used as the distance increment for each expansion. The path cost f(n) of the node after expansion in each direction is then calculated to determine the optimal path. The expansion directions are as follows:

[0095] α all ={[1,0];[0,1];[-1,0];[0,-1];[1,1];[-1,-1];[1,-1];[-1,1]}

[0096] Among them, α all Represents all extension directions. {[1,0]; [0,1]; [-1,0]; [0,-1]; [1,1]; [-1,-1]; [1,-1]; [-1,1]} represent the north, east, south, west, northeast, southwest, northwest, and southeast directions respectively in the North Celestial East coordinate system.

[0097] When the node is expanded to the end point, the node sequence from the starting point to the end point is obtained by backtracking according to the correspondence between each node and its "parent node", which constitutes the reconnaissance waypoint between the current equivalent task point and the previous equivalent task point.

[0098] (3.1) Expand node step size.

[0099] In order to ensure that the UAV has sufficient maneuvering distance on a single route, the turning radius R zw As the node expansion step, it is calculated as follows:

[0100]

[0101] step=R zw

[0102] Among them, V ref represents the expected flight speed of the drone, g is the acceleration due to gravity, and φ is the expected roll angle when turning.

[0103] (3.2) Expand node direction.

[0104] Considering that it is difficult for drones to make sharp turns during actual flight, the waypoint design should consider that the continuous route must meet the constraint of an angle of no less than 90°. That is, when the heuristic algorithm performs node expansion, the direction strategy is as follows:

[0105]

[0106] Among them, α wpt (n) represents the expansion direction of node n.

[0107] (3.3) Obstacle avoidance.

[0108] A safe distance is reserved at the edge of the no-fly zone. Each time a node is expanded, the distance between the node and the center of the no-fly zone is determined. The node's safety is determined based on the distance. Only nodes that are safe are considered valid nodes.

[0109]

[0110] Where d(n,k) represents the distance between node n and the kth no-fly zone, obs(k,1), obs(k,2), and obs(k,3) represent the north position, east position, and radius of the kth no-fly zone, respectively. safe Indicates that a safe distance is reserved, flag safe (n)=1 indicates that node n is safe.

[0111] (3.4) Terminal circling standby route planning.

[0112] UAVs have the characteristics of long flight time. In order to enable the UAV to receive new mission instructions after completing the current reconnaissance mission, after completing the approach reconnaissance route planning for the last equivalent mission point, the terminal flight speed direction θ is considered. v (i.e. the direction of the reconnaissance route), with 2R zw Plan a rectangular circling waiting route for the side length, θ v The calculation is as follows:

[0113]

[0114] where wpt(t,1) and wpt(t,2) represent the northing and easting positions of the last reconnaissance waypoint to the last equivalent mission point, respectively; wpt(t-1,1) and wpt(t-1,2) represent the northing and easting positions of the second-to-last reconnaissance waypoint to the last equivalent mission point, respectively.

[0115] The circling route is calculated as follows:

[0116] wpt(t+1,:)=wpt(t,:)+2R zw [cos(θ v ),sin(θ v )]

[0117] wpt(t+2,:)=wpt(t+1,:)+2R zw [cos(θ v +90),sin(θ v +90)]

[0118] wpt(t+3,:)=wpt(t+2,:)+2R zw [cos(θ v +180),sin(θ v +180)]

[0119] wpt(t+4,:)=wpt(t+3,:)+2R zw [cos(θ v +270),sin(θ v +270)]

[0120] Where wpt(t,:) represents the North Celestial East coordinate of the t-th reconnaissance waypoint.

[0121] Finally, the multi-equivalent mission point approach reconnaissance route wpt is obtained.

[0122] In summary, this method first screens densely populated mission points within a region based on the detection performance of the drone's onboard detection device to obtain an equivalent mission point set, thus avoiding repeated reconnaissance of multiple dense mission points. A greedy algorithm is then used to determine the reconnaissance order based on the location distribution relationship between the drone and the mission point set, minimizing inefficient flight ranges. Finally, based on the reconnaissance order, maneuverability constraints are added to the heuristic route planning method to complete close-in reconnaissance route planning for mission points within the region. This UAV multi-mission point close-in reconnaissance route planning method boasts high reliability, strong real-time performance, strong engineering practicality, and a wide range of application scenarios. It can be extended to various military and civilian unmanned autonomous flight platforms for use in scenarios such as regional situational awareness and search and rescue, and has broad application prospects.

[0123] See also Figures 2 to 5 ,in Figure 2 This is the result of route planning using the present invention in one embodiment of the present invention, that is, first pre-processing of dense task points is performed to convert them into equivalent task points; secondly, the order of approach reconnaissance of equivalent task points is decided; finally, the improved heuristic A* method is used to plan the approach reconnaissance route, and the aircraft hovers at the end and waits.

[0124] Figure 3 This is the route planning result of dense mission point preprocessing + equivalent mission point approach reconnaissance order decision + traditional heuristic A* method.

[0125] Figure 4 This is the route planning result of dense mission point preprocessing + traditional heuristic A* method.

[0126] Figure 5 This is the route planning result of the traditional heuristic A* method.

[0127] from Figures 2 to 5 It can be seen that the method of the present invention can effectively reduce the planned flight distance and improve the reconnaissance efficiency.

[0128] 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 aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned 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 various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. An online route planning method for UAV multi-mission point close reconnaissance, characterized by: include: First, all the task points to be surveyed in the mission area are determined. Based on the distribution of the task points and the detection width of the UAV's onboard detection device, the dense task points are merged and equivalent task points are determined to obtain an equivalent task point set. Secondly, according to the position distribution relationship between the UAV and the equivalent task point set, a greedy algorithm is used to determine the reconnaissance order of the equivalent task points; Finally, according to the reconnaissance order of equivalent mission points, the heuristic A* method is used to increase the maneuverability constraints of the UAV's turning radius and flight speed direction, and obstacle avoidance is performed considering the no-fly zones along the way, so as to complete the close reconnaissance route planning for all equivalent mission points in the mission area.

2. The online route planning method for UAV multi-mission point close reconnaissance according to claim 1 is characterized in that: According to the distribution of task points and the detection width of the UAV airborne detection device, the dense task points are merged and equivalent task points are determined to obtain the equivalent task point set, including: Calculate the relative distance dist_t2t(i,j) between every two task points i and j in the task area; Based on the relative distance dist_t2t(i,j) between task points i and j, if dist_t2t(i,j)≤ωW sm , then task points i, j are dense task points, and task point j is regarded as the dense point of task point i; where ω is the preset proportional coefficient, D sm The detection width of the UAV's onboard detection device; Calculate the number of dense points corresponding to each task point in turn and construct a dense point information matrix; Sort the dense point information matrix in descending order to obtain the order of task point numbers from the largest to the smallest number of dense points. t ; Sort by task point number t , by order t (1)To order t (n) The corresponding task point number is calculated in sequence to calculate the equivalent task point; where order t (1) order t (n) represents the order of the task points with the most and least dense points, respectively, and n is the total number of task points; For task points with dense points, the mean of the north and east positions of all task points is calculated as the coordinates of the equivalent task point; for task points without dense points, the equivalent task point of the task point is itself. Construct an equivalent task point set from all equivalent task points.

3. The online route planning method for UAV multi-mission point close reconnaissance according to claim 2 is characterized in that: In the process of calculating equivalent task points, if the dense point sequence corresponding to task point A contains dense point C, and the dense point sequence corresponding to task point B also contains dense point C, and the number of dense points corresponding to A is greater than the number of dense points corresponding to B, then when calculating the equivalent task points of task point A and all its dense points, dense point C is considered to participate in the calculation, while when calculating the equivalent task points of task point B and its dense point sequence, dense point C is not considered.

4. The online route planning method for UAV multi-mission point close reconnaissance according to claim 1 is characterized in that: According to the position distribution relationship between the UAV and the equivalent mission point set, a greedy algorithm is used to determine the reconnaissance order of the equivalent mission points, including: Construct a relative distance matrix, which includes the distance between the current position of the UAV and each equivalent mission point, as well as the distance between every two equivalent mission points; Use a greedy algorithm to find the optimal solution based on the relative distance matrix: At the beginning, the equivalent mission point closest to the UAV is selected as the first approach point; When the first approach point is determined, the next equivalent task point is selected as the approach point in sequence based on the principle that the current equivalent task point is closest to other equivalent task points, until all equivalent task points are traversed, and finally the equivalent task point set after order decision is obtained.

5. The online route planning method for UAV multi-mission point close reconnaissance according to claim 1 is characterized in that: The heuristic A* method sets the heuristic cost function, node expansion direction, and node expansion step size, and expands the nodes from the starting point to the end point under the constraints, and finally backtracks to obtain the optimal route; the starting point is the previous equivalent mission point or the starting point of the drone, and the end point is the current equivalent mission point; At each step of node expansion, the node with the smallest cost function is used as the valid node, and the sum of the cumulative distance from the starting point to the current node and the Euclidean distance between the current node and the end point is used; when expanding the node at each step, the expansion direction is selected, and the expansion step size is used as the distance increment for each expansion. Then, the path cost of the node after expansion in each direction is calculated, and the optimal path is determined; when the node is expanded to the end point, based on the recorded correspondence between each node and its "parent node", the node sequence from the starting point to the end point is backtracked to form the reconnaissance waypoints between the current equivalent task point and the previous equivalent task point.

6. The online route planning method for UAV multi-mission point close reconnaissance according to claim 5 is characterized in that: With turning radius R zw As the node expansion step, it is calculated as follows: step=R zw Among them, V ref represents the expected flight speed of the drone, g is the acceleration due to gravity, and φ is the expected roll angle when turning.

7. The online route planning method for UAV multi-mission point close reconnaissance according to claim 5 is characterized in that: When performing node expansion, the direction strategy is as follows: Among them, α wpt (n) represents the expandable direction of node n; [1,0]; [0,1]; [-1,0]; [0,-1]; [1,1]; [-1,-1]; [1,-1]; [-1,1] represent the north, east, south, west, northeast, southwest, northwest, and southeast directions respectively in the North-Sky-East coordinate system.

8. The online route planning method for UAV multi-mission point close reconnaissance according to claim 5 is characterized in that: A safe distance is reserved at the edge of the no-fly zone. Each time a node is expanded, the distance between the node and the center of the no-fly zone is determined. The node's safety is determined based on the distance. Only nodes that are safe are considered valid nodes. Where d(n,k) represents the distance between node n and the kth no-fly zone, obs(k,1), obs(k,2), and obs(k,3) represent the north position, east position, and radius of the kth no-fly zone, respectively. safe Indicates that a safe distance is reserved, flag safe (n)=1 indicates that node n is safe.

9. The online route planning method for UAV multi-mission point close reconnaissance according to claim 5 is characterized in that: After completing the close reconnaissance route planning for the last equivalent mission point, consider the terminal flight speed direction θ v , with 2R zw Plan a rectangular circling waiting route for the side length, θ v The calculation is as follows: Where wpt(t,1) and wpt(t,2) represent the northing and easting positions of the last reconnaissance waypoint to the last equivalent mission point, respectively; wpt(t-1,1) and wpt(t-1,2) represent the northing and easting positions of the second-to-last reconnaissance waypoint to the last equivalent mission point, respectively; The circling route is calculated as follows: wpt(t+1,:)=wpt(t,:)+2R zw [cos(θ v ),sin(θ v )] wpt(t+2,:)=wpt(t+1,:)+2R zw [cos(θ v +90),sin(θ v +90)] wpt(t+3,:)=wpt(t+2,:)+2R zw [cos(θ v +180),sin(θ v +180)] wpt(t+4,:)=wpt(t+3,:)+2R zw [cos(θ v +270),sin(θ v +270)] Where wpt(t,:) represents the North Celestial East coordinate of the t-th reconnaissance waypoint.

10. A terminal device comprising a processor, a memory, and a computer program stored in the memory; characterized in that: When the processor executes the computer program, it implements the online route planning method for multi-mission point close reconnaissance of the unmanned aerial vehicle according to any one of claims 1 to 9.

Images