Flight route planning method based on improved A* algorithm

Abstract

The invention relates to a flight route planning method based on an improved A * algorithm, and the method comprises the following steps: building an appropriate space grid model according to given map data, and determining the size of a minimum grid; assigning values to grids covered by threats according to given threat data; the grid assignment value of the terrain threat is 1, the grid assignment value of the radar or air defense weapon threat is (0, 1), and the grid assignment value of the threat-free grid is 0; the grid where the starting point is located is determined as a starting node, and the grid where the target point is located is determined as a target node; an open table and a close table are established; putting the initial node into an open table, and starting to search; the method solves the problem of limitation of two-dimensional research, cost value and node expansion in the application of a conventional A * algorithm in the flight route planning technology, gives consideration to threat factors and aircraft maneuvering characteristics in a three-dimensional space, and plans a flight three-dimensional route meeting conditions. And the balance between the shortest path and the passing of the risk area is automatically obtained.

Description

Technical Field

[0001] The invention relates to a flight-assisted route planning technology, and in particular to a route planning method based on an improved A* algorithm. Background Art

[0002] With the development of computer technology and artificial intelligence, flight route planning has gradually turned to rely on computers. At present, flight route planning algorithms based on artificial intelligence are mainly divided into two categories: static global route planning and real-time local route planning. Static global route planning is planned and designed offline based on the confirmed information of the flight environment, and then the pre-planned optimal path is loaded on the aircraft, and the aircraft automatically flies along the predetermined route. This type of planning is generally completed before takeoff, and the real-time requirements are not high. Real-time local route planning plans and designs the route within the corresponding time after the feedback of the sensor on the environmental changes is updated. This type of planning has high real-time requirements and is one of the most effective means to increase the probability of survival.

[0003] Although AI-based flight route planning technology has been widely used and received attention, there are still some challenges and problems. The main difficulties currently exist include safety, route planning optimality, and route feasibility. First of all, from the perspective of safety, compared with high-altitude flight, low-altitude flight can use terrain shielding and reflected ground clutter to reduce the probability of being detected, but some detection signals will still be reflected back, so it is very important to design a flight route with good concealment. There are usually two ways of concealment used by people: one is to keep the planned route away from the threat source; the other is to use terrain and other factors to avoid the planned path as much as possible. Planning optimality refers to achieving the global optimum, that is, the shortest distance and the least time under the condition of ensuring safety. Limited by fuel and flight mission restrictions, the flight time must also be limited to a certain time. Therefore, the planned path needs to have the shortest flight distance under the premise of meeting safety. Finally, for route feasibility, the physical limitations of the aircraft must be considered when planning the route, otherwise the aircraft will not be able to fly according to the generated trajectory. The physical limitations of the aircraft mainly restrict the route: maximum turning angle, maximum climb / dive angle, minimum route segment length, minimum flight altitude, etc.

[0004] The A* algorithm is a heuristic search algorithm that selects the best path by evaluating the cost of the current node and the estimation of the heuristic function. It has the advantages of high efficiency, space and time optimization, etc. In flight route planning, the A* algorithm can be used to solve the shortest path problem, path planning, conflict avoidance and other problems.

[0005] At present, route planning based on the A* algorithm has been widely used and studied. For example, in aircraft route planning, the A* algorithm can be used to solve the shortest path or the optimal path to reduce flight time and energy consumption. In the field of drone express delivery, the A* algorithm can be used to plan the shortest or optimal delivery path to improve delivery efficiency. In addition, in the field of intelligent transportation, the A* algorithm can also be used for vehicle path planning, traffic signal control and other issues. However, route planning based on the A* algorithm still has some challenges and problems. For example, in a complex environment, how to fully consider various constraints and design effective heuristic functions to guide the search direction is a key issue.

[0006] Currently, flight route planning technology is mostly used in two-dimensional scenes, and there are many limitations in practical applications. Since the distance factor is often only one of the factors in path planning, there are other factors that need to be considered, such as obstacles such as threats, so different types of cost values ​​must be used in the algorithm. In addition, when the traditional A* algorithm expands nodes, it expands outward one layer with the current node as the center, that is, the eight nodes adjacent to the current node are the points to be expanded. However, during the actual navigation of the aircraft, the aircraft's own motion constraints need to be considered, and node expansion cannot be performed blindly. Summary of the invention

[0007] Aiming at the limitations of two-dimensional research, cost value and node expansion in the application of traditional A* algorithm in flight route planning technology, a flight route planning method based on improved A* algorithm is proposed. Considering the threat factors in three-dimensional space and the maneuverability of aircraft, a three-dimensional flight route that meets the conditions is planned. The revised A* algorithm will automatically strike a balance between the shortest path and passing through the risk area.

[0008] The technical solution of the present invention is:

[0009] A flight route planning method based on an improved A* algorithm comprises the following steps:

[0010] Step 1: Establish a suitable spatial grid model based on the given map data and determine the size of the minimum grid;

[0011] Step 2: According to the given threat data, assign values ​​to the grids covered by the threat; the grids of terrain threat are assigned a value of 1, the grids of radar or air defense weapon threat are assigned a value of (0,1), and the grids of no threat are assigned a value of 0;

[0012] Step 3: Determine the grid where the starting point is located as the starting node, and the grid where the target point is located as the target node; create an open table and a close table;

[0013] Step 4: Put the starting node into the open table and start searching;

[0014] Step 5: Loop through the following steps to expand nodes and find nodes:

[0015] S5.1: Pop up the first node in the open list, which is also the node with the smallest cost function value. Record this node as node A;

[0016] S5.2: Put node A into the close table, indicating that node A is the node currently being searched;

[0017] S5.3: Expand the three adjacent nodes around node A. For each adjacent node, first determine whether the grid where the node is located is 1. If it is 1, it means that the node is impassable and the node is no longer expanded. If the grid where the node is located is (0,1), it means that the node is passable, that is, the node can be expanded. The expanded node is recorded as B, and its cost function is f(B). The operation on node B is divided into the following three cases:

[0018] Case 1: If node B already appears in the open table, it means that node B has been expanded to through other nodes. At this time, it is necessary to determine which node is better for node B to expand through, and then save the best expansion in the open table; the judgment is as follows: f_new(B) and f_lod(B) are used to represent the cost function value of the newly expanded node B and the cost function value of the expanded node B respectively; if f_new(B)<f_old(B), it means that the current expansion of node B is better than the previous expansion of node B, and the old cost function value of node B in the open table is updated with the new cost function value of node B, and the parent node of node B in the open table is changed to node A, and the open table is reordered; if f_new(B)>f_old(B), no changes are made to node B in the open table;

[0019] Case 2: If node B already appears in the close table, it means that the node has not only been expanded to it via other nodes, but has also been found in the iteration process. At this time, it is necessary to determine whether the current way to reach node B is less expensive or the previous way to search for node B is less expensive. The judgment method is the same as case 1, except that if f_new(B)<f_old(B), not only the cost function value of node B and the corresponding parent node need to be updated, but node B also needs to be removed from the close table, re-added to the open table, and the open table needs to be sorted.

[0020] Case 3: If node B does not appear in either the open list or the close list, it means that node B is a newly expanded node. Add the newly expanded node B to the open list and reorder the open list.

[0021] S5.4: If the node popped out from the open list is the target node, it means that a safe path from the starting point to the target point has been found, and the loop of step 5 is terminated;

[0022] S5.5: Starting from the target node, trace back upward through the parent node of each node, and directly trace back to the starting node, so as to obtain all the nodes on the path from the starting node to the target node; during the backtracking process, if multiple nodes are on the same straight line, only the nodes at both ends of the straight line are saved, and finally a route including the starting point, the target point and the flight turning point is generated; the open table and the close table are deleted, and the algorithm ends;

[0023] S5.6: If the target node is not popped up until the open table is empty, it means that there is no path from the start node to the target node. The open table and the close table are deleted and the algorithm ends.

[0024] Furthermore, the improved A* algorithm needs to gradually diffuse nodes from the starting node, and each time a node with the lowest current cost is selected for expansion; the diffusion of the improved A* algorithm needs to consider the motion constraints of the aircraft, and the aircraft is regarded as a particle model with velocity, P represents the current position of the aircraft, the short horizontal line with an arrow represents the current direction of the aircraft, θ represents the maximum turning angle of the aircraft, and v represents the current speed of the aircraft; on this basis, the aircraft can only appear on the arc segment with P as the center, v*dt as the radius, and an angle of 2θ at the next unit time dt;

[0025] According to the aircraft motion model, the improved A* algorithm adopts three-neighborhood node expansion. Each time a node is expanded, only the nodes in the front-front direction, front-up direction, and front-down direction of the aircraft's current speed direction are expanded, while the nodes in the up-up direction, down-down direction, back-back direction, back-up direction, and back-down direction will no longer be expanded. Each time a node is expanded, the improved algorithm needs to save the expanded direction of the node as the basis for the next node expansion.

[0026] In the process of expanding nodes, the improved A* algorithm introduces two lists, called the open list and the close list. The open list is used to store nodes that have not yet been expanded and are waiting to be expanded, and the close list is used to store nodes that have been searched. The process of node expansion is to take out the best node from the open list each time and put it into the close list, and then use this node as the current node to expand the new node.

[0027] Furthermore, to improve the A* algorithm, an evaluation function needs to be established:

[0028] f(n)=g(n)+h(n)

[0029] g(n) is the cost function, which represents the cost from the starting point to the current node n. The actual cost at the starting point is 0. The cost function of the improved A* algorithm is shown as follows:

[0030]

[0031] Where g(0) = 0;

[0032] Taking the three-neighborhood node diffusion graph as an example, node B is set as the starting point; when the diffusion nodes are D and E, they are adjacent to B on the same side; when the diffusion node is C, it is adjacent to B on the diagonal side; the constants 10 and 14 are used because computers process integer data faster;

[0033] h(n) is the heuristic function, which indicates the cost that may be spent from the current node to the target node, and reflects the heuristic information of the cost function. The heuristic function of the improved A* algorithm is shown as follows:

[0034]

[0035] Where (x n ,y n ) represents the position of the current node n, (x end ,y end ) represents the position of the target node, which is multiplied by 10 to offset the effect of increasing the cost function g(n) by 10 times;

[0036] Th(n) represents the threat value of the current node n, k represents the weight of the distance cost, and (1-k) represents the weight of the threat cost. The importance of safety and distance in the pathfinding process is determined by adjusting the size of k.

[0037] f(n) is an evaluation function, which indicates the cost required to reach the target node from the starting node through node n. The sizes of f(n) are compared and the node n with the smallest f(n) is selected as the valid node.

[0038] Further, the basic principle of the A* algorithm is improved:

[0039] The basic principle of the improved A* algorithm is consistent with that of the A* algorithm, which is a heuristic graph search algorithm. The A* algorithm establishes two tables, open and close, and sets node attributes for nodes in the discretized space area. Nodes waiting to be expanded are placed in the open table, and nodes that have been expanded are placed in the close table. During the automatic pathfinding process, the open table and the close table are in dynamic change, and the same node may circulate in the open table and the close table multiple times. During the node diffusion process, the A* algorithm sets a heuristic cost function and uses the cost function to find the target node. After finding the target node, the final path is obtained by path backtracking.

[0040] Furthermore, adjusting the size of k determines the importance of safety and distance in the path-finding process. When the k value is relatively large, it means that the distance weight between the current node and the target node is greater, that is, the current plan pays more attention to the flight time required for planning the flight route, which is generally suitable for tasks that require aircraft to reach the battlefield quickly; when the k value is relatively small, it means that the current node threat weight is greater, which means that the current plan pays more attention to the safety of the planned flight route, which is generally suitable for reconnaissance and patrol tasks.

[0041] The beneficial effects of the present invention are:

[0042] The present invention takes full account of the special constraints of flight route planning in three-dimensional space and proposes an improved A* algorithm. The improved A* algorithm combines the threat factors in three-dimensional space and the aircraft's maneuverability, considers the actual flight conditions of the aircraft when expanding nodes, and plans a flight route trajectory that meets the conditions, which will automatically strike a balance between the shortest path and passing through the risk area. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a particle model diagram of the aircraft of the present invention;

[0044] Figure 2 This is the three-neighborhood node expansion graph of the present invention. DETAILED DESCRIPTION

[0045] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0046] (1) Improved basic principles of A* algorithm

[0047] The basic principle of the improved A* algorithm is consistent with the A* algorithm, which is a heuristic graph search algorithm. The A* algorithm establishes two tables, open and close. For the nodes in the discretized space area, the node attributes are set, and the nodes waiting to be expanded are placed in the open table, and the nodes that have been expanded are placed in the close table. During the automatic path finding process, the open table and the close table are in dynamic change, and the same node may be circulated in the open table and the close table multiple times. During the node diffusion process, the A* algorithm sets a heuristic cost function and uses the cost function to find the target node. After finding the target node, the final path is obtained by path backtracking.

[0048] (2) Node expansion

[0049] The improved A* algorithm needs to gradually diffuse nodes from the starting node, and each time a node with the lowest current cost is selected for expansion. The diffusion of the improved A* algorithm needs to consider the motion constraints of the aircraft, and regard the aircraft as a particle model with velocity, such as Figure 1 shown.

[0050] P represents the current position of the aircraft, the short horizontal line with an arrow represents the current direction of the aircraft, θ represents the maximum turning angle of the aircraft, and v represents the current speed of the aircraft. Based on this, the aircraft can only appear on the arc segment with P as the center, v*dt as the radius, and an angle of 2θ at the next unit time dt.

[0051] According to the aircraft motion model, the improved A* algorithm uses three-neighborhood node expansion. Each time a node is expanded, only the nodes in the front-front direction, front-up direction, and front-down direction of the aircraft's current speed direction are expanded, while the nodes in the up-up direction, down-down direction, back-back direction, back-up direction, and back-down direction will no longer be expanded. Each time a node is expanded, the improved algorithm needs to save the expanded direction of the node as the basis for the next node expansion.

[0052] Specific as Figure 2 As shown. Assume that the aircraft starts to fly from node A to node B. When the node is expanded at the current node B, the forward direction along the current speed of the aircraft will be expanded to node C, the forward upward direction along the current speed of the aircraft will be expanded to node D, and the forward downward direction along the current speed of the aircraft will be expanded to node E. When nodes D, C, and E are expanded in the same way, nodes L, K, J, F, G, H, and I are obtained respectively.

[0053] In the process of expanding nodes, the improved A* algorithm introduces two lists, called the open list and the close list. The open list is used to store nodes that have not been expanded and are waiting to be expanded, and the close list is used to store nodes that have been searched. The process of node expansion is to take out the best node from the open list each time and put it into the close list, and then use this node as the current node to expand the new node.

[0054] (3) Cost Function

[0055] To improve the A* algorithm, an evaluation function needs to be established:

[0056] f(n)=g(n)+h(n)

[0057] g(n) is the cost function, which represents the cost from the starting point to the current node n. The actual cost at the starting point is 0. The cost function of the improved A* algorithm is shown as follows:

[0058]

[0059] Where g(0)=0.

[0060] Taking the three-neighborhood node diffusion graph as an example, node B is set as the starting point. When the diffusion nodes are D and E, they are adjacent to B on the same side; when the diffusion node is C, it is adjacent to B on the diagonal side. Constants 10 and 14 are used because computers process integer data faster.

[0061] h(n) is the heuristic function, which indicates the cost that may be spent from the current node to the target node, reflecting the heuristic information of the cost function. The heuristic function of the improved A* algorithm is shown as follows:

[0062]

[0063] Where (x n ,y n ) represents the position of the current node n, (x end ,y end ) represents the position of the target node, which is multiplied by 10 to offset the effect of increasing the cost function g(n) by 10 times.

[0064] Th(n) represents the threat value of the current node n, k represents the weight of the distance cost, and (1-k) represents the weight of the threat cost. The importance of safety and distance in the pathfinding process is determined by adjusting the size of k. When the k value is relatively large, it means that the distance weight between the current node and the target node is greater, that is, the current plan pays more attention to the flight time required for the planned flight route, which is generally suitable for tasks that require the aircraft to reach the battlefield quickly; when the k value is relatively small, it means that the threat weight of the current node is greater, which means that the current plan pays more attention to the safety of the planned flight route, which is generally suitable for reconnaissance, patrol and other tasks.

[0065] f(n) is an evaluation function, which indicates the cost required to reach the target node from the starting node through node n. The sizes of f(n) are compared and the node n with the smallest f(n) is selected as the valid node.

[0066] (4) Algorithm steps

[0067] Step 1: Establish a suitable spatial grid model based on the given map data and determine the size of the minimum grid.

[0068] Step 2: According to the given threat data, assign values ​​to the grids covered by the threat. The grids with terrain threat are assigned a value of 1, the grids with radar or air defense weapon threat are assigned a value of (0,1), and the grids with no threat are assigned a value of 0.

[0069] Step 3: Determine the grid where the starting point is located as the starting node, and the grid where the target point is located as the target node. Create an open table and a close table.

[0070] Step 4: Put the starting node into the open table and start searching.

[0071] Step 5: Loop through the following steps to expand nodes and find nodes:

[0072] S5.1: Pop out the first node in the open table. This node is also the node with the smallest cost function value. This node is recorded as node A.

[0073] S5.2: Put node A into the close table, indicating that node A is the node currently being searched.

[0074] S5.3: Expand the three adjacent nodes around node A. For each adjacent node, first determine whether the grid where the node is located is 1. If it is 1, it means that the node is not passable and the node is no longer expanded; if the grid where the node is located is (0,1), it means that the node is passable, that is, the node can be expanded. The node to be expanded is denoted as B, and its cost function is f(B). The operation on node B is divided into the following three cases:

[0075] ① If node B already appears in the open table, it means that node B has been expanded to through other nodes. At this time, it is necessary to determine which node is better for node B to expand, and then save the best expansion in the open table. The judgment is as follows: f_new(B) and f_old(B) represent the cost function value of the newly expanded node B and the cost function value of the expanded node B respectively. If f_new(B)<f_old(B), it means that the current expansion of node B is better than the previous expansion of node B. The old cost function value of node B in the open table is updated with the new cost function value of node B, and the parent node of node B in the open table is changed to node A, and the open table is reordered; if f_new(B)>f_old(B), no changes are made to node B in the open table.

[0076] ②If node B already appears in the close table, it means that the node has not only been expanded to it via other nodes, but has also been found in the iteration process. At this time, it is necessary to determine whether the current way to reach node B is less expensive or the way to search for node B in the past is less expensive. The judgment method is the same as ①, except that if f_new(B)<f_old(B), not only do you need to update the cost function value of node B and the corresponding parent node, but you also need to remove node B from the close table, re-add it to the open table, and sort the open table.

[0077] ③If node B appears neither in the open table nor in the close table, it means that node B is a newly expanded node. Add the newly expanded node B to the open table and reorder the open table.

[0078] S5.4: If the node popped out from the open list is the target node, it means that a safe path from the starting point to the target point has been found, and the loop of step 5 is terminated.

[0079] S5.5: Starting from the target node, trace back through the parent node of each node and directly trace back to the start node, and all nodes on the path from the start node to the target node can be obtained. During the backtracking process, if multiple nodes are on the same straight line, only the nodes at both ends of the straight line are saved, and finally a route containing the start point, target point and flight turning point is generated. Delete the open table and close table, and the algorithm ends.

[0080] S5.6: If the target node is not popped up until the open table is empty, it means that there is no path from the start node to the target node. The open table and the close table are deleted and the algorithm ends.

[0081] The above-mentioned embodiment only expresses one implementation mode of the present invention, and its description is relatively specific and detailed, but it cannot be understood as limiting the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.

Claims

1. A flight route planning method based on an improved A* algorithm, characterized in that: The following steps are involved: Step 1: Establish a suitable spatial grid model based on the given map data and determine the size of the minimum grid; Step 2: According to the given threat data, assign values ​​to the grids covered by the threat; the grids of terrain threat are assigned a value of 1, the grids of radar or air defense weapon threat are assigned a value of (0,1), and the grids of no threat are assigned a value of 0; Step 3: Determine the grid where the starting point is located as the starting node, and the grid where the target point is located as the target node; create an open table and a close table; Step 4: Put the starting node into the open table and start searching; Step 5: Loop through the following steps to expand nodes and find nodes: S5.1: Pop up the first node in the open list, which is also the node with the smallest cost function value. Record this node as node A; S5.2: Put node A into the close table, indicating that node A is the node currently being searched; S5.3: Expand the three adjacent nodes around node A. For each adjacent node, first determine whether the grid where the node is located is 1. If it is 1, it means that the node is impassable and the node is no longer expanded. If the grid where the node is located is (0,1), it means that the node is passable, that is, the node can be expanded. The expanded node is recorded as B, and its cost function is f(B). The operation on node B is divided into the following three cases: Case 1: If node B already appears in the open table, it means that node B has been expanded to through other nodes. At this time, it is necessary to determine which node is better for node B to expand through, and then save the best expansion in the open table; the judgment is as follows: f_new(B) and f_old(B) are used to represent the cost function value of the newly expanded node B and the cost function value of the expanded node B respectively; if f_new(B)<f_old(B), it means that the current expansion of node B is better than the previous expansion of node B, and the old cost function value of node B in the open table is updated with the new cost function value of node B, and the parent node of node B in the open table is changed to node A, and the open table is reordered; if f_new(B)>f_old(B), no changes are made to node B in the open table; Case 2: If node B already appears in the close table, it means that the node has not only been expanded to it via other nodes, but has also been found in the iteration process. At this time, it is necessary to determine whether the current way to reach node B is less expensive or the previous way to search for node B is less expensive. The judgment method is the same as case 1, except that if f_new(B)<f_old(B), not only the cost function value of node B and the corresponding parent node need to be updated, but node B also needs to be removed from the close table, re-added to the open table, and the open table needs to be sorted. Case 3: If node B does not appear in either the open list or the close list, it means that node B is a newly expanded node. Add the newly expanded node B to the open list and reorder the open list. S5.4: If the node popped out from the open list is the target node, it means that a safe path from the starting point to the target point has been found, and the loop of step 5 is terminated; S5.5: Starting from the target node, trace back upward through the parent node of each node, and directly trace back to the starting node, so as to obtain all the nodes on the path from the starting node to the target node; during the backtracking process, if multiple nodes are on the same straight line, only the nodes at both ends of the straight line are saved, and finally a route including the starting point, the target point and the flight turning point is generated; the open table and the close table are deleted, and the algorithm ends; S5.6: If the target node is not popped up until the open table is empty, it means that there is no path from the start node to the target node. The open table and the close table are deleted and the algorithm ends.

2. The flight route planning method based on the improved A* algorithm according to claim 1, characterized in that: The improved A* algorithm needs to gradually spread nodes from the starting node, and each time a node with the lowest current cost is selected for expansion; the diffusion of the improved A* algorithm needs to consider the motion constraints of the aircraft, and the aircraft is regarded as a particle model with speed. P represents the current position of the aircraft, the short horizontal line with an arrow represents the current direction of the aircraft, θ represents the maximum turning angle of the aircraft, and v represents the current speed of the aircraft; on this basis, the aircraft can only appear on the arc segment with P as the center, v*dt as the radius, and an angle of 2θ at the next unit time dt; According to the aircraft's motion model, the improved A* algorithm uses three-neighborhood node expansion. Each time a node is expanded, only the nodes in the front-front direction, front-up direction, and front-down direction of the aircraft's current speed direction are expanded, while the nodes in the up-up direction, down-down direction, back-back direction, back-up direction, and back-down direction will no longer be expanded. The improved algorithm needs to save the expanded direction of the node each time it expands a node, which will serve as the basis for the next node expansion. In the process of expanding nodes, the improved A* algorithm introduces two lists, called open list and close list. The open list is used to store nodes that have not been expanded and are waiting to be expanded, and the close list is used to store nodes that have been searched. The process of node expansion is to take out the best node from the open table each time and put it into the close table, and then use this node as the current node to expand the new node.

3. The flight route planning method based on the improved A* algorithm according to claim 1, characterized in that: To improve the A* algorithm, we need to establish an evaluation function: f(n)=g(n)+h(m) g(n) is the cost function, which represents the cost from the starting point to the current node n. The actual cost at the starting point is 0. The cost function of the improved A* algorithm is shown as follows: Where g(0) = 0; Taking the three-neighborhood node diffusion graph as an example, node B is set as the starting point; When the diffusion nodes are D and E, they are adjacent to B on the same side; When the diffusion node is C, it is diagonally adjacent to B. The constants 10 and 14 are used because computers process integer data faster. h(n) is the heuristic function, which indicates the cost that may be spent from the current node to the target node, and reflects the heuristic information of the cost function; the heuristic function of the improved A* algorithm is shown as follows: Where (x n ,y n ) represents the position of the current node n, (x end ,y end ) represents the position of the target node, which is multiplied by 10 to offset the effect of increasing the cost function g(n) by 10 times; Th(n) represents the threat value of the current node n, k represents the weight of the distance cost, and (1-k) represents the weight of the threat cost. The importance of safety and distance in the pathfinding process is determined by adjusting the size of k. f(n) is an evaluation function, which indicates the cost required to reach the target node from the starting node through node n. The sizes of f(n) are compared and the node n with the smallest f(n) is selected as the valid node.

4. The flight route planning method based on the improved A* algorithm according to claim 1, characterized in that: Improved A* algorithm basic principle: The basic principle of the improved A* algorithm is consistent with the A* algorithm, which is a heuristic graph search algorithm. The A* algorithm establishes two tables, open and close, and sets node attributes for nodes in the discretized space area. Nodes waiting to be expanded are placed in the open table, and nodes that have been expanded are placed in the close table. During the automatic pathfinding process, the open table and the close table are in dynamic change, and the same node may circulate in the open table and the close table multiple times. During the node diffusion process, the A* algorithm sets a heuristic cost function and uses the cost function to find the target node. After finding the target node, the final path is obtained by path backtracking.

5. The flight route planning method based on the improved A* algorithm according to claim 3, characterized in that: Adjusting the size of k determines the importance of safety and distance in the path-finding process. When the k value is relatively large, it means that the distance weight between the current node and the target node is greater, that is, the current plan pays more attention to the flight time required for the planned flight route, which is generally suitable for tasks that require aircraft to reach the battlefield quickly; when the k value is relatively small, it means that the current node threat weight is greater, which means that the current plan pays more attention to the safety of the planned flight route, which is generally suitable for reconnaissance and patrol tasks.

Images