A multi-route planning method for anti-ship missiles

By establishing the benchmark unit vector and boundary vector model, calculating and screening the possible subnodes of missiles, the efficiency and avoidance problems of A* search algorithm in three-dimensional maps and multi-way planning are solved, and efficient route planning of missiles in complex environments is realized. and avoid effects.

CN114674332BActive Publication Date: 2024-11-15NAVAL UNIV OF ENG PLA
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210264723.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-17
Publication Date
2024-11-15
Estimated Expiration
2042-03-17

AI Technical Summary

Technical Problem

The existing A* search algorithm is difficult to effectively deal with the evasion problem between three-dimensional maps, multi-angle and multi-way routes in missile route planning, resulting in excessive computing space and time occupancy, and it is difficult to adapt to the evasion of complex terrain and threat areas.

Method used

By establishing the route starting reference unit vector model, horizontal boundary vector model and vertical boundary vector model, the possible subnode vector is calculated, and the optimal subnode vector is screened using bilinear interpolation and optimization model to realize missile multi-way planning and transformation Air Defense Threat Zone Avoidance is a terrain Avoidance problem, and the detection method of points in any plane polygon is adopted to solve the problem of horizontal route avoidance.

Benefits of technology

The multi-way planning of a single platform for single targets is realized. The parent node quickly adjusts its heading and quickly turns to the child nodes, providing a practical engineering method different from the traditional expansion method, and improving the avoidance effect of missiles in complex environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114674332B_ABST
    Figure CN114674332B_ABST
Patent Text Reader

Abstract

The invention discloses a multi-route planning method for an anti-ship missile, comprising the following steps: 1) calculating a route starting reference unit vector; 2) calculating a horizontal boundary vector; 4) calculating a vertical boundary vector; 5) calculating a possible sub-node vector according to the horizontal boundary vector and the vertical boundary vector; 6) screening the possible sub-node vectors to obtain a screened possible sub-node vector table; 7) establishing an optimization model, and inputting the screened possible sub-node vector table into the optimization model to obtain an optimal sub-node vector #imgabs0#; 8) judging whether the optimal sub-node vector #imgabs1# meets an iteration end condition, if so, outputting the optimal sub-node vector #imgabs2# as a route, otherwise, returning to step 2). The invention constructs terrain avoidance, air defense threat zone avoidance and horizontal route avoidance models, which can not only realize the multi-route planning of concealed attacks of three methods, but also realize the multi-route planning of attacking island and reef targets of three methods, with high work efficiency and strong engineering operability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the application field of missile weapon systems, and in particular to a multi-route planning method for anti-ship missiles. Background Art

[0002] Aircraft path planning (PathPlanning) refers to finding the optimal flight trajectory from the initial point to the target point and satisfying certain specific indicators under specific constraints. Path planning is basically divided into four categories: artificial potential field method; mathematics-based planning method, including nonlinear programming algorithm and dynamic programming algorithm; graphics-based method, including A* algorithm, Voronoi diagram method, rapidly expanding random tree; intelligent optimization algorithm, including genetic algorithm, simulated annealing algorithm, particle swarm algorithm and ant colony algorithm.

[0003] At present, many foreign applications in route planning are based on the A* search algorithm. The A* search algorithm is a heuristic optimal search method. In each iteration, the node with the smallest heuristic function value in the search space is selected for expansion. Starting from the starting node, it continuously searches for nodes that are expected to lead to the target point with the minimum cost, and gives priority to expanding these nodes that can minimize the heuristic function value, thereby forming a node set. Then, the ordered connection of certain nodes in the set is the optimal route. The heuristic function is obtained by weighted calculation of the actual cost function value from the starting point to the current node and the estimated cost function value from the current node to the target point, and its calculation depends on the heuristic information.

[0004] The cost function of the A* search algorithm is f(x) = a·g(x) + b·h(x), where g(x) represents the actual cost from the starting position to the current position x; h(x) represents the estimated cost from the current position to the target point; a and b represent the weighted coefficients of the actual cost and the estimated cost, which are generally 1. Each expansion step of the A* search will select the node with the smallest f(x) and insert it into the possible path list. It can be proved that as long as the actual cost from x to the target point is always greater than or equal to the estimated cost, the A* search can find an optimal path.

[0005] Since each newly generated node in the A* search algorithm represents an almost infinite number of possible paths, it will also take up a lot of computing space and time. At the same time, the existing A* search algorithm route planning generally targets two-dimensional maps, single routes, and finding the shortest path, and there is little research on avoidance planning between three-dimensional maps, multiple angles, and multiple routes. Summary of the invention

[0006] The object of the present invention is to provide a method for multi-route planning of an anti-ship missile, comprising the following steps:

[0007] 1) Establish a route start reference unit vector model to calculate the route start reference unit vector;

[0008] The steps to establish the route start reference unit vector model include:

[0009] 1.1) Establish the simulation coordinate system oxyz and the waypoint coordinate system OXYZ; where the origin o is the position of the observation ship's search radar; the ox axis points to the north, the oz axis is vertically upward; the oy axis forms a right-hand system; O is the waypoint, OZ is vertically upward, OX points to the north, and OY is perpendicular to OXZ;

[0010] 1.2) Let the reference attack direction unit vector be in, The starting point of the route; is the end point of the route; q x ,q y ,q z for (q x ,q y ,q z ) is the reference attack direction, that is, the vector from the start point of the route to the end point of the route. axis, Axis and The projection coordinates of the axes;

[0011] 1.3) Establish a route start reference unit vector model to calculate the route start reference unit vector for positive and negative attack angles;

[0012] The route start reference unit vector model is as follows:

[0013]

[0014] In the formula, θ 攻击 is the angle of attack; is the route starting reference unit vector; (x, y, z) is the vector exist axis, Axis and The projected coordinates of the axes.

[0015] 2) Establishing a horizontal boundary vector model to calculate the horizontal boundary vector;

[0016] The horizontal boundary vector model is as follows:

[0017]

[0018] In the formula, θ 水 is the horizontal turning limit angle; is the horizontal boundary vector; is the base unit vector.

[0019] 4) Establishing a vertical boundary vector model to calculate the vertical boundary vector;

[0020] Preferably, the vertical boundary vector model is as follows:

[0021]

[0022] In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

[0023] Preferably, the vertical boundary vector model is as follows:

[0024]

[0025]

[0026] In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

[0027] Preferably, the vertical boundary vector model is as follows:

[0028]

[0029]

[0030] In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

[0031] 5) Calculate possible child node vectors according to the horizontal boundary vector and the vertical boundary vector; the possible child nodes include the next possible waypoint after the current waypoint;

[0032] Calculate possible child node vectors, the steps include:

[0033] 5.1) Calculate the initial child node vector Right now:

[0034]

[0035] In the formula, are the upper and lower bounds of the horizontal vector; are the upper and lower bounds of the vertical vector;

[0036] 5.2) Calculate the child node vector Right now:

[0037]

[0038] 5.3) Normalize child node vectors get:

[0039]

[0040] In the formula, is the normalized child node vector;

[0041] 5.4) Calculate the (i-1)*5+jth possible child node vector Right now:

[0042]

[0043] Where h is the step length; l+1=(i-1)*5+j; Possible child nodes.

[0044] 6) Screening possible child node vectors to obtain a screened possible child node vector table;

[0045] The steps of screening possible child node vectors include:

[0046] 6.1) The possible child node vector to be screened is Calculate the linear interpolation temp1 and linear interpolation temp2 of bilinear interpolation, that is:

[0047]

[0048]

[0049] In the formula, (x i-1 ,y j-1 ,z(x i-1 ,y j-1 ))、(x i ,y j-1 ,z(x i ,y j-1 ))、(x i-1 ,y j ,z(x i-1 ,y j )) and (x i ,y j ,z(x i ,y j )) is the adjacent grid point data of the possible child node of the digital map; the adjacent grid points of the possible child node are grid points used to represent the surface height of the possible child node;

[0050]

[0051] 6.2) Calculation Corresponding ground height in digital map Right now:

[0052]

[0053] 6.3) Add n points between two nodes and get the vector corresponding to the mth node among the n points Right now:

[0054]

[0055] Where, 0<m≤n;

[0056] 6.4) will satisfy and The possible child node vector of is set to zero vector, and the possible child node vector table is obtained

[0057] 7) Establish an optimization model and input the screened possible child node vector table into the optimization model to obtain the optimal child node vector

[0058] The optimization model is as follows:

[0059]

[0060]

[0061]

[0062]

[0063] In the formula, is the optimal child node vector. is the objective function of the optimization model; is the cost function; From the starting position to the position The true cost of From the starting position to the position estimated cost.

[0064] 8) Determine the optimal child node vector Whether the iteration end condition is met, if so, the optimal child node vector Output as route, otherwise, return to step 2).

[0065] The iteration end conditions include condition 1 and condition 2; if the optimal child node vector satisfies condition 1 or condition 2, the iteration ends;

[0066] Condition one:

[0067]

[0068] Condition two:

[0069]

[0070] The technical effect of the present invention is unquestionable. Compared with the prior art, the advantages of the present invention are as follows: 1) the route starting reference unit vectors of positive and negative attack angles are obtained by referencing the attack direction unit vector, so as to realize multi-route planning of a single platform for a single target; 2) three methods of rotating the reference unit vector to obtain the horizontal boundary vector and the vertical boundary vector, and a method of expanding the boundary vector to obtain the possible child node vector, are provided, and a new engineering practical method of extending the parent node to the child node is provided, which is characterized in that the parent node quickly adjusts the heading to quickly turn to the child node, which is different from the expansion method of gradually turning to the child node around the center of curvature at the parent node; 3) the problem of avoiding the air defense threat zone is converted into a problem of terrain avoidance, and a new method of avoiding the terrain and threat zone is provided by screening the child nodes with a height lower than the terrain surface, which is different from the method of detecting the intersection of the connecting line of adjacent waypoints with any edge of the terrain or threat area polygon; 4) a detection method based on the point within the arbitrary polygon of the plane is adopted to solve the problem of horizontal route avoidance, and the avoidance effect is better. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 is the horizontal boundary vector model diagram; Figure 2 is the vertical boundary vector model method 1; Figure 3 is the vertical boundary vector model method 2; Figure 4 is the vertical boundary vector model method 3; Figure 5 is a possible child node vector model graph; Figure 6 Flowchart of the anti-ship missile route planning method for improving the A* search algorithm; Figure 7 This is the route planning result using method 1 when the step length is 6; Figure 8 This is a bird's-eye view of the route planning result using method 1 when the step size is 6; Fig. 9 This is the route planning result using method 1 when the step length is 8; Fig.10 This is a bird's-eye view of the route planning result using method 1 when the step length is 8;

[0072] Fig.11 This is the route planning result using method 1 when the step size is 10; Fig.12 This is a bird's-eye view of the route planning results using method 1 when the step size is 10; Fig.13 This is the route planning result using method 2 when the step size is 11; Fig.14 This is a bird's-eye view of the route planning result using method 2 when the step size is 11; Fig.15 This is the route planning result using method 2 when the step length is 12; Fig.16This is a bird's-eye view of the route planning result using method 2 when the step length is 12; Fig.17 This is the route planning result using method 2 when the step length is 13; Fig.18 The top view of the route planning result using method 2 when the step length is 13; Fig.19 The route planning result using method 3 when the step length is 6; Fig. 20 The top view of the route planning result using method 3 when the step length is 6;

[0073] Fig.21 The route planning result using method 3 when the step length is 8; Fig. 22 The top view of the route planning result using method 3 when the step length is 8; Fig.23 The route planning result using method 3 when the step size is 10; Fig.24 The top view of the route planning result using method 3 when the step size is 10; Fig.25 This is the route planning result of method 1. Fig.26 This is a bird's-eye view of the route planning results of method 1. Fig. 27 This is the 3-route planning result of method 1; Fig.28 This is a bird's-eye view of the 3-route planning result of method 1; Fig.29 This is the 4-route planning result of method 1; Fig.30 This is a bird's-eye view of the 4-route planning result of method 1;

[0074] Fig.31 This is the 5-route planning result of method 1; Fig.32 This is a bird's-eye view of the 5-route planning result of method 1; Fig.33 This is the route planning result of method 2; Fig.34 This is a bird's-eye view of the route planning results of Method 2; Fig.35 This is the route planning result of method 2. Fig.36 This is a bird's-eye view of the 3-route planning result of method 2; Fig.37 This is the 4-route planning result of method 2; Fig.38 This is a bird's-eye view of the 4-route planning result of method 2; Fig.39 This is the 5-route planning result of method 2; Fig.40 This is a bird's-eye view of the 5-route planning result of method 2;

[0075] Fig.41 This is the route planning result of method 3. Fig.42 This is a bird's-eye view of the route planning results of method 3. Fig.43 This is the route planning result of method 3; Fig.44 This is a bird's-eye view of the route planning results of method 3; Fig.45 This is the route planning result of method 3. Fig.46 This is a bird's-eye view of the 4-route planning result of method 3; Fig.47 This is the route planning result of method 3. Fig.48 This is a bird's-eye view of the 5-route planning result of method 3;

[0076] Figure 49(a) is multi-route planning I of method 1; Figure 49(b) is multi-route planning II of method 1; Figure 49(c) is multi-route planning III of method 1; Figure 49(d) is multi-route planning IV of method 1;

[0077] FIG. 50( a ) is a top view I of multi-route planning of method 1; FIG. 50( b ) is a top view II of multi-route planning of method 1; FIG. 50( c ) is a top view III of multi-route planning of method 1; FIG. 50( d ) is a top view IV of multi-route planning of method 1;

[0078] Figure 51(a) is multi-route planning I of method 2; Figure 51(b) is multi-route planning II of method 2; Figure 51(c) is multi-route planning III of method 2; Figure 51(d) is multi-route planning IV of method 2;

[0079] FIG. 52( a ) is a top view I of multi-route planning of method 2; FIG. 52( b ) is a top view II of multi-route planning of method 2; FIG. 52( c ) is a top view III of multi-route planning of method 2; FIG. 52( d ) is a top view IV of multi-route planning of method 2;

[0080] Figure 53(a) is multi-route planning I of method 3; Figure 53(b) is multi-route planning II of method 3; Figure 53(c) is multi-route planning III of method 3; Figure 53(d) is multi-route planning IV of method 3;

[0081] FIG. 54( a ) is a top view I of multi-route planning of method 3; FIG. 54( b ) is a top view II of multi-route planning of method 3; FIG. 54( c ) is a top view III of multi-route planning of method 3; FIG. 54( d ) is a top view IV of multi-route planning of method 3;

[0082] Fig.55 This is a flowchart for multi-route planning of anti-ship missiles.

[0083] The above subject matter of the present invention is limited to the following embodiments. Without departing from the above technical concept of the present invention, various substitutions and changes can be made according to common technical knowledge and conventional means in the field, which should all be included in the protection scope of the present invention.

[0084] Embodiment 1:

[0085] A multi-route planning method for an anti-ship missile comprises the following steps:

[0086] 1) Establish a route start reference unit vector model to calculate the route start reference unit vector;

[0087] The steps to establish the route start reference unit vector model include:

[0088] 1.1) Establish the simulation coordinate system oxyz and the waypoint coordinate system OXYZ; where the origin o is the position of the observation ship's search radar; the ox axis points to the north, the oz axis is vertically upward; the oy axis forms a right-hand system; O is the waypoint, OZ is vertically upward, OX points to the north, and OY is perpendicular to OXZ;

[0089] 1.2) Let the reference attack direction unit vector be in, The starting point of the route; is the end point of the route; q x ,q y ,q z for (q x ,q y ,q z ) is the reference attack direction, that is, the vector from the start point of the route to the end point of the route. axis, Axis and The projection coordinates of the axes;

[0090] 1.3) Establish the route starting reference unit vector model to calculate the route starting reference unit vector for positive and negative attack angles It is the root of the equation group (1), the formula is similar, and the solution is only a numerical solution. The characteristic of the present invention is that it uses an implicit expression of the equation group, which is different from the explicit iteration of other iterative methods.

[0091] The route start reference unit vector model is as follows:

[0092]

[0093] In the formula, θ 攻击 is the angle of attack; is the route starting reference unit vector; (x, y, z) is the vector exist axis, Axis and Axis (that is Figure 1 to Figure 5 The projection coordinates of the X, Y, and Z axes are shown.

[0094] 2) Establishing a horizontal boundary vector model to calculate the horizontal boundary vector;

[0095] The horizontal boundary vector model is as follows:

[0096]

[0097] In the formula, θ 水 is the horizontal turning limit angle; is the horizontal boundary vector; is the base unit vector.

[0098] 4) Establish a vertical boundary vector model to calculate the vertical boundary vector; the vertical boundary vector model includes the following three types:

[0099] The first vertical boundary vector model:

[0100]

[0101] In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

[0102] The second vertical boundary vector model is as follows:

[0103]

[0104]

[0105] In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

[0106] The third vertical boundary vector model is as follows:

[0107]

[0108]

[0109] In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

[0110] 5) Calculate possible child node vectors according to the horizontal boundary vector and the vertical boundary vector; the possible child nodes include the next possible waypoint after the current waypoint;

[0111] Calculate possible child node vectors, the steps include:

[0112] 5.1) Calculate the initial child node vector Right now:

[0113]

[0114] In the formula, are the upper and lower bounds of the horizontal vector; are the upper and lower bounds of the vertical vector;

[0115] 5.2) Calculate the child node vector Right now:

[0116]

[0117] 5.3) Normalize child node vectors get:

[0118]

[0119] In the formula, is the normalized child node vector;

[0120] 5.4) Calculate the (i-1)*5+jth possible child node vector Right now:

[0121]

[0122] Where h is the step length; l+1=(i-1)*5+j; Possible child nodes.

[0123] 6) Screening possible child node vectors to obtain a screened possible child node vector table;

[0124] The steps of screening possible child node vectors include:

[0125] 6.1) The possible child node vector to be screened is Calculate the linear interpolation temp1 and linear interpolation temp2 of bilinear interpolation, that is:

[0126]

[0127]

[0128] In the formula, (x i-1 ,y j-1 ,z(x i-1 ,y j-1 ))、(x i ,y j-1 ,z(x i ,y j-1 ))、(x i-1 ,y j ,z(x i-1 ,y j )) and (x i ,y j ,z(x i ,y j )) is the adjacent grid point data of the possible child node of the digital map; the adjacent grid points of the possible child node are grid points used to represent the surface height of the possible child node; Digital maps only have grid point data, and there is no data between grid points. These four points are the four points closest to the node, that is, these four points are used to represent the topographic height of the node. The child node may be higher than this topographic height, otherwise the child node will be removed.

[0129] 6.2) Calculation Corresponding ground height in digital map Right now:

[0130]

[0131] 6.3) Add n points between two nodes and get the coordinates of the mth node among the n points Right now:

[0132]

[0133] Where, 0<m≤n;

[0134] 6.4) will satisfy and The possible child node vector of is set to zero vector, and the possible child node vector table is obtained

[0135] 7) Establish an optimization model and input the screened possible child node vector table into the optimization model to obtain the optimal child node vector

[0136] The optimization model is as follows:

[0137]

[0138]

[0139]

[0140]

[0141] In the formula, is the optimal child node vector. is the objective function of the optimization model; f is the cost function; From the starting position to the position The true cost of From the starting position to the position estimated cost.

[0142] 8) Determine the optimal child node vector Whether the iteration end condition is met, if so, the optimal child node vector Output as route, otherwise, return to step 2).

[0143] The iteration end conditions include condition 1 and condition 2; if the optimal child node vector satisfies condition 1 or condition 2, the iteration ends;

[0144] Condition one:

[0145]

[0146] Condition two:

[0147]

[0148] Condition 1 is the condition for the end of the iteration of the starting planning at the starting point, and condition 2 is the condition for the end of the iteration of the starting planning at the end point.

[0149] Embodiment 2:

[0150] A multi-route planning method for an anti-ship missile mainly includes the following contents:

[0151] When establishing the coordinate system, only the motion of the missile's center of mass is considered, and the rotation around the center of mass is not considered. The simulation coordinate system oxyz is established, o takes the position of the observation ship's search radar, the ox axis points to the north, and the oz axis is vertically upward; the oy axis forms a right-handed system, and the simulation coordinate system is regarded as an inertial coordinate system.

[0152] Assume the waypoint coordinate system OXYZ (which is in translation relationship with the simulation coordinate system) to obtain the node vector and displacement of each waypoint. O is the waypoint, OZ is vertically upward, OX points to the north, and OY is perpendicular to OXZ, which conforms to the right-hand rule. The vector digital map grid data is stored in three m×n matrices datax(m,n), datay(m,n) and dataz(m,n). Among them, the matrix datax(m,n) is a column vector of n m-dimensional elements with the same elements, storing the ordinate of the n-column space vector; the matrix datay(m,n) is a row vector of m n-dimensional elements with the same elements, storing the abscissa of the m-row space vector; the matrix dataz(m,n) stores the height of the m-row and n-column space vector. Then the actual geographic coordinates corresponding to the grid (i,j) on the vector digital map are [i,j,k]=(datax(i,j),datay(i,j),dataz(i,j)).

[0153] 1. Route start reference unit vector model

[0154] Known route start point End point There are two types of route start planning and route end point start planning. Generally, route end point start planning is adopted. is the reference attack direction unit vector (azimuth, elevation angle), the attack angle θ 攻击 (Vector With vector The angle between the clockwise and counterclockwise directions is negative, and the counterclockwise direction is positive, corresponding to the attacking ship’s port side being positive and the starboard side being negative). Then the starting reference unit vector of a route is By listing the equations, we can get the route starting reference unit vector for positive and negative attack angles.

[0155]

[0156] 2. Horizontal Boundary Vector Model

[0157] Known route start point End point The lth route node vector is The base unit vector of the node Horizontal turning limit angle θ 水 , let the horizontal boundary vector be List the equations and solve the horizontal vector upper bound The Nether The horizontal boundary vector model is as follows Figure 1 shown.

[0158]

[0159] The reference unit direction vector rotates around the Z axis by θ 水 , the height remains unchanged.

[0160] 3. Vertical Boundary Vector Model

[0161] Given the vertical turning limit angle θ 垂 , let the vertical boundary vector be Find the upper bound of the vertical vector by writing a system of equations The Nether

[0162] Method 1: While the horizontal boundary vector remains unchanged, the vector direction of the vertical boundary vector rotates around the Y axis by θ 垂 , vertical boundary vector model method 1 as Figure 2 Shown

[0163]

[0164] The trajectory is relatively jumpy and the solution is simple. The reference unit direction vector rotates around the Y axis by θ 垂 , the distance to the vertical plane remains unchanged, but there is a situation where the surface obstacles cannot be avoided.

[0165] Method 2: While the horizontal boundary vector remains unchanged, the vertical boundary vector rotates around the node by θ in the projection plane of XOZ 垂 , the cross product vector is in the XOZ plane. Vertical boundary vector model method 2 is as follows Figure 3 shown.

[0166]

[0167] The unit direction vector rotates around the node by θ in the plane perpendicular to the plumb plane (projection) 垂 There is only one plane that passes through a line that is not perpendicular to the known plane and is perpendicular to the plane; if the unit direction vector is perpendicular to the plumb plane, that is, (0,1,0), there are infinite solutions. The trajectory is relatively smooth, with both horizontal and vertical turns.

[0168] Method 3: When the horizontal boundary vector remains unchanged, the vertical boundary vector rotates around the node by θ in the projection plane of the horizontal plane 垂 , the cross product vector is in the XOY plane. Vertical boundary vector model method 3 is as follows Figure 4 shown.

[0169]

[0170] In method 3, the unit direction vector rotates around the node in the vertical (projection) plane of the horizontal plane. The result is similar to method 1. The trajectory is more jumpy, the X and Y axis motions are coupled, and the obstacle crossing ability is strong.

[0171] 4. Possible child node vector model

[0172] From the current node as the origin, the step length h and the vertical turning limit angle θ 限 A graph can be obtained, and the formula is used to process it, dividing the graph into 16 equal parts, 25 points. The next node will select the optimal solution from these 25 nodes. After the missile flies to the designated node position, it starts to select the next node, and it will start from the current node. A region to be planned is constructed by the vertical and horizontal turning limit angles and step lengths. It can be drawn that this region is part of the sphere. Then it is divided into 4 parts in the vertical direction and 4 parts in the horizontal direction. In this way, an area to be planned is turned into a 4*4 sub-area, and each of its vertices is regarded as a node to be planned and left until the next step to calculate the cost value of the node. This step directly transforms the previous two-dimensional search into a three-dimensional search. The horizontal and vertical planning can be analogized to the three-dimensional search. The area to be planned in the next step is also determined by the constraints first. After the information data of the area is obtained, it is divided into several independent areas to obtain the required nodes and label them. Generally speaking, the more parts are divided, the more accurate the results will be. Of course, this will also greatly increase the computing time and storage space, especially the computing time. Adding a node may not have much effect, but from the perspective of the overall route planning, it will affect the combat efficiency, so choosing a suitable number of divisions is also very important. The possible sub-node vector model diagram is as follows Figure 5 shown.

[0173] Initial child nodes vector:

[0174]

[0175] Child nodes vector:

[0176]

[0177] Normalized child node direction vector:

[0178]

[0179] The step length h gives the (i-1)*5+jth possible child node vector

[0180]

[0181] 5. Filter subnode vector model

[0182] 5.1 Terrain Avoidance

[0183] Filter the possible subnode vector model above the surface. Suppose the possible subnode vector to be filtered is Possible child node adjacent grid points (x i-1 ,y j-1 ,z(x i-1 ,y j-1 ))、(x i ,y j-1 ,z(x i ,y j-1 ))、(x i-1 ,y j ,z(x i-1 ,y j )) and (x i ,y j ,z(x i ,y j )),have but:

[0184]

[0185]

[0186] Relying on the bilinear interpolation method, we can find Corresponding terrain surface height in digital map

[0187]

[0188] Bundle The child node vector is set to zero vector, and the possible child node vector table is obtained

[0189] Only through the possible child nodes vector The judgment of whether it is above the ground is not perfect. In fact, it is necessary to judge whether all points of the route between the previous node and the possible child node to be screened are above the ground, but the amount of calculation is unacceptable. It is possible to add 2 points, 3 points, or even n points between two nodes to judge the feasibility of this route.

[0190] (1) Take the bisection point

[0191] When n = 2, the bipartite coordinates are

[0192]

[0193] From equation (10) to equation (12), we get Bundle and The child node vector is set to zero vector, and the possible child node vector table is obtained

[0194] (2) Take the three-point point

[0195] When n=3, the coordinates of the three-point point are

[0196]

[0197]

[0198] From equation (10) to equation (12), we get and Bundle and The child node vector is set to zero vector, and the possible child node vector table is obtained

[0199] (3) Take n points

[0200] The coordinates of the mth (0<m≤n) node in the n points are:

[0201]

[0202] From equation (10) to equation (12), we get Bundle and The child node vector is set to zero vector, and the possible child node vector table is obtained

[0203] 5.2 Avoidance of air defense threat zones

[0204] The air defense threat zone can be composed of radar detection threat zone, fire resistance zone, shooting permission zone, etc. Because the area within the fire resistance zone is an absolute no-fly zone, the air defense threat zone can be treated as equivalent to the fire resistance zone when constructing the model. Considering that the movement of the enemy's fire resistance zone is relatively static relative to the flight speed of our missiles, the enemy's air defense threat zone and terrain threat zone can be regarded as known fixed threats after simplifying the fire resistance zone. Since it is difficult to determine the detection angle of the enemy's air defense fire when our missile is about to enter its threat zone, for the safety and concealment of missile flight, the air defense threat zone is regarded as a threat zone T with a radius of r with a certain origin as the center. Among them, (x, y) are the coordinates of the air defense threat zone, r is the fire resistance radius, α1 and α2 are the initial angle and the end angle respectively, where α2-α1=360°.

[0205] It is known that in the map, 0≤i≤m, 0≤j≤n, the complete coordinates corresponding to the grid point (i,j) on the digital map are (datax(i,j),datay(i,j),dataz(i,j)), and the complete coordinates corresponding to the coordinates (x,y) of the air defense threat zone are (datax(x,y),datay(x,y),dataz(x,y)). If ml is the length from point (i,j) to (x,y), ml=[(datax(i,j)-datax(x,y)) 2 +(datay(i,j)-datay(x,y)) 2 ] 0.5 , the altitude of the air defense threat zone

[0206]

[0207] Formula (17) can be decomposed into formulas (18) to (20), assuming

[0208]

[0209] Establish the discriminant matrix E = (e ij ) m×n and

[0210]

[0211]

[0212] The air defense threat zone is

[0213]

[0214] 5.3 Horizontal route avoidance

[0215] When planning multiple routes, generally, the low-attack-angle route is planned first and then the high-attack-angle route. To avoid the horizontal plane intersection of the high-attack-angle route and the low-attack-angle route, a route pipeline of a certain width can be set for the low-attack-angle route, and the high-attack-angle route can be planned outside the route pipeline, thereby achieving horizontal route avoidance.

[0216] Assume that the planned low-angle-of-attack route i contains k waypoints, which are (excluding the starting point of the planned route end ), whose horizontal coordinate is (P i1x ,P i1y ),(P i2x ,P i2y ),…,(P ikx ,P iky ). If the width of the airway pipe is 2d, then (P i1x ,P i1y +d), (P i2x ,P i2y +d),…,(P ikx ,P iky +d), (P ikx ,P iky -d), (P i(k-1)x ,P i(k-1)y -d),…,(P i1x ,P i1y -d), (P i1x ,P i1y +d) Forming route pipeline polygon

[0217] Pair Vector Depend on

[0218]

[0219]

[0220]

[0221]

[0222] Calculate the quadrant number vector

[0223]

[0224] Among them, & represents the logical "and", q i1 ∈{0,1,2,3}.

[0225] Find vector and The cross product vector

[0226]

[0227] Cross Product The corresponding symbol vector

[0228]

[0229] Find vector and The dot product vector

[0230] D i1 =xv i ·xv i+1 +yv j ·yv j+1 i=1,…,2k (24)

[0231] Find the quadrant vector δ i ′1∈{-3,-2,-1,0,1,2,3}, where δ i ′1=q (i+1)1 -q i1 ,i=1,2,…,2k。 The corrected quadrant vector is obtained δ i1 ∈{-2,-1,0,1,2},i=1,2,…,2k.

[0232]

[0233] For any detection point on the horizontal plane Detects the closed airway pipe polygon Within the route, above the border or outside.

[0234] Detection point In closed airway pipeline polygon The logical criterion in within the route (excluding the boundary) is:

[0235]

[0236] Detection point In closed airway pipeline polygon The logical criterion on the route boundary is:

[0237]

[0238] After the detection is completed, the height of the grid points within the planned route pipeline boundary and within the boundary is set to the maximum value of the terrain and the air defense threat zone height.

[0239] 6. Optimal Model

[0240] From the cost function of the A* search algorithm, we can get

[0241]

[0242] Right now

[0243]

[0244] Starting point planning

[0245]

[0246] End-start planning

[0247]

[0248] Get the best child node vector

[0249] 7. Iterate the model

[0250] Starting point planning

[0251]

[0252] End-start planning

[0253]

[0254] If equation (32) or equation (33) does not meet the conditions, the six steps of horizontal boundary vector, vertical boundary vector, possible child node vector, screening child node vector, optimization and iteration are repeated until the optimal solution is found to complete the route planning. If equation (32) or equation (33) meets the conditions, the route planning is completed. The process of the anti-ship missile route planning method based on the improved A* search algorithm is as follows Figure 6 shown.

[0255] Embodiment 2:

[0256] A multi-route planning method for anti-ship missiles applied to covert attacks of anti-ship missiles includes:

[0257] Operational scenario 1: The grid position (row number, column number) target indication provided by the Red side reconnaissance aircraft locks the grid position (1109,1069) of the Blue side ship at sea, and the grid position (489,449) of a certain type of Red side missile frigate near the island reef is used as a reference for the attack direction The unit vector is the starting reference unit vector (the attack angle is 0rad) and an anti-ship missile is launched to attack the blue ship. When planning the route, pay attention to avoiding the blue radar threat area to improve concealment. Parameter setting:

[0258] Anti-ship missile parameters: θ 限 =π / 6, the number of nodes k≤15;

[0259] Digital map parameters: the entire area is a grid (1-1836, 1-1676), the island and reef sea areas are grids (460-918, 420-838), and the rest are sea areas. The grid accuracy is 0.1km, and the highest altitude is 1.4488km.

[0260] Radar threat zone parameters: R1 = 3km, W1 = (559,539); W2 = (659,569), R2 = 4km; W3 = (759,739), R3 = 2km; W4 = (609,739), R4 = 3.5km; W5 = (659,499), R5 = 4.2km;

[0261] Horizontal and vertical boundary vector segmentation parameters: S=4, M=4.

[0262] (1) Method 1

[0263] Choosing different step lengths will have a greater impact on the results. Selecting step length h = 6, the route planning results Figure 7 and Figure 8 Selecting step length h = 8, the route planning result is as follows Fig. 9 and Fig.10 , select step size h = 10 to get the route planning result as follows Fig.11 and Fig.12 . (Unit: km)

[0264] (2) Method 2

[0265] After inputting the blue side ship position, the red side ship position and various constraints (missile limit angle, step length h = 11), method 2 is selected for route planning to obtain the result Fig.13 and Fig.14 Choosing different step lengths will have a significant impact on the results. If h≤10, the planning will fail. The following step length h=12 is selected to obtain the route planning result as follows Fig.15 and Fig.16 , select step size h = 13 to get the route planning result as follows Fig.17 and Fig.18 .

[0266] (3) Method 3

[0267] After inputting the blue side ship position, the red side ship position and various constraints (missile limit angle, step length h = 6), method 3 is selected for route planning to obtain the result Fig.19 and Fig. 20 Selecting different step lengths will have different effects on the results. The following step length h = 8 is selected to obtain the route planning result as follows: Fig.21 and Fig. 22 , select the step size h = 10 to get the route planning result as follows Fig.23 and Fig.24 .

[0268] Embodiment 3:

[0269] A multi-route planning method for anti-ship missiles applied to covert attacks of anti-ship missiles includes:

[0270] Operational scenario 2: The Red reconnaissance aircraft provides target indication and locks the Blue ship's sea grid position (1109,1069). The Red missile frigate is located at the grid position (489,449) near the islands and reefs. It fires 2 to 5 anti-ship missiles at the Blue ship. During the route planning process, it is necessary to avoid the Blue radar threat area to improve concealment. The parameter settings are the same as those in scenario 1.

[0271] (1) Method 1

[0272] The reference attack direction azimuth is The attack angles of the 2 routes are The attack angles of the three routes are 0, The attack angles of the 4 routes are The attack angles of the 5 routes are 0, Planning Figure 25 to Figure 32 shown.

[0273] (2) Method 2

[0274] The reference attack direction azimuth is The attack angles of the 2 routes are The attack angles of the three routes are The attack angles of the 4 routes are The attack angles of the 5 routes are Planning Figures 33 to 40 shown.

[0275] (3) Method 3

[0276] The reference attack direction azimuth is The attack angles of the 2 routes are The attack angles of the three routes are 0, The attack angles of the 4 routes are The attack angles of the 5 routes are 0, Planning Figures 41 to 48 shown.

[0277] Embodiment 4:

[0278] A multi-route planning method for anti-ship missiles used to attack island and reef targets includes:

[0279] Operational scenario 3: The Red reconnaissance aircraft provides target indication, locks the Blue ship at the grid position (489,449) near the island and reef, and the Red ship at the grid position (1509,1469), and launches 2 to 5 anti-ship missiles with different attack angles to attack the Blue ship. In the process of route planning, pay attention to avoid the Blue air defense threat zone to improve concealment. Parameter settings are the same as those in scenario 1.

[0280] (1) Method 1

[0281] The reference attack direction azimuth is The attack angles of the 2 routes are The attack angles of the three routes are 0, The attack angles of the 4 routes are The attack angles of the 5 routes are 0, After inputting 2 to 5 reasonable attack angles, route planning is performed and the planned routes are obtained on the digital map, as shown in Figures 49 and 50. Each path can effectively avoid the air defense threat area and the terrain threat area, and the route planning is completed under the constraints of various constraints.

[0282] (2) Method 2

[0283] The reference attack direction azimuth is The attack angles of the 2 routes are The attack angles of the three routes are The attack angles of the 4 routes are The attack angles of the 5 routes are Launch 2 to 5 anti-ship missiles with different attack angles to attack the blue side's ships. As shown in Figures 51 and 52, the planned routes can be obtained on the digital map, and radar threat areas and terrain threat areas can be avoided. However, only some attack angles can be successfully planned within the attack angle limit range, and most of the missile's high attack angle route is in level flight over the sea, with few obstacle avoidance sections; the small attack angle route has a larger launch angle in the starting route space, which is suitable for vertical launch.

[0284] (3) Method 3

[0285] The reference attack direction azimuth is The attack angles of the 2 routes are The attack angles of the three routes are 0, The attack angles of the 4 routes are The attack angles of the 5 routes are 0, As shown in Figures 53 and 54, 2 to 5 planned routes can be obtained on the digital map, and the anti-ship missile can complete the route planning under the constraints of various constraints. Although the route of the penetration section after crossing the obstacle is relatively flat, the route of the launch section is very steep, which is suitable for vertical launch, and the cruise section has large fluctuations.

[0286] Simulations and comparisons were conducted for multi-route planning of different modes of anti-ship missiles. Method 1 has strong applicability for both concealed launch planning and island and reef target attack planning of anti-ship missile route planning, and has strong horizontal and vertical planning capabilities. The initial stage adopts obstacle crossing mode, mainly to prevent obstacles from crossing. Method 2 tends to plan in the horizontal direction, with small height changes throughout the whole process, smooth trajectory, and strong concealment, but it is necessary to avoid route crossing and has high requirements for attack angles. If these requirements cannot be met, convergence will fail. Method 3 is more inclined to vertical planning, has strong obstacle avoidance capabilities, and has a greater degree of route jumping in the initial stage than Method 1, but has high requirements for the maneuverability of the missile. Simulations and comparisons were conducted for the corresponding modes of multi-route planning for concealed attack and island and reef target attack of anti-ship missiles. Under methods 1 and 3, during covert attacks, only the initial route section jumps, the cruising route is relatively flat, and the horizontal route intersection is not obvious. When attacking island and reef targets, most of the routes are relatively jumpy, and only a small section after crossing the obstacle is flat and the horizontal route intersection is obvious. Under method 2, the course change of the entire route is large during covert attacks, and when attacking island and reef targets, a large course change occurs only in the obstacle-circling section.

Claims

1. A multi-route planning method for anti-ship missiles, characterized in that: The following steps are involved: 1) Establish a route start reference unit vector model to calculate the route start reference unit vector; 2) Establishing a horizontal boundary vector model to calculate the horizontal boundary vector; 4) Establishing a vertical boundary vector model to calculate the vertical boundary vector; 5) Calculate possible child node vectors according to the horizontal boundary vector and the vertical boundary vector; the possible child nodes include the next possible waypoint after the current waypoint; 6) Screening possible child node vectors to obtain a screened possible child node vector table; 7) Establish an optimization model and input the screened possible child node vector table into the optimization model to obtain the optimal child node vector 8) Determine the optimal child node vector Whether the iteration end condition is met, if so, the optimal child node vector Output as route, otherwise, return to step 2); The steps to establish the route start reference unit vector model include: s1) Establish the simulation coordinate system oxyz and the waypoint coordinate system OXYZ; where the origin o is the position of the observation ship's search radar; the ox axis points to the north, the oz axis is vertically upward; the oy axis forms a right-hand system; O is the waypoint, OZ is vertically upward, OX points to the north, and OY is perpendicular to OXZ; s2) Let the reference attack direction unit vector be in, The starting point of the route; is the end point of the route; q x ,q y ,q z is the reference attack direction, that is, the vector from the start point of the route to the end point of the route. axis, Axis and The projection coordinates of the axes; s3) Establishing a route start reference unit vector model to calculate the route start reference unit vector for positive and negative attack angles; The route start reference unit vector model is as follows: In the formula, θ 攻击 is the angle of attack; is the route starting reference unit vector; (x, y, z) is the vector exist axis, Axis and The projected coordinates of the axes.

2. The method for multi-route planning of an anti-ship missile according to claim 1, characterized in that: The horizontal boundary vector model is as follows: In the formula, θ 水 is the horizontal turning limit angle; is the horizontal boundary vector; is the base unit vector.

3. The anti-ship missile multi-route planning method according to claim 1, characterized in that: The vertical boundary vector model is as follows: In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

4. The anti-ship missile multi-route planning method according to claim 1, characterized in that: The vertical boundary vector model is as follows: In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

5. The method for multi-route planning of an anti-ship missile according to claim 1, characterized in that: The vertical boundary vector model is as follows: In the formula, θ 垂 The vertical turning limit angle; is the vertical boundary vector.

6. The method for multi-route planning of an anti-ship missile according to claim 1, characterized in that: Calculate possible child node vectors, the steps include: 1) Calculate the initial child node vector Right now: In the formula, are the upper and lower bounds of the horizontal vector; are the upper and lower bounds of the vertical vector; 2) Calculate the child node vector Right now: 3) Normalize child node vector get: In the formula, is the normalized child node vector; 4) Calculate the (i-1)*5+jth possible child node vector Right now: Where h is the step length; l+1=(i-1)*5+j; Possible child nodes.

7. The anti-ship missile multi-route planning method according to claim 1, characterized in that: The steps of screening possible child node vectors include: 1) The possible child node vector to be screened is Calculate the linear interpolation temp1 and linear interpolation temp2 of bilinear interpolation, that is: In the formula, (x i-1 ,y j-1 ,z(x i-1 ,y j-1 ))、(x i ,y j-1 ,z(x i ,y j-1 ))、(x i-1 ,y j ,z(x i-1 ,y j )) and (x i ,y j ,z(x i ,y j )) is the adjacent grid point data of the possible child node of the digital map; the adjacent grid points of the possible child node are grid points used to represent the surface height of the possible child node; 2) Calculation Corresponding ground height in digital map Right now: 3) Add n points between two nodes and get the corresponding m-th node vector among the n points Right now: Where, 0<m≤n; 4) Will satisfy and The possible child node vector of is set to zero vector, and the possible child node vector table is obtained 8. The method for multi-route planning of an anti-ship missile according to claim 1, characterized in that: The optimization model is as follows: In the formula, is the optimal child node vector; is the objective function of the optimization model; is the cost function; From the starting position to the position The true cost of From the starting position to the position estimated cost.

9. The method for multi-route planning of an anti-ship missile according to claim 1, characterized in that: The iteration end conditions include condition 1 and condition 2; if the optimal child node vector satisfies condition 1 or condition 2, the iteration ends; Condition one: Condition two: Where h is the step size.