A method and system for optimizing the design of the appearance of a large quad-rotor UAV

Abstract

The present invention discloses a method and system for optimizing the appearance of a large four-axis multi-rotor unmanned aerial vehicle, which relates to the technical field of unmanned aerial vehicle manufacturing, including: collecting public control point information on the appearance of a large four-axis multi-rotor unmanned aerial vehicle through surveying and mapping tools; obtaining a function model of the appearance of the large four-axis multi-rotor unmanned aerial vehicle through quadratic polynomial fitting; randomly generating a discrete point set, inserting the points in the discrete point set into a Delaunay triangle one by one and connecting them to obtain a Delaunay triangulation; constructing the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation; performing mesh optimization on the Voronoi diagram; performing material mapping operations on the optimized Voronoi diagram; obtaining an allowable error range and judging whether the final appearance modeling and the public control point information meet the allowable error range. By setting an appearance information processing module and an appearance modeling optimization module, an appearance model is constructed, the processing accuracy is effectively improved, and the error of the appearance modeling of the large four-axis multi-rotor unmanned aerial vehicle is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle manufacturing, and in particular to an appearance optimization design method and system for a large four-axis multi-rotor unmanned aerial vehicle. Background Art

[0002] A quadcopter is also called a quadrotor aircraft or a quadcopter helicopter, or quadcopter for short. This quadcopter is a multi-rotor aircraft. The four propellers of a quadcopter are simple mechanisms directly connected to the motor. The cross-shaped layout allows the aircraft to obtain the force to rotate the fuselage by changing the motor speed, thereby adjusting its own posture.

[0003] In the design process of the appearance of a large quadcopter multi-rotor drone using existing technology, it is difficult to construct the appearance model and the data information processing accuracy is low, which can easily lead to large design errors in the final appearance. Summary of the invention

[0004] In order to solve the above technical problems, a method and system for optimizing the appearance design of a large four-axis multi-rotor UAV is provided. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A method for optimizing the appearance of a large quad-rotor UAV comprises:

[0007] Collecting public control point information on the exterior of the large quad-rotor drone using a surveying tool, wherein the public control point information includes abscissa, ordinate and vertical coordinates;

[0008] Performing data processing according to the public control point information, and obtaining a function model of the appearance of the large quad-axis multi-rotor drone by quadratic polynomial fitting;

[0009] According to the function model of the appearance of the large quad-rotor UAV, a discrete point set is randomly generated, wherein the horizontal coordinates of all discrete points in the discrete point set are not repeated and are arranged in ascending order according to the horizontal coordinates, a Delaunay triangle is constructed, and the Delaunay triangle contains all the discrete points in the discrete point set, and the points in the discrete point set are inserted into the Delaunay triangle one by one and connected to obtain a Delaunay triangulation network;

[0010] According to the Delaunay triangulation, perpendicular bisectors of the sides of all triangles in the Delaunay triangulation are constructed, which are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone;

[0011] Obtaining the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, and performing grid optimization on the Voronoi diagram of the appearance of the large quad-rotor UAV according to the smoothness and continuity requirements of the appearance of the large quad-rotor UAV to obtain an optimized Voronoi diagram;

[0012] Performing material mapping operation on the optimized Voronoi diagram to obtain a final appearance shape;

[0013] Obtain the allowable error range and determine whether the final shape and the public control point information meet the allowable error range. If so, output the final shape. If not, output that the final shape optimization error is large and recalculate the final shape.

[0014] Preferably, the data processing according to the public control point information and obtaining the function model of the appearance of the large quad-axis multi-rotor drone by quadratic polynomial fitting specifically include:

[0015] Setting the kernel function formula according to the public control point information;

[0016] Derived the vertical coordinate value fitting function of all points on the shape according to the kernel function formula;

[0017] According to the vertical coordinate values ​​of the common control points, the undetermined coefficient matrix of the appearance model is calculated using the least square method;

[0018] According to the undetermined coefficient matrix of the appearance model and the kernel function formula, the vertical coordinate value fitting function of all points on the appearance, the function model of the appearance of the large quad-axis multi-rotor drone is obtained using the matrix form of the error equation;

[0019] The kernel function formula is:

[0020] Q(x,y;x i ,y i )=[(xx i ) 2 +(yy i ) 2 +q 2 ] b ,

[0021] Where Q is the kernel function, q is the smoothing factor used to correct the kernel function formula, b is 1 / 2 or -1 / 2, when it is 1 / 2 it is called a positive shape, when it is -1 / 2 it is called an inverted shape, (x i ,y i ) are the horizontal and vertical coordinates of the common control point, and (x, y) are the horizontal and vertical coordinates of the unknown point on the shape;

[0022] The fitting function of the vertical coordinate values ​​of all points on the shape is:

[0023]

[0024] In the formula, H is the vertical coordinate value fitting function, a i is the unknown coefficient, n is the sum of the appearances;

[0025] The undetermined coefficient matrix of the appearance model is:

[0026] A=(Q(x,y;x i ,y i ) T Q(x,y;x i ,y i )) -1 Q(x,y;x i ,y i ) T H(x,y),

[0027] Where A is the unknown coefficient matrix of the appearance model;

[0028] The matrix form of the error equation is:

[0029] V=Q(x,y;x i ,y i )AH(x,y),

[0030] Where V is the matrix form of the error equation.

[0031] Preferably, inserting the points in the discrete point set into the Delaunay triangle one by one and connecting the lines to obtain the Delaunay triangulation specifically includes:

[0032] Construct a Delaunay triangle, wherein the Delaunay triangle includes all discrete points in the discrete point set and stores them in a Delaunay triangle linked list;

[0033] Insert the points in the discrete point set into the Delaunay triangle one by one, find its circumscribed circle in the Delaunay triangle linked list, and the triangle containing the inserted point is called the influence triangle of the discrete point;

[0034] Delete the common edges of the affected triangles and connect the insertion point to all vertices of the affected triangles;

[0035] According to the optimization criteria, the newly formed local triangles are optimized and the formed triangles are stored in the Delaunay triangle linked list;

[0036] The optimization criterion is to optimize the newly formed triangle, combine two triangles with common sides into a polygon, and use the maximum empty circle criterion to determine whether the fourth vertex is inside the circumscribed circle of the triangle. If so, the diagonal lines are swapped and the local optimization process is output to end. If not, the newly formed triangle is deleted.

[0037] The maximum empty circle criterion is that the circumscribed circle of any triangle does not contain other points.

[0038] Preferably, the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone. Specifically, it includes:

[0039] All discrete points form a Delaunay triangulation network, and the discrete points and the formed triangles are numbered, and the center of the circumscribed circle of each triangle is calculated and recorded;

[0040] According to the Delaunay triangle linked list, get at most three adjacent triangles whose three sides are common to the current triangle;

[0041] Connect the circumcenter of the adjacent triangle obtained with the circumcenter of the current triangle and store them in the Voronoi edge list;

[0042] According to the Delaunay triangle linked list, obtain the edge of the current triangle that does not have any adjacent triangles, and store the perpendicular ray of the edge into the Voronoi edge linked list;

[0043] The Voronoi diagram of the appearance of the large quad-rotor drone is obtained according to all the edges in the Voronoi edge list.

[0044] Preferably, the obtaining of the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, and performing grid optimization on the Voronoi diagram of the appearance of the large quad-rotor UAV according to the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, to obtain the optimized Voronoi diagram specifically includes:

[0045] The cotangent weight of each point in the Voronoi diagram of the large quad-rotor drone is calculated according to the cotangent weight formula of the mesh surface;

[0046] Substituting the cotangent weight into the Laplace coordinate formula, the Laplace coordinates of each point in the Voronoi diagram of the appearance of the large quad-rotor drone are obtained;

[0047] Placing the Laplace coordinates in a spatial rectangular coordinate system to obtain an optimized Voronoi diagram;

[0048] The mesh surface cotangent weight formula is:

[0049] ω ij =cotα+cotβ,

[0050] In the formula, ω ij is the cotangent weight of point i and its adjacent point j, α and β are the two opposite angles of the two adjacent triangles where the line connecting point i and its adjacent point j is located;

[0051] The Laplace coordinate formula is:

[0052]

[0053] In the formula, δ i is the Laplace coordinate of point i, v i ,v j are the corresponding coordinate values ​​of point i and its adjacent point j respectively.

[0054] A large quad-axis multi-rotor UAV appearance optimization design system is used to implement the large quad-axis multi-rotor UAV appearance optimization design method as described above, comprising:

[0055] An appearance information processing module, the appearance information processing module is used to collect public control point information on the appearance of the large four-axis multi-rotor drone through a surveying and mapping tool, perform data processing according to the public control point information, obtain a function model of the appearance of the large four-axis multi-rotor drone through quadratic polynomial fitting, and randomly generate a discrete point set according to the function model of the appearance of the large four-axis multi-rotor drone;

[0056] An appearance modeling optimization module, the appearance modeling optimization module is used to construct a Delaunay triangle, construct the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation network based on the Delaunay triangulation network, obtain the appearance modeling smoothness and continuity requirements of the large four-axis multi-rotor UAV, according to the appearance modeling smoothness and continuity requirements of the large four-axis multi-rotor UAV, mesh optimize the Voronoi diagram of the appearance of the large four-axis multi-rotor UAV, perform material mapping operations on the optimized Voronoi diagram, obtain the allowable error range and determine whether the final appearance modeling and the common control point information are within the allowable error range.

[0057] Preferably, the appearance information processing module includes:

[0058] An information collection unit, the information collection unit is used to collect public control point information on the appearance of the large four-axis multi-rotor drone through a surveying and mapping tool, the public control point information includes abscissa, ordinate and vertical coordinate;

[0059] A data processing unit, the data processing unit is used to perform data processing according to the public control point information, and obtain a function model of the appearance of the large four-axis multi-rotor drone through quadratic polynomial fitting;

[0060] A data generation unit, wherein the data generation unit is used to randomly generate a discrete point set according to a function model of the appearance of the large four-axis multi-rotor drone, wherein the horizontal coordinates of all discrete points in the discrete point set are not repeated and are arranged in ascending order according to the horizontal coordinates.

[0061] Preferably, the appearance optimization module includes:

[0062] A preliminary drawing unit, wherein the preliminary drawing unit is used to construct a Delaunay triangle, wherein the Delaunay triangle includes all discrete points in the discrete point set, and the points in the discrete point set are inserted into the Delaunay triangle one by one and connected to obtain a Delaunay triangulation network;

[0063] A refinement drawing unit, wherein the refinement drawing unit is used to construct perpendicular bisectors of the sides of all triangles in the Delaunay triangulation according to the Delaunay triangulation, which are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone;

[0064] An optimization drawing unit, wherein the optimization drawing unit is used to obtain the smoothness and continuity requirements of the appearance of the large four-axis multi-rotor drone, and according to the smoothness and continuity requirements of the appearance of the large four-axis multi-rotor drone, a grid optimization is performed on the Voronoi diagram of the appearance of the large four-axis multi-rotor drone to obtain an optimized Voronoi diagram;

[0065] A material mapping unit, the material mapping unit is used to perform a material mapping operation on the optimized Voronoi diagram to obtain a final appearance shape;

[0066] A data detection unit is used to obtain an allowable error range and determine whether the final shape modeling and the public control point information meet the allowable error range. If so, the final shape modeling is output; if not, the final shape modeling optimization error is large and the final shape modeling is recalculated.

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] The present invention proposes an appearance optimization design scheme for a large quad-axis multi-rotor unmanned aerial vehicle. The scheme is based on collecting public control point information on the appearance of the large quad-axis multi-rotor unmanned aerial vehicle through surveying and mapping tools, obtaining a function model of the appearance of the large quad-axis multi-rotor unmanned aerial vehicle through quadratic polynomial fitting, randomly generating a discrete point set, constructing a Delaunay triangle, inserting the points in the discrete point set into the Delaunay triangle one by one and connecting them to obtain a Delaunay triangulation, constructing the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation, performing grid optimization on the Voronoi diagram of the appearance of the large quad-axis multi-rotor unmanned aerial vehicle, performing material mapping operations on the optimized Voronoi diagram, obtaining an allowable error range and judging whether the final appearance modeling and the public control point information meet the allowable error range, effectively improving processing accuracy and reducing the error of the appearance modeling of the large quad-axis multi-rotor unmanned aerial vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a schematic diagram of the flow of the appearance optimization design method of the large four-axis multi-rotor UAV of the present invention;

[0070] Figure 2A schematic diagram of a function model flow of the present invention for performing data processing according to the common control point information and obtaining the appearance of a large quad-axis multi-rotor UAV through quadratic polynomial fitting;

[0071] Figure 3 The present invention is to insert the points in the discrete point set into the Delaunay triangle one by one and connect them to obtain the Delaunay triangulation process diagram;

[0072] Figure 4 The perpendicular bisectors of the sides of all triangles in the Delaunay triangulation network of the present invention are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram process flow diagram of the appearance of the large four-axis multi-rotor drone;

[0073] Figure 5 The present invention performs grid optimization on the Voronoi diagram of the appearance of the large four-axis multi-rotor drone to obtain a schematic diagram of the optimized Voronoi diagram process. DETAILED DESCRIPTION

[0074] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0075] Reference Figure 1 As shown, a method for optimizing the appearance of a large quad-rotor UAV comprises:

[0076] The public control point information on the appearance of the large quad-rotor UAV is collected through surveying and mapping tools. The public control point information includes the horizontal coordinate, the vertical coordinate and the vertical coordinate;

[0077] Data processing is performed based on the public control point information, and a function model of the appearance of the large quad-rotor UAV is obtained through quadratic polynomial fitting;

[0078] According to the function model of the appearance of the large quad-rotor UAV, a discrete point set is randomly generated. The horizontal coordinates of all discrete points in the discrete point set are not repeated and are arranged in ascending order of the horizontal coordinates. A Delaunay triangle is constructed. The Delaunay triangle contains all the discrete points in the discrete point set. The points in the discrete point set are inserted into the Delaunay triangle one by one and connected to obtain the Delaunay triangulation.

[0079] According to the Delaunay triangulation, the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation are constructed, which are recorded as Voronoi edges. The continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone.

[0080] Obtaining the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, and performing grid optimization on the Voronoi diagram of the appearance of the large quad-rotor UAV according to the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, to obtain an optimized Voronoi diagram;

[0081] Perform material mapping operation on the optimized Voronoi diagram to obtain the final appearance;

[0082] Get the allowable error range and determine whether the final shape and the public control point information meet the allowable error range. If so, output the final shape. If not, output the final shape optimization error is large and recalculate the final shape.

[0083] Reference Figure 2 As shown, data processing is performed according to the public control point information, and the function model of the appearance of the large four-axis multi-rotor drone is obtained by quadratic polynomial fitting, which specifically includes:

[0084] Setting the kernel function formula according to the public control point information;

[0085] According to the kernel function formula, the vertical coordinate value fitting function of all points on the shape is derived;

[0086] According to the vertical coordinate values ​​of the common control points, the undetermined coefficient matrix of the appearance model is calculated using the least square method;

[0087] According to the undetermined coefficient matrix of the appearance model and the kernel function formula, the vertical coordinate values ​​of all points on the appearance are fitted, and the function model of the appearance of the large quad-rotor UAV is obtained using the matrix form of the error equation;

[0088] The kernel function formula is:

[0089] Q(x,y;x i ,y i )=[(xx i ) 2 +(yy i ) 2 +q 2 ] b ,

[0090] Where Q is the kernel function, q is the smoothing factor used to correct the kernel function formula, b is 1 / 2 or -1 / 2, when it is 1 / 2 it is called a positive shape, when it is -1 / 2 it is called an inverted shape, (x i ,y i ) are the horizontal and vertical coordinates of the common control point, and (x, y) are the horizontal and vertical coordinates of the unknown point on the shape;

[0091] The fitting function of the vertical coordinate values ​​of all points on the shape is:

[0092]

[0093] In the formula, H is the vertical coordinate value fitting function, a i is the unknown coefficient, n is the sum of the appearances;

[0094] The undetermined coefficient matrix of the appearance model is:

[0095] A=(Q(x,y;x i ,y i ) T Q(x,y;x i ,y i )) -1 Q(x,y;x i ,y i ) T H(x,y),

[0096] Where A is the unknown coefficient matrix of the appearance model;

[0097] The matrix form of the error equation is:

[0098] V=Q(x,y;x i ,y i )AH(x,y),

[0099] Where V is the matrix form of the error equation.

[0100] In machine learning algorithms, linear models based on nonlinear functions of data are very common. This method can be as efficient as a linear model, while making the model applicable to a wider range of data. Polynomial fitting is the simplest of these algorithms. If you want to fit a parabola instead of a plane, you need to calculate a linear combination of the quadratic terms of the input variable x.

[0101] Reference Figure 3 As shown, the points in the discrete point set are inserted into the Delaunay triangle one by one and connected, and the Delaunay triangulation network is obtained, which specifically includes:

[0102] Construct a Delaunay triangle, which contains all the discrete points in the discrete point set and stores them in the Delaunay triangle linked list;

[0103] Insert the points in the discrete point set into the Delaunay triangle one by one, find its circumscribed circle in the Delaunay triangle linked list, and the triangle containing the inserted point is called the influence triangle of the discrete point;

[0104] Delete the common edges of the affected triangles and connect the insertion point to all vertices of the affected triangles;

[0105] According to the optimization criteria, the newly formed local triangles are optimized and the formed triangles are stored in the Delaunay triangle linked list;

[0106] The optimization criterion is to optimize the newly formed triangle, combine two triangles with common edges into a polygon, and use the maximum empty circle criterion to determine whether the fourth vertex is inside the circumscribed circle of the triangle. If so, swap the diagonals and output the end of the local optimization process. If not, delete the newly formed triangle.

[0107] The maximum empty circle criterion is that the circumscribed circle of any triangle does not contain other points.

[0108] The advantages of Delaunay triangulation are good structure, simple data structure, low data redundancy, high storage efficiency, harmony with irregular ground features, ability to represent linear features and superimpose regional boundaries of arbitrary shapes, easy updating, adaptability to data of various distribution densities, etc. Its limitation is that the algorithm implementation is relatively complex and difficult, but there are already many mature implementation algorithms.

[0109] Reference Figure 4 As shown in the figure, the perpendicular bisectors of the edges of all triangles in the Delaunay triangulation are constructed, which are recorded as Voronoi edges. The continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone. Specifically, it includes:

[0110] All discrete points form a Delaunay triangulation network, and the discrete points and the formed triangles are numbered, and the center of the circumscribed circle of each triangle is calculated and recorded;

[0111] According to the Delaunay triangle linked list, get at most three adjacent triangles whose three sides are common to the current triangle;

[0112] Connect the circumcenter of the adjacent triangle obtained with the circumcenter of the current triangle and store them in the Voronoi edge list;

[0113] According to the Delaunay triangle linked list, obtain the edge of the current triangle that does not have any adjacent triangles, and store the perpendicular ray of the edge into the Voronoi edge linked list;

[0114] According to all the edges in the Voronoi edge list, we can get the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone.

[0115] The Voronoi diagram plays an important role in computational geometry. Due to its characteristic that the distance between the area divided by the point set is the shortest, it has a wide range of applications in geography, meteorology, crystallography, aerospace, nuclear physics, robotics and other fields, such as avoiding obstacles and finding the best path in an obstacle point set.

[0116] Reference Figure 5 As shown in the figure, the Voronoi diagram of the appearance of the large quad-rotor UAV is grid optimized, and the optimized Voronoi diagram specifically includes:

[0117] The cotangent weight of each point in the Voronoi diagram of the large quad-rotor drone is calculated according to the cotangent weight formula of the mesh surface;

[0118] Substitute the cotangent weight into the Laplace coordinate formula to obtain the Laplace coordinates of each point in the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone;

[0119] Placing the Laplace coordinates in a spatial rectangular coordinate system yields the optimized Voronoi diagram;

[0120] The formula for the cotangent weight of the mesh surface is:

[0121] ω ij =cotα+cotβ,

[0122] In the formula, ω ij is the cotangent weight of point i and its adjacent point j, α and β are the two opposite angles of the two adjacent triangles where the line connecting point i and its adjacent point j is located;

[0123] The Laplace coordinate formula is:

[0124]

[0125] In the formula, δ i is the Laplace coordinate of point i, v i ,v j are the corresponding coordinate values ​​of point i and its adjacent point j respectively.

[0126] One requirement of mesh deformation is to ensure that local detail features remain unchanged as much as possible. Conventional Cartesian coordinates only save the spatial information of vertices. Introducing Laplace coordinates can ensure that local details remain unchanged during mesh deformation.

[0127] A large quad-axis multi-rotor UAV appearance optimization design system is used to implement the large quad-axis multi-rotor UAV appearance optimization design method as described above, comprising:

[0128] The appearance information processing module is used to collect the public control point information on the appearance of the large four-axis multi-rotor UAV through surveying and mapping tools, perform data processing according to the public control point information, obtain the function model of the appearance of the large four-axis multi-rotor UAV through quadratic polynomial fitting, and randomly generate a discrete point set according to the function model of the appearance of the large four-axis multi-rotor UAV;

[0129] The appearance modeling optimization module is used to construct a Delaunay triangle, construct the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation network based on the Delaunay triangulation network, obtain the smoothness and continuity requirements of the appearance modeling of the large four-axis multi-rotor UAV, and optimize the mesh of the Voronoi diagram of the appearance of the large four-axis multi-rotor UAV according to the smoothness and continuity requirements of the appearance modeling of the large four-axis multi-rotor UAV, perform material mapping operations on the optimized Voronoi diagram, obtain the allowable error range and determine whether the final appearance modeling and the public control point information are within the allowable error range.

[0130] The appearance information processing module includes:

[0131] An information collection unit, which is used to collect public control point information on the appearance of the large four-axis multi-rotor drone through surveying and mapping tools. The public control point information includes abscissa, ordinate and vertical coordinates;

[0132] A data processing unit, the data processing unit is used to process data according to the public control point information, and obtain a function model of the appearance of the large four-axis multi-rotor UAV through quadratic polynomial fitting;

[0133] The data generation unit is used to randomly generate a discrete point set according to the function model of the appearance of the large four-axis multi-rotor drone. The horizontal coordinates of all discrete points in the discrete point set are not repeated and are arranged in ascending order according to the horizontal coordinates.

[0134] The appearance optimization module includes:

[0135] A preliminary drawing unit is used to construct a Delaunay triangle. The Delaunay triangle contains all the discrete points in the discrete point set. The points in the discrete point set are inserted into the Delaunay triangle one by one and connected to obtain a Delaunay triangulation network.

[0136] A refinement drawing unit is used to construct the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation according to the Delaunay triangulation, which are recorded as Voronoi edges. The continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone.

[0137] The optimization drawing unit is used to obtain the smoothness and continuity requirements of the appearance of the large four-axis multi-rotor UAV. According to the smoothness and continuity requirements of the appearance of the large four-axis multi-rotor UAV, the Voronoi diagram of the appearance of the large four-axis multi-rotor UAV is grid-optimized to obtain the optimized Voronoi diagram;

[0138] A material mapping unit is used to perform material mapping operations on the optimized Voronoi diagram to obtain the final appearance;

[0139] The data detection unit is used to obtain the allowable error range and determine whether the final appearance modeling and the public control point information meet the allowable error range. If so, the final appearance modeling is output; if not, the final appearance modeling optimization error is large and the final appearance modeling is recalculated.

[0140] Furthermore, the present solution also proposes a storage medium on which a computer-readable program is stored. When the computer-readable program is called, the above-mentioned large four-axis multi-rotor UAV appearance optimization design method is executed.

[0141] It is understandable that the storage medium may be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid state drive (SSD).

[0142] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for optimizing the appearance of a large quad-rotor UAV, characterized in that: include: Collecting public control point information on the exterior of the large quad-rotor drone using a surveying tool, wherein the public control point information includes abscissa, ordinate and vertical coordinates; Performing data processing according to the public control point information, and obtaining a function model of the appearance of the large quad-axis multi-rotor drone by quadratic polynomial fitting; According to the function model of the appearance of the large quad-rotor UAV, a discrete point set is randomly generated, wherein the horizontal coordinates of all discrete points in the discrete point set are not repeated and are arranged in ascending order according to the horizontal coordinates, a Delaunay triangle is constructed, and the Delaunay triangle contains all the discrete points in the discrete point set, and the points in the discrete point set are inserted into the Delaunay triangle one by one and connected to obtain a Delaunay triangulation network; According to the Delaunay triangulation, perpendicular bisectors of the sides of all triangles in the Delaunay triangulation are constructed, which are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone; Obtaining the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, and performing grid optimization on the Voronoi diagram of the appearance of the large quad-rotor UAV according to the smoothness and continuity requirements of the appearance of the large quad-rotor UAV to obtain an optimized Voronoi diagram; Performing material mapping operation on the optimized Voronoi diagram to obtain a final appearance shape; Obtain the allowable error range and determine whether the final shape and the public control point information meet the allowable error range. If so, output the final shape. If not, output that the final shape optimization error is large and recalculate the final shape.

2. The appearance optimization design method of a large quad-axis multi-rotor UAV according to claim 1 is characterized in that: The data processing according to the public control point information and obtaining the function model of the appearance of the large four-axis multi-rotor drone by quadratic polynomial fitting specifically include: Setting the kernel function formula according to the public control point information; Derived the vertical coordinate value fitting function of all points on the shape according to the kernel function formula; According to the vertical coordinate values ​​of the common control points, the undetermined coefficient matrix of the appearance model is calculated using the least square method; According to the undetermined coefficient matrix of the appearance model and the kernel function formula, the vertical coordinate value fitting function of all points on the appearance, the function model of the appearance of the large quad-axis multi-rotor drone is obtained using the matrix form of the error equation; The kernel function formula is: Q(x,y;x i ,and i )=(xx i ) 2 +(yy i ) 2 +q 2b , Where Q is the kernel function, q is the smoothing factor used to correct the kernel function formula, b is 1 / 2 or -1 / 2, when it is 1 / 2 it is called a positive shape, when it is -1 / 2 it is called an inverted shape, (x i ,y i ) are the horizontal and vertical coordinates of the common control point, and x, y are the horizontal and vertical coordinates of the unknown point on the shape; The fitting function of the vertical coordinate values ​​of all points on the shape is: In the formula, H is the vertical coordinate value fitting function, a i is the unknown coefficient, n is the sum of the appearances; The undetermined coefficient matrix of the appearance model is: A=(Q(x,y;x i ,and i ) T Q(x,y;x i ,and i )) -1 Q(x,y;x i ,and i ) T H(x,y) Where A is the unknown coefficient matrix of the appearance model; The matrix form of the error equation is: V=Q(x,y;x i ,y i )A-H(x,y), Where V is the matrix form of the error equation.

3. The appearance optimization design method of a large quad-axis multi-rotor UAV according to claim 2 is characterized in that: The step of inserting the points in the discrete point set into the Delaunay triangle one by one and connecting the lines to obtain the Delaunay triangulation specifically includes: Construct a Delaunay triangle, wherein the Delaunay triangle includes all discrete points in the discrete point set and stores them in a Delaunay triangle linked list; Insert the points in the discrete point set into the Delaunay triangle one by one, find its circumscribed circle in the Delaunay triangle linked list, and the triangle containing the inserted point is called the influence triangle of the discrete point; Delete the common edges of the affected triangles and connect the insertion point to all vertices of the affected triangles; According to the optimization criteria, the newly formed local triangles are optimized and the formed triangles are stored in the Delaunay triangle linked list; The optimization criterion is to optimize the newly formed triangle, combine two triangles with common sides into a polygon, and use the maximum empty circle criterion to determine whether the fourth vertex is inside the circumscribed circle of the triangle. If so, the diagonal lines are swapped and the local optimization process is output to end. If not, the newly formed triangle is deleted. The maximum empty circle criterion is that the circumscribed circle of any triangle does not contain other points.

4. The appearance optimization design method of a large quad-axis multi-rotor UAV according to claim 3 is characterized in that: The perpendicular bisectors of the sides of all triangles in the Delaunay triangulation are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone. Specifically, it includes: All discrete points form a Delaunay triangulation network, and the discrete points and the formed triangles are numbered, and the center of the circumscribed circle of each triangle is calculated and recorded; According to the Delaunay triangle linked list, get at most three adjacent triangles whose three sides are common to the current triangle; Connect the circumcenter of the adjacent triangle obtained with the circumcenter of the current triangle and store them in the Voronoi edge list; According to the Delaunay triangle linked list, obtain the edge of the current triangle that does not have any adjacent triangles, and store the perpendicular ray of the edge into the Voronoi edge linked list; The Voronoi diagram of the appearance of the large quad-rotor drone is obtained based on all the edges in the Voronoi edge list.

5. The appearance optimization design method of a large quad-rotor UAV according to claim 4 is characterized in that: The obtaining of the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, and performing grid optimization on the Voronoi diagram of the appearance of the large quad-rotor UAV according to the smoothness and continuity requirements of the appearance of the large quad-rotor UAV, to obtain the optimized Voronoi diagram specifically includes: The cotangent weight of each point in the Voronoi diagram of the large quad-rotor drone is calculated according to the cotangent weight formula of the mesh surface; Substituting the cotangent weight into the Laplace coordinate formula, the Laplace coordinates of each point in the Voronoi diagram of the appearance of the large quad-rotor drone are obtained; Placing the Laplace coordinates in a spatial rectangular coordinate system to obtain an optimized Voronoi diagram; The mesh surface cotangent weight formula is: ω ij =cotα+cotβ, In the formula, ω ij is the cotangent weight of point i and its adjacent point j, α and β are the two opposite angles of the two adjacent triangles where the line connecting point i and its adjacent point j is located; The Laplace coordinate formula is: In the formula, δ i is the Laplace coordinate of point i, v i ,v j are the corresponding coordinate values ​​of point i and its adjacent point j respectively.

6. A system for optimizing the appearance of a large quad-rotor UAV, used to implement the method for optimizing the appearance of a large quad-rotor UAV as claimed in any one of claims 1 to 5, characterized in that: include: An appearance information processing module, the appearance information processing module is used to collect public control point information on the appearance of the large four-axis multi-rotor drone through a surveying and mapping tool, perform data processing according to the public control point information, obtain a function model of the appearance of the large four-axis multi-rotor drone through quadratic polynomial fitting, and randomly generate a discrete point set according to the function model of the appearance of the large four-axis multi-rotor drone; An appearance modeling optimization module, the appearance modeling optimization module is used to construct a Delaunay triangle, construct the perpendicular bisectors of the sides of all triangles in the Delaunay triangulation network based on the Delaunay triangulation network, obtain the appearance modeling smoothness and continuity requirements of the large four-axis multi-rotor UAV, according to the appearance modeling smoothness and continuity requirements of the large four-axis multi-rotor UAV, mesh optimize the Voronoi diagram of the appearance of the large four-axis multi-rotor UAV, perform material mapping operations on the optimized Voronoi diagram, obtain the allowable error range and determine whether the final appearance modeling and the common control point information are within the allowable error range.

7. The appearance optimization design system of a large quad-axis multi-rotor UAV according to claim 6 is characterized in that: The appearance information processing module includes: An information collection unit, the information collection unit is used to collect public control point information on the appearance of the large four-axis multi-rotor drone through a surveying and mapping tool, the public control point information includes abscissa, ordinate and vertical coordinate; A data processing unit, the data processing unit is used to perform data processing according to the public control point information, and obtain a function model of the appearance of the large four-axis multi-rotor drone through quadratic polynomial fitting; A data generation unit is used to randomly generate a discrete point set according to a function model of the appearance of the large four-axis multi-rotor drone, wherein the horizontal coordinates of all discrete points in the discrete point set are not repeated and are arranged in ascending order according to the horizontal coordinates.

8. The appearance optimization design system of a large quad-axis multi-rotor UAV according to claim 7, characterized in that: The appearance optimization module includes: A preliminary drawing unit, wherein the preliminary drawing unit is used to construct a Delaunay triangle, wherein the Delaunay triangle includes all discrete points in the discrete point set, and the points in the discrete point set are inserted into the Delaunay triangle one by one and connected to obtain a Delaunay triangulation network; A refinement drawing unit, wherein the refinement drawing unit is used to construct perpendicular bisectors of the sides of all triangles in the Delaunay triangulation according to the Delaunay triangulation, which are recorded as Voronoi edges, and the continuous polygons formed by the Voronoi edges are recorded as the Voronoi diagram of the appearance of the large quadcopter multi-rotor drone; An optimization drawing unit, wherein the optimization drawing unit is used to obtain the smoothness and continuity requirements of the appearance of the large four-axis multi-rotor drone, and according to the smoothness and continuity requirements of the appearance of the large four-axis multi-rotor drone, a grid optimization is performed on the Voronoi diagram of the appearance of the large four-axis multi-rotor drone to obtain an optimized Voronoi diagram; A material mapping unit, the material mapping unit is used to perform a material mapping operation on the optimized Voronoi diagram to obtain a final appearance shape; A data detection unit is used to obtain an allowable error range and determine whether the final shape modeling and the public control point information meet the allowable error range. If so, the final shape modeling is output; if not, the final shape modeling optimization error is large and the final shape modeling is recalculated.

Images