Path planning method and system for navigation guidance of low-altitude aircraft

Abstract

The invention relates to a path planning method and system for navigation guidance of a low-altitude aircraft, and the method comprises the steps: calculating a probability density value of pre-processed prior flight data based on kernel density estimation, mapping the normalized probability density value to a grid map of a rasterized environment model, generating a flight path safety experience distribution map, and carrying out the navigation guidance of the low-altitude aircraft. And expanding the boundary of the obstacle area based on the safety distance constraint matrix, generating a restricted area matrix, generating an initial path according to the bidirectional A *-KDE algorithm cost function, the flight line safety experience distribution map and the restricted area matrix, removing redundant nodes from the initial path, and carrying out smoothing processing to obtain a flight route. The flight route is obtained by using the prior flight data, the deviation between the flight route and the actual navigation path can be reduced, the obstacle area boundary is expanded based on the safety distance constraint matrix, the obstacle safety distance is fully considered, and the risk that the low-altitude aircraft collides with the obstacle can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of flight control for low-altitude aircraft, and particularly to a path planning method and system for guiding the navigation of low-altitude aircraft. Background Art

[0002] Path planning for low-altitude aircraft is to plan an optimal or safe path from the starting point to the ending point for the low-altitude aircraft to ensure that the low-altitude aircraft can complete the task safely and efficiently. In the related low-altitude aircraft path planning technology, the deviation between the planned path and the actual navigation path is significant. And the planned path is prone to approaching obstacles, with the risk of the low-altitude aircraft colliding with obstacles. Summary of the Invention

[0003] The present invention provides a path planning method and system for guiding the navigation of low-altitude aircraft, aiming to solve at least one of the technical problems existing in the prior art.

[0004] The technical solution of the present invention is a path planning method for guiding the navigation of low-altitude aircraft, including:

[0005] Obtaining the prior flight data of the low-altitude aircraft in the target area and preprocessing the prior flight data;

[0006] Based on kernel density estimation, calculating the probability density value for the preprocessed prior flight data, normalizing the probability density value, and mapping the normalized probability density value into the grid map of the rasterized environment model to generate a flight route safety experience distribution map;

[0007] Expanding the boundary of the obstacle area based on the safety distance constraint matrix to generate a restricted area matrix including a safety buffer zone;

[0008] Generating an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix;

[0009] Removing redundant nodes from the initial path and performing smoothing processing to obtain a flight route.

[0010] According to some embodiments of the present invention, preprocessing the prior flight data includes:

[0011] Performing data cleaning and data preprocessing on the prior flight data, and filling in the missing sampling points by the cubic spline interpolation method.

[0012] According to some embodiments of the present invention, based on kernel density estimation, calculating a probability density value for the preprocessed prior flight data, normalizing the probability density value, and mapping the normalized probability density value into the grid map of the rasterized environment model to generate a flight route safety experience distribution map, including:

[0013] Mapping the trajectory points of the preprocessed prior flight data into the three-dimensional cubes of the rasterized environment model, and calculating the probability density value of the three-dimensional cube based on the kernel density estimation;

[0014] Normalizing the probability density value, and mapping the normalized probability density value into the grid map of the rasterized environment model to generate the flight route safety experience distribution map.

[0015] According to some embodiments of the present invention, the kernel density estimation is expressed as:

[0016]

[0017] In the formula, p(G) is the probability density value of the prior flight data at the position of the three-dimensional cube G, n is the number of trajectory points in the dataset of the prior flight data, h is the bandwidth, d is the dimension of the data, K(·) is the kernel function, and X i is the i-th trajectory point in the dataset of the prior flight data;

[0018] The normalization process is expressed as:

[0019]

[0020] In the formula, d norm (G) is the probability density value after normalization of the prior flight data at the three-dimensional cube G, and p max = max(p(G)) is the maximum value of the probability density value.

[0021] According to some embodiments of the present invention, generating an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix includes:

[0022] Generating an initial path node according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix;

[0023] Adopting a bidirectional search mechanism to synchronously perform forward and backward path searches on the initial path node, and generating the initial path when the forward and backward paths meet.

[0024] According to some embodiments of the present invention, removing redundant nodes from the initial path and performing smoothing processing to obtain a flight route, including:

[0025] Using a line-of-sight detection algorithm to remove the redundant nodes from the initial path;

[0026] Using a Bessel curve to perform smoothing processing on the initial path after removing the redundant nodes to obtain the flight route.

[0027] According to some embodiments of the present invention, the using a line-of-sight detection algorithm to remove the redundant nodes from the initial path includes:

[0028] Based on the line-of-sight detection algorithm, extracting nodes from the initial path;

[0029] Starting from the first node, traversing each node, and determining whether the line of sight between the current node and the subsequent node is blocked by an obstacle or a dangerous node that does not satisfy the safety distance constraint matrix;

[0030] If the line of sight between the current node and the subsequent node is unblocked, then

[0031] Connecting the current node and the subsequent node to obtain a connection line;

[0032] Calculating the vertical distance from the intermediate original node between the current node and the subsequent node to the connection line to obtain an offset distance;

[0033] Determining whether the offset distance exceeds a deviation threshold;

[0034] If the offset distance does not exceed the deviation threshold, then taking the intermediate original node as the redundant node, connecting the current node and the subsequent node, and removing the redundant node; otherwise, giving up connecting the current node and the subsequent node and retaining the intermediate original node.

[0035] According to some embodiments of the present invention, if the offset distance does not exceed the deviation threshold, it is expressed as:

[0036]

[0037] In the formula, δ is the deviation threshold.

[0038] According to some embodiments of the present invention, the cost function of the bidirectional A*-KDE algorithm is expressed as:

[0039]

[0040] Wherein, g(n) is the actual cost of the path, h(n) is the heuristic cost, P(n) is the cost of the kernel density potential field, F(neighbor(n)) is the cost function of the bidirectional A*-KDE algorithm, and α is the weight parameter of the cost of the kernel density potential field.

[0041] The technical solution of the present invention also relates to an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements a path planning method for low-altitude aircraft navigation guidance as described above.

[0042] The technical solution of the present invention also relates to a computer system, which includes a storage medium. The storage medium stores a computer program, and when the computer program is executed by a processor, it implements a path planning method for low-altitude aircraft navigation guidance as described above.

[0043] The beneficial effects of the present invention include: obtaining the prior flight data of low-altitude aircraft in the target area, preprocessing the prior flight data, then calculating the probability density value of the preprocessed prior flight data based on kernel density estimation, normalizing the probability density value, mapping the normalized probability density value into the grid map of the rasterized environment model to generate a flight route safety experience distribution map, expanding the boundary of the obstacle area based on the safety distance constraint matrix to generate a restricted area matrix including a safety buffer zone, generating an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map and the restricted area matrix, removing redundant nodes from the initial path and smoothing it to obtain a flight route. Using the prior flight data of low-altitude aircraft to obtain the flight route can reduce the deviation between the flight route and the actual navigation path, and expanding the boundary of the obstacle area based on the safety distance constraint matrix fully considers the safety distance of the obstacle, which can reduce the risk of low-altitude aircraft colliding with obstacles.

[0044] In addition, the additional aspects and advantages of the present invention will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present invention. Description of the Drawings

[0045] Figure 1 is an alternative flowchart of a path planning method for low-altitude aircraft navigation guidance in an embodiment of the present invention.

[0046] Figure 2 is an alternative flowchart of generating a flight route safety experience distribution map based on kernel density estimation in an embodiment of the present invention.

[0047] Figure 3It is an optional flowchart of obtaining a flight route according to an initial path in an embodiment of the present invention. Detailed implementation manners

[0048] The concept, specific structure and technical effects of the present invention will be clearly and completely described below in conjunction with embodiments and drawings to fully understand the purpose, solution and effects of the present invention. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0049] It should be noted that, unless otherwise specified, when a feature is referred to as "fixed" or "connected" to another feature, it can be directly fixed or connected to the other feature, or indirectly fixed or connected to the other feature. In addition, the up, down, left, right, top, bottom, etc. used in the present invention are only relative to the mutual positional relationship of the components of the present invention in the drawings.

[0050] In addition, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of the present invention. The terms used in the description of the present specification are only for describing specific embodiments, rather than for limiting the present invention. The term "and / or" used herein includes any combination of one or more of the related listed items.

[0051] It should be understood that although terms such as first, second, and third may be used in the present invention to describe various elements, these elements should not be limited to these terms. These terms are only used to distinguish elements of the same type from each other. For example, without departing from the scope of the present invention, the first element may also be referred to as the second element, and similarly, the second element may also be referred to as the first element.

[0052] Refer to Figures 1 to 3 , in some embodiments, the technical solution of the present invention is a path planning method for low-altitude aircraft navigation guidance, including but not limited to steps 101 to 105, and each step will be introduced below in sequence.

[0053] Step 101: Obtain the prior flight data of the low-altitude aircraft in the target area and preprocess the prior flight data.

[0054] In a specific embodiment, the low-altitude aircraft is a drone. Obtain the prior flight data of the drone in the target area and preprocess the prior flight data.

[0055] In some embodiments, before obtaining the prior flight data of the low-altitude aircraft in the target area and preprocessing the prior flight data, the path planning method for low-altitude aircraft navigation guidance further includes rasterizing the target area to build a rasterized environment model.

[0056] Specifically, for rasterized space modeling, rasterization divides the target area into 100×100 grids. Here, in addition to 100, other positive integers can also be used, and the specific value is selected according to the actual situation, which is not specifically limited in the present invention. Spatial rasterization is to divide three-dimensional cubes, that is, to divide the target area into multiple three-dimensional cubes.

[0057] In a specific embodiment, before rasterized space modeling of the target area to obtain a rasterized environment model, the path planning method for low-altitude aircraft navigation further includes confirming the target area based on the starting point and ending point of the low-altitude aircraft flight.

[0058] In a specific embodiment, the target area is rasterized space modeled using a three-dimensional cube division method to obtain a rasterized environment model.

[0059] In a specific embodiment, the actual map data of the target area is obtained, and the actual map data of the target area is processed using gray-scale image processing technology to generate a map after gray-scale processing. Then, the map after gray-scale processing is rasterized, and the obstacle information in the rasterized processed map is mapped into the rasterized environment model.

[0060] Specifically, in the rasterized environment model, the black three-dimensional cube represents the position and size of the obstacle, and the white three-dimensional cube represents the area where the low-altitude aircraft can pass. The prior flight data is historical flight data.

[0061] In some embodiments, preprocessing the prior flight data includes: performing data cleaning and data preprocessing on the prior flight data, and filling in the missing sampling points by cubic spline interpolation. It can be understood that performing data cleaning and data preprocessing on the prior flight data optimizes the quality of the prior flight data and improves the accuracy of the flight route.

[0062] In a specific embodiment, performing data cleaning and data preprocessing on the prior flight data includes: standardizing the formats of the timestamp, longitude and latitude, flight speed, and heading fields of the prior flight data; removing abnormal position information and abnormal data points of speed and heading.

[0063] Step 102: Based on kernel density estimation, calculate the probability density value for the preprocessed prior flight data, perform normalization processing on the probability density value, and map the normalized probability density value into the grid map of the rasterized environment model to generate a flight route safety experience distribution map.

[0064] Specifically, Kernel Density Estimation (KDE) is a non-parametric method for estimating probability density functions. It should be noted that the probability density values after normalization are mapped into each three-dimensional cube of the rasterized environmental model's raster map. Among them, the probability density value represents safety experience, and the larger the probability density value, the safer it is. The raster map contains probability density values, forming a safety experience distribution map of flight routes.

[0065] Refer to Figure 2 , in some embodiments, based on kernel density estimation, calculate the probability density value for the preprocessed prior flight data, perform normalization processing on the probability density value, and map the normalized probability density value into the raster map of the rasterized environmental model to generate a safety experience distribution map of flight routes, including but not limited to the following steps 201 to 202.

[0066] Step 201: Map the trajectory points of the preprocessed prior flight data into the three-dimensional cubes of the rasterized environmental model, and calculate the probability density value of the three-dimensional cube based on kernel density estimation.

[0067] Step 202: Perform normalization processing on the probability density value, and map the normalized probability density value into the raster map of the rasterized environmental model to generate a safety experience distribution map of flight routes.

[0068] Specifically, the coordinates of the prior flight data are three-dimensional coordinates. The probability density value of the low-altitude aircraft flight route is statistically analyzed with the three-dimensional cube as the basic unit. If the trajectory points near the three-dimensional cube are concentrated, the probability density value of the three-dimensional cube is higher; if there are fewer trajectory points near the three-dimensional cube or the three-dimensional cube is far from the trajectory points, the probability density value of the three-dimensional cube is lower.

[0069] In some embodiments, the kernel density estimation is expressed as:

[0070]

[0071] In the formula, p(G) is the probability density value of the prior flight data at the position of the three-dimensional cube G, n is the number of trajectory points in the dataset of the prior flight data, h is the bandwidth, d is the dimension of the data, K(·) is the kernel function, and X i is the i-th trajectory point in the dataset of the prior flight data;

[0072] The normalization processing is expressed as:

[0073]

[0074] In the formula, d norm (G) is the probability density value of the prior flight data at the three-dimensional cube G after normalization processing, p max= max(p(G)) is the maximum value of the probability density.

[0075] Specifically, X is the currently observed trajectory point, which is the point for which the density needs to be estimated. K(·) is the kernel function, and a Gaussian kernel can also be used. After using the Gaussian kernel function, the expression form of kernel density estimation is:

[0076]

[0077] In the formula, σ is the potential field diffusion coefficient, exp(·) is the exponential function. For example, exp(m) represents the m-th power of the base e of the natural logarithm. T represents the transpose operation, and (X - X i ) T is the transpose of the vector (X - X i ), and the transpose is to convert a column vector into a row vector.

[0078] Step 103: Expand the boundary of the obstacle area based on the safety distance constraint matrix to generate a restricted area matrix including a safety buffer.

[0079] In a specific embodiment, the safety distance constraint matrix A nn is a matrix of size n×n, which is intended to expand the obstacle area in the flight path to form a restricted area matrix including a safety buffer. Specifically, by a preset safety distance d safe adjust the size of the safety distance constraint matrix A nn to ensure that during the node expansion process, only those nodes that maintain a preset safety distance from the obstacles are selected for expansion. The relationship between the safety distance constraint matrix A nn and the preset safety distance d safe is expressed as:

[0080] A nn , n = 2·d safe + 1,

[0081] When d safe takes 1, the grid units around the obstacle will not satisfy the safety distance constraint.

[0082] Specifically, after the target area is rasterized, multiple three-dimensional cubes are formed. The obstacles will not be selected for path expansion. When d safe takes 1, a three-dimensional cube adjacent to the obstacle will not be used as a node for path selection either, so that the generated path has a certain distance from the obstacle.

[0083] Step 104: Generate an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix.

[0084] In a specific embodiment, a cost function of the bidirectional A*-KDE algorithm is designed to fuse the actual cost of the path, the heuristic cost, and the cost of the kernel density potential field, guiding the path to preferentially select high-security-density areas for path search. Specifically, the bidirectional A*-KDE algorithm means a combination of the A* algorithm and the KDE algorithm.

[0085] In some embodiments, an initial path is generated according to the cost function of the bidirectional A*-KDE algorithm, the safety experience distribution map of the flight route, and the restricted area matrix, including:

[0086] Generating an initial node of the path according to the cost function of the bidirectional A*-KDE algorithm, the safety experience distribution map of the flight route, and the restricted area matrix;

[0087] Adopting a bidirectional search mechanism to perform forward and backward path searches on the initial node of the path synchronously, and generating an initial path when the forward and backward paths meet.

[0088] Specifically, when the forward and backward paths meet, a path backtracking algorithm is used to generate the initial path, and the coordinates and index information of the meeting node are recorded. The bidirectional search mechanism significantly shortens the calculation time of path planning.

[0089] Step 105: Removing redundant nodes from the initial path and performing smoothing processing to obtain a flight route.

[0090] Referring to Figure 3 , in some embodiments, removing redundant nodes from the initial path and performing smoothing processing to obtain a flight route includes, but is not limited to, the following steps 301 to 302.

[0091] Step 301: Using the line-of-sight detection algorithm to remove redundant nodes from the initial path.

[0092] Step 302: Using a Bezier curve to perform smoothing processing on the initial path after removing redundant nodes to obtain a flight route.

[0093] In a specific embodiment, the path curvature continuity of the initial path after removing redundant nodes is optimized through Bezier curve smoothing processing.

[0094] Specifically, the basic idea of the line-of-sight detection (LOS) algorithm is to determine whether there is a straight-line path without object obstruction between an observation point and a target point. The line-of-sight detection algorithm guides the system to adjust the heading and speed by calculating the virtual line between the current position and the target path and the angle deviation between this virtual line and the forward direction, so as to stay on the predetermined path. Using a Bezier curve to perform smoothing optimization on the initial path after removing redundant nodes improves the continuity, stability, and safety of the flight route.

[0095] In some embodiments, a line-of-sight detection algorithm is used to remove redundant nodes from the initial path, including:

[0096] Based on the line-of-sight detection algorithm, extract nodes from the initial path;

[0097] Starting from the first node, traverse each node and determine whether the line of sight between the current node and the subsequent node is blocked by an obstacle or a dangerous node that does not satisfy the safety distance constraint matrix;

[0098] If the line of sight between the current node and the subsequent node is unblocked, then

[0099] Connect the current node and the subsequent node to obtain a connection line;

[0100] Calculate the vertical distance from the intermediate original node between the current node and the subsequent node to the connection line to obtain an offset distance;

[0101] Determine whether the offset distance exceeds the deviation threshold;

[0102] If the offset distance does not exceed the deviation threshold, then use the intermediate original node as a redundant node, connect the current node and the subsequent node, and remove the redundant node; otherwise, abandon connecting the current node and the subsequent node and retain the intermediate original node.

[0103] In a specific embodiment, if the line of sight between the current node and the subsequent node is blocked, it is impossible to further connect the current node and the subsequent node, and the operation of removing redundant nodes is terminated. At this time, according to the actual situation, it can be determined whether to perform the next smoothing operation, or to return to the previous step and regenerate the initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix.

[0104] In a specific embodiment, when traversing to the last node, the operation of removing redundant nodes is terminated.

[0105] Specifically, a dangerous node that does not satisfy the safety distance constraint matrix is a node whose distance from an obstacle does not satisfy the safety distance constraint matrix, and it is impossible to expand the boundary of the obstacle area to include a safety buffer zone for the node. It can be understood that removing redundant nodes from the initial path optimizes the length of the flight route.

[0106] In some embodiments, if the offset distance does not exceed the deviation threshold, it is expressed as:

[0107]

[0108] In the formula, δ is the deviation threshold, E, F, and C are the normal vectors of the three dimensions of the initial path after removing redundant nodes, D is a constant, x i , y i , z iThey are the x, y, and z coordinates of the intermediate original node respectively. Specifically, this formula is a constraint formula that restricts the maximum offset distance between the path optimized by the LOS algorithm and the initial path.

[0109] In some embodiments, the cost function of the bidirectional A*-KDE algorithm is expressed as:

[0110]

[0111] In the formula, g(n) is the actual cost of the path, h(n) is the heuristic cost, P(n) is the cost of the kernel density potential field, F(neighbor(n)) is the cost function of the bidirectional A*-KDE algorithm, α is the weight parameter of the cost of the kernel density potential field, d safe is the preset safety distance, and neighbor(n) represents the neighboring nodes of the current node n.

[0112] Specifically, the actual cost of the path represents the length of the path from the starting point to the current node, the heuristic cost represents the estimated cost from the current node to the target node, the cost of the kernel density potential field represents the cost of the potential field generated by kernel density estimation, and the weight parameter of the cost of the kernel density potential field is used to adjust the balance between the path length and safety.

[0113] In a specific embodiment, the actual cost of the path g(n) is expressed as:

[0114]

[0115] In the formula, pos x (n), pos y (n) and pos z (n) are the x, y, and z coordinates of the current node n respectively, and parent(n) represents the parent node of the current node n.

[0116] In a specific embodiment, the heuristic cost h(n) is used to estimate the shortest path cost from the current node to the target node. Combining the Manhattan distance and the diagonal distance, it is expressed as:

[0117]

[0118] In the formula, dx = |pos x (goal) - pos x (n)|, dy = |pos y (goal) - pos y (n)|, dz = |pos z (goal) - pos z (n)|, and goal represents the position of the target node.

[0119] In a specific embodiment, the kernel density potential field cost P(n) guides the path to avoid low-density regions through the normalized probability density value of kernel density estimation. The exponential form of the Gaussian distribution can achieve rapid decay of density, effectively amplify the advantages of high-density regions, suppress the risk of path selection in low-density regions, and avoid excessive fluctuations numerically, which is helpful for the stability of path planning. The controllable parameters A and σ enable the path planning method for low-altitude aircraft navigation to flexibly adapt to different navigation scenarios. The design of the potential field cost function adopts the form of Gaussian distribution and is expressed as:

[0120]

[0121] In the formula, exp(·) is the exponential function, A is the potential field intensity coefficient, which controls the influence degree of the density field on the path. In high-risk scenarios, A can be increased to strengthen the penalty effect of low-density regions. σ is the potential field diffusion coefficient, which determines the range of influence of the density field. A smaller σ limits the influence range of high density and emphasizes locality; while a larger σ expands the influence range of density and is suitable for more global path search requirements. d(n) is the normalized probability density value at position n, which is calculated by KDE.

[0122] In a specific embodiment, the Bézier curve can smoothly transition the path by adjusting the control points and is expressed as:

[0123]

[0124] In the formula, B(t) is the Bézier curve, t represents a parameter value on the Bézier curve, which is used to control the position of the Bézier curve. The expression of the Bézier curve interpolates within the interval [0,1] of the Bézier curve using this parameter t.

[0125] Specifically, n is the highest degree of the Bézier curve, which is equal to the number of control points minus 1. For example, if the Bézier curve has 4 control points, then n = 3, which is called a cubic Bézier curve. In, P i is the i-th control point; is the binomial coefficient, which is a mathematical calculation, that is

[0126]

[0127] It should be noted that the Bézier Curve is a vector curve defined by control points.

[0128] Among them, control points: In a Bézier curve, control points are the points that define the shape of the Bézier curve. The Bézier curve does not pass directly through these control points, but they control the bending and path of the Bézier curve. The starting control point and the ending control point are the endpoints of the Bézier curve. The Bézier curve starts from the starting control point and finally reaches the ending control point. Intermediate control points (for quadratic or higher-order Bézier curves) determine the degree and direction of bending of the Bézier curve. By adjusting the positions of these control points, the shape of the Bézier curve can be changed. For example, for a quadratic Bézier curve, there are three control points: a starting point, an ending point, and an intermediate control point. For a cubic Bézier curve, there are four control points: a starting point, an ending point, and two intermediate control points.

[0129] In a specific embodiment, the initial path after removing redundant nodes is segmented to obtain segmented paths. At the same time, in combination with the minimum safety distance detection, it is ensured that the segmented paths meet the minimum safety distance requirements; the order of the Bézier curve is confirmed according to the number of path turning points of the segmented paths; if the segmented paths cross obstacles and dangerous nodes, the control points of the Bézier curve are increased; the segmented paths are smoothed and optimized using the Bézier curve, and the smoothed and optimized segmented paths are connected to obtain a flight route. Among them, the minimum safety distance is the deviation threshold δ.

[0130] In a specific embodiment, path initial nodes are generated according to the cost function of the bidirectional A*-KDE algorithm, the safety experience distribution map of the flight route, and the restricted area matrix, including:

[0131] Traverse each node. The cost function of the bidirectional A*-KDE algorithm comprehensively considers the actual cost of the path, the heuristic distance, and the potential influence of the density field. It fuses the actual cost of the path, the heuristic cost, and the potential cost of the kernel density field, guides the path search to preferentially select high-safety-density areas of the safety experience distribution map of the flight route, and strictly restricts the expansion direction of the path points through the restricted area matrix to restrict the selection of dangerous nodes that do not meet the safety distance constraint matrix. Finally, path initial nodes are obtained, ensuring the safety and rationality of the path search.

[0132] It can be understood that the safety experience distribution map of the flight route is generated using kernel density estimation, and the cost function of the bidirectional A*-KDE algorithm is used to guide the path search to preferentially select high-safety-density areas of the safety experience distribution map of the flight route, narrowing the path search range, reducing the computational complexity, and being able to avoid potential risk areas.

[0133] In a possible implementation manner, the path planning method for low-altitude aircraft navigation guidance is applied to the main control unit of a low-altitude aircraft and works in coordination with the flight control system to achieve precise control of the flight state of the low-altitude aircraft and efficient execution of tasks.

[0134] In a specific implementation, the path planning method for low-altitude aircraft navigation guidance is applied to the main control unit of an unmanned aerial vehicle (UAV) and works in coordination with the flight control system to achieve precise control of the UAV's flight state and efficient execution of tasks.

[0135] It can be seen that by performing grid-based spatial modeling on the target area to obtain a grid-based environmental model, acquiring the prior flight data of low-altitude aircraft in the target area, preprocessing the prior flight data, calculating the probability density value of the preprocessed prior flight data based on kernel density estimation, normalizing the probability density value, mapping the normalized probability density value to the grid map of the grid-based environmental model to generate a flight route safety experience distribution map, expanding the boundary of the obstacle area based on the safety distance constraint matrix to generate a restricted area matrix including a safety buffer zone, generating an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix, and removing redundant nodes and smoothing the initial path to obtain a flight route. Using the prior flight data of low-altitude aircraft to obtain the flight route can reduce the deviation between the flight route and the actual navigation path, and expanding the boundary of the obstacle area based on the safety distance constraint matrix fully considers the safety distance of obstacles and can reduce the risk of low-altitude aircraft colliding with obstacles.

[0136] An embodiment of the present invention also provides an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the above-mentioned path planning method for low-altitude aircraft navigation guidance. This electronic device can be any intelligent terminal including a computer, etc.

[0137] An embodiment of the present invention also provides a computer system. The computer system includes a storage medium that stores a computer program, and when the computer program is executed by a processor, it implements the above-mentioned path planning method for low-altitude aircraft navigation guidance.

[0138] It should be recognized that the method steps in the embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or computer instructions stored in a non-transitory computer-readable memory. The method can use standard programming techniques. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if necessary, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, for this purpose, the program can run on a dedicated integrated circuit programmed for this purpose.

[0139] In addition, the operations of the processes described herein can be performed in any suitable order, unless otherwise indicated herein or otherwise clearly contradicted by the context. The processes described herein (or variations and / or combinations thereof) can be performed under the control of one or more computer systems configured with executable instructions and can be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) collectively executed on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions executable by one or more processors.

[0140] Further, the method can be implemented in any type of computing platform operably connected to a suitable one, including but not limited to personal computers, minicomputers, mainframes, workstations, network or distributed computing environments, separate or integrated computer platforms, or communicating with charged particle tools or other imaging devices, etc. Aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into the computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it can be read by a programmable computer and can be used to configure and operate the computer to perform the processes described herein when the storage medium or device is read by the computer. In addition, the machine-readable code, or portions thereof, can be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the above-described steps in combination with a microprocessor or other data processor, the inventions described herein include these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention can also include the computer itself.

[0141] The computer program can be applied to input data to perform the functions described herein, thereby transforming the input data to generate output data stored in non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the transformed data represents physical and tangible objects, including a specific visual depiction of the physical and tangible objects generated on the display.

[0142] As described above, these are only the preferred embodiments of the present invention. The present invention is not limited to the above-described embodiments. As long as it achieves the technical effects of the present invention by the same means, any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention. Within the scope of protection of the present invention, its technical solutions and / or implementation manners can have various different modifications and changes.

Claims

1. A path planning method for low-altitude aircraft navigation guidance, characterized in that, Including: Obtain the prior flight data of low-altitude aircraft in the target area, and preprocess the prior flight data; Based on kernel density estimation, calculate the probability density value for the preprocessed prior flight data, perform normalization processing on the probability density value, and map the normalized probability density value into the grid map of the rasterized environment model to generate a flight route safety experience distribution map; Expand the boundary of the obstacle area based on the safety distance constraint matrix to generate a restricted area matrix including a safety buffer; Generate an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix; Remove redundant nodes from the initial path and perform smoothing processing to obtain a flight route.

2. A path planning method for low-altitude aircraft navigation guidance according to claim 1, characterized in that, Preprocessing the prior flight data includes: Perform data cleaning and data preprocessing on the prior flight data, and fill in the missing sampling points by cubic spline interpolation.

3. A path planning method for low-altitude aircraft navigation guidance according to claim 1, characterized in that, The method of calculating the probability density value for the preprocessed prior flight data based on kernel density estimation, performing normalization processing on the probability density value, and mapping the normalized probability density value into the grid map of the rasterized environment model to generate a flight route safety experience distribution map includes: Map the trajectory points of the preprocessed prior flight data into the three-dimensional cubes of the rasterized environment model, and calculate the probability density value of the three-dimensional cube based on the kernel density estimation; Perform normalization processing on the probability density value, and map the normalized probability density value into the grid map of the rasterized environment model to generate the flight route safety experience distribution map.

4. A path planning method for low-altitude aircraft navigation guidance according to claim 3, characterized in that The kernel density estimation is expressed as: Wherein, p(G) is the probability density value of the prior flight data at the position of the cubic block G, n is the number of trajectory points in the dataset of the prior flight data, h is the bandwidth, d is the dimension of the data, K(·) is the kernel function, and X i is the i-th trajectory point in the dataset of the prior flight data; The normalization processing is expressed as: where d norm (G) is the probability density value after the normalization process of the prior flight data at the three-dimensional cube G, p max = max(p(G)) is the maximum value of the probability density value.

5. A path planning method for low-altitude aircraft navigation guidance according to claim 1, characterized in that, The method of generating an initial path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix includes: Generate an initial node of the path according to the cost function of the bidirectional A*-KDE algorithm, the flight route safety experience distribution map, and the restricted area matrix; Adopt a bidirectional search mechanism to synchronously perform forward and backward path searches on the initial path node. When the forward and backward paths meet, generate the initial path.

6. A path planning method for low-altitude aircraft navigation guidance according to claim 1, characterized in that, Removing redundant nodes from the initial path and performing smoothing processing to obtain a flight route includes: Use the line-of-sight detection algorithm to remove the redundant nodes from the initial path; Use a Bessel curve to smooth the initial path after removing the redundant nodes to obtain the flight route.

7. A path planning method for low-altitude aircraft navigation guidance according to claim 6, characterized in that, The method of using the line-of-sight detection algorithm to remove the redundant nodes from the initial path includes: Extract nodes from the initial path based on the line-of-sight detection algorithm; Starting from the first node, traverse each node, and judge whether the line of sight between the current node and the subsequent node is blocked by an obstacle or a dangerous node that does not meet the safety distance constraint matrix; If the line of sight between the current node and the subsequent node is unblocked, then Connect the current node and the subsequent node to obtain a connection line; Calculate the vertical distance from the intermediate original node between the current node and the subsequent node to the connection line to obtain an offset distance; Determine whether the offset distance exceeds a deviation threshold; If the offset distance does not exceed the deviation threshold, use the intermediate original node as the redundant node, connect the current node and the subsequent node, and remove the redundant node; otherwise, abandon connecting the current node and the subsequent node and retain the intermediate original node.

8. A path planning method for low-altitude aircraft navigation guidance according to claim 7, characterized in that, If the offset distance does not exceed the deviation threshold, it is expressed as: In the formula, δ is the deviation threshold.

9. A path planning method for low-altitude aircraft navigation guidance according to claim 1, characterized in that The cost function of the bidirectional A*-KDE algorithm is expressed as: In the formula, g(n) is the actual cost of the path, h(n) is the heuristic cost, P(n) is the cost of the kernel density potential field, F(neighbor(n)) is the cost function of the bidirectional A*-KDE algorithm, and α is the weight parameter of the cost of the kernel density potential field.

10. A computer system, the computer system comprising a storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements a path planning method for low-altitude aircraft navigation guidance according to any one of claims 1 to 9.

Images