A method and system for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data

Abstract

The present invention relates to the field of unmanned aerial vehicle (UAV) flight control, and particularly to a method and system for reconstructing the flight trajectory of a UAV based on big data. The method includes: obtaining the flight state data of the UAV and the three-dimensional modeling of the surrounding environment to identify the positions of obstacle points; constructing a local environment area centered on the current position of the UAV, screening out candidate nodes that meet the conditions, and calculating the comprehensive priority of each candidate node using the evaluation function of the A* algorithm; calculating the estimated impact degree by analyzing the tortuosity of the UAV's past path and the stability of the current flight state, and adjusting the comprehensive priority of the candidate nodes based on the estimated impact degree, and performing trajectory reconstruction using the A* algorithm according to the adjusted comprehensive priority. By analyzing the flight stability and path tortuosity of the UAV, the path planning is dynamically adjusted to reduce the collision risk and improve the flight safety.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) flight control. In particular, it relates to a method and system for reconstructing the flight trajectory of a UAV based on big data. Background Art

[0002] A UAV is a device capable of autonomous flight in the air. It completes specific tasks by carrying various sensors and actuators. The UAV has the characteristics of small size, light weight, flexible operation, low cost, etc., which makes it widely used in multiple fields. The flight control system of the UAV is its core part, which controls the flight state of the UAV by receiving sensor data and executing instructions.

[0003] UAV flight plays a crucial role in the field of modern technology. It is widely used in multiple fields such as agricultural monitoring, power line inspection, security monitoring, and logistics distribution. The flexibility and efficiency of the UAV enable it to perform tasks in complex environments without relying on human labor, greatly improving work efficiency and safety. The autonomous flight ability and the ability to respond to emergencies of the UAV make it an indispensable part of modern technology. In the agricultural field, UAVs can be used for farmland monitoring to help farmers better manage crops; in the logistics field, UAVs can quickly deliver goods to improve efficiency; in the security field, UAVs can conduct patrol and monitoring to ensure safety.

[0004] However, traditional path planning methods, such as the A* algorithm, have certain limitations in practical applications. The A* algorithm mainly focuses on the static distance between nodes and insufficiently considers the actual flight state of the UAV (such as speed, acceleration, and attitude angle). In addition, in complex and dynamic environments, relying solely on whether there are obstacles at nodes for path planning may not be able to fully handle emergencies, thus increasing flight risks. Therefore, in order to improve the flight safety and reliability of the UAV in complex environments, it is necessary to improve the path planning algorithm so that it can better adapt to the dynamic characteristics of the UAV and environmental changes. Summary of the Invention

[0005] To solve the problem of how to adjust the path planning in real time during the flight of the UAV to cope with environmental changes and the uncertainty of the flight state, the present invention provides solutions in the following aspects.

[0006] In a first aspect, a method for reconstructing an unmanned aerial vehicle (UAV) flight trajectory based on big data includes: obtaining UAV flight state data, performing three-dimensional modeling on the environment around the UAV to obtain the positions of obstacle points; constructing a local environment area centered on the current position of the UAV, screening eligible candidate nodes based on the local environment area, and calculating the comprehensive priority of each candidate node using the evaluation function of the A* algorithm; obtaining the degree of tortuosity of the control points in the UAV's past path and the stability of the flight state at the current position of the UAV, taking the ratio between the degree of tortuosity and the stability as the estimated influence degree, adjusting the comprehensive priority of the candidate nodes based on the estimated influence degree, and performing trajectory reconstruction using the A* algorithm according to the adjusted comprehensive priority.

[0007] By obtaining the UAV's flight state data and performing three-dimensional modeling to identify the positions of obstacles, the A* algorithm is used to efficiently screen eligible candidate nodes in the local environment. Through the calculation of the comprehensive priority, the accuracy and optimality of path planning are ensured. By analyzing the degree of tortuosity of the past path and the stability of the current flight state, the estimated influence degree is calculated to adjust the node priority, thereby optimizing the flight trajectory. This not only effectively improves the flight safety of the UAV in complex environments but also significantly optimizes the efficiency of path planning, reduces energy consumption, enhances the system's ability to respond to emergencies, and ensures that the UAV can complete various tasks stably and reliably.

[0008] Preferably, the flight state data includes: speed data and angle data. The speed data includes: linear velocity, angular velocity, and acceleration. The angle data includes: roll angle, pitch angle, and yaw angle.

[0009] Preferably, the screening of eligible candidate nodes includes:

[0010] The local environment area contains a preset number of adjacent points around the UAV. Taking the directions of the preset number of adjacent points around as the ray directions respectively, calculate the angle between each ray direction and the current flight direction of the UAV.

[0011] Select the points with an angle less than the preset angle as candidate nodes. Among the candidate nodes, calculate the distance between each candidate node and the current position of the UAV, sort them from smallest to largest, and select the preset number of candidate nodes closest to the current position of the UAV from the sorted distance list as the candidate nodes.

[0012] By screening eligible candidate nodes in the local environment, sorting them according to the distance from the current position of the UAV, and selecting the nearest preset number of nodes as the candidate nodes, the local optimality of path planning is ensured. This effectively improves the flight safety and path planning efficiency of the UAV in complex environments.

[0013] Preferably, the comprehensive priority includes:

[0014] Taking the current position of the drone as the target node, calculate the actual flight distance from the starting point to the target node to obtain the actual cost, and calculate the estimated cost from the target node to the end point. Among them, calculate the sum of the lengths of pairwise adjacent flight paths from the starting point to the current position of the drone to obtain the actual cost, and the estimated cost is obtained by calculating the straight-line distance from the end point to the target node according to the Euclidean distance;

[0015] Take the sum between the actual cost and the estimated cost as the comprehensive priority of the target node.

[0016] By calculating the actual flight distance of the drone from the starting point to the current position as the actual cost, and calculating the estimated cost from the target node to the end point according to the Euclidean distance, and taking the sum of the two as the comprehensive priority, the path planning is effectively optimized, ensuring the safety and efficiency of the drone flight.

[0017] Preferably, the stability degree includes:

[0018] Obtain the flight state data corresponding to the preset number of control points before the current position of the drone. Among them, the flight state data includes: speed data and angle data;

[0019] Taking any speed data as the reference speed data, calculate the difference between each reference speed data and the mean value of the reference speed data respectively, and divide it by the average value of the sum of the standard deviations to obtain the comprehensive speed fluctuation degree; taking any angle data as the reference angle data, use the information entropy function to calculate the information entropy of each reference angle data, and perform the average value of the sum to obtain the comprehensive angle complexity;

[0020] Take the sum of 1, the comprehensive speed fluctuation degree, and the comprehensive angle complexity respectively, and use the logarithmic function to take the reciprocal of the logarithmic function as the stability degree of the current drone.

[0021] By analyzing the speed and angle data in the past path of the drone, calculate the comprehensive speed fluctuation degree and the comprehensive angle complexity, and then evaluate the flight stability degree of the drone. This method effectively improves the flight stability of the drone in complex environments and the reliability of path planning.

[0022] Preferably, the tortuosity degree includes:

[0023] Obtain the control points with the preset nearest number before the current position of the drone for least squares fitting to obtain a fitted curve. Extract the information of the first-order, second-order, and third-order derivatives from the fitted curve to calculate the curvature and torsion of each control point. Take the ratio of the curvature of each control point to the average curvature of the area where the drone has passed historically as the relative bending degree, and take the ratio of the torsion of each control point to the maximum torsion of the area where the drone has passed historically as the relative twisting degree; Take the average of the sum of the relative bending degrees and relative twisting degrees of all control points as the overall complexity of all control points;

[0024] Take the ratio between the turning angle and the pitch angle of each control point, perform exponential decay using the negative exponential function, and take the product of the exponentially decayed ratio and the overall complexity as the tortuosity degree.

[0025] By least squares fitting the control points in the past path of the drone, a fitted curve is obtained, and the curvature and torsion of each control point are calculated using the information of the first-order, second-order, and third-order derivatives of the curve. By comparing with historical data to evaluate the relative bending and twisting degrees, and combining with the exponential decay processing of the turning angle, the overall tortuosity degree of the path is finally determined, so as to optimize the flight trajectory of the drone and improve its adaptability and flight safety in complex environments.

[0026] Preferably, the turning angle includes:

[0027] In the initial path, use the arccosine function to calculate the ratio between the dot product of the vectors between two adjacent control points and the product of the magnitudes of the adjacent vectors to obtain the degree of direction change between two adjacent control points, which is the turning angle.

[0028] Preferably, adjusting the comprehensive priority of candidate nodes based on the estimated influence degree of the drone includes:

[0029] Multiply the estimated influence degree of each node by the obstacle avoidance cost to correct the comprehensive priority of each node to obtain the adjusted comprehensive priority.

[0030] Preferably, the obstacle avoidance cost includes:

[0031] Take the reciprocal of the sum of the distance between each node and the nearest obstacle in the 3D modeling and the hyperparameter as the obstacle avoidance cost.

[0032] In a second aspect, a drone flight trajectory reconstruction system based on big data includes: a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned drone flight trajectory reconstruction method based on big data is implemented.

[0033] The present invention has the following effects:

[0034] 1. By obtaining the flight status data of the drone in real time and performing three-dimensional modeling of the surrounding environment, the present invention can detect obstacles in a timely manner and avoid them. At the same time, by analyzing the flight stability and path tortuosity of the drone, the path planning is dynamically adjusted, reducing the collision risk caused by environmental changes and flight status uncertainties, and improving the flight safety of the drone.

[0035] 2. Based on the evaluation function and comprehensive priority adjustment of the A* algorithm, the present invention can optimize the flight path of the drone, reduce unnecessary flight distance and energy consumption. Through least squares fitting and curvature and torsion calculations, the smoothness of the path is further optimized, reducing energy loss during flight and improving flight efficiency.

[0036] 3. By dynamically adjusting the path planning according to the flight status of the drone and environmental changes, the present invention enhances the adaptability of the system to complex environments and emergencies. Through big data analysis and real-time adjustment, the robustness of the system is improved, ensuring that the drone can complete flight tasks stably and reliably under various complex conditions. Brief Description of the Drawings

[0037] Figure 1 is a flowchart of the method from step S1 to step S3 in a method for reconstructing the flight trajectory of a drone based on big data according to an embodiment of the present invention.

[0038] Figure 2 is a structural block diagram of a system for reconstructing the flight trajectory of a drone based on big data according to an embodiment of the present invention. Detailed Embodiment

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention.

[0040] Referring to Figure 1 , a method for reconstructing the flight trajectory of a drone based on big data includes steps S1 to S3, which are specifically as follows:

[0041] S1: Obtain the flight status data of the drone, perform three-dimensional modeling on the surrounding environment of the drone, and obtain the positions of obstacle points.

[0042] Among them, the flight status data includes, but is not limited to: the speed data and angle data of the drone. The speed data includes: linear velocity, angular velocity, and acceleration. The angle data includes: roll angle, pitch angle, and yaw angle. The preset acquisition frequency is 10 Hz. The six-degree-of-freedom pose, speed, and acceleration data of the drone are calculated in real time by using Kalman filter to fuse IMU, GPS, and barometer data; a dynamic obstacle grid map is constructed by jointly calibrating lidar point cloud and visual SLAM.

[0043] Construct a three-dimensional model of the environment around the UAV by using sensor data (such as lidar point cloud, visual SLAM, etc.). Divide the three-dimensional space into individual three-dimensional grids (voxels), and each three-dimensional grid represents a node.

[0044] Each node contains the following information: the coordinates of the node in the three-dimensional space and whether the node is occupied by an obstacle. It should be noted that the node is the basic unit of path planning, representing the points that the UAV may pass through. Through the status information of the nodes, the UAV can understand the layout of the surrounding environment and the distribution of obstacles.

[0045] Centered on the current position of the UAV, set a local range and filter out eligible candidate nodes. These nodes represent the possible positions that the UAV may fly to at the next moment. The candidate nodes are selected from the nodes, which are candidate points that meet the dynamic constraints of the UAV and the obstacle avoidance requirements. Use the A* algorithm to obtain the comprehensive priority of each candidate node, and determine which candidate node is the control point according to the comprehensive priority. The control point is the key point selected from the nodes and is used to define the flight path of the UAV.

[0046] It should be noted that in the process of obtaining the UAV flight trajectory control points by the A* algorithm, the stability of the UAV flight state is often ignored. If the UAV flight is not stable enough and there is a certain deviation from the preset route, it may collide with the obstacles in the scene and cause danger; therefore, judge the smoothness of the flight by the recent flight state of the UAV, and obtain the path tortuosity according to the control points passed by the current UAV flight, and calculate the adaptive preference degree of each node accordingly, so as to ensure the safe operation of the UAV.

[0047] S2: Construct a local environment area centered on the current position of the UAV, filter out eligible candidate nodes based on the local environment area, and use the evaluation function of the A* algorithm to calculate the comprehensive priority of each candidate node.

[0048] Filter out eligible candidate nodes, including:

[0049] The local environment area contains a preset number of adjacent points around the UAV. Respectively, take the directions of the preset number of adjacent points around as the ray directions, and calculate the angle between each ray direction and the current flight direction of the UAV;

[0050] Select the points with an angle less than the preset angle as candidate nodes. Among the candidate nodes, calculate the distance between each candidate node and the current position of the UAV, and sort them from small to large. Select the preset number of candidate nodes closest to the current position of the UAV from the sorted distance list as the candidate nodes.

[0051] Exemplarily, the local environment area is cubic blocks of a size, where 26 adjacent points (including the central point) around the drone are included in the cubic block area. Respectively, taking the directions of the 26 adjacent points around as the ray directions, according to the maximum pitch angle in the dynamic constraints being , the included angle between the ray direction and the flight direction of the drone is less than , and the 9 candidate nodes with the smallest distance are recorded as the nodes to be selected.

[0052] Comprehensive priority, including:

[0053] Taking the current position of the drone as the target node, calculating the actual flight distance from the starting point to the target node to obtain the actual cost, and calculating the estimated cost from the target node to the end point. Among them, the actual cost is obtained by calculating the sum of the flight path lengths between two adjacent points from the starting point to the current position of the drone, and the estimated cost is obtained by calculating the straight-line distance from the end point to the target node according to the Euclidean distance;

[0054] Taking the sum of the actual cost and the estimated cost as the comprehensive priority of the target node.

[0055] Specifically, the comprehensive priority satisfies the following relational expression:

[0056] ;

[0057] In the formula, represents the comprehensive priority of node , represents the actual cost of node from the starting point, represents the estimated cost of node from the end point.

[0058] Among them, the estimated cost is also the heuristic function of the A* algorithm. The main role of the heuristic function is to provide an estimated cost from the current node to the target node to guide the search process, which can help the algorithm quickly converge to the optimal path and reduce unnecessary searches. The A* algorithm is a well-known technology in the field and will not be described in detail.

[0059] It should be noted that whether the flight state of the drone is stable during operation greatly affects the decision-making in the process of trajectory reconstruction of the drone. Therefore, it is necessary to evaluate in combination with the stability of the area where the drone has flown in the recent period. The specific steps are as follows:

[0060] S3: Obtain the tortuosity degree of the control points in the past path of the drone and the stability degree of the flight state of the current position of the drone, take the ratio between the tortuosity degree and the stability degree as the estimated influence degree, and adjust the comprehensive priority of the nodes to be selected based on the estimated influence degree. Use the A* algorithm for trajectory reconstruction according to the adjusted comprehensive priority.

[0061] The stability level includes:

[0062] Obtain the flight state data corresponding to the preset number of control points before the current position of the UAV. Among them, the flight state data includes: speed data and angle data;

[0063] Taking any speed data as the reference speed data, calculate the difference between each reference speed data and the mean value of the reference speed data, and divide it by the average value of the sum of the standard deviations to obtain the comprehensive speed fluctuation degree; taking any angle data as the reference angle data, use the information entropy function to calculate the information entropy of each reference angle data, and perform the average value of the sum to obtain the comprehensive angle complexity;

[0064] Sum up 1, the comprehensive speed fluctuation degree, and the comprehensive angle complexity respectively, and use the logarithmic function. Take the reciprocal of the logarithmic function as the stability level of the current UAV.

[0065] Exemplarily, take the flight state data corresponding to the first 30 control points before the current position of the UAV, including the speed data and angle data of the first 30 control points. It should also be noted that when obtaining control points from the starting point through the A* algorithm, if the number of control points flown before the current position of the UAV is less than 30, only analyze the flight state data of the number of control points flown before the current position of the UAV.

[0066] Specifically, the stability level of the UAV satisfies the following relational expression:

[0067] ;

[0068] In the formula, represents the stability level of the current operation of the UAV, represents the number of reference data, represents the th item of reference speed data of the th control point, represents the average value of the th item of reference speed data, represents the standard deviation of the th item of reference speed data, represents the th item of reference angle data, represents the logarithmic function with the natural constant as the base, represents the information entropy function.

[0069] That is to say, represents the 3 speed data of the UAV ( , corresponding to linear velocity, angular velocity, and acceleration respectively). The higher the comprehensive speed fluctuation degree, the more unstable the flight of the UAV. Represent three angular data of the drone ( , corresponding to the pitch angle, roll angle, and yaw angle respectively). If the changes in all three attitude angles are complex (high information entropy), it indicates that the flight attitude of the drone is unstable. If the changes in all three attitude angles are simple (low information entropy), it indicates that the flight attitude of the drone is stable.

[0070] By adding the speed fluctuation and the angle complexity, a comprehensive instability is obtained. Adding 1 to the denominator in the formula can ensure that the result of the formula is always positive, avoiding the situation of negative values or zero, and can more clearly distinguish the stability degrees in different flight states.

[0071] In addition, in another embodiment, it further includes:

[0072] Obtain the flight state data corresponding to a preset number of control points before the current position of the drone, where the flight state data includes: speed data and angular data;

[0073] Taking any speed data as the reference speed data, calculate the difference between each reference speed data and the mean value of the reference speed data, and divide it by the average value of the sum of the standard deviations to obtain the comprehensive speed fluctuation degree; taking any angular data as the reference angular data, calculate the difference between each reference angular data and the mean value of the reference angular data, and divide it by the average value of the sum of the standard deviations to obtain the comprehensive angle complexity;

[0074] Sum the values of 1, the comprehensive speed fluctuation degree, and the comprehensive angle complexity respectively, and use the logarithmic function, and take the reciprocal of the logarithmic function as the stability degree of the current drone.

[0075] Specifically, the stability degree of the drone satisfies the following relational expression:

[0076] ;

[0077] In the formula, represents the stability degree of the current operation of the drone, represents the number of reference data, represents the th item of reference speed data of the th control point, represents the average value of the th item of reference speed data, represents the standard deviation of the th item of reference speed data, represents the th item of reference angular data, represents the logarithmic function with the natural constant as the base, represents the information entropy function.

[0078] It should be noted that in addition to judging the smoothness based on the reference data of the drone in the past period, the tortuosity of the path corresponding to the control points in the past period should also be judged as the tortuosity of the path at the next moment, and the probability of possible fluctuations of the drone in the next step should be analyzed.

[0079] The tortuosity includes:

[0080] Obtain the preset nearest number of control points before the current position of the drone for least squares fitting to obtain a fitting curve. Extract the information of the first, second, and third derivatives from the fitting curve to calculate the curvature and torsion of each control point. Take the ratio of the curvature of each control point to the average curvature of the area where the drone has passed in history as the relative bending degree, and take the ratio of the torsion of each control point to the maximum torsion of the area where the drone has passed in history as the relative twisting degree; Take the average of the sum of the relative bending degrees and relative twisting degrees of all control points as the overall complexity of all control points;

[0081] Take the ratio between the turning angle and the pitch angle of each control point, perform exponential decay using the negative exponential function, and take the product of the exponentially decayed ratio and the overall complexity as the tortuosity.

[0082] That is to say, the preset nearest number is 10, which means obtaining the 10 control points closest to the current sequential index of the drone control points. It should be noted that the preset nearest number of control points is for fitting the flight path to facilitate the calculation of the curvature and torsion of each control point, but when calculating the tortuosity of the path at the next moment, it is still analyzed based on the 30 control points before the current position of the drone. The maximum pitch angle in the dynamic constraints is , and the pitch angle is used as the reference angle, which reflects the dynamic constraints of the drone. A larger indicates that the drone can withstand a greater change in pitch angle, so the influence of the turning angle on flight stability is relatively small. The turning angle is weighted through the negative exponential function. The larger the turning angle, the smaller the weight, indicating that the tortuosity of the path at this point is higher and the influence on flight stability is greater.

[0083] Specifically, the tortuosity satisfies the following relational expression:

[0084] ;

[0085] Among them, represents the tortuosity of the drone's path at the next moment, represents the number of control points, represents the curvature of the th control point, The torsion of a control point, represents the average curvature of the area traversed by the drone's history, represents the maximum torsion of the area traversed by the drone's history, represents the turning angle of the control point, represents the pitch angle in the dynamic constraints, represents the exponential function with the natural number as the base.

[0086] That is to say, Part calculates the degree of tortuosity by synthesizing the curvature and torsion at each control point in the past path. The greater the curvature and torsion, the greater the degree of tortuosity of the path; in the formula, Part measures the degree of tortuosity through the turning angle of the path control points. The larger the turning angle, the smoother the path is at this point, and the smaller the degree of tortuosity.

[0087] If the curvature at the current control point is greater than the historical average curvature, it means that the bending degree of the path at this point is higher than the situation usually encountered by the drone. If the curvature at the current control point is equal to the historical average curvature, it means that the bending degree of the path at this point is comparable to the situation usually encountered by the drone. If the curvature at the current control point is less than the historical average curvature, it means that the bending degree of the path at this point is lower than the situation usually encountered by the drone.

[0088] If the torsion at the current control point is greater than the historical maximum torsion, it means that the degree of distortion of the path at this point exceeds the maximum degree of distortion encountered in the drone's historical flight. If the torsion at the current control point is equal to the historical maximum torsion, it means that the degree of distortion of the path at this point is comparable to the maximum degree of distortion encountered in the drone's historical flight. If the torsion at the current control point is less than the historical maximum torsion, it means that the degree of distortion of the path at this point is lower than the maximum degree of distortion encountered in the drone's historical flight.

[0089] It should be noted that the calculation of the curvature and torsion of the control points is a well-known technology to those skilled in the art and will not be described in detail.

[0090] The turning angle includes:

[0091] In the initial path, the dot product of the vectors between two adjacent control points is calculated using the inverse cosine function, and the ratio to the product of the magnitudes of the adjacent vectors is obtained to get the degree of direction change between two adjacent control points, which is the turning angle.

[0092] Specifically, the turning angle satisfies the following relational expression:

[0093] ;

[0094] In the formula, represents the The turning angle of a control point represents the arccosine function represents the vector of the represents the vector of the control point, and represents the magnitude of the vector.

[0095] That is to say, represents the degree of direction change of the path between two adjacent control points. The larger the turning angle, the higher the degree of tortuosity of the path at the control point. The result of the dot product of vectors reflects the similarity between two vectors.

[0096] Take the reciprocal of the sum of the distance between each node and the nearest obstacle in 3D modeling and the hyperparameter as the obstacle avoidance cost.

[0097] Specifically, the obstacle avoidance cost satisfies the following relational expression:

[0098] ;

[0099] In the formula, represents the obstacle avoidance cost of node , represents the distance from node to the nearest obstacle, and represents the hyperparameter.

[0100] Exemplarily, to prevent the denominator from being zero.

[0101] Multiply the estimated influence degree of each node by the obstacle avoidance cost to correct the comprehensive priority of each node, and obtain the adjusted comprehensive priority.

[0102] Specifically, the adjusted comprehensive priority satisfies the following relational expression:

[0103] w ;

[0104] In the formula, represents the adjusted comprehensive priority of node , represents the comprehensive priority of node , represents the estimated influence degree, and represents the obstacle avoidance cost of node .

[0105] The present invention also provides a UAV flight trajectory reconstruction system based on big data. As Figure 2As shown, the system includes a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data according to the first aspect of the present invention is implemented. The system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface, whose settings and functions are known in the art, and thus will not be elaborated herein.

[0106] It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent shall be subject to the appended claims.

Claims

1. A method for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data, characterized in that, Including: Obtain the flight state data of the drone, perform three-dimensional modeling on the environment around the drone, and obtain the positions of obstacle points; Construct a local environment area centered on the current position of the drone, screen out candidate nodes that meet the conditions based on the local environment area, and calculate the comprehensive priority of each candidate node using the evaluation function of the A* algorithm; Obtain the tortuosity of the control points in the past path of the drone and the stability of the flight state at the current position of the drone, use the ratio between the tortuosity and the stability as the estimated influence degree, adjust the comprehensive priority of the candidate nodes based on the estimated influence degree, and perform trajectory reconstruction using the A* algorithm according to the adjusted comprehensive priority; The stability includes: Obtain the flight state data corresponding to a preset number of control points before the current position of the drone, where the flight state data includes: speed data and angle data; Taking any speed data as the reference speed data, calculate the difference between each reference speed data and the mean value of the reference speed data respectively, and divide by the average value of the sum of the standard deviations to obtain the comprehensive speed fluctuation degree; taking any angle data as the reference angle data, use the information entropy function to calculate the information entropy of each reference angle data, and perform the average value of the sum to obtain the comprehensive angle complexity; Add the sum of 1, the comprehensive speed fluctuation degree, and the comprehensive angle complexity, and use the logarithmic function to take the reciprocal of the logarithmic function as the stability of the current drone.

2. The method for reconstructing the flight trajectory of a drone based on big data according to claim 1, wherein, The flight state data includes: speed data and angle data, the speed data includes: linear velocity, angular velocity, and acceleration, and the angle data includes: roll angle, pitch angle, and yaw angle.

3. A method for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data according to claim 1, characterized in that, The screening out of candidate nodes that meet the conditions includes: The local environment area contains a preset number of adjacent points around the drone. Taking the directions of the preset number of adjacent points around as the ray directions respectively, calculate the included angle between each ray direction and the current flight direction of the drone; Select the points with an included angle less than the preset angle as candidate nodes. Among the candidate nodes, calculate the distance between each candidate node and the current position of the drone, and sort them from small to large. Select the preset number of candidate nodes closest to the current position of the drone from the sorted distance list as the candidate nodes.

4. A method for reconstructing an unmanned aerial vehicle flight trajectory based on big data according to claim 1, characterized in that, The comprehensive priority includes: Taking the current position of the drone as the target node, calculate the actual flight distance from the starting point to the target node to obtain the actual cost, and calculate the estimated cost from the target node to the end point. Among them, calculate the sum of the lengths of the pairwise adjacent flight paths of the drone from the starting point to the current position to obtain the actual cost, and the estimated cost is obtained by calculating the straight-line distance from the end point to the target node according to the Euclidean distance; Take the sum of the actual cost and the estimated cost as the comprehensive priority of the target node.

5. A method for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data according to claim 1, characterized in that, The tortuosity includes: Obtain a preset number of the closest control points to the current position of the drone for least squares fitting to obtain a fitted curve. Extract information on the first-order, second-order, and third-order derivatives from the fitted curve for calculating the curvature and torsion of each control point. Take the ratio of the curvature of each control point to the average curvature of the area where the drone has passed historically as the relative bending degree, and take the ratio of the torsion of each control point to the maximum torsion of the area where the drone has passed historically as the relative distortion degree; Take the average of the sum of the relative bending degrees and relative distortion degrees of all control points as the overall complexity of all control points. Take the ratio between the turning angle and the pitch angle of each control point, perform exponential decay using the negative exponential function, and take the product of the exponentially decayed ratio and the overall complexity as the tortuosity.

6. The method for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data according to claim 5, wherein The turning angle includes: In the initial path, use the inverse cosine function to calculate the ratio between the dot product of the vectors between two adjacent control points and the product of the magnitudes of the adjacent vectors to obtain the degree of direction change between two adjacent control points, which is the turning angle.

7. A method for reconstructing the flight trajectory of an unmanned aerial vehicle based on big data according to claim 1, characterized in that The adjustment of the comprehensive priority of candidate nodes based on the estimated influence degree includes: Correct the comprehensive priority of each node by multiplying the estimated influence degree of each node by the obstacle avoidance cost to obtain the adjusted comprehensive priority.

8. A method for reconstructing the flight trajectory of a drone based on big data according to claim 7, characterized in that, The obstacle avoidance cost includes: Take the reciprocal of the sum of the distance between each node and the nearest obstacle in the three-dimensional modeling and the hyperparameter as the obstacle avoidance cost.

9. An unmanned aerial vehicle flight trajectory reconstruction system based on big data, characterized in that, Include: A processor and a memory, where the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for reconstructing the flight trajectory of a drone based on big data according to any one of claims 1-8 is implemented.

Images