Unmanned aerial vehicle indoor navigation simulation method and system based on laser SLAM

Abstract

The invention provides an unmanned aerial vehicle indoor navigation simulation method and system based on laser SLAM (Simultaneous Localization and Mapping). The method comprises the following steps: carrying out distortion removal on environmental point cloud data to obtain a low-precision odometer; estimating a linear velocity and an angular velocity by using a low-precision speedometer, and performing map construction according to the distortion-free point cloud after distortion correction to obtain a high-precision speedometer and environmental map information; based on the pose information provided by the high-precision speedometer, fusing the low-precision speedometer and the high-precision speedometer to obtain high-precision pose information; based on the environmental map information and the high-precision pose information, performing path search by adopting a mixed A * path search algorithm to generate an initial path; and taking the key node of the initial path as a control point, performing trajectory fitting by using a third-order B spline curve, and performing smoothing and safety optimization on the trajectory through an optimization function to generate a final flight trajectory. According to the invention, the human input and equipment deployment cost are reduced, and the flight efficiency and the task completion rate are improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an indoor navigation simulation method and system for unmanned aerial vehicles based on laser SLAM. Background Technology

[0002] Currently, most drone systems still rely on GPS for navigation and positioning. In environments where GPS signals are limited, such as indoors, tunnels, and mines, their autonomous navigation capabilities are affected, failing to meet the needs of precise operations and continuous operation, thus limiting the further application of drones in indoor scenarios.

[0003] Currently, most unmanned aerial vehicle (UAV) systems still primarily rely on GPS signals for navigation and positioning. However, in complex and enclosed environments such as indoor spaces, tunnels, and mines, GPS signals are often severely interfered with or even completely disabled, leading to a sharp decline in UAV positioning accuracy and significantly limiting their autonomous navigation capabilities, making it difficult to maintain stable operation over extended periods. This limitation not only affects the UAV's ability to perform precise operations and continuous mission execution in complex environments but also restricts its promotion and application in typical indoor scenarios such as security patrols, disaster relief, and warehousing and logistics. Therefore, researching high-precision positioning and mapping technologies in GPS-free environments, combined with efficient and robust path planning and obstacle avoidance algorithms, has become a key research direction for improving the autonomous navigation capabilities of UAVs and expanding their application scope. Summary of the Invention

[0004] Therefore, the purpose of this invention is to provide a laser SLAM-based indoor navigation simulation method and system for unmanned aerial vehicles (UAVs) to at least address the shortcomings of the aforementioned technologies.

[0005] This invention proposes a laser SLAM-based indoor navigation simulation method for unmanned aerial vehicles (UAVs), comprising: Environmental point cloud data is acquired by scanning with lidar, and distortion removal is performed on the environmental point cloud data to obtain a low-precision odometer at the first hertz. The low-precision odometer is used to estimate linear velocity and angular velocity to achieve motion distortion correction of each frame of environmental point cloud data, and a map is constructed based on the distortion-corrected distortion-free point cloud to obtain a high-precision odometer at the second hertz and the corresponding environmental map information, wherein the first hertz is greater than the second hertz. Based on the pose information provided by the high-precision odometer, the low-precision odometer and the high-precision odometer are fused to obtain high-precision pose information at the second hertz. Based on the environmental map information and the high-precision pose information, a hybrid A* path search algorithm is used to perform path search in order to generate an initial path that satisfies the kinematic constraints of the UAV. The key nodes of the initial path are used as control points. A third-order B-spline curve is used for trajectory fitting, and the trajectory is smoothed and optimized for safety using an optimization function to generate the final flight trajectory.

[0006] Furthermore, based on the environmental map information and the high-precision pose information, the step of using a hybrid A* path search algorithm to perform path search to generate an initial path that satisfies the kinematic constraints of the UAV includes: Based on the high-precision pose information, the spatial position and orientation of the UAV are obtained to determine the starting position of the path search and the position feedback when tracking the path. Based on the UAV's kinematic model and the starting position, feasible trajectory segments per unit time are calculated using dynamic integration, and collision detection is performed based on the current local map. A cost function is then used to guide the path towards the target point.

[0007] Furthermore, the formula for calculating the cost function is as follows:

[0008] In the formula, This represents the cumulative actual cost from the initial state to the current state. The path the drone took from the initial state to the current state traversed a total of [number missing]. There are discrete grid cells, and the control input within each grid cell is... The corresponding flight time is , This represents the estimated cost from the current state to the final state, where... Euclidean distance. This refers to the distance to Manhattan.

[0009] Furthermore, the steps of using the key nodes of the initial path as control points, fitting the trajectory with a third-order B-spline curve, and smoothing and optimizing the trajectory for safety using an optimization function to generate the final flight trajectory include: The key nodes of the initial path are used as control points, and the shortest distance between the control points and obstacles is calculated using the convex hull property of the uniform third-order B-spline curve. Euclidean distance field is introduced to obtain the distance information from any point in space to the nearest obstacle. Define a smoothing cost function and use gradient descent to iteratively optimize the control points to minimize the total cost function, thereby generating the final flight trajectory.

[0010] Furthermore, the formula for calculating the smoothing cost function is as follows:

[0011] In the formula, For point arrive and The straight-line distance, when The smaller the value, the more points... arrive and The closer the straight-line distance, the better.

[0012] This invention also proposes an indoor navigation simulation system for unmanned aerial vehicles based on laser SLAM, comprising: The data acquisition module is used to acquire environmental point cloud data by scanning with lidar and to remove distortion from the environmental point cloud data in order to obtain a low-precision odometer at the first hertz. The data processing module is used to estimate linear velocity and angular velocity using the low-precision odometer to achieve motion distortion correction of each frame of environmental point cloud data, and to construct a map based on the distortion-corrected distortion-free point cloud to obtain a high-precision odometer at the second hertz and the corresponding environmental map information, wherein the first hertz is greater than the second hertz. The data fusion module is used to fuse the low-precision odometer and the high-precision odometer based on the pose information provided by the high-precision odometer to obtain high-precision pose information at the second hertz. The path search module is used to perform path search using a hybrid A* path search algorithm based on the environmental map information and the high-precision pose information, so as to generate an initial path that satisfies the kinematic constraints of the UAV. The trajectory fitting module is used to use the key nodes of the initial path as control points, perform trajectory fitting using third-order B-spline curves, and optimize the trajectory for smoothing and safety through an optimization function to generate the final flight trajectory.

[0013] Furthermore, the path search module is specifically used for: Based on the high-precision pose information, the spatial position and orientation of the UAV are obtained to determine the starting position of the path search and the position feedback when tracking the path. Based on the UAV's kinematic model and the starting position, feasible trajectory segments per unit time are calculated using dynamic integration, and collision detection is performed based on the current local map. A cost function is then used to guide the path towards the target point.

[0014] Furthermore, the trajectory fitting module is specifically used for: The key nodes of the initial path are used as control points, and the shortest distance between the control points and obstacles is calculated using the convex hull property of the uniform third-order B-spline curve. Euclidean distance field is introduced to obtain the distance information from any point in space to the nearest obstacle. Define a smoothing cost function and use gradient descent to iteratively optimize the control points to minimize the total cost function, thereby generating the final flight trajectory.

[0015] The present invention also proposes a storage medium on which a computer program is stored, which, when executed by a processor, implements the above-described laser SLAM-based indoor navigation simulation method for unmanned aerial vehicles.

[0016] The present invention also proposes a computer, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described laser SLAM-based indoor navigation simulation method for unmanned aerial vehicles.

[0017] The laser SLAM-based indoor navigation simulation method and system for UAVs in this invention can achieve high-precision positioning and consistent map construction in GPS-free environments. By introducing loop closure detection, it effectively suppresses cumulative errors, ensuring mapping accuracy and global consistency over long-term operation. During path planning, by combining the UAV's kinematic constraints and an improved cost function, an executable initial path can be generated. Furthermore, third-order B-spline curves and Euclidean distance fields are used for trajectory smoothing and obstacle avoidance optimization, thereby obtaining a continuous, safe, and smooth final flight trajectory. It exhibits good real-time performance and robustness in complex environments, adapting to dynamic obstacles and environmental changes, and improving the safety and reliability of UAVs during mission execution. It effectively overcomes the limitations of traditional solutions that rely on manual remote control or the deployment of positioning base stations, reducing manpower input and equipment deployment costs, and improving flight efficiency and mission completion rate. Attached Figure Description

[0018] Figure 1 This is a flowchart of the UAV indoor navigation simulation method based on laser SLAM in the first embodiment of the present invention; Figure 2 This is a schematic diagram of the convex hull properties of a third-order B-spline in the first embodiment of the present invention; Figure 3 This is a geometrical schematic diagram of the smoothing cost function in the first embodiment of the present invention; Figure 4 This is a structural block diagram of the UAV indoor navigation simulation system based on laser SLAM in the second embodiment of the present invention; Figure 5 This is a structural block diagram of the computer in the third embodiment of the present invention.

[0019] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0020] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0022] Example 1 Please see Figure 1 The figure shows a laser SLAM-based indoor navigation simulation method for unmanned aerial vehicles (UAVs) according to the first embodiment of the present invention. The method specifically includes steps S101 to S106: S101, environmental point cloud data is acquired by scanning with lidar, and distortion removal is performed on the environmental point cloud data to obtain a low-precision odometer at the first hertz. In practice, the 3D LiDAR on the drone performs a full-range scan of the surrounding environment at a frequency of 10Hz to obtain laser point cloud data. The point cloud is divided into edge points and planar points based on its curvature characteristics, and redundant and noise points are filtered out using a curvature threshold to improve subsequent computational efficiency. After distortion correction and feature extraction, the first... The data from the second scan is assembled into a complete point cloud frame. Based on feature point matching, pose changes are calculated to obtain a low-precision odometer with a frequency of 10Hz.

[0023] S102, the low-precision odometer is used to estimate the linear velocity and angular velocity in order to realize the motion distortion correction of each frame of environmental point cloud data, and the map is constructed based on the distortion-corrected distortion-free point cloud to obtain the second-frequency high-precision odometer and the corresponding environmental map information, wherein the first frequency is greater than the second frequency. In practice, a 10Hz low-precision odometry is used to estimate the linear and angular velocities of the UAV at the current moment, and these are applied to point cloud motion compensation. Temporal interpolation is performed point-by-point on each frame of the point cloud to obtain a distortion-free point cloud. The corrected point cloud is projected onto the global coordinate system, and combined with a sliding window mechanism and voxel filtering, redundant points are effectively removed, generating a sparse local sub-map, avoiding storage redundancy and computational overhead.

[0024] During the mapping process, the UAV pose is iteratively updated using point-to-line and point-to-area constraints combined with nonlinear least squares optimization, ensuring geometric consistency between the current frame and historical maps. The optimized result is updated once per second to output a high-precision odometer at 1Hz, while simultaneously generating a corresponding high-precision environmental map. Through this process, the system achieves progressive modeling and correction of the environment, providing the UAV with reliable positioning and environmental awareness capabilities.

[0025] S103, based on the pose information provided by the high-precision odometer, the low-precision odometer and the high-precision odometer are fused to obtain high-precision pose information of the second hertz. In practical implementation, based on the pose provided by the aforementioned 1Hz high-precision odometer, the low-precision and high-precision odometers are fused together, and a stitching method is used to achieve high-frequency pose output. That is, using the 1Hz high-precision pose as a reference, and through the formula... ,in, It is a high-precision pose output at 10Hz after fusion. It is the high-precision pose transformation matrix output by the global mapping module at 1Hz corresponding to the k-th frame. This is the relative pose transformation output by the 10Hz odometer module at time t. The final real-time output is a high-precision continuous pose at a frequency of 10Hz, balancing the high frequency of the odometer with the high accuracy of pose estimation, thus enabling real-time mapping and precise positioning of the LiDAR.

[0026] S104, Based on the environmental map information and the high-precision pose information, a hybrid A* path search algorithm is used to perform path search in order to generate an initial path that satisfies the kinematic constraints of the UAV. Furthermore, step S104 specifically includes steps S1041 to S1042: S1041, Based on the high-precision pose information, obtain the spatial position and orientation of the UAV to determine the starting position of the path search and the position feedback when tracking the path; S1042, based on the kinematic model of the UAV and the starting position, the feasible trajectory segment per unit time is calculated using dynamic integration, and collision detection is performed according to the current local map. The cost function is then used to guide the path to the target point.

[0027] In practical implementation, during UAV path planning, 1Hz environmental map information is used to provide local environmental map information at the current moment. Through the point cloud mapping process of LiDAR, the current obstacle distribution and passable area are updated every second. This latest map information is used in the path search process to ensure that the generated path accurately avoids obstacles. A 10Hz high-precision odometry is used to provide the UAV's current position and attitude information. During path search and execution, pose estimation based on the 10Hz frequency output is used to obtain the UAV's spatial position and orientation in real time, which is used to determine the starting point of the path search and position feedback during path tracking.

[0028] Furthermore, the hybrid A* path search algorithm can continuously expand feasible paths based on the current state of the UAV through forward integration. Starting from the current starting position, the algorithm selects several control inputs according to the UAV's kinematic characteristics, using different speeds and steering angles, and calculates the executable trajectory per unit time through dynamic integration. These trajectory segments all satisfy the UAV's kinematic constraints, ensuring that each step can be executed during actual flight. During trajectory expansion, each trajectory segment undergoes collision detection based on the current local map to avoid collisions with obstacles. The cost function is designed as follows:

[0029] In the formula, This represents the cumulative actual cost from the initial state to the current state. The drone's path from the initial state to the current state traversed a total of [number missing] [location missing]. There are discrete grid cells, and the control input within each grid cell is... The corresponding flight time is Therefore, this cost measure the combined cost of controlling energy consumption and time consumption in the path. The estimated cost from the current state to the final state is denoted as , where Euclidean distance. The distance is the Manhattan distance, and the average of the two values ​​is taken to balance the conservatism of route planning with the accuracy of the inspiration.

[0030] The cost function restricts path generation towards the destination, reducing invalid path expansion or deviation from the target. The first part measures the control energy consumption and flight time cost, while the second part uses Euclidean distance and Manhattan distance as heuristic functions to guide the search towards the target point. Finally, an initial path that balances energy consumption, time, and feasibility is obtained through backtracking.

[0031] S105, the key nodes of the initial path are used as control points, the trajectory is fitted using a third-order B-spline curve, and the trajectory is smoothed and optimized for safety using an optimization function to generate the final flight trajectory.

[0032] Furthermore, step S105 specifically includes steps S1051 to S1052: S1051, the key nodes of the initial path are used as control points, and the shortest distance between the control points and obstacles is calculated using the convex hull property of the uniform third-order B-spline curve. Euclidean distance field is introduced to obtain the distance information from any point in space to the nearest obstacle. S1052, define a smoothing cost function and use gradient descent to iteratively optimize the control points to minimize the total cost function, thereby generating the final flight trajectory.

[0033] In practical implementation, the key nodes of the path generated by the hybrid A are used as initial control points. The entire path is then fitted using the formula of a uniform third-order B-spline to generate a result that satisfies the given conditions. For continuous trajectories, third-order B-spline curves better balance trajectory smoothness, flexible adjustment, and practicality.

[0034] Furthermore, such as Figure 2 As shown, to ensure the safety and smoothness of the trajectory in complex environments, the following optimization function was designed. .in, It is a collision cost calculation that utilizes the convex hull property of B-splines to calculate the shortest distance between control points and obstacles, and introduces a Euclidean distance field to obtain real-time distance information from any point in space to the nearest obstacle. Specifically .

[0035]

[0036] in, Control points Euclidean distance to the nearest obstacle To set the minimum safe distance, the function It is a continuously differentiable distance potential function. This function introduces a large cost when the control point is close to the obstacle, thereby pushing the control point away from the obstacle area, while it has no effect outside the safe distance range, thus avoiding the negative impact on the trajectory of segments far from the obstacle.

[0037] Specifically, construct a smooth cost function. Specifically:

[0038] In the formula, For point arrive and The straight-line distance, when The smaller the value, the more points... arrive and The closer the straight-line distance, the more the trajectory tends to be a straight line, thus making it smoother.

[0039] like Figure 3 As shown, the specific geometric meaning is a point. arrive and The straight-line distance. When The smaller the value, the more points... arrive and The closer the straight-line distance, the more the trajectory tends to be a straight line, resulting in a smoother path.

[0040] Specifically, the goal of trajectory optimization is to minimize the total cost function while maintaining trajectory feasibility. To achieve this goal, gradient descent is used to iteratively update the control points. Since both cost functions mentioned above are continuously differentiable functions of the control points, the gradient direction is calculated efficiently, and the minimum value is continuously approached along the gradient direction with a preset step size. Under the premise of satisfying dynamic constraints and boundary conditions, gradient descent is used to optimize the total cost function to a local minimum, thereby generating a smooth and safe flight trajectory.

[0041] The method in this embodiment can achieve high-precision, real-time pose estimation and map building in a 3D environment, while balancing low computational resource consumption and good system scalability. It is suitable for resource-constrained UAV platforms and performs better in terms of 3D map building, real-time performance, and stability. In summary, the laser SLAM-based UAV indoor navigation simulation method described in the above embodiments of the present invention can achieve high-precision positioning and consistent map construction in GPS-free environments. By introducing loop closure detection, it effectively suppresses accumulated errors, ensuring mapping accuracy and global consistency over long-term operation. During path planning, by combining the UAV's kinematic constraints with an improved cost function, an executable initial path can be generated. Furthermore, third-order B-spline curves and Euclidean distance fields are used for trajectory smoothing and obstacle avoidance optimization, resulting in a continuous, safe, and smooth final flight trajectory. It exhibits good real-time performance and robustness in complex environments, adapting to dynamic obstacles and environmental changes, thus improving the safety and reliability of the UAV during mission execution. It effectively overcomes the limitations of traditional solutions that rely on manual remote control or the deployment of positioning base stations, reducing manpower and equipment deployment costs, and improving flight efficiency and mission completion rate.

[0042] Example 2 In another aspect, this invention also proposes an indoor navigation simulation system for unmanned aerial vehicles based on laser SLAM. Please refer to [link / reference needed]. Figure 4The figure shows a laser SLAM-based indoor navigation simulation system for unmanned aerial vehicles (UAVs) according to a second embodiment of the present invention. The system includes: The data acquisition module 11 is used to acquire environmental point cloud data by scanning with lidar and to remove distortion from the environmental point cloud data in order to obtain a low-precision odometer at the first hertz. Data processing module 12 is used to estimate linear velocity and angular velocity using the low-precision odometer to achieve motion distortion correction of each frame of environmental point cloud data, and to construct a map based on the distortion-corrected distortion-free point cloud to obtain a high-precision odometer at the second hertz and the corresponding environmental map information, wherein the first hertz is greater than the second hertz. The data fusion module 13 is used to fuse the low-precision odometer and the high-precision odometer based on the pose information provided by the high-precision odometer to obtain high-precision pose information at the second hertz. The path search module 14 is used to perform path search based on the environmental map information and the high-precision pose information, using a hybrid A* path search algorithm to generate an initial path that satisfies the kinematic constraints of the UAV. Furthermore, the path search module 14 is specifically used for: Based on the high-precision pose information, the spatial position and orientation of the UAV are obtained to determine the starting position of the path search and the position feedback when tracking the path. Based on the UAV's kinematic model and the starting position, feasible trajectory segments per unit time are calculated using dynamic integration, and collision detection is performed based on the current local map. A cost function is then used to guide the path towards the target point.

[0043] The trajectory fitting module 15 is used to use the key nodes of the initial path as control points, perform trajectory fitting using a third-order B-spline curve, and perform trajectory smoothing and safety optimization through an optimization function to generate the final flight trajectory.

[0044] Furthermore, the trajectory fitting module 15 is specifically used for: The key nodes of the initial path are used as control points, and the shortest distance between the control points and obstacles is calculated using the convex hull property of the uniform third-order B-spline curve. Euclidean distance field is introduced to obtain the distance information from any point in space to the nearest obstacle. Define a smoothing cost function and use gradient descent to iteratively optimize the control points to minimize the total cost function, thereby generating the final flight trajectory.

[0045] The functions or operation steps implemented by the above modules and units are largely the same as those in the above method embodiments, and will not be repeated here.

[0046] The UAV indoor navigation simulation system based on laser SLAM provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the system embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0047] Example 3 This invention also proposes a computer, please refer to [link / reference]. Figure 5 The computer shown in the third embodiment of the present invention includes a memory 10, a processor 20, and a computer program 30 stored in the memory 10 and executable on the processor 20. When the processor 20 executes the computer program 30, it implements the above-described laser SLAM-based UAV indoor navigation simulation method.

[0048] The memory 10 includes at least one type of storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 10 can be an internal storage unit of a computer, such as the computer's hard disk. In other embodiments, the memory 10 can be an external storage device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Furthermore, the memory 10 can include both internal and external storage units of the computer. The memory 10 can be used not only to store application software and various types of data installed on the computer, but also to temporarily store data that has been output or will be output.

[0049] In some embodiments, the processor 20 may be an electronic control unit (ECU), a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip, used to run program code stored in the memory 10 or process data, such as executing access restriction programs.

[0050] It should be pointed out that, Figure 5 The structure shown does not constitute a limitation on the computer. In other embodiments, the computer may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0051] This invention also proposes a storage medium storing a computer program that, when executed by a processor, implements the laser SLAM-based indoor navigation simulation method for unmanned aerial vehicles as described above.

[0052] Those skilled in the art will understand that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0053] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0054] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0055] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0056] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A laser SLAM-based indoor navigation simulation method for unmanned aerial vehicles (UAVs), characterized in that, include: Environmental point cloud data is acquired by scanning with lidar, and distortion removal is performed on the environmental point cloud data to obtain a low-precision odometer at the first hertz. The low-precision odometer is used to estimate linear velocity and angular velocity to achieve motion distortion correction of each frame of environmental point cloud data, and a map is constructed based on the distortion-corrected distortion-free point cloud to obtain a high-precision odometer at the second hertz and the corresponding environmental map information, wherein the first hertz is greater than the second hertz. Based on the pose information provided by the high-precision odometer, the low-precision odometer and the high-precision odometer are fused to obtain high-precision pose information at the second hertz. Based on the environmental map information and the high-precision pose information, a hybrid A* path search algorithm is used to perform path search in order to generate an initial path that satisfies the kinematic constraints of the UAV. The key nodes of the initial path are used as control points. A third-order B-spline curve is used for trajectory fitting, and the trajectory is smoothed and optimized for safety using an optimization function to generate the final flight trajectory.

2. The UAV indoor navigation simulation method based on laser SLAM according to claim 1, characterized in that, Based on the environmental map information and the high-precision pose information, the steps of using a hybrid A* path search algorithm to generate an initial path that satisfies the kinematic constraints of the UAV include: Based on the high-precision pose information, the spatial position and orientation of the UAV are obtained to determine the starting position of the path search and the position feedback when tracking the path. Based on the UAV's kinematic model and the starting position, feasible trajectory segments per unit time are calculated using dynamic integration, and collision detection is performed based on the current local map. A cost function is then used to guide the path towards the target point.

3. The UAV indoor navigation simulation method based on laser SLAM according to claim 2, characterized in that, The formula for calculating the cost function is as follows: In the formula, This represents the cumulative actual cost from the initial state to the current state. The path the drone took from the initial state to the current state traversed a total of [number missing]. There are discrete grid cells, and the control input within each grid cell is... The corresponding flight time is , This represents the estimated cost from the current state to the final state, where... Euclidean distance. This refers to the distance to Manhattan.

4. The UAV indoor navigation simulation method based on laser SLAM according to claim 1, characterized in that, The steps for generating the final flight trajectory include: using key nodes of the initial path as control points, fitting the trajectory with a third-order B-spline curve, and smoothing and optimizing the trajectory for safety using an optimization function. The key nodes of the initial path are used as control points, and the shortest distance between the control points and obstacles is calculated using the convex hull property of the uniform third-order B-spline curve. Euclidean distance field is introduced to obtain the distance information from any point in space to the nearest obstacle. Define a smoothing cost function and use gradient descent to iteratively optimize the control points to minimize the total cost function, thereby generating the final flight trajectory.

5. The UAV indoor navigation simulation method based on laser SLAM according to claim 4, characterized in that, The formula for calculating the smoothing cost function is as follows: In the formula, For point arrive and The straight-line distance, when The smaller the value, the more points... arrive and The closer the straight-line distance, the better.

6. A laser SLAM-based indoor navigation simulation system for unmanned aerial vehicles, characterized in that, include: The data acquisition module is used to acquire environmental point cloud data by scanning with lidar and to remove distortion from the environmental point cloud data in order to obtain a low-precision odometer at the first hertz. The data processing module is used to estimate linear velocity and angular velocity using the low-precision odometer to achieve motion distortion correction of each frame of environmental point cloud data, and to construct a map based on the distortion-corrected distortion-free point cloud to obtain a high-precision odometer at the second hertz and the corresponding environmental map information, wherein the first hertz is greater than the second hertz. The data fusion module is used to fuse the low-precision odometer and the high-precision odometer based on the pose information provided by the high-precision odometer to obtain high-precision pose information at the second hertz. The path search module is used to perform path search using a hybrid A* path search algorithm based on the environmental map information and the high-precision pose information, so as to generate an initial path that satisfies the kinematic constraints of the UAV. The trajectory fitting module is used to use the key nodes of the initial path as control points, perform trajectory fitting using third-order B-spline curves, and optimize the trajectory for smoothing and safety through an optimization function to generate the final flight trajectory.

7. The UAV indoor navigation simulation system based on laser SLAM according to claim 6, characterized in that, The path search module is specifically used for: Based on the high-precision pose information, the spatial position and orientation of the UAV are obtained to determine the starting position of the path search and the position feedback when tracking the path. Based on the UAV's kinematic model and the starting position, feasible trajectory segments per unit time are calculated using dynamic integration, and collision detection is performed based on the current local map. A cost function is then used to guide the path towards the target point.

8. The UAV indoor navigation simulation system based on laser SLAM according to claim 6, characterized in that, The trajectory fitting module is specifically used for: The key nodes of the initial path are used as control points, and the shortest distance between the control points and obstacles is calculated using the convex hull property of the uniform third-order B-spline curve. Euclidean distance field is introduced to obtain the distance information from any point in space to the nearest obstacle. Define a smoothing cost function and use gradient descent to iteratively optimize the control points to minimize the total cost function, thereby generating the final flight trajectory.

9. A storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the UAV indoor navigation simulation method based on laser SLAM as described in any one of claims 1 to 5.

10. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the UAV indoor navigation simulation method based on laser SLAM as described in any one of claims 1 to 5.

Images