Unmanned surrounding sensing aircraft method based on laser radar

Abstract

The invention discloses an unmanned surrounding sensing aircraft method based on a laser radar. The three-dimensional space information of the surrounding environment is obtained in real time by emitting the laser beam and receiving the returned reflection signal. These lidar devices can cover various directions around the aircraft and can scan at high frequency to capture more details and changes. In order to realize efficient data processing and perceptual analysis, the method also comprises a powerful processing unit which is used for receiving and processing mass data transmitted by the laser radar equipment. The processing unit adopts an advanced algorithm and a mode recognition technology, can convert data acquired by the laser radar into available information, and generates a surrounding environment map for path planning and obstacle avoidance of an aircraft. The application of the method is expected to improve the safety, reliability and intelligent level of the unmanned tractor.

Description

Technical Field

[0001] The present invention relates to the field of unmanned driving, and in particular to a laser radar-based unmanned driving surrounding perception aircraft method. Background Art

[0002] With the development of unmanned driving technology, unmanned aircraft are increasingly used in the aviation field. However, there are still many challenges in achieving the safety and reliability of unmanned tractors, one of which is to ensure accurate perception of the surrounding environment and obstacle avoidance capabilities. To address this problem, current solutions mainly rely on sensor technology.

[0003] As a commonly used environmental perception sensor, LiDAR obtains 3D structural information of the surrounding environment by emitting laser beams and measuring their return time. However, the current application of LiDAR systems in autonomous driving still faces some limitations, such as accuracy, processing efficiency, and reliability in complex environments, which still have room for improvement. Summary of the invention

[0004] The present invention provides a method for an unmanned aircraft to perceive its surroundings based on a laser radar, which makes full use of three-dimensional point cloud data acquired by a laser radar sensor to achieve accurate perception of the aircraft's surroundings, target recognition, and obstacle avoidance decisions.

[0005] Specifically, the method of the present invention comprises the following steps: first, one or more laser radar sensors are installed on the unmanned tractor to obtain point cloud data of the surrounding environment. These laser radar sensors can be Velodyne VLP-16 multi-line laser radar or Velodyne HDL-64E rotating laser radar to obtain high-resolution and omnidirectional environmental perception.

[0006] Then, using the 3D point cloud data acquired by the LiDAR sensor, point cloud segmentation is performed to divide the point cloud data into different clusters. This can be achieved through clustering algorithms such as distance-based clustering (such as the DBSCAN algorithm) or density-based clustering (such as the MeanShift algorithm), which clusters point clouds on the same object into a group.

[0007] Next, feature extraction is performed on each cluster to identify the type and location of the surrounding flying objects. Feature extraction can classify and locate objects based on features such as point cloud shape, motion trajectory, color or texture. For example, normal features, point cloud geometry descriptors (such as PFH descriptors) and deep learning methods (such as PointNet) can be used for feature extraction and classification.

[0008] Finally, according to the position and motion status of the identified nearby vehicles, decisions and planning are made to achieve obstacle avoidance and path optimization of the drone. Decisions and planning can be made based on information such as the distance, speed, and relative position between the perceived flying object and the drone, ensuring that the drone safely and effectively bypasses the flying object. For example, a dynamic path planning algorithm (such as the A* algorithm or the RRT* algorithm) can be used to plan an obstacle avoidance path.

[0009] The advantage of the present invention is that by making full use of the point cloud data obtained by the laser radar sensor, the method of the present invention can achieve all-round perception and accurate recognition of the surrounding environment of the unmanned aircraft. Compared with traditional sensor technology, laser radar can provide high-precision three-dimensional spatial information, thereby enhancing the aircraft's ability to recognize the surrounding environment.

[0010] Through the steps of point cloud segmentation and feature extraction, the method of the present invention can accurately separate objects in the environment and identify their types and locations. This provides an effective basis for the decision-making and planning of the aircraft, thereby realizing the functions of obstacle avoidance and path optimization. By avoiding collisions with other flying objects and selecting the best path, the unmanned aircraft can complete the flight mission safely and efficiently.

[0011] In addition, the method of the present invention can also be combined with multi-sensor fusion technology to further improve the perception ability and decision-making accuracy of unmanned driving. By fusing data with other sensors (such as RGB cameras, millimeter-wave radars or inertial measurement units), more comprehensive and robust environmental perception can be achieved, enhancing the safety and reliability of tractors in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 : Shows the rotating lidar sensor on the unmanned tractor, which is used to obtain point cloud data of the surrounding environment.

[0013] Figure 2 : The process of point cloud segmentation, feature extraction and feature extraction is demonstrated to segment the point cloud data in the environment into different clusters.

[0014] Figure 3 : It shows the process of the unmanned tractor making decisions and planning based on the perceived information to achieve obstacle avoidance and path optimization. DETAILED DESCRIPTION

[0015] The following will describe the implementation of the present invention in conjunction with specific examples. The implementation of the present invention can be appropriately changed and improved according to specific needs, and only one example implementation is provided below.

[0016] The example implementation includes installing a rotating laser radar sensor on an unmanned tractor to obtain three-dimensional point cloud data of the surrounding environment. The laser radar sensor generates point cloud data by rotating and scanning and measuring the distance and position information of surrounding objects. These point cloud data are pre-processed and filtered for subsequent environment perception and target recognition.

[0017] In terms of point cloud data processing, point cloud data is clustered and grouped through point cloud segmentation algorithms, and point clouds belonging to the same object are grouped together. Clustering algorithms can group point clouds according to features such as distance or density between point clouds to achieve point cloud data segmentation.

[0018] Next, feature extraction is performed on each clustered point cloud group. Feature extraction can be based on features such as shape, motion trajectory, color or texture of the point cloud. By analyzing the features of each point cloud group, the surrounding flying objects can be classified and located.

[0019] Based on the position and motion of the identified objects, the control unit of the unmanned tractor can make decisions and plans. Based on information such as the relative position and distance between the aircraft and the vehicle, the decision-making and planning algorithms can generate obstacle avoidance strategies and path planning for the unmanned tractor. By adjusting the heading, speed and attitude of the tractor, the unmanned vehicle can safely bypass obstacles and choose the optimal driving path.

[0020] In addition, the unmanned tractor can also transmit the aircraft information it senses to the ground control terminal through the wireless communication system. The ground control terminal can receive and analyze the object information in real time and perform remote monitoring and dispatch as needed. This can further improve the operational safety and reliability of the unmanned tractor.

[0021] In summary, the present invention provides a method and system for unmanned aircraft surrounding perception based on laser radar. By making full use of the point cloud data obtained by the laser radar sensor, all-round perception and accurate identification of the aircraft's surrounding environment are achieved. The method is simple, efficient, and reliable, and can meet the needs of unmanned aircraft in different application scenarios. Therefore, the present invention has broad application prospects and commercial value.

[0022] Through the implementation of the present invention, the unmanned tractor can accurately perceive and identify the surrounding environment through the laser radar sensor, thereby improving the safety and reliability of the aircraft. The unmanned tractor can avoid collisions with other objects in complex environments and choose the best path to move.

Claims

1. A method for an unmanned surrounding perception aircraft based on laser radar, comprising the following steps: a. Install one or more lidar sensors on the unmanned aircraft to obtain 3D point cloud data of the surrounding environment; b. Use the three-dimensional point cloud data obtained by the lidar sensor to perform point cloud segmentation and divide the point cloud data into different clusters; c. Extract features from each cluster to identify the type and location of surrounding flying objects; d. Make decisions and plans based on the identified position and motion status of the flying object to achieve obstacle avoidance and path optimization for the unmanned aircraft.

2. The method according to claim 1, wherein: The laser radar sensor in step a can use a multi-line laser radar or a rotating laser radar to obtain omnidirectional environmental perception. The laser radar sensor can scan and measure the distance and position information of surrounding objects to generate point cloud data.

3. The method according to claim 1 or 2, wherein: The point cloud segmentation in step b can use a clustering algorithm, such as distance-based clustering or density-based clustering, to divide the point cloud data into subsets with similar features. In this way, the point clouds on the same object can be grouped together to facilitate subsequent object recognition and positioning.

4. The method according to claim 1 or 2, wherein: The feature extraction in step c may include object classification and location based on features such as point cloud shape, motion trajectory, color or texture. By performing feature extraction on each cluster, different types of flying objects can be identified, such as other aircraft, drones, buildings or terrain.

5. The method according to claim 1, wherein: The decision and planning in step d can make obstacle avoidance decisions and path planning based on the distance, speed and relative position between the perceived flying object and the unmanned aircraft to ensure that the unmanned aircraft safely and effectively bypasses the flying object. The decision and planning algorithm can consider factors such as route optimization, slope adjustment and flight speed to provide the best flight path.

6. The method according to claim 1, wherein: The unmanned aircraft may further include a control unit for receiving and executing the path planning result generated in step d to control the movement of the aircraft. The control unit may include components such as a flight controller, an electric motor controller, and a steering gear controller to accurately control the attitude, speed, and heading of the aircraft.

7. The method according to claim 1, wherein: The sensed flying object information can also be transmitted to the ground control end through the wireless communication system, so as to remotely monitor and dispatch the operation status of the unmanned aircraft. The ground control end can receive and analyze the flying object information, conduct remote guidance and monitoring, and ensure the safety and reliability of the aircraft during operation.

8. The method according to claim 1, wherein: It is also possible to integrate the drone's surrounding perception information with the navigation system and autonomous flight control system to achieve comprehensive flight environment perception and autonomous decision-making capabilities. By combining surrounding perception information with the flight control system, the drone can make corresponding decisions based on real-time perception data, such as adjusting flight speed, changing course, or performing obstacle avoidance operations to ensure a safe flight path and avoid collisions with other flying objects.

9. The method according to claim 1, wherein: Multi-sensor fusion can also be used to combine lidar sensors with other sensors (such as cameras, radars or inertial measurement units) to comprehensively utilize data from different sensors for surrounding perception. Multi-sensor fusion can improve perception accuracy and robustness, further enhancing the safety and reliability of unmanned aircraft.

Images