Automatic obstacle avoidance system of unmanned aerial vehicle

Abstract

The invention discloses an unmanned aerial vehicle automatic obstacle avoidance system, and belongs to the technical field of unmanned aerial vehicle control. The system comprises a power system, a radar scanning system and a camera imaging system. The power system comprises a vertical power unit, a horizontal power unit and an adjusting power unit and is used for providing flight power for the unmanned aerial vehicle. A radar scanning system scans an airspace in front of a predetermined flight path of the unmanned aerial vehicle, weighted averaging is performed on multiple radar scanning data to obtain airspace information T, and impurity information is filtered out through screening. A camera imaging system shoots a real-time scene of an airspace in front of a predetermined flight path, and airspace information F is determined. And establishing a three-dimensional coordinate by taking the unmanned aerial vehicle as a coordinate origin, respectively converting the airspace information T and F into coordinate information, and determining real-time airspace information by judging a matching degree. The system further comprises an image processing center and a mode switching system, the mode switching system can be switched between a manual mode and an automatic driving mode, and when it is detected that the preset flight path condition is poor, the control system automatically changes the flight path. The flight safety and the automation degree of the unmanned aerial vehicle can be improved.

Description

Technical Field

[0001] This invention relates to the field of drone obstacle avoidance technology, specifically to an automatic obstacle avoidance system for drones. Background Technology

[0002] With the rapid development of drone technology, drones have been widely used in aerial photography, logistics delivery, agricultural plant protection, and security monitoring. However, the ability of drones to fly autonomously and avoid obstacles in complex environments remains a key technical challenge restricting their further application. Existing drone obstacle avoidance systems mainly rely on a single sensor for obstacle detection, which suffers from insufficient detection accuracy and poor reliability in complex flight environments.

[0003] Currently, obstacle avoidance technology for unmanned aerial vehicles (UAVs) mainly employs two methods: radar detection and visual recognition. Chinese patent CN111399535A discloses a UAV obstacle avoidance method. This method acquires radar detection data and camera visual data, and based on the detection and visual data, determines if there are obstacles within a predetermined airspace along the UAV's flight path. If so, it causes the UAV to pause immediately and execute the corresponding obstacle avoidance mode [CN111399535A]. Chinese patent CN119200625A discloses a dynamic planning method for UAV patrol paths in a three-dimensional scene. This method obtains distance measurements in front of the UAV and measured depth measurements below the UAV by scanning the depth information of the flight area in real time. Based on the distance and depth measurements, it determines the collision risk and performs automatic obstacle avoidance [CN119200625A]. Chinese patent CN103984357B discloses an automatic obstacle avoidance flight system for unmanned aerial vehicles based on a panoramic stereo imaging device. It uses a catadioptric panoramic stereo imaging device to collect three-dimensional scenes from all directions and uses the geographic coordinate information provided by the navigation module to detect obstacles and determine the effective flight path [CN103984357B].

[0004] In addition, some technical solutions attempt to improve obstacle avoidance performance by using multi-sensor fusion. Chinese patent CN108334103B discloses a multi-distance obstacle avoidance method for UAVs, which acquires obstacle information at multiple distance levels in the UAV's flight airspace and uses multiple obstacle avoidance mechanisms in parallel to predictively avoid large obstacles and moving obstacles within a medium distance range on the flight path [CN108334103B]. Chinese patent CN108227738B discloses an obstacle avoidance method for UAVs, which uses millimeter-wave radar to detect obstacles, combines video images captured by a front-facing camera to analyze obstacle types, and comprehensively analyzes obstacle shape and distance information to construct a grid map for path planning [CN108227738B].

[0005] However, existing technologies still suffer from the following problems: First, the solution combining a dedicated remote controller with a drone suffers from limited display functionality, limited control system processing power, and an inability to support complex flight maneuvers. Second, current drone autopilot control systems can only fly along predetermined flight paths, lacking flexible path adjustment capabilities. Most importantly, relying solely on radar warnings results in inaccurate airspace information, making them susceptible to environmental interference and equipment errors, and unable to effectively handle obstacle avoidance. These problems severely impact the flight safety and reliability of drones in complex environments, limiting the further development and application of autonomous drone flight technology. Summary of the Invention

[0006] (a) Technical problems to be solved To address the shortcomings of existing technologies that combine dedicated remote controllers with drones, such as limited remote controller display functions, limited control system processing capabilities, and inability to support complex flight maneuvers, as well as the limitations of current drone autopilot control systems that can only fly along predetermined flight paths and rely solely on radar warnings resulting in inaccurate airspace information and ineffective obstacle avoidance, this paper proposes an automatic obstacle avoidance system for drones. This system achieves warnings through a combination of radar scanning and camera imaging, significantly improving the accuracy of airspace information and providing reliable assurance for unmanned autopilot control.

[0007] (II) Technical Solution The technical solution adopted by the present invention to solve its technical problem is: to provide an automatic obstacle avoidance system for unmanned aerial vehicles (UAVs), including a power system, wherein the power system includes a vertical power unit, a horizontal power unit and an adjustment power unit, for providing flight power for the UAV; A radar scanning system is used to scan the airspace ahead of the UAV's predetermined flight path to obtain radar scanning data t of the airspace. The radar scanning data t generated multiple times is weighted and averaged to obtain the airspace information T of the predetermined flight path. A camera imaging system is used to capture real-time images of the airspace ahead of a UAV's predetermined flight path, and to determine the airspace information F of the predetermined flight path based on the real-time images. The data processing system is used to establish three-dimensional coordinates with the UAV as the coordinate origin, convert airspace information T and airspace information F into first coordinate information and second coordinate information respectively, determine the matching degree between the first coordinate information and the second coordinate information, and determine the real-time airspace information of the predetermined flight path based on the matching degree. The control system is used to control the drone to automatically avoid obstacles based on the processed airspace information.

[0008] Preferably, the radar scan data t is filtered to remove impurity information, wherein the radar scan data t is generated by the radar scanning system in the nth scan. n The radar scanning system generates radar scanning data t during its (n+1)th scan. n+1 The radar scan data t n+1 The shared information is t n+1 -A1+B1, for spatial information t n and sharing information n+1 -A is used for matching to obtain the matching value. The true spatial information T of the spatial domain is then determined based on the matching value. n .

[0009] Furthermore, the spatial information T n By matching the three-dimensional coordinates and taking the moving UAV as the origin, the airspace information T is obtained. n The three-dimensional coordinate information Q1, the spatial information T n In three-dimensional coordinates, it is a dynamic straight line, which is K-correlated with the UAV's predetermined flight path.

[0010] Optionally, A represents the airspace information added by the UAV on the predetermined flight path from the nth scan to the (n+1)th scan, and the time from the nth scan to the (n+1)th scan is t1.

[0011] Preferably, the camera imaging system performs imaging synchronously while the radar scanning system is scanning, and transmits the spatial information F generated from each imaging session. n By matching the three-dimensional coordinates, the spatial information F is obtained. n The three-dimensional coordinate information Q2, and the spatial information F n The three-dimensional coordinate information Q2 and the spatial information T n The three-dimensional coordinate information Q1 is compared to obtain the matching value P. The error between Q1 and Q2 is set to not exceed Y. If P < Y, then the spatial information F is determined. n With spatial information T n Consistent; if P > Y, then determine the spatial information F. n With spatial information T n If they are inconsistent, then the spatial information F is judged. n With spatial information T n The accuracy.

[0012] Furthermore, it also includes an image processing center that will process the acquired spatial information F n The image is sent to an image processing center, which needs to determine the spatial information F. n The authenticity of the spatial information F is determined when judging its validity. n When the authenticity is poor and P < Y, then the spatial information T is judged.n The authenticity is also poor, so we need to reacquire the spatial information T. n and airspace information F n When judging the spatial information F n When the authenticity of the information is different and P > Y, then the spatial information T is judged. n If the authenticity is optimal, then the spatial information F is reacquired. n When judging the spatial information F n When the authenticity is good and P < Y, then the spatial information T is judged. n The authenticity of this airspace information T is excellent. n and airspace information F n It can be used directly; when judging the spatial information F n When the authenticity is optimal and P > Y, then the spatial information T is judged. n The authenticity is poor, so spatial information T is reacquired. n .

[0013] Optionally, the airspace information includes moving object information, weather information, and obstacle information.

[0014] Preferably, it also includes a mode switching system, which is used to switch the flight mode of the UAV, including manual mode and autopilot mode. When the flight mode is manual mode, the control system is turned off.

[0015] Furthermore, when the flight mode is autopilot mode, the control system is activated. When the airspace information acquired by the radar scanning system and camera imaging system shows poor conditions for the predetermined flight path, the control system controls the power system to change the predetermined flight path of the UAV.

[0016] (III) Beneficial Effects Compared with the prior art, the present invention provides an automatic obstacle avoidance system for unmanned aerial vehicles (UAVs), which has the following beneficial effects: The beneficial effects of this invention are as follows: Compared with the prior art, this invention uses both radar scanning system and camera imaging system to compare and warn, which greatly improves the accuracy of airspace information and provides reliable protection for the unmanned autonomous driving control of UAVs; through real-time comparison and updating of data, it can quickly filter data that meets the conditions, improve data accuracy, and achieve effective automatic obstacle avoidance; by establishing a three-dimensional coordinate system and matching degree judgment mechanism, it effectively solves the problem of inaccurate information from a single radar warning system, and significantly improves the reliability and safety of UAV automatic obstacle avoidance. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1: This example provides an automatic obstacle avoidance system for unmanned aerial vehicles (UAVs). The system mainly includes a power system, a radar scanning system, a camera imaging system, an image processing center, a mode switching system, and a control system.

[0019] The propulsion system, as the core driving component of the UAV, includes a vertical propulsion unit, a horizontal propulsion unit, and an adjustable propulsion unit. The vertical propulsion unit provides upward lift through multiple propellers, enabling the UAV to take off, land, and hover vertically. The horizontal propulsion unit provides thrust for forward, backward, and lateral movement. The adjustable propulsion unit adjusts the power output of each propulsion unit to achieve precise attitude control and flight trajectory adjustment in three-dimensional space. These three propulsion units work together to provide comprehensive flight propulsion support for the UAV.

[0020] The radar scanning system employs millimeter-wave radar technology to scan the airspace ahead of the UAV's predetermined flight path. The system performs continuous scans at fixed time intervals, generating radar scan data t for each scan. The system then performs a weighted average of the multiple generated radar scan data t to obtain the airspace information T for the predetermined flight path. During data processing, the system filters out impurities from the radar scan data t. Specifically, the radar scanning system generates radar scan data t during the nth scan. n The (n+1)th scan generates radar scan data t. n+1 Radar scan data t n+1 The shared information is t n+1 -A+B. The system targets spatial information t. n and sharing information n+1 -A is used for matching to obtain the matching value. The true spatial information T of the spatial domain is then determined based on the matching value. n Here, A represents the airspace information added by the UAV along the predetermined flight path from the nth scan to the (n+1)th scan, and the time from the nth scan to the (n+1)th scan is t1.

[0021] The camera imaging system works synchronously with the radar scanning system to capture real-time images of the airspace ahead of the UAV's predetermined flight path. The system images simultaneously with the radar scanning system, and uses image recognition and analysis technology to determine the airspace information F of the predetermined flight path based on the real-time image. The system then processes the airspace information F generated from each imaging session. nBy matching with three-dimensional coordinates, spatial information F is obtained. n The three-dimensional coordinate information Q2.

[0022] The system establishes a three-dimensional coordinate system with the UAV as the origin, and transforms the spatial information T and F into first coordinate information and second coordinate information, respectively. n By matching the three-dimensional coordinates and taking the moving UAV as the origin, the spatial information T is obtained. n The three-dimensional coordinate information Q1. Spatial information T. n In three-dimensional coordinates, it is represented as a dynamic straight line, which is K-correlated with the UAV's predetermined flight path.

[0023] The system determines the real-time airspace information of the predetermined flight path by judging the matching degree between the first coordinate information and the second coordinate information. Specifically, the system will use the airspace information F n The three-dimensional coordinate information Q2 and the spatial information T n The system compares the three-dimensional coordinate information Q1 with the target coordinate information to obtain a matching value P. The system sets the error between Q1 and Q2 to not exceed Y. If P is less than Y, then the spatial information F is determined. n With spatial information T n Consistent; if P is greater than Y, then determine the spatial information F. n With spatial information T n Inconsistency leads to the determination of spatial information F n The accuracy of the spatial information Tn.

[0024] The image processing center receives spatial information F from the camera imaging system. n And determine its authenticity. When determining the spatial information F n When the authenticity of the information is poor and P is less than Y, then the spatial information T is judged. n The accuracy is also poor, at which point the system re-acquires the spatial information T. n and airspace information F n When judging the spatial information F n If the authenticity of the information is poor and P is greater than Y, then the spatial information T is judged. n If the accuracy is optimal, the system then re-acquires the spatial information F. n When judging the spatial information F n If the authenticity is good and P is less than Y, then the spatial information T is judged. n The authenticity of this airspace information T is excellent. n and airspace information F n It can be used directly; when judging the spatial information F n If the authenticity is good and P is greater than Y, then the spatial information T is judged. nThe authenticity is poor, at which point the system re-acquires the spatial information T. n .

[0025] The airspace information acquired by the system includes moving object information, weather information, and obstacle information. Moving object information covers the position, speed, and trajectory of dynamic targets such as other aircraft and birds; weather information includes meteorological parameters that affect flight safety, such as wind speed, wind direction, and visibility; and obstacle information includes the position and size data of static obstacles such as buildings, mountains, and utility poles.

[0026] The mode switching system is used to switch the drone's flight mode, which includes manual mode and autopilot mode. When the flight mode is manual, the control system is off, and the drone is manually controlled by the operator. When the flight mode is autopilot, the control system is on, and the drone flies autonomously according to a preset program.

[0027] The control system plays a central role in autopilot mode. When the airspace information acquired by the radar scanning system and camera imaging system indicates poor conditions for the predetermined flight path, the control system immediately analyzes the position, size, and movement trend of obstacles, calculates the optimal obstacle avoidance path, and controls the propulsion system to change the UAV's predetermined flight path. By adjusting the power output of the vertical propulsion unit, horizontal propulsion unit, and regulating propulsion unit, the control system enables the UAV to perform rapid obstacle avoidance maneuvers, ensuring flight safety.

[0028] The entire system achieves comprehensive monitoring and accurate judgment of the flight airspace through dual perception of radar scanning and camera imaging, combined with intelligent analysis by the image processing center. This provides the UAV with reliable automatic obstacle avoidance capabilities, greatly improving the safety and intelligence level of UAV flight.

[0029] Example 2: This example provides an electronic device including at least one processor and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor. When executed by the at least one processor, the computer program enables the at least one processor to perform various functions applied to the flight control system of an electric unmanned aerial vehicle.

[0030] The electronic device's processor employs a high-performance multi-core processor architecture, possessing powerful parallel computing capabilities, enabling it to simultaneously process multiple sensor data streams and complex flight control algorithms. The memory utilizes high-speed DDR4 RAM with a capacity of 8GB, ensuring smooth system operation and real-time data processing. The memory and processor are connected via a high-speed bus, with a data transfer rate reaching 3200MHz, guaranteeing efficient data exchange.

[0031] The computer program stored in the memory contains the complete control logic for the drone's automatic obstacle avoidance system. This program first controls the power system, which includes a vertical power unit, a horizontal power unit, and an adjustment power unit. The vertical power unit consists of four motor-driven propellers, providing the drone with lift and descent power; the horizontal power unit achieves forward, backward, left, and right movement by adjusting the speed differences of each propeller; and the adjustment power unit achieves fine-tuning and stabilization of the drone's attitude by precisely controlling the power output of each motor.

[0032] The radar scanning system's control program is responsible for scanning the airspace ahead of the UAV's predetermined flight path, obtaining radar scan data t. The program performs a weighted average of multiple generated radar scan data t to obtain the airspace information T for the predetermined flight path. For the radar scan data t, the program executes a filtering algorithm to remove impurities. The radar scanning system generates radar scan data t during its nth scan. n The (n+1)th scan generates radar scan data t. n+1 Radar scan data t n+1 The shared information is t n+1 -A+B. The program targets spatial information t. n and sharing information n+1 -A is used for matching to obtain the matching value. The true spatial information T of the spatial domain is then determined based on the matching value. n Where A represents the airspace information added by the UAV on the predetermined flight path from the nth scan to the (n+1)th scan, and the time from the nth scan to the (n+1)th scan is t1.

[0033] The program matches the spatial information Tn with three-dimensional coordinates, using the moving UAV as the origin, to obtain the spatial information T. n The three-dimensional coordinate information Q1. Spatial information T. n In three-dimensional coordinates, it is represented as a dynamic straight line, which is K-correlated with the UAV's planned flight path.

[0034] The camera imaging system's control program captures real-time images of the airspace ahead of the UAV's predetermined flight path and determines the airspace information F of the predetermined flight path based on these real-time images. The camera imaging system simultaneously performs imaging while the radar scanning system is scanning, and records the airspace information F generated from each imaging session. n By matching with three-dimensional coordinates, spatial information F is obtained. n The program will use the three-dimensional coordinate information Q2. The program will then use the spatial information F. n The three-dimensional coordinate information Q2 and the spatial information T n The three-dimensional coordinate information Q1 is compared to obtain the matching value P. The program is set that the error between Q1 and Q2 does not exceed Y. If P is less than Y, then the spatial information F is judged. n With spatial information Tn Consistent; if P is greater than Y, then determine the spatial information F. n With spatial information T n Inconsistency leads to the determination of spatial information F n With spatial information T n The accuracy.

[0035] The electronic device also integrates an image processing center module, and the program will acquire the spatial information F n The data is sent to the image processing center for analysis. The image processing center determines the spatial information F. n The authenticity of the spatial information F is determined when judging its validity. n When the authenticity of the information is poor and P is less than Y, then the spatial information T is judged. n The authenticity is also poor, so we need to reacquire the spatial information T. n and airspace information F n When judging the spatial information F n If the authenticity of the information is poor and P is greater than Y, then the spatial information T is judged. n If the authenticity is optimal, then the spatial information F is reacquired. n When judging the spatial information F n If the authenticity is good and P is less than Y, then the spatial information T is judged. n The authenticity of this airspace information T is excellent. n and airspace information F n It can be used directly; when judging the spatial information F n If the authenticity is good and P is greater than Y, then the spatial information T is judged. n The authenticity is poor, so spatial information T is reacquired. n .

[0036] The program establishes a three-dimensional coordinate system with the UAV as the origin, transforming airspace information T and F into first and second coordinate information, respectively. An algorithm determines the matching degree between the first and second coordinate information, and based on this matching degree, determines the real-time airspace information for the predetermined flight path. Airspace information includes moving object information, weather information, and obstacle information; the program can classify, process, and comprehensively analyze these different types of information.

[0037] The electronic device also includes a control program for a mode switching system, which is used to switch the drone's flight modes. Flight modes include manual mode and autopilot mode. When the flight mode is manual, the control system disables the automatic obstacle avoidance function, and the flight is controlled by a human operator. When switching to autopilot mode, the program activates the full automatic obstacle avoidance system, enabling intelligent flight of the drone.

[0038] This electronic device, through its integrated hardware platform and software system, provides complete support for the drone flight control system, ensuring safe flight and precise obstacle avoidance in complex environments.

[0039] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An automatic obstacle avoidance system for unmanned aerial vehicles (UAVs), characterized in that, include: A power system, comprising a vertical power unit, a horizontal power unit, and an adjustable power unit, is used to provide flight power for the UAV. A radar scanning system is used to scan the airspace ahead of the UAV's predetermined flight path to obtain radar scanning data t of the airspace. The radar scanning data t generated multiple times is weighted and averaged to obtain the airspace information T of the predetermined flight path. A camera imaging system is used to capture real-time images of the airspace ahead of a UAV's predetermined flight path, and to determine the airspace information F of the predetermined flight path based on the real-time images. A three-dimensional coordinate system is established with the UAV as the origin. The airspace information T and F are converted into first coordinate information and second coordinate information, respectively. The matching degree between the first coordinate information and the second coordinate information is judged, and the real-time airspace information of the predetermined flight path is determined based on the matching degree.

2. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 1, characterized in that: The radar scan data t is filtered to remove impurities, including: The radar scanning system generates radar scanning data t during the nth scan. n The radar scanning system generates radar scanning data t during its (n+1)th scan. n+1 The radar scan data t n+1 The shared information is t n+1 -A1+B1, for spatial information t n and sharing information n+1 -A is used for matching to obtain the matching value. The true spatial information T of the spatial domain is then determined based on the matching value. n .

3. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 2, characterized in that: The spatial information T n By matching the three-dimensional coordinates and taking the moving UAV as the origin, the airspace information T is obtained. n The three-dimensional coordinate information Q1, the spatial information T n In three-dimensional coordinates, it is a dynamic straight line, which is K-correlated with the UAV's predetermined flight path.

4. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 2, characterized in that: A represents the airspace information added by the UAV on the predetermined flight path from the nth scan to the (n+1)th scan, and the time from the nth scan to the (n+1)th scan is t1.

5. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 1, characterized in that: The camera imaging system performs imaging synchronously with the radar scanning system, and transmits the spatial information F generated from each imaging session. n By matching the three-dimensional coordinates, the spatial information F is obtained. n The three-dimensional coordinate information Q2, and the spatial information F n The three-dimensional coordinate information Q2 and the spatial information T n The three-dimensional coordinate information Q1 is compared to obtain the matching value P. The error between Q1 and Q2 is set to not exceed Y. If P < Y, then the spatial information F is determined. n With spatial information T n Consistent; if P > Y, then determine the spatial information F. n With spatial information T n If they are inconsistent, then the spatial information F is judged. n With spatial information T n The accuracy.

6. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 5, characterized in that: It also includes an image processing center, which will acquire the spatial information F n The image is sent to an image processing center, which needs to determine the spatial information F. n The authenticity of the spatial information F is determined when judging its validity. n When the authenticity is poor and P < Y, then the spatial information T is judged. n The authenticity is also poor, so we need to reacquire the spatial information T. n and airspace information F n When judging the spatial information F n When the authenticity of the information is different and P > Y, then the spatial information T is judged. n If the authenticity is optimal, then the spatial information F is reacquired. n When judging the spatial information F n When the authenticity is good and P < Y, then the spatial information T is judged. n The authenticity of this airspace information T is excellent. n and airspace information F n It can be used directly; when judging the spatial information F n When the authenticity is optimal and P > Y, then the spatial information T is judged. n The authenticity is poor, so spatial information T is reacquired. n .

7. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 1, characterized in that: The airspace information includes information on moving objects, weather information, and obstacle information.

8. The automatic obstacle avoidance system for unmanned aerial vehicles according to claim 1, characterized in that: It also includes a mode switching system, which is used to switch the flight mode of the UAV. The flight mode includes manual mode and autopilot mode. When the flight mode is manual mode, the control system is turned off.

9. An automatic obstacle avoidance system for unmanned aerial vehicles according to claim 8, characterized in that: When the flight mode is autopilot mode, the control system is activated. When the airspace information acquired by the radar scanning system and camera imaging system shows poor conditions for the predetermined flight path, the control system controls the power system to change the predetermined flight path of the UAV.

10. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to execute the flight control system for an electric unmanned aerial vehicle as described in any one of claims 1-8.