Unmanned aerial vehicle automatic homing system and method

Abstract

The invention discloses an automatic homing system for an unmanned aerial vehicle, and the system comprises a multi-mode positioning module, an intelligent sensing module, an automatic homing module, an obstacle avoidance module and a signal enhancement module, and guarantees the safe homing of the unmanned aerial vehicle under the conditions of complex and changeable obstacle distribution and unconfused signals.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and in particular to an automatic homing system and method for unmanned aerial vehicles. Background Art

[0002] In recent years, drones have been widely used in many fields such as aerial photography, agricultural plant protection, power inspection, and express delivery due to their flexibility, convenience, and low operating costs. As the usage scenarios become increasingly complex and diverse, whether drones can smoothly and reliably return to the starting point or designated nest after the flight mission has completed has become a key issue that needs to be solved urgently.

[0003] Traditional drone homing often relies on a single positioning method, such as simple GPS positioning, which is prone to positioning deviation or even signal loss in environments such as urban high-rise buildings and valleys with dense forests where signal obstruction is severe. Faced with complex and changeable obstacle distribution and unstable communication conditions, the existing homing system lacks effective response strategies, which often leads to drone homing failure, causing equipment damage and even safety accidents. Summary of the invention

[0004] In order to overcome the above technical defects, the present invention provides a UAV automatic homing system and method.

[0005] In order to solve the above problems, the present invention is implemented according to the following technical solutions:

[0006] In a first aspect, the present invention provides an automatic homing system for a drone, characterized in that it comprises:

[0007] A multi-modal positioning module, which is integrated with a global satellite navigation unit, an inertial navigation unit, a visual positioning unit, and an auxiliary positioning unit based on a ground base station. The multi-modal positioning module is used to achieve positioning in various complex scenarios;

[0008] Intelligent perception module, which consists of a deep learning algorithm, several high-definition cameras and a map database. The intelligent perception module uses a deep learning algorithm to perform in-depth analysis on the images collected by several high-definition cameras, identify various landforms, and timely update the data in the map database to provide geographic reference for drone navigation;

[0009] An automatic homing module, which includes a nest recognition unit and a machine learning algorithm. The nest recognition unit ensures that the drone can accurately identify and dock with the nest, and the machine learning algorithm is used to optimize the drone's descent method and docking location selection;

[0010] An obstacle avoidance module, the obstacle avoidance module includes a laser radar and a visual sensor, and processes the data collected by the laser radar and the visual sensor to generate a real-time updated obstacle database, the obstacle database is used to improve the accuracy of obstacle avoidance;

[0011] The signal enhancement module is provided with a signal interference detection unit for real-time monitoring and adjusting the signal strength.

[0012] In combination with the first aspect, the present invention also provides a first implementation manner of the first aspect, wherein the positioning vision unit is installed at the bottom of the drone, and is used to capture ground feature points or preset landmarks, compare them with pre-stored map data through an image matching algorithm, calculate the relative position offset, and assist in correcting the positioning error.

[0013] In combination with the first aspect, the present invention also provides a second implementation of the first aspect, wherein the map database adopts a distributed storage architecture, stores basic geographic location information, and can construct a local map based on data sensed during the flight of the drone.

[0014] In combination with the first aspect, the present invention also provides a third implementation of the first aspect, wherein the nest identification unit is used to identify a specific electromagnetic signal emitted by the nest, and the UAV calculates the distance and direction of the nest based on the strength and phase parameters of the electromagnetic signal.

[0015] In combination with the first aspect, the present invention provides a fourth implementation of the first aspect, wherein the obstacle database is used to record the position, size, and motion status information of the obstacle.

[0016] In combination with the first aspect, the present invention further provides a fifth implementation of the first aspect, wherein the automatic homing system for the unmanned aerial vehicle further comprises: a power management module, the power management module comprising a power monitoring unit and a signal detection unit, and sending power warning information to the flight control processing module when the power is lower than a preset threshold, and the flight control processing module executes a homing program according to the power warning information;

[0017] Communication module: The communication module is used for data communication between the UAV and the machine nest and the ground control center.

[0018] In a second aspect, a method for automatically homing a drone is provided, characterized in that it comprises the following steps:

[0019] Step 1: Drone triggers homing command

[0020] The drone homing command is triggered when the battery level is below a threshold, the mission is completed, the drone fails, or an emergency occurs;

[0021] Step 2: The drone identifies nearby nests

[0022] The drone determines location information through a multi-modal positioning module; the drone accesses a built-in landmark image database or communicates with a ground control center to identify the nearest available nest;

[0023] Step 3: The drone sets a homing route

[0024] The drone’s intelligent perception module assists the positioning module, avoiding obstacles and dangerous areas through cameras and target recognition algorithms; the drone’s obstacle avoidance module marks obstacle locations on the landmark library and updates the obstacle avoidance path in real time;

[0025] Step 4: Land the drone and reset the system

[0026] When the drone approaches the nest, the automatic homing module identifies the type of drone and plans the docking location and descent method for the drone; the drone executes a precise landing procedure, including adjusting the flight attitude and speed to ensure a safe landing; after the drone lands, the execution unit reset procedure, including turning off the flight mode, performing hardware self-test and data recording; the drone's power management module monitors the power after landing and initiates emergency procedures when necessary.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The present invention discloses an automatic homing system for unmanned aerial vehicles, comprising: a multimodal positioning module, the multimodal positioning module is integrated in a global satellite navigation unit, an inertial navigation unit, a visual positioning unit and an auxiliary positioning unit based on a ground base station, the multimodal positioning module is used to realize real-time positioning in various complex scenes; an intelligent perception module, the intelligent perception module is composed of a deep learning algorithm, a plurality of high-definition cameras and a map database, the intelligent perception module performs in-depth analysis on the images collected by the high-definition camera through the deep learning algorithm, identifies various landforms and updates the data in the map database, and provides a geographical reference for the navigation of the unmanned aerial vehicle; an automatic homing module, the automatic homing module comprises a machine nest recognition unit and a machine learning algorithm, the machine nest recognition unit ensures that the unmanned aerial vehicle can accurately identify and dock with the machine nest, and the machine learning algorithm is used to optimize the descent mode and docking position selection of the unmanned aerial vehicle; an obstacle avoidance module, the obstacle avoidance module comprises a laser radar and a visual sensor, the data collected by the laser radar and the visual sensor are processed to generate a real-time updated obstacle database, and the obstacle database is used to improve the accuracy of obstacle avoidance; a signal enhancement module, the signal enhancement module is provided with a signal interference detection unit, which is used to monitor and adjust the signal strength in real time.

[0029] The present invention can ensure that the UAV can automatically home in various complex scenarios, ensure the flight safety of the UAV through the obstacle avoidance module, and prevent the UAV from losing its signal and causing the loss of the UAV by the signal enhancement module. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The specific embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings, wherein:

[0031] Figure 1 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0032] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0033] The automatic homing system for unmanned aerial vehicles described in the present invention is characterized in that the automatic homing unit for unmanned aerial vehicles comprises:

[0034] A multi-modal positioning module, which is integrated with a global satellite navigation unit, an inertial navigation unit, a visual positioning unit, and an auxiliary positioning unit based on a ground base station. The multi-modal positioning module is used to achieve real-time positioning in various complex scenarios;

[0035] Specifically, the multimodal positioning module uses a multimodal fusion method to enable the drone to adapt to the positioning needs in different scenarios. When the drone enters complex terrain such as urban canyons, mountainous forests, etc. where satellite signals are blocked, the inertial navigation unit measures the acceleration and angular velocity of the drone to maintain positioning for a short period of time; the visual positioning unit captures ground feature points or preset landmarks, uses image matching algorithms to compare with pre-stored map data, calculates relative position offsets, and assists in correcting positioning errors; the auxiliary positioning unit of the ground base station further enhances the accuracy and stability of positioning in specific operating areas, such as industrial parks and large farms. Using multiple positioning systems to work together, exchange data in real time, and compare and analyze the location information collected by each unit through the built-in intelligent fusion algorithm, ensure that the drone can obtain reliable real-time positioning at any time, and provide a solid foundation for subsequent flight decisions.

[0036] Intelligent perception module, which is composed of a deep learning algorithm, several high-definition cameras and a map database. The intelligent perception module uses a deep learning algorithm to perform in-depth analysis on the images captured by the high-definition camera, identifies various landforms, updates the data in the map database, and provides geographic reference for the navigation of the drone;

[0037] Specifically, several high-definition cameras are installed around the drone to collect all-round images of the drone during flight and transmit them to the drone's processing unit; the deep learning algorithm conducts in-depth analysis of the collected images and can identify various landform features such as mountains, rivers, buildings, roads, etc. Through continuous learning and comparison, it can match the identified landform information with the existing data in the map database, and update the map database in a timely manner once differences or new geographical features are found.

[0038] An automatic homing module, which includes a nest recognition unit and a machine learning algorithm. The nest recognition unit ensures that the drone can accurately identify and dock with the nest, and the machine learning algorithm is used to optimize the drone's descent mode and docking location selection;

[0039] Specifically, the robot algorithm is trained based on a large amount of historical flight data and simulation experiment data to learn the optimal landing method for different drone models and different environmental conditions. In open areas with strong winds, the robot algorithm will plan a landing path with a slightly larger inclination angle, a faster descent speed but a stable landing path, and choose a docking position in the nest with good wind protection; in narrow spaces, the landing path is smoother and more accurate, ensuring that the drone is safely docked at the designated location and avoids collisions with surrounding obstacles.

[0040] An obstacle avoidance module, the obstacle avoidance module includes a laser radar and a visual sensor, and processes the data collected by the laser radar and the visual sensor to generate a real-time updated obstacle database, the obstacle database is used to improve the accuracy of obstacle avoidance;

[0041] A signal enhancement module is provided with a signal interference detection unit, and the signal interference detection unit is used to monitor and adjust the signal strength in real time.

[0042] Specifically, the signal enhancement module is also equipped with a signal adjustment mechanism. Once signal interference or weakening is detected, the signal enhancement module immediately initiates the corresponding adjustment strategy. Automatically switch the communication frequency band to avoid interference sources; or enhance the signal transmission power through the power amplifier to ensure stable and reliable communication between the drone and the ground control center, the machine nest, etc.

[0043] In a specific implementation, the positioning vision unit is installed at the bottom of the drone to capture ground feature points or preset landmarks, compare them with pre-stored map data through an image matching algorithm, calculate the relative position offset, and assist in correcting the positioning error.

[0044] In a specific implementation, the map data adopts a distributed storage architecture, which not only stores basic geographic location information, but also dynamically constructs a local map based on the data sensed during the flight of the drone.

[0045] Specifically, the local map is the current drone operation area and the environment around the drone. The map contains information such as obstacles encountered during the operation and temporary landmarks, which can facilitate the navigation of the drone in complex and changeable local environments.

[0046] In a specific implementation, the obstacle database is used to record the location, size, and motion status information of obstacles.

[0047] Specifically, the UAV can adjust its flight attitude and flight path according to the obstacle information in the obstacle library data.

[0048] In a specific implementation, the nest identification unit is used to identify a specific electromagnetic signal emitted by the nest, and the drone calculates the distance and direction of the nest based on the strength and phase parameters of the electromagnetic signal.

[0049] Specifically, the machine nest emits a specific electromagnetic wave signal. After the drone's receiving device captures the electromagnetic signal, it calculates the distance and direction of the machine nest through a built-in triangulation algorithm based on the strength, phase and other parameters of the electromagnetic signal. In combination with visual recognition technology, a high-definition camera is used to identify and verify the appearance features of the machine nest, further improving the accuracy of machine nest recognition and avoiding misjudgment.

[0050] In a specific implementation, the drone automatic homing system further includes:

[0051] A power management module, the power management module includes a power monitoring unit and a signal detection unit, and sends power warning information to the flight control processing module when the power is lower than a preset threshold, and the flight control processing module executes a homing program according to the power warning information;

[0052] Specifically, the power management module samples and analyzes the battery voltage, current and other parameters at regular intervals, and calculates the remaining power and the expected flight time through the built-in power estimation model; when the power is lower than the preset threshold, the power monitoring unit sends a power warning message to the flight control processing module; the UAV begins to search for the distance to the nearest nest to determine whether it can be reached. If not, the flight control processing module will immediately interrupt the current task after receiving the warning and start the automatic homing program to ensure that the UAV returns to the nest safely before the power is exhausted; if the task can continue to be completed, it will be given priority.

[0053] Communication module: The communication module is used for data communication between the UAV and the machine nest and the ground control center.

[0054] Example 2

[0055] like Figure 1 As shown, a method for automatic homing of a drone includes the following steps:

[0056] S1: Drone triggers homing command

[0057] S2: Drone identifies nearby nests

[0058] S3: Drones plan their homing routes

[0059] S4: Drone landing and system reset

[0060] In a specific implementation, the drone homing command is triggered by the battery level being lower than a threshold, mission completion, self-failure, or emergency situation;

[0061] Specifically, when the power management module detects that the power is lower than the preset threshold, it automatically triggers the homing command; after the drone completes the tasks of aerial photography covering the designated area, delivering materials to the target location, and traversing the inspection route according to the preset task flow, it automatically switches to the homing mode and prepares to return to the nest. The drone's built-in fault detection system monitors the drone's operating status in real time. Once a fault is detected, such as abnormal motor speed, excessive deviation of flight control data, etc., it immediately triggers the homing command to make the drone return safely in the event of a fault as much as possible to avoid losses caused by a crash; when encountering severe weather that exceeds the drone's flight safety standards, or receiving an emergency recall command from the ground control center, the drone will also quickly start the homing program; in these cases, the drone will start the homing command.

[0062] In a specific implementation, the drone determines location information through a multi-modal positioning module; the drone accesses a built-in landmark image database or communicates with a ground control center to identify the nearest available nest;

[0063] Specifically, the drone determines its current location information through the multimodal positioning module, including latitude and longitude, altitude, and orientation relative to surrounding landmarks. Based on its own location and combined with the built-in map database, it preliminarily screens candidate areas of nearby nests; selects the nest closest to the drone, and the drone's receiving module identifies the specific electromagnetic waves emitted by the selected nest to determine the location of the nest; finally, the drone accesses the built-in landmark image database, uses a high-definition camera to capture images of the surrounding environment, and uses image recognition technology to find landmark features related to the nest, such as unique buildings and signboards around the nest, to further narrow the search range of the nest.

[0064] In one specific implementation, the drone’s intelligent perception module assists the positioning module to avoid obstacles and dangerous areas through cameras and target recognition algorithms; the drone’s obstacle avoidance module marks obstacle locations on a landmark library and updates the obstacle avoidance path in real time;

[0065] In detail, the drone formulates the homing route including intelligent perception and obstacle avoidance coordination and dynamic flight path adjustment.

[0066] Specifically, the drone's intelligent perception module assists the positioning module, using high-definition cameras and target recognition algorithms to scan the front and surrounding environment in real time to identify obstacles and dangerous areas. The obstacle avoidance module accurately marks the location of obstacles on the map based on the landmark library and data collected by the lidar and visual sensors, and combines the drone's current position, speed, direction, and nest position information to update the obstacle avoidance path in real time through the path planning algorithm. During the flight, as the environment changes, the drone continuously updates the obstacle database and map information, and adjusts the homing route in real time. During the flight, the drone periodically re-evaluates the safety of the flight path. If the original path is found to be risky, it will immediately re-plan and guide the drone to switch to a new homing path.

[0067] In a specific implementation, when the drone approaches the nest, the automatic homing module identifies the type of drone and plans the docking location and descent method for the drone; the drone executes a precise landing procedure, including adjusting the flight attitude and speed to ensure a safe landing; after the drone lands, the unit reset procedure is executed, including turning off the flight mode, performing hardware self-test and data recording.

[0068] Specifically, when the drone approaches the nest, the automatic homing module is activated. The drone performs precise landing according to the planned landing procedure, and the flight control system accurately adjusts the flight attitude, including pitch angle, roll angle, yaw angle, etc., while controlling the descent speed to ensure that the drone lands smoothly and accurately at the designated docking position. During the landing process, the visual positioning system and the nest recognition system are used to monitor the relative position of the drone and the nest in real time, and fine-tune and correct to ensure high landing accuracy. After the drone lands, the system reset procedure is executed immediately. Turn off the flight mode and stop the power output; then perform a hardware self-test to detect the hardware such as the motor, battery, sensor, and communication module to check whether there are any abnormal conditions, such as overheating, looseness, short circuit, etc., and record the self-test results; record the data of this flight, including flight trajectory, power consumption, obstacle information encountered, etc. The flight data will be used for subsequent flight analysis, system optimization, and troubleshooting.

[0069] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Therefore, any modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention shall still fall within the scope of the technical solution of the present invention.

Claims

1. An automatic homing system for unmanned aerial vehicles, characterized in that: The drone automatic homing unit comprises: A multi-modal positioning module, which is integrated with a global satellite navigation unit, an inertial navigation unit, a visual positioning unit, and an auxiliary positioning unit based on a ground base station. The multi-modal positioning module is used to achieve real-time positioning in various complex scenarios; Intelligent perception module, which is composed of a deep learning algorithm, several high-definition cameras and a map database. The intelligent perception module uses a deep learning algorithm to perform in-depth analysis on the images captured by the high-definition camera, identifies various landforms, updates the data in the map database, and provides geographic reference for the navigation of the drone; An automatic homing module, which includes a nest recognition unit and a machine learning algorithm. The nest recognition unit ensures that the drone can accurately identify and dock with the nest, and the machine learning algorithm is used to optimize the drone's descent mode and docking location selection; An obstacle avoidance module, the obstacle avoidance module includes a laser radar and a visual sensor, and processes the data collected by the laser radar and the visual sensor to generate a real-time updated obstacle database, the obstacle database is used to improve the accuracy of obstacle avoidance; The signal enhancement module is provided with a signal interference detection unit for real-time monitoring and adjusting the signal strength.

2. The automatic homing system for unmanned aerial vehicles according to claim 1, characterized in that: The positioning vision unit is installed at the bottom of the drone and is used to capture ground feature points or preset landmarks, compare them with pre-stored map data through an image matching algorithm, calculate the relative position offset, and assist in correcting the positioning error.

3. The automatic homing system for unmanned aerial vehicles according to claim 1, characterized in that: The map data adopts a distributed storage architecture, which not only stores basic geographic location information, but also dynamically constructs a local map based on the data sensed during the flight of the drone.

4. The automatic homing system for unmanned aerial vehicles according to claim 1, characterized in that: The nest identification unit is used to identify the specific electromagnetic signal emitted by the nest, and the UAV calculates the distance and direction of the nest based on the strength and phase parameters of the electromagnetic signal.

5. The automatic homing system for unmanned aerial vehicles according to claim 1, characterized in that: The obstacle database is used to record the location, size, and motion status information of obstacles.

6. The automatic homing system for unmanned aerial vehicles according to claim 1, characterized in that: The UAV automatic homing system also includes: A power management module, the power management module includes a power monitoring unit and a signal detection unit, and sends power warning information to the flight control processing module when the power is lower than a preset threshold, and the flight control processing module executes a homing program according to the power warning information; Communication module: The communication module is used for data communication between the UAV and the machine nest and the ground control center.

7. A method for automatic homing of a drone, implemented based on the automatic homing unit of the drone according to any one of claims 1 to 6, characterized in that: The following steps are involved: Step 1: Drone triggers homing command The drone homing command is triggered when the battery level is below a threshold, the mission is completed, the drone fails, or an emergency occurs; Step 2: The drone identifies nearby nests The drone determines location information through a multi-modal positioning module; the drone accesses a built-in landmark image database or communicates with a ground control center to identify the nearest available nest; Step 3: The drone sets a homing route The drone’s intelligent perception module assists the positioning module, avoiding obstacles and dangerous areas through cameras and target recognition algorithms; the drone’s obstacle avoidance module marks obstacle locations on the landmark library and updates the obstacle avoidance path in real time; Step 4: Land the drone and reset the system When the drone approaches the nest, the automatic homing module identifies the type of drone and plans the docking location and descent method for the drone; the drone executes a precise landing procedure, including adjusting the flight attitude and speed to ensure a safe landing; After the drone lands, the unit reset procedure is executed, including turning off the flight mode, performing hardware self-test and data recording; the drone's power management module monitors the power after landing and initiates emergency procedures when necessary.

Images