Unmanned aerial vehicle path planning method based on deep learning

Abstract

The invention relates to the technical field of unmanned aerial vehicle systems, and discloses an unmanned aerial vehicle path planning method based on deep learning, and the method comprises the following working steps: 1, obtaining the flight environment data of an unmanned aerial vehicle in real time through a data collection module; 2, transmitting the collected data to a data storage module for storage, and providing the data for a data training module to use; and step 3, the data training module trains the stored data by using a deep learning algorithm, and optimizes a path planning model in the artificial intelligence module. According to the method, intelligent planning and optimization of the path of the unmanned aerial vehicle are realized through a deep learning technology, and the flight safety, flexibility and robustness of the unmanned aerial vehicle in a complex environment are improved in combination with mechanisms such as real-time data acquisition, safety early warning, system self-optimization and upgrading and the like. Through the 5G communication and data encryption technology, the high efficiency and safety of data transmission are ensured, and the method has wide application prospects in the fields of logistics, agriculture, environment monitoring, search and rescue and the like.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle systems, and in particular, relates to a unmanned aerial vehicle path planning method based on deep learning. Background Art

[0002] With the rapid development of drone technology, it has shown great potential in many fields such as aerial photography, environmental monitoring, and logistics distribution. However, the complex and changing flight environment has put forward higher requirements on the path planning ability of drones. Traditional path planning methods are difficult to cope with real-time changing environmental factors, which limits the application scope and efficiency of drones.

[0003] In view of this, the present invention is proposed. Summary of the invention

[0004] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0005] A UAV path planning method based on deep learning includes the following working steps:

[0006] Step 1: Obtain UAV flight environment data in real time through the data acquisition module;

[0007] Step 2: The collected data is transferred to the data storage module for storage and used by the data training module;

[0008] Step 3: The data training module uses deep learning algorithms to train the stored data and optimize the path planning model in the artificial intelligence module;

[0009] Step 4: The artificial intelligence module generates the optimal path planning strategy based on the current environment data and the optimized model;

[0010] Step 5: Send the path planning instructions to the drone through the 5G communication module, and receive the drone status information in real time to achieve closed-loop control;

[0011] Step 6: Use the weather warning module to obtain and analyze weather data in the flight area in real time, predict weather change trends, evaluate the impact of weather on drone flight, and issue warning signals or adjust flight plans when necessary;

[0012] Step 7: The location warning module obtains and processes the terrain data of the flight area, builds a three-dimensional environmental model, evaluates the collision risk between the drone and the terrain or obstacles, and issues a warning signal or automatically adjusts the flight path when potential risks are detected;

[0013] Step 8: During the execution of the drone, feedback information is continuously collected and the model is adjusted through the self-optimization unit to achieve self-improvement and growth of the system.

[0014] As a preferred embodiment of the present invention, the artificial intelligence module is used to execute a deep learning algorithm, including but not limited to a convolutional neural network (CNN), a recurrent neural network (RNN) or a deep reinforcement learning model, to process and analyze the environmental data collected by the drone and generate an optimal path planning strategy; the data acquisition module is configured with a sensor group, including but not limited to GPS, a lidar, and a camera, for real-time collection of data such as the drone's location information, surrounding environment images, and obstacle information.

[0015] As a preferred embodiment of the present invention, the data storage module is used to store the data collected by the data acquisition module, the training model of the artificial intelligence module, the historical path planning records and learning results, and supports efficient data retrieval and management; the data training module works together with the artificial intelligence module to train and optimize the deep learning model using the data in the storage module to improve the accuracy and adaptability of path planning. The data training module supports both online learning and offline training modes to meet the needs of different scenarios.

[0016] As a preferred embodiment of the present invention, the 5G communication module provides a high-speed, low-latency data transmission channel to ensure real-time communication between the UAV and the ground control station or other UAVs, and realizes data sharing, command transmission and collaborative operation; the self-optimization unit automatically adjusts the parameters and structure of the deep learning model based on historical path planning effect feedback to achieve self-optimization and performance improvement.

[0017] As a preferred embodiment of the present invention, the environmental adaptability learning unit analyzes data features under different flight environments, learns and adapts to complex and changeable flight conditions, and improves the flexibility and robustness of path planning.

[0018] As a preferred embodiment of the present invention, it also includes a data security and encryption module, which includes a data encryption unit, an access control unit, and a security audit unit. The data encryption unit is responsible for encrypting all sensitive data, including but not limited to the original data collected by the drone, training models, path planning strategies, etc., to ensure the security of data during transmission and storage.

[0019] As a preferred embodiment of the present invention, the access control unit implements a strict access control policy, and only authorized users or systems can access specific data or perform specific operations to prevent unauthorized access and data leakage; the security audit unit records system operation logs, including but not limited to data access, modification, and deletion behaviors, in order to conduct security audits and track potential security threats.

[0020] As a preferred embodiment of the present invention, it also includes a path planning active update and upgrade module, which includes a dynamic environment monitor, an intelligent path optimizer, a path planning strategy library, and an upgrade management module, wherein the dynamic environment monitor continuously monitors changes in the environment around the drone, including newly emerging obstacles, weather conditions, and airspace restrictions, and provides real-time data support for active updates of path planning; the intelligent path optimizer integrates a variety of path optimization algorithms and deep learning models, and automatically selects a suitable algorithm for path optimization based on the dynamic environment monitoring results, ensuring that the drone can flexibly respond to environmental changes; the path planning strategy library maintains a library containing a variety of path planning strategies, including basic strategies, advanced strategies, and customized strategies for specific scenarios, and supports dynamic loading and updating of strategies; the upgrade management module is responsible for the software upgrade of the path planning system itself, including the update of the deep learning model, the improvement of the path optimization algorithm, etc., and ensures that the system is always kept up to date through online or offline upgrades.

[0021] As a preferred embodiment of the present invention, the weather warning module includes a meteorological data interface, a weather analysis engine, and a warning decision unit, wherein the meteorological data interface is connected to the meteorological data service provider to obtain real-time weather data on the UAV flight path, including but not limited to temperature, humidity, wind speed, wind direction, rainfall probability, and thunderstorm warning; the weather analysis engine uses meteorological data and deep learning models to accurately analyze the weather conditions in the flight area, predict weather change trends in the future, and evaluate the impact on UAV flight; warning decision unit: based on the weather analysis results, automatically determine whether it is necessary to adjust the flight plan or issue a warning signal. In the event of severe weather, it will recommend alternative routes or suggest suspending the flight.

[0022] As a preferred embodiment of the present invention, the position warning module includes a terrain data interface, a three-dimensional environment modeling unit, a collision risk assessor, and a position warning decision unit, wherein the terrain data interface is connected to a high-precision terrain database to obtain terrain information on the UAV flight path, including but not limited to the location and height of mountains, rivers, lakes, and buildings; the three-dimensional environment modeling unit: based on the terrain data, constructs a three-dimensional environment model of the flight area to accurately simulate the terrain undulations and obstacle distribution; the collision risk assessor combines the flight altitude, speed, direction and three-dimensional environment model of the UAV to assess the risk of collision between the UAV and the terrain or obstacles, and calculates the safe flight boundary; when the position warning decision unit detects a potential collision risk, it immediately sends a warning signal to the UAV, and automatically adjusts the flight path or recommends obstacle avoidance measures to ensure the safe flight of the UAV.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. The deep learning-based UAV path planning method of the present invention integrates advanced artificial intelligence, data acquisition, data storage, data training and 5G communication technology to construct an efficient, intelligent and adaptive UAV path planning system, which not only improves the path planning capability of UAVs in complex environments, but also realizes the self-improvement and continuous growth of the system, and has broad application prospects and important social value.

[0025] 2. The deep learning-based UAV path planning method of the present invention can realize safety warning and risk assessment. The weather warning module obtains and analyzes the weather data of the flight area in real time, predicts weather change trends, evaluates the impact of weather on flight, and issues warnings or adjusts flight plans when necessary; the position warning module constructs a three-dimensional environmental model, evaluates the collision risk of the UAV with terrain or obstacles, and issues warnings or automatically adjusts the path when potential risks are detected.

[0026] 3. The deep learning-based UAV path planning method of the present invention can achieve system self-improvement and upgrading. The self-optimization unit automatically adjusts the parameters and structure of the deep learning model based on historical path planning effect feedback to achieve self-optimization; the path planning active update and upgrade module includes a dynamic environment monitor, an intelligent path optimizer, a path planning strategy library and an upgrade management module to ensure that the system can flexibly respond to environmental changes and keep up to date.

[0027] 4. The method of the present invention realizes intelligent planning and optimization of drone paths through deep learning technology, and combines real-time data collection, safety warning, system self-optimization and upgrading mechanisms to improve the flight safety, flexibility and robustness of drones in complex environments. At the same time, through 5G communication and data encryption technology, the efficiency and security of data transmission are ensured. Therefore, the method of the present invention has broad application prospects in logistics, agriculture, environmental monitoring, search and rescue and other fields.

[0028] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In the attached picture:

[0030] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The following embodiments are used to illustrate the present invention.

[0032] A UAV path planning method based on deep learning, such as Figure 1As shown, the following working steps are included:

[0033] Step 1: Obtain UAV flight environment data in real time through the data acquisition module;

[0034] Step 2: The collected data is transferred to the data storage module for storage and used by the data training module;

[0035] Step 3: The data training module uses deep learning algorithms to train the stored data and optimize the path planning model in the artificial intelligence module;

[0036] Step 4: The artificial intelligence module generates the optimal path planning strategy based on the current environment data and the optimized model;

[0037] Step 5: Send the path planning instructions to the drone through the 5G communication module, and receive the drone status information in real time to achieve closed-loop control;

[0038] Step 6: Use the weather warning module to obtain and analyze weather data in the flight area in real time, predict weather change trends, evaluate the impact of weather on drone flight, and issue warning signals or adjust flight plans when necessary;

[0039] Step 7: The location warning module obtains and processes the terrain data of the flight area, builds a three-dimensional environmental model, evaluates the collision risk between the drone and the terrain or obstacles, and issues a warning signal or automatically adjusts the flight path when potential risks are detected;

[0040] Step 8: During the execution of the drone, feedback information is continuously collected and the model is adjusted through the self-optimization unit to achieve self-improvement and growth of the system.

[0041] Specifically, the artificial intelligence module is used to execute deep learning algorithms, including but not limited to convolutional neural networks (CNN), recurrent neural networks (RNN) or deep reinforcement learning models, to process and analyze environmental data collected by the drone and generate an optimal path planning strategy; the data acquisition module is configured with a sensor group, including but not limited to GPS, lidar, and cameras, which are used to collect real-time data such as the drone's location information, surrounding environment images, and obstacle information.

[0042] Specifically, the data storage module is used to store the data collected by the data acquisition module, the training model of the artificial intelligence module, historical path planning records and learning results, and supports efficient data retrieval and management; the data training module works together with the artificial intelligence module to train and optimize the deep learning model using the data in the storage module to improve the accuracy and adaptability of path planning. The data training module supports both online learning and offline training modes to meet the needs of different scenarios.

[0043] Among them, the 5G communication module provides a high-speed, low-latency data transmission channel to ensure real-time communication between the drone and the ground control station or other drones, and realize data sharing, command transmission and collaborative operation; the self-optimization unit automatically adjusts the parameters and structure of the deep learning model based on historical path planning effect feedback to achieve self-optimization and performance improvement. The environmental adaptability learning unit analyzes the data characteristics under different flight environments, learns and adapts to complex and changeable flight conditions, and improves the flexibility and robustness of path planning.

[0044] Furthermore, it also includes a data security and encryption module, which includes a data encryption unit, an access control unit, and a security audit unit, wherein the data encryption unit is responsible for encrypting all sensitive data, including but not limited to the original data collected by the drone, training models, path planning strategies, etc., to ensure the security of data during transmission and storage. The access control unit implements a strict access control strategy, and only authorized users or systems can access specific data or perform specific operations to prevent unauthorized access and data leakage; the security audit unit records system operation logs, including but not limited to data access, modification, and deletion behaviors, in order to conduct security audits and track potential security threats.

[0045] Furthermore, it also includes a path planning active update and upgrade module, which includes a dynamic environment monitor, an intelligent path optimizer, a path planning strategy library, and an upgrade management module. The dynamic environment monitor continuously monitors changes in the environment around the drone, including newly emerging obstacles, weather conditions, and airspace restrictions, and provides real-time data support for active updates of path planning; the intelligent path optimizer integrates a variety of path optimization algorithms and deep learning models, and automatically selects a suitable algorithm for path optimization based on the dynamic environment monitoring results, ensuring that the drone can flexibly respond to environmental changes; the path planning strategy library maintains a library containing a variety of path planning strategies, including basic strategies, advanced strategies, and customized strategies for specific scenarios, and supports dynamic loading and updating of strategies; the upgrade management module is responsible for software upgrades of the path planning system itself, including updates of deep learning models, improvements to path optimization algorithms, etc., and ensures that the system is always kept up to date through online or offline upgrades.

[0046] The weather warning module includes a meteorological data interface, a weather analysis engine, and a warning decision unit, wherein the meteorological data interface is connected to the meteorological data service provider to obtain real-time weather data on the UAV flight path, including but not limited to temperature, humidity, wind speed, wind direction, rainfall probability, and thunderstorm warning; the weather analysis engine uses meteorological data and deep learning models to accurately analyze the weather conditions in the flight area, predict weather change trends in the future, and evaluate the impact on UAV flight; the warning decision unit: based on the weather analysis results, automatically determines whether it is necessary to adjust the flight plan or issue a warning signal. In the event of severe weather, it will recommend alternative routes or suggest suspending the flight.

[0047] The position warning module includes a terrain data interface, a three-dimensional environment modeling unit, a collision risk assessor, and a position warning decision unit, wherein the terrain data interface is connected to a high-precision terrain database to obtain terrain information on the UAV's flight path, including but not limited to the location and height of mountains, rivers, lakes, and buildings; the three-dimensional environment modeling unit: based on the terrain data, constructs a three-dimensional environmental model of the flight area to accurately simulate the undulations of the terrain and the distribution of obstacles; the collision risk assessor combines the UAV's flight altitude, speed, direction and three-dimensional environmental model to assess the risk of collision between the UAV and the terrain or obstacles, and calculates the safe flight boundary; when the position warning decision unit detects a potential collision risk, it immediately sends a warning signal to the UAV, and automatically adjusts the flight path or recommends obstacle avoidance measures to ensure the safe flight of the UAV.

[0048] The deep learning-based UAV path planning method of the present invention constructs an efficient, intelligent and adaptive UAV path planning system by integrating advanced artificial intelligence, data acquisition, data storage, data training and 5G communication technology. It not only improves the path planning capability of UAVs in complex environments, but also realizes the self-improvement and continuous growth of the system. It has broad application prospects and important social value.

[0049] The deep learning-based UAV path planning method of the present invention can achieve safety warning and risk assessment. The weather warning module obtains and analyzes weather data of the flight area in real time, predicts weather change trends, evaluates the impact of weather on flight, and issues warnings or adjusts flight plans when necessary; the position warning module constructs a three-dimensional environmental model, evaluates the collision risk of the UAV with terrain or obstacles, and issues warnings or automatically adjusts the path when potential risks are detected.

[0050] The deep learning-based UAV path planning method of the present invention can achieve system self-improvement and upgrading. The self-optimization unit automatically adjusts the parameters and structure of the deep learning model based on historical path planning effect feedback to achieve self-optimization; the path planning active update and upgrade module includes a dynamic environment monitor, an intelligent path optimizer, a path planning strategy library and an upgrade management module to ensure that the system can flexibly respond to environmental changes and keep the latest status.

[0051] The method of the present invention realizes intelligent planning and optimization of drone paths through deep learning technology, and combines real-time data collection, safety warning, system self-optimization and upgrading mechanisms to improve the flight safety, flexibility and robustness of drones in complex environments. At the same time, through 5G communication and data encryption technology, the efficiency and security of data transmission are ensured. Therefore, the method of the present invention has broad application prospects in logistics, agriculture, environmental monitoring, search and rescue and other fields.

[0052] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A UAV path planning method based on deep learning, characterized in that: The following steps are included: Step 1: Obtain UAV flight environment data in real time through the data acquisition module; Step 2: The collected data is transferred to the data storage module for storage and used by the data training module; Step 3: The data training module uses deep learning algorithms to train the stored data and optimize the path planning model in the artificial intelligence module; Step 4: The artificial intelligence module generates the optimal path planning strategy based on the current environment data and the optimized model; Step 5: Send the path planning instructions to the drone through the 5G communication module, and receive the drone status information in real time to achieve closed-loop control; Step 6: Use the weather warning module to obtain and analyze weather data in the flight area in real time, predict weather change trends, evaluate the impact of weather on drone flight, and issue warning signals or adjust flight plans when necessary; Step 7: The location warning module obtains and processes the terrain data of the flight area, builds a three-dimensional environmental model, evaluates the collision risk between the drone and the terrain or obstacles, and issues a warning signal or automatically adjusts the flight path when potential risks are detected; Step 8: During the execution of the drone, feedback information is continuously collected and the model is adjusted through the self-optimization unit to achieve self-improvement and growth of the system.

2. The deep learning-based UAV path planning method according to claim 1, characterized in that: The artificial intelligence module is used to execute deep learning algorithms, including but not limited to convolutional neural networks (CNN), recurrent neural networks (RNN) or deep reinforcement learning models, to process and analyze environmental data collected by the drone and generate an optimal path planning strategy; the data acquisition module is configured with a sensor group, including but not limited to GPS, lidar, and cameras, which are used to collect real-time data such as the drone's location information, surrounding environment images, and obstacle information.

3. The UAV path planning method based on deep learning according to claim 1, characterized in that: The data storage module is used to store the data collected by the data acquisition module, the training model of the artificial intelligence module, the historical path planning records and learning results, and supports efficient data retrieval and management; the data training module works together with the artificial intelligence module to train and optimize the deep learning model using the data in the storage module to improve the accuracy and adaptability of path planning. The data training module supports both online learning and offline training modes to meet the needs of different scenarios.

4. The UAV path planning method based on deep learning according to claim 1, characterized in that: The 5G communication module provides a high-speed, low-latency data transmission channel to ensure real-time communication between the drone and the ground control station or other drones, and realize data sharing, command transmission and collaborative operation; The self-optimization unit automatically adjusts the parameters and structure of the deep learning model based on historical path planning effect feedback to achieve self-optimization and performance improvement.

5. The UAV path planning method based on deep learning according to claim 1, characterized in that: The environmental adaptability learning unit analyzes data features under different flight environments, learns and adapts to complex and changeable flight conditions, and improves the flexibility and robustness of path planning.

6. The UAV path planning method based on deep learning according to claim 1, characterized in that: It also includes a data security and encryption module, which includes a data encryption unit, an access control unit, and a security audit unit. The data encryption unit is responsible for encrypting all sensitive data, including but not limited to the original data collected by the drone, training models, path planning strategies, etc., to ensure the security of data during transmission and storage.

7. The deep learning-based UAV path planning method according to claim 6, characterized in that: The access control unit implements a strict access control policy, and only authorized users or systems can access specific data or perform specific operations to prevent unauthorized access and data leakage; the security audit unit records system operation logs, including but not limited to data access, modification, and deletion behaviors, in order to conduct security audits and track potential security threats.

8. The UAV path planning method based on deep learning according to claim 1, characterized in that: It also includes a path planning active update and upgrade module, which includes a dynamic environment monitor, an intelligent path optimizer, a path planning strategy library, and an upgrade management module. The dynamic environment monitor continuously monitors changes in the environment around the drone, including new obstacles, weather conditions, and airspace restrictions, providing real-time data support for active updates of path planning. The intelligent path optimizer integrates multiple path optimization algorithms and deep learning models, and automatically selects the appropriate algorithm for path optimization based on the dynamic environment monitoring results, ensuring that the drone can flexibly respond to environmental changes; the path planning strategy library maintains a library containing multiple path planning strategies, including basic strategies, advanced strategies, and customized strategies for specific scenarios, and supports dynamic loading and updating of strategies; The upgrade management module is responsible for the software upgrade of the path planning system itself, including the update of the deep learning model, the improvement of the path optimization algorithm, etc. It ensures that the system is always kept up to date through online or offline upgrades.

9. The UAV path planning method based on deep learning according to claim 1, characterized in that: The weather warning module includes a meteorological data interface, a weather analysis engine, and a warning decision unit, wherein the meteorological data interface is connected to the meteorological data service provider to obtain real-time weather data on the UAV flight path, including but not limited to temperature, humidity, wind speed, wind direction, rainfall probability, and thunderstorm warning; the weather analysis engine uses meteorological data and deep learning models to accurately analyze the weather conditions in the flight area, predict weather change trends in the future, and evaluate the impact on UAV flight; the warning decision unit: based on the weather analysis results, automatically determines whether it is necessary to adjust the flight plan or issue a warning signal. In the event of severe weather, it will recommend alternative routes or suggest suspending the flight.

10. The deep learning-based UAV path planning method according to claim 1, characterized in that: The position warning module includes a terrain data interface, a three-dimensional environment modeling unit, a collision risk assessor, and a position warning decision unit, wherein the terrain data interface is connected to a high-precision terrain database to obtain terrain information on the UAV's flight path, including but not limited to the location and height of mountains, rivers, lakes, and buildings; the three-dimensional environment modeling unit: based on the terrain data, constructs a three-dimensional environmental model of the flight area to accurately simulate the undulations of the terrain and the distribution of obstacles; the collision risk assessor combines the UAV's flight altitude, speed, direction and three-dimensional environmental model to assess the risk of collision between the UAV and the terrain or obstacles, and calculates the safe flight boundary; when the position warning decision unit detects a potential collision risk, it immediately sends a warning signal to the UAV, and automatically adjusts the flight path or recommends obstacle avoidance measures to ensure the safe flight of the UAV.

Images