Collision avoidance control is a key element of the intelligent operation of unmanned aerial vehicles. It is composed of the perception subsystem, decision subsystem and control subsystem. It is dependent on the analysis and processing of the data received by the optical flow, inclinometer/accelerometer, IR/Ultrasound, ultrasonic sensors and other sensors, in order to avoid obstacles in the flight, and optimize the route in the environment.
Perception subsystem
The perception subsystem is the basis and foundation of collision avoidance control. It is used to obtain the data of external environment and track targets through a variety of sensors. According to the characteristics of the sensors, the perception subsystem generally includes 3D imaging and lidar systems, IR/ultrasound system, ultrasonic system, inertial navigation system, barometer and other sensors and systems.
3D Imaging System
3D Imaging System is a ideal system for collision avoidance control. It can obtain the data of external environment that include the feature, pose and velocity of the obstacle through 3D, stereo, monocular imaging sensors such as cameras and infrared imagers, which provide the reliable information for the obstacle avoidance control of unmanned aerial vehicles.
Lidar system
Compared with the active imaging system, lidar is a kind of passive imaging system based on laser distance detection, which can send laser signals to the object, measure the distance between the UAV and the obstacle, and provide the range data for obstacle avoidance control.
IR/Ultrasound System
The IR/ultrasound system is an ideal solution for close range target detection. The principle of this system is similar to the bat echolocation system. It uses two cameras to track the target in 3D space in the short distance, and then use the infrared and ultrasonic devices to measure the relative distance and relative angular information between the UAV and the target.
Ultrasonic System
The ultrasonic system consists of several ultrasonic transducers, which can measure the distance of the obstacle use the time of flight of the ultrasonic signal. This system is suitable for measuring the obstacle which is close to the unmanned aerial vehicle.
Inertial Measurement Unit
The Inertial Measurement Unit generally contains accelerometers, inclinometers and gyroscopes, and is used to measure the attitude and position of the unmanned aerial vehicle. The data provided by the inertial measurement unit can effectively supplement the data of the optical sensors and other sensors, enhancing the reliability and accuracy of the collision avoidance control algorithm.
Barometer
Barometer is used to measure the relative altitude, which can effectively supplement the data of the optical sensors and other sensors, not only increasing the reliability and accuracy of the obstacle avoidance control, but also providing the appropriate information for the UAV to maintain its relative altitude and speed.
Decision subsystem
The decision subsystem is mainly based on the data of the perception system, then judge the path of UAV and make decision. According to the features of different sensors, the decision process can be divided into classification and tracking, mapping, route planning and other processes.
Classification and tracking
Based on the data of the perception subsystem, the decision subsystem can self-judge the type of the obstacle and the track the obstacle. This process can help the unmanned aerial vehicle to accurately recognize the characteristics of the obstacle, such as size, shape, distance, speed and other information.
Mapping
Then, the decision subsystem will use the data of the perception subsystem to map its surrounding area, generate a 3D image of the obstacle and its environment in the local coordinate system, and update the map to make sure that the information of the environment is accurate.
Route Planning
After getting the data of the environment, the decision subsystem can find the optimal route automatically through the path planning algorithm, and then customize the navigation path in automatic flight.
Control System
The control system is responsible for the implementation of the navigation path and the implementation of obstacle avoidance action. It processes the navigation path obtained by the decision subsystem, controls the attitude of the unmanned aerial vehicle and the execution of command, making sure the unmanned aerial vehicle complete the collision avoidance control in the guidance of the navigation path.
In conclusion, collision avoidance control is a key element of the intelligent operation of unmanned aerial vehicles, which is composed of the perception subsystem, decision subsystem and control subsystem. Through the analysis the data from a variety of sensors, the collision avoidance control can accurately recognize the environment and avoid the obstacle, providing a safe and efficient operation for the unmanned aerial vehicles.
