Design Of Micro-attack UAV

The miniature multi-rotor UAV has a simple structure,can carry out three-dimensional space movement in any direction without rotating the fuselage,and combines the machine visionattacking type micro multi-rotor UAV,which has great application prospects in military and security.This project focuses on four key technologies of micro-attack drones:system design,integrated navigation,target recognition,and control systems.The main work and innovations of the project are as follows:Firstly,due to the micro-multi-rotor UAV with small load and limited anti-interference ability,the launch structure based on friction wheel is designed,which is light in weight,controllable in launch speed and small in recoil.In addition,a high-performance,ultra-low-power flight control system based on the MSP432 and an embedded machine vision system based on the Raspberry Pi are designed.Then,because the working environment of the UAV is complex and variable,and the precision strike has high requirements on navigation accuracy and robustness,an IMU/magnetometer/optical flow/GNSS integrated navigation system is designed.The calibration algorithm is designed based on the error model of IMU and magnetometer,and the quaternion attitude fusion algorithm based on UKF is designed.Its fusion precision and robustness are better than complementary filter fusion and EKF attitude fusion algorithm.In addition,according to the different characteristics of barometer,accelerometer,optical flow and GNSS,the KF-based height,velocity and position fusion algorithm is designed to reduce the noise of the four sensors andimprove the robustness of integrated navigationNext,according to the variable working environment of the recognition object and the high real-time requirements of the system,a recognition algorithm based on target feature matching is designed.The acquired image is corrected,locally enhanced,target segmented,refined,and a series of processes are used to obtain the feature parameters of the target to be identified,and the target is accurately identified by matching the H₁moment.In addition,the image decoupling is used to eliminate the influence of the attitude change of the aircraft on the target position.Finally,a cascade PID flight controller and a Yuntai controller were designed.By establishing a target tracking attack model,a target tracking algorithm based on KF prediction is designed to improve the target hit rate.The experimental results show that the IMU/magnetometer/optical flow/GNSS integrated navigation system has strong adaptability,and strong robustness.In addition,the recognition algorithm of target feature matching has good adaptability to illumination intensity transformation,rotation and scale transformation.Finally,the flight control and pan/tilt controller of the cascade PID can achieve the target tracking well,and there will be no loss of the target.With The KF algorithm predicting and optimizing in advance,The success rate of the system static strike is 81.9%,and the dynamic success rate is 71.4%.