Research On Obstacle Avoidance Of Quadcopter Based On ToF And Visual Detection

The quadcopter is a classic type of multi-rotor aircraft.By changing the speed of four propeller motors,the motion of six degrees of freedom can be carried out.Its architecture is simple,easy to control,easy to take off,and low cost and power consumption.Therefore,after years of research and development,its control and flight technology has been mature,and has played a unique advantage in many fields,such as military and civil aviation,such as aerial photography.Because of its flexible flight,it can work in many environments and spaces.With the continuous improvement of endurance,the four rotor aircraft will be widely applied and will have good application prospects.Through the analysis of structural characteristics of four rotor aircraft,air dynamics in the model,based on the powerful function of STM32 based on embedded MCU as the core of the vehicle control unit;in attitude measurement,using the latest research and development of the company and series products MEMS LSM6DS3TR inertial measurement unit is introduced,and the gyroscope and accelerometer calibration to improve the measurement accuracy of three axis attitude angle and acceleration,lay the foundation for stable flight control aircraft;the attitude sensor and magnetometer measured data through.data fusion,complementary filtering algorithm,for the attitude solution using four element method to obtain the Euler angle.In the aspect of obstacle detection,through the study on non-contact detection principle,and the current sensor of the latest international comparison and selection,the combination of the two sensors to show their respective advantages,so the company continued to use the VL53L0X TOF laser sensor based on the principle of the latest development,which has strong anti-interference the performance and accuracy of measurement.In the basis of feasibility research and application performance,at the same time to ensure the accuracy of detection,reduce the operation and processing of the data,for the monocular vision based visual detection technology,through the calibration of the camera qualitatively,and the establishment of pinhole imaging model,can achieve the image acquisition,provide convenience for eye recognition.Through the hardware assembly and software compilation and debugging of the four rotor aircraft,the distance precision can be completed in the static condition.After setting up the obstacle avoidance conditions,we can make the corresponding decision and obstacle avoidance decisions for the aircraft when detecting obstacles,so as to ensure that the aircraft can avoid dangerous objects.Finally,through the analysis and modification of the existing problems in the research process,the accuracy of the detection can be further improved and the performance of obstacle avoidance can be improved.