Research On Key Technology Of Quadcopter Control System

The quadcopter,as a branch of rotor-typed unmanned aerial vehicle,becomes the research hotspot in the field of MAV with it's characteristics of vertical take-off and landing,insensibility to environment varying,high mobility and stability,easy operation.The quadcopter system is highly-coupled,under-actuated and non-linear inherently,which poses a great challenge to the micro-processor,the sensor,the mechanism,the navigation and control algorithm in terms of system stability.Therefore,the research on the quadcopter is surely meaningful and it provides the development of the fore-mentioned fields with an efficient platform.This paper launches the research from quadcopter control system,and focus on the key technology of quadcopter control system.The key technology of quadcopter control system mainly divided into three aspects: flying platform building,navigation algorithm design and control algorithm research.Firstly,we study the mathematical model of quadcopter,and then the paper describes the Euler angle and quaternion attitude description method,studies the quadcopter dynamic model to establish the quadcopter attitude model and the position model.The quadcopter flying platform is one of the key technique of quadcopter that includes hardware platform design and software platform design.The hardware platform selects STM32F4 as main control chip,the MEMS gyroscope,accelerometer and compass as the attitude sensor,while it adds berometer and GPS as the position sensor.The software platform?C/OS-III as application platform which improves the real-time performance and stability of the quadcopter.Aimed at the shortcomings of high accelerometer noise,this paper design a spring-mass vibration isolation system to eliminate the vibration of the body vibration.Further more,this paper designed calibration system for gyroscope,accelerometer,magnetic sensor calibration and calibration that improved the measurement accuracy.The quadcopter flying platform is one of the key technique of quadcopter that includes hardware platform design and software platform design.The hardware platform selects STM32F4 as main control chip,the MEMS gyroscope and the accelerometer as the main sensor,while it adds the geomagnetic compass,berometer and GPS as the auxiliary sensor.The software platform ? C/OS-III as application platform which improves the real-time performance and stability of the quadcopter.Aimed at the shortcomings of high accelerometernoise,this paper design a spring-mass vibration isolation system to eliminate the vibration of the body vibration.Further more,this paper designed calibration system for gyroscope,accelerometer,magnetic sensor calibration and calibration that improved the measurement accuracy.Further more,for quadcopter navigation algorithm,the accumulation error of the estimate posture of traditional single sensor and position is getting bigger and bigger.Therefore,this paper adopted Square-Root Unscented Kalman Filter,and used gyroscope and accelerometer as the data source to establish the system state updating equation.At the same time,the system measurement equation was established by using magnetometer,GPS and barometer as the observation sensor.Through experiment verify the validity and accuracy of the algorithm.Finally,we design two controller,based on the simplification of the quadcopter flight model.They are analyzed from the characteristics of cascade and nonholonomic system,and the classical PID method is used to control the design.The nonlinear control is made by the idea of Backstepping which breaks down a complex nonlinear system into subsystems that do not exceed the order of the system.Then Lyapunov function and intermediate virtual control are designed for each subsystem and the corresponding Lyapunov function is selected for each subsystem,find the actual control of the system input through a step by step iteration.It analyzed and verified the control effect through the simulation.