Research On Attitude Algorithm And Control System Of Spherical Robot

The driving unit,sensor and control unit of spherical robot are all contained in the spherical shell,which has the complex characteristics of nonlinearity,strong coupling and multi variable,and is an ideal platform for the research of multi-sensor data fusion and control theory.This thesis focuses on the structure design,kinematics and dynamics modeling,attitude calculation and control system design of spherical robot.Firstly,based on the movement mode of center of mass migration,a spherical robot based on omnidirectional wheel drive platform is designed and its system design is introduced in detail.The spherical robot has three symmetrical omnidirectional wheels.By controlling the omnidirectional wheel moving inside the sphere,the whole center of mass will be offset,and the ball robot can move rapidly and comprehensively.Then,the kinematics / dynamics model of the spherical robot platform is built.The plane sphere model is used to describe the sphere in the inertial coordinate system.The angular velocity of the driving platform is transformed into the inertial coordinate system by rotating matrix.The kinematic model of the driving platform of the spherical robot is established.Based on Lagrange equation,spherical robot is divided into three parts: spherical shell,driving platform and omnidirectional wheel.According to the virtual work principle,the dynamic model of spherical robot is established.The simulation results show that the model of spherical robot is reasonable.Secondly,in order to improve the accuracy and speed of the attitude solution of spherical robot,a new attitude algorithm based on the fusion of complementary filtering and particle filter is proposed.For the attitude calculation of yaw angle,considering the limitation of the control unit of spherical robot,in order to avoid the influence of the noise of the magnetometer on the pitch angle and roll angle,the magnetometer noise is observed by information entropy,and then the magnetometer data is filtered to obtain the yaw angle.The weight value of the complementary filter is dynamically adjusted according to the noise size of the magnetometer.For the attitude solution of pitch angle and roll angle,particle filter and complementary filtering algorithm are combined to reduce the errors caused by the temperature drift and vibration of gyroscope,so as to obtain the real-time position and attitude information of spherical robot quickly and accurately,and improve the control accuracy of spherical robot.The simulation results show that the attitude algorithm designed in this thesis is effective.Finally,a spherical robot control system based on model reference adaptive control is implemented.Based on the experimental platform of spherical robot system,the experiment results show that the spherical robot system designed in this thesis is feasible and effective.